This article was downloaded by: [Department Of Fisheries]

On: 28 February 2013, At: 19:47

Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,
37-41 Mortimer Street, London W1T 3JH, UK

North American Journal of Aquaculture

Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/unaj20

Pumping Performance of a Slow-Rotating Paddlewheel

for Split-Pond Aquaculture Systems
a a
Travis W. Brown & Craig S. Tucker
a
U.S. Department of Agriculture, Agricultural Research Service, National Warmwater
Aquaculture Center, Post Office Box 38, Stoneville, Mississippi, 38776, USA
Version of record first published: 23 Jan 2013.

To cite this article: Travis W. Brown & Craig S. Tucker (2013): Pumping Performance of a Slow-Rotating Paddlewheel for Split-
Pond Aquaculture Systems, North American Journal of Aquaculture, 75:2, 153-158

To link to this article: http://dx.doi.org/10.1080/15222055.2012.743935

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions

This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to
anyone is expressly forbidden.

The publisher does not give any warranty express or implied or make any representation that the contents
will be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses should
be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims,
proceedings, demand, or costs or damages whatsoever or howsoever caused arising directly or indirectly in
connection with or arising out of the use of this material.
 North American Journal of Aquaculture 75:153–158, 2013
American Fisheries Society 2013
ISSN: 1522-2055 print / 1548-8454 online
DOI: 10.1080/15222055.2012.743935
TECHNICAL NOTE
Pumping Performance of a Slow-Rotating Paddlewheel
for Split-Pond Aquaculture Systems
Travis W. Brown and Craig S. Tucker*
U.S. Department of Agriculture, Agricultural Research Service, National Warmwater Aquaculture Center,
Post Office Box 38, Stoneville, Mississippi 38776, USA
Downloaded by [Department Of Fisheries] at 19:47 28 February 2013
Abstract
Commercial catfish farmers are intensifying production by
retrofitting ponds with variations of the partitioned aquaculture
system. The split-pond system is the most common variation used
commercially. The split-pond consists of a small fish-holding basin
connected to a waste treatment lagoon by two conduits. Water is
circulated between the two basins to remove fish waste and pro-
vide oxygenated water to the fish-holding basin. Although much
research has been devoted to algal and fish production dynamics
in variations of the partitioned aquaculture system, little informa-
tion is available on basic engineering considerations for devices
to circulate water in these systems. This study evaluated perfor-
mance characteristics for a slow-rotating paddlewheel pump that
looked at the relationships among power input, rotational speed
(circular tip velocity), water velocity, and water flow rate. Rota-
tional speeds of 1.0, 2.0, 3.0, and 4.0 rpm were evaluated in open
channels and in channels with fish barriers. Measured power input
was greater than the calculated power input for all four rotational
speeds and ranged from 0.11 to 1.80 hp. Water flow rate ranged
from 4,548–19,330 gal/min and water discharge per unit power
input decreased dramatically as rotational speed increased. Instal-
lation of fish barriers decreased channel open area and the resulting
frictional losses reduced water flow rates. Results from this study
provide initial pump performance data for designing split-pond
aquaculture systems.
U.S. fish farmers produced 699 million lb of Channel Cat-
fish Ictalurus punctatus and hybrid catfish (Channel Catfish ×
Blue Catfish I. furcatus) in 2003, representing more than half
of total U.S. aquaculture production (USDA 2004). Since 2003,
catfish production has declined by approximately 50% as a result
of high feed costs, competition from lower-priced imports, and
economically attractive land-use alternatives for farmers (USDA
2012). In an effort to remain competitive in the face of adverse
economic conditions, some catfish farmers have started using
intensive, outdoor culture systems based on the partitioned aqua-
*Corresponding author: craig.tucker@ars.usda.gov
Received September 6, 2012; accepted October 19, 2012
culture system (PAS) developed at Clemson University (Brune
et al. 2003, 2004).
The PAS physically separates pond-fish culture into distinct
physical, chemical, and biological processes, which are then
linked by a homogenous water velocity field maintained by
highly efficient, slow-rotating paddlewheels (SRPs). In practice,
the PAS is an outdoor, photoautotrophic recirculating aquacul-
ture system that combines the fish management advantages of
high-density raceway culture with the enhanced waste treatment
and oxygen production of high-rate algal oxidation ditches. De-
spite impressive catfish production in pilot-scale facilities, the
PAS has not been widely adopted for commercial catfish farm-
ing, perhaps because original designs required new facility con-
struction and intensive management oversight.
Two modifications of the PAS—in-pond raceways and split-
ponds—use the original PAS concept and apply it to commercial
settings by using existing earthen ponds as starting points for
construction. In-pond raceways consist of multiple fish culture
raceways constructed inside a pond in which water is circu-
lated between the raceways and the pond using SRPs, large
airlift pumps, or both (Brown et al. 2011; Brune et al. 2012;
Masser 2012). Split-ponds are constructed by dividing an ex-
isting catfish pond into two unequal basins with an earthen
levee. The smaller of the two basins (usually about 15–20% of
total water area) holds fish, and the larger basin (the lagoon)
treats fish waste and produces much of the oxygen for the sys-
tem through photosynthesis. Water is circulated between the la-
goon and fish-holding basin using high-volume pumps (Tucker
and Kingsbury 2010; Brune et al. 2012). Aquaculture perfor-
mance at pilot and commercial scales has been impressive in
in-pond raceways and split-ponds, with net annual production
(based on the total water area) ranging from 15,000 to more
than 22,000 lb/acre. Presently, the split-pond system is the most
commonly used PAS variant in commercial catfish farming, with
153
 154 BROWN AND TUCKER
more than 1,300 acres of ponds in use in Mississippi, Arkansas,
and Alabama.
The two critical design parameters for split-ponds are water
flow rate between the two basins and the amount of aeration
required in the fish-holding area (Brune et al. 2012). During
daylight and early evening, oxygenated water from the lagoon
is pumped through the fish-holding basin and return flow to
the lagoon removes fish metabolic wastes. At night, when dis-
solved oxygen concentrations decrease in the lagoon, circulation
between the two basins ceases and oxygen is provided by me-
chanical aerators in the fish-holding area. No attempt is made
to manage dissolved oxygen in the lagoon. In practice, pump
and aerator operation are controlled by oxygen sensors located
in both sections of the split-pond.
Split-pond aeration requirements are calculated by deter-
mining the amount of aerator oxygen transfer needed to meet
Downloaded by [Department Of Fisheries] at 19:47 28 February 2013
fish respiratory demands. Other sources and sinks of oxygen,
such as free water-surface diffusion and plankton and benthic
metabolism are assumed to be insignificant at the high loading
rates for fish biomass in the fish-holding area. Channel Catfish
respiratory rates can be calculated from projected maximum
fish biomass, average fish size, and water temperature using
the equation developed by Boyd et al. (1978). Standard oxygen
transfer rates are available for several commonly used catfish
pond aerators (Boyd and Ahmad 1987; Boyd 1998), which can
then be corrected to field conditions (Boyd and Watten 1989) to
determine aeration requirements.
Pumping rate estimates are based on the assumption that fish
oxygen requirements are met during daylight and early evening
by oxygen in water pumped into the fish-holding area from the
lagoon (Brune et al. 2012). A simple mass balance is used to cal-
culate flow rate (volume / time) by dividing fish respiratory rate
(oxygen mass / time) by the minimum desired dissolved oxy-
gen concentration (oxygen mass / volume). Required water flow
varies with time as fish grow and water temperature changes.
Pumps used in the original PAS moved water continuously
throughout the system and through fish-holding cells. Large
SRPs were identified as the most efficient and reliable method
of pumping large water volumes against low head. Paddlewheel
pumps in the PAS operate in open channels and have relatively
shallow wetted depths (1.5 ft). Because of the shallow wetted
depth, relatively long paddles are needed to obtain the high water
flow rates required in the PAS (for example, the required flow
rate in a 2.0-acre PAS is approximately 24,000 gal/min). Long,
shallow paddles cannot be used in split-ponds because water
conduits (open channels or culverts) between the lagoon and the
fish-holding area must be as narrow as possible to reduce costs
and to allow bridges or other connections to cross the conduit so
that feed trucks, grass mowers, and other farm equipment have
access across the levee. Accordingly, the paddlewheel pumps
used in current split-ponds have greater wetted depth (3.7 ft) to
reduce conduit width while maintaining high water flows.
The design of any system using pumped fluids relies on
performance data to determine the best pump type and relation-
ships among power requirement, pump speed, flow rate, and
other operational characteristics. Performance data are abun-
dant for most pump types, but no information is available for
the SRP pumps used in split-ponds. Such information is needed
as commercialization of the split-pond concept increases. This
study therefore determined performance characteristics for a
SRP pump that examined the relationships among power input,
rotational speed (circular tip velocity), water velocity, and water
flow rate.
An important aspect of split-pond operation is preventing fish
from escaping the fish holding area by using screens or barriers
across the two conduits connecting the fish-holding basin and
the lagoon. Fish barriers reduce channel cross-sectional area
and restrict water flow to some degree depending on the type of
screen and opening size. Therefore, we also evaluated changes
in water flow caused by two different types of fish barriers in
the open channels.
METHODS
This study was conducted in 2012 at the Thad Cochran Na-
tional Warmwater Aquaculture Center, Stoneville, Mississippi.
The split-pond system (Figure 1) consisted of a fish-holding
basin (0.15 acre, 4.9 ft average water depth), a waste-treatment
lagoon (0.55 acre, 3.0 ft average water depth), two open chan-
nels connecting the two basins, and one SRP pump. Channels
had concrete foundations and cinder-block walls. The channel
fitted with the SRP pump was 16.1 ft long × 10.3 ft wide, and
the other channel was 10.3 ft long × 10.0 ft wide. Both channels
had a total wall height of 6.0 ft and an average water depth of
approximately 4.0 ft. No fish were present in the system during
this study.
The SRP pump was constructed from six sections of mild-
steel plate (10 ft long × 4 ft wide × 0.02 ft thick), which were
FIGURE 1. The split-pond aquaculture system with slow-rotating paddle-
wheel (on left), concrete sluiceway channels, fish culture basin (foreground),
and waste treatment lagoon (background). [Figure available in color online.]
 TECHNICAL NOTE
evenly distributed and welded to a 0.72-ft-diameter central hub.
A central shaft (0.25 ft diameter) was fabricated from cold-rolled
steel and supported on double-row, spherical roller bearings
(Link-Belt, Indianapolis, Indiana). The central shaft attached to
the hub, which was designated as the hub assembly. Correct shaft
diameter was estimated using the following equation (Oberg
et al. 2008):
 determine power applied directly to the SRP shaft. The estimated
38P
D =3 , head loss and was estimated using equations presented by Brune
N
where D is the diameter of the shaft in inches, P is the power
transmitted in horsepower (hp), and N is the angular velocity
of the shaft in revolutions per minute (rpm). Struts (0.83 ft ×
0.02 ft, mild-steel plate sections) were welded in between the
paddles every 3.3 ft. The paddlewheel with hub assembly and
Downloaded by [Department Of Fisheries] at 19:47 28 February 2013
supports had an overall diameter of 8.71 ft and weighed approx-
imately 4,413 lb (Figure 2). Clearance between the bottom of
the channel and the outermost part (tip) of the paddles and outer
channel wall and the sides of the paddles of the SRP pump were
0.30 ft and 0.17 ft, respectively.
The SRP pump was powered by a 5.0-hp, three-phase electric
motor (Blador, Fort Smith, Arizona) that had a rated rotational
speed of 1,750 rpm. The motor was attached to an enclosed
gear drive (Nord Gear Corporation, Waunakee, Wisconsin) that
reduced shaft speed by a factor of 50:1. An 84-tooth sprocket
was attached to the output end of the gear drive, and a roller
chain (80H) connected the sprocket of the gear drive to a 20-
tooth sprocket (1:4.2) of the central shaft of a second gear drive.
The second gear drive (Falk, Milwaukee, Wisconsin) further
reduced shaft speed by a factor of 25:1. The combination of
motor, gear drives, and sprockets with chain produced a paddle-
wheel rotational speed of approximately 2.94 rpm at 30 Hz. A
general purpose open-loop vector AC drive (Saftronics, model
FIGURE 2. The slow-rotating paddlewheel used to circulate water in a split-
pond aquaculture system. [Figure available in color online.]
155
GP10E1ST32005, Fort Myers, Florida), designated as variable
frequency drive, was installed to allow control of the paddle-
wheel rotational speed.
The actual power supplied to the electric motor was obtained
directly from the variable frequency drive. Power input is most
important from an economic standpoint and therefore was the
power criterion used in this study. Thus, no attempt was made to
power required for maintaining flow in channels is due to friction
et al. (2004). The channel slope can be calculated by using the
Manning equation:
 2
(V)(n)
Channel slope = S = ,
(1.49)(R) 3
2
where S is the difference in water level between the front and
back of the SRP (ft/ft), V is the water velocity (ft/s), n is the
Manning’s coefficient (0.02), and R is the average water depth
of the SRP (ft). The channel slope was then used to determine
the required head:
Hf = L × S,
where Hf is the predicted frictional water head loss and L is
the channel length (ft). Additional friction head loss was also
calculated by,
Hb = V2/2g,
where Hb is the friction head due to bends and g is the accel-
eration due to gravity. The total head loss was then calculated
for each desired velocity by,
HT = Hf + Hb ,
where HT is the total required head that must be met across
the paddle of the SRP in order to achieve the desired velocity
through the channel. The estimated power required to achieve
the desired head was then calculated by,
(HT )(F)(W)
hp =
33,000(E)
where F is the water flow rate (gal/min), W is the specific weight
of water (8.34 lb/ft3), and E is the estimated pumping effi-
ciency (50%). Net torque (ft-lb) was estimated by multiplying
the wetted paddlewheel surface area (36.6 ft2) by half the dis-
tance of the wetted area to the shaft (2.53 ft) by the force of
the water pressure (lb/ft2) acting on the paddle, as determined
by HT .
Rotational speeds (1.0, 2.0, 3.0, and 4.0 rpm) of the SRP
pump were verified using a stopwatch and average water
velocity was determined for each rotational speed using a current
 156 BROWN AND TUCKER
meter (General Oceanics, model 2035 MKIV, Miami, Florida).
Five points were selected at equal distances across the open
channel of the inflow side to the fish culture basin and mea-
surements (n = 1,800) were made at 20, 50, and 80% of wa-
ter depth—a combination of two methods recommended by
Bankston and Baker (1995). Average water velocity was mul-
tiplied by the width of the channel and average water depth to
determine water flow rate.
Water velocity was measured with and without fish barriers in
the open channels to determine changes in water flow caused by
friction as water flowed through the screens. One set of barriers
(one for each channel) was constructed from polymer-coated
steel-mesh wire (0.08 ft × 0.08 ft openings) mounted to a square
aluminum tubing frame (0.10 ft2, wall thickness of 0.01 ft). The
second set of barriers tested used the same aluminum frame as
the first barrier type, but the mesh material was expanded metal
Downloaded by [Department Of Fisheries] at 19:47 28 February 2013
with 0.02 ft × 0.08 ft openings.
RESULTS AND DISCUSSION
Water flow rate at the four rotational speeds of 1.0, 2.0, 3.0,
and 4.0 rpm ranged from 4,548 to 19,330 gal/min with a mea-
sured power input of 0.11–1.80 hp (Figure 3). Measured power
input was greater than the estimated power requirement at all
water flow rates. This was probably due to mechanical losses
associated with the gear drives and the sprocket and chain as-
sembly. Brune et al. (2004) estimated a power requirement of ap-
proximately 0.5 hp to match a water flow rate of 23,578 gal/min
(47,156 gal·min−1·hp−1) in a PAS. However, he suggested that
the design should be selected for a “worst case” situation and
corrected for mechanical losses.
The SRP is an efficient pump relative to other devices used
for horizontal water circulation in aquaculture. Tucker and Har-
FIGURE 3. Measured power input and estimated power input to circulate
water with a slow-rotating paddlewheel in a split-pond aquaculture system. All
values are for open channels.
greaves (2004) calculated a water flow of approximately 10,000
gal/min for a 10-hp paddlewheel aerator (1,000 gal·min−1·hp−1
at full load). Brown and Torrans (unpublished) designed and
fabricated an airlift, U-tube-type aerator with a water discharge
of 5,550–8,882 gal/min or 1,526–896 gal·min−1·hp−1. A hori-
zontal axial-flow water circulator designed by Howerton et al.
(1994) had a maximum discharge rate per unit of power in-
put of approximately 18,000 gal·min−1·hp−1, which is com-
parable to the results we observed at 3.0 rpm with the SRP
pump.
When rotational speed of the SRP was accelerated, water
velocity and flow rate increased (Figure 4). Channel water ve-
locity ranged from 0.26 to 1.09 ft/s within the rotational speeds
tested. However, there was a dramatic decrease in water dis-
charge per unit power input (gal·min−1·hp−1) as rotational speed
increased. For example, at 1.0 rpm the water discharge per
unit power input was 40,729 gal·min−1·hp−1 compared with
10,749 gal·min−1·hp−1 at 4.0 rpm. Brune et al. (2012) also found
that slow-speed paddlewheels required excessive power input
at higher water velocities (approximately 0.50 ft/s or greater).
Howerton et al. (1994) observed a decrease in circulator effi-
ciency with increased discharge rates using a horizontal axial-
flow water circulator. In addition, water velocity had a direct
linear relationship to circular tip velocity (Figure 5). These re-
sults agree with the model by Drapcho (1993):
Vw = 0.6224 Vt ,
where Vw is water velocity and Vt is circulator tip velocity.
The amount of power required to circulate water in a split-
pond with a SRP pump depends on the maximum water flow
rate required by the system, which is a function of paddlewheel
FIGURE 4. Water discharge per unit power input and channel water velocity in
relation to rotational speed of a slow-rotating paddlewheel (SRP) in a split-pond
aquaculture system.

Downloaded by [Department Of Fisheries] at 19:47 28 February 2013
FIGURE 5. Relationship of water velocity and circular tip velocity of a slow-
rotating paddlewheel (SRP) in a split-pond aquaculture system. All values are
for open channels.
size and rotational speed. Because SRPs operate for long pe-
riods in the split-pond (12–18 h/d during midsummer), design
should account for the decreased pumping efficiency as rota-
tional speed increases. That is, SRP pumps should be sized
to achieve targeted flow rates using rotational speeds of 1.0–
2.0 rpm rather than attempting to use an undersized device at
very high rotational speeds. Elevated rotational speed for this
particular unit increased torque (Table 1). Field observations
indicate that the SRP pump we tested should not operate above
2.0 rpm for extended periods of time to minimize the likeli-
hood of paddlewheel cavitation, shaft torque surge, and reduced
operational life. In addition, correct design of SRPs should be
taken into consideration to reduce the possibility of mechanical
failure.
The presence of fish barriers reduced water flow, and flow
reductions increased as SRP rotational speed (water velocity) in-
creased (Figure 6). Fish barriers constructed of polymer-coated
steel-mesh wire reduced water flow rate less than barriers con-
structed out of expanded metal. For example, at 4.0 rpm, flow
rate was reduced from 19,330 to 17,320 gal/min for channels
with polymer-coated steel-mesh wire and to 14,847 gal/min for
channels with expanded metal barriers. This trend was expected
TABLE 1. Total required head (HT ) and resulting net torque on the shaft
of a slow-rotating paddlewheel at different rotational speeds in a split-pond
aquaculture system.
Rotational speed (rpm) HT (ft) Torque (ft-lb)
1.0 0.003
2.0 0.014
3.0 0.030
4.0 0.054
TECHNICAL NOTE 157
FIGURE 6. Resulting water flow rate at different rotational speeds of a slow-
rotating paddlewheel (SRP) using open channels (OC), polymer coated steel-
mesh wire fish barriers (PCSMW), or expanded metal fish barriers (EM) installed
in a split-pond aquaculture system.
based on differences in mesh open area in the steel-mesh wire
and expanded metal. The polymer-coated steel-mesh wire has
about 80% open area whereas the expanded metal has about
58% open area. The frame of the fish barriers further restricted
the open surface area of the channel by approximately 11%. The
total cross-sectional area of the open channel was 39.6 ft2, and
the combination of frame and barrier material reduced the open
surface area to 28.0 ft2 (71% of the open channel area) for the
polymer-coated steel-mesh wire and to 20.4 ft2 (52% of the open
channel area) for expanded metal. Though not reported, Brown
et al. (2011) observed reduced water flow in in-pond raceways
after installing smaller mesh (0.04 ft × 0.08 ft) fish barriers
before stocking fish.
Fish barriers should be installed in the sluiceway of the chan-
nels with a maximum mesh size to minimize the reduction in
water flow caused by frictional losses associated with reduced
open area in the channel. Biofouling or fouling with grass or
other debris will further reduce open channel area. Flow re-
ductions and labor needed for cleaning the fish barriers can be
reduced by using screens with large open areas to remove larger
debris before water flows through the fish barriers. Various types
of bar screens have been developed for use at the headworks of
wastewater treatment plants to prefilter the waste stream before
it enters the treatment plant (Vesilind 2003). Similar devices
should be developed for split-ponds to collect floating debris,
which would allow for easier maintenance of fish barriers.
In summary, SRP pumps operated at 1.0–2.0 rpm are highly
efficient, although efficiency decreases dramatically as rota-
tional speed increases. The SRP that we tested, with a rotational
17
speed of 1.0 rpm, will move approximately 4,500 gal/min at
78
an annual operating expense of US$30.24 at $0.12/kW-h. This
171
flow rate will support approximately 23,000 lb of catfish accord-
313
ing to the simple mass balance described in the Introduction.
 158 BROWN AND TUCKER
Information in this study can be used to design SRP pumps
for split-ponds and to adjust water flows throughout the sea-
son by using variable-speed motors to change rotational speed
and flow rates in response to fish growth. Future work should
address direct-drive systems for SRP pumps, which should im-
prove mechanical efficiency and increase operational life. Stud-
ies that focus on channel width, rotational speed of the SRP, and
associated water flow rate as it relates to power input would also
be useful.
ACKNOWLEDGMENTS
We thank all those who have taken the time to critically
review this manuscript as well as those who assisted with fund-
ing to support this research. James Santucci and Billy Rut-
land designed and fabricated the SRP pump used in this study
Downloaded by [Department Of Fisheries] at 19:47 28 February 2013
and their assistance throughout the study is gratefully acknowl-
edged. Mention of trade names or commercial products in this
publication is solely for the purpose of providing specific infor-
mation and does not imply recommendation or endorsement by
the U.S. Department of Agriculture.
REFERENCES
Bankston, J. D., Jr., and F. E. Baker. 1995. Open channel flow in aquaculture.
Southern Regional Aquaculture Center, Mississippi State University, Publi-
cation 374, Stoneville.
Boyd, C. E. 1998. Pond water aeration systems. Aquacultural Engineering 18:9–
40.
Boyd, C. E., and T. Ahmad. 1987. Evaluation of aerators for Channel Cat-
fish farming. Alabama Agricultural Experiment Station Auburn University
Bulletin 584.
Boyd, C. E., R. P. Romaire, and E. Johnston. 1978. Predicting early morning
dissolved oxygen concentrations in Channel Catfish ponds. Transactions of
the American Fisheries Society 107:484–492.
Boyd, C. E., and B. J. Watten. 1989. Aeration systems in aquaculture. Reviews
in Aquatic Sciences 1:425–472.
Brown, T. W., J. A. Chappell, and C. E. Boyd. 2011. A commercial-scale,
in-pond raceway system for ictalurid catfish production. Aquacultural Engi-
neering 44:72–79.
Brune, D. E., G. Schwartz, A. G. Eversole, J. A. Collier, and T. E. Schwedler.
2003. Intensification of pond aquaculture and high rate photosynthetic sys-
tems. Aquacultural Engineering 28:65–86.
Brune, D. E., G. Schwartz, A. G. Eversole, J. A. Collier, and T. E. Schwedler.
2004. Partitioned aquaculture systems. Pages 561–584 in C. S. Tucker and
J. A. Hargreaves, editors. Biology and culture of Channel Catfish. Elsevier,
New York.
Brune, D. E., C. Tucker, M. Massingill, and J. Chappell. 2012. Partitioned
aquaculture systems. Pages 308–340 in J. H. Tidwell, editor. Aquaculture
production systems. Wiley-Blackwell Scientific Publications, Ames, Iowa.
Drapcho, C. M. 1993. Modeling of algal productivity and diel oxygen pro-
files in the partitioned aquaculture system. Doctoral dissertation. Clemson
University, Clemson, South Carolina.
Howerton, R. D., C. E. Boyd, and B. J. Watten. 1994. Design and performance
of a horizontal, axial-flow water circulator. Journal of Applied Aquaculture
3:163–184.
Masser, M. P. 2012. In-pond raceways. Pages 387–393 in J. H. Tidwell, editor.
Aquaculture production systems. Wiley-Blackwell Scientific Publications,
Ames, Iowa.
Oberg, E., F. D. Jones, H. L. Horton, and H. H. Ryffel. 2008. Machinery’s
handbook, 28th edition. Industrial Press, New York.
Tucker, C. S., and J. A. Hargreaves. 2004. Pond water quality. Pages 215–278
in C. S. Tucker and J. A. Hargreaves, editors. Biology and culture of Channel
Catfish. Elsevier, New York.
Tucker, C. S., and S. K. Kingsbury. 2010. High-density split-pond systems offer
high output, low maintenance. Global Aquaculture Advocate 13:64–65.
USDA (U.S. Department of Agriculture). 2004. Catfish production.
USDA, National Agricultural Statistics Service, Washington, D.C. Avail-
able: usda01.library.cornell.edu/usda/nass/CatfProd//2000s/2004/CatfProd-
02-05-2004.pdf. (August 2012).
USDA (U.S. Department of Agriculture). 2012. Catfish production.
USDA, National Agricultural Statistics Service, Washington, D.C. Avail-
able: usda01.library.cornell.edu/usda/nass/CatfProd//2010s/2012/CatfProd-
01-27-2012.pdf. (August 2012).
Vesilind, P. A., editor. 2003. Wastewater treatment plant design. Water Environ-
ment Federation, Alexandria, Virginia.
This article was downloaded by: [Department Of Fisheries]
On: 28 February 2013, At: 21:53
Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,
37-41 Mortimer Street, London W1T 3JH, UK

North American Journal of Aquaculture

Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/unaj20

Characterization of Bacteria Isolated from Landlocked

Fall Chinook Salmon Eggs from Lake Oahe, South
Dakota
a a b
David J. Bergmann , Alicia Brakke & Michael E. Barnes
a
Department of Science, Black Hills State University, 1100 University Boulevard, Spearfish,
South Dakota, 57799, USA
b
McNenny State Fish Hatchery, South Dakota Game, Fish, and Parks, 19619 Trout Loop,
Spearfish, South Dakota, 57783, USA
Version of record first published: 07 Feb 2013.

To cite this article: David J. Bergmann , Alicia Brakke & Michael E. Barnes (2013): Characterization of Bacteria Isolated from
Landlocked Fall Chinook Salmon Eggs from Lake Oahe, South Dakota, North American Journal of Aquaculture, 75:2, 159-163

To link to this article: http://dx.doi.org/10.1080/15222055.2012.756439

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions

This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to
anyone is expressly forbidden.

The publisher does not give any warranty express or implied or make any representation that the contents
will be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses should
be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims,
proceedings, demand, or costs or damages whatsoever or howsoever caused arising directly or indirectly in
connection with or arising out of the use of this material.
 North American Journal of Aquaculture 75:159–163, 2013

C American Fisheries Society 2013

ISSN: 1522-2055 print / 1548-8454 online

DOI: 10.1080/15222055.2012.756439

COMMUNICATION

Characterization of Bacteria Isolated from Landlocked Fall

Chinook Salmon Eggs from Lake Oahe, South Dakota
David J. Bergmann and Alicia Brakke
Department of Science, Black Hills State University, 1100 University Boulevard, Spearfish,
South Dakota 57799, USA

Michael E. Barnes*
McNenny State Fish Hatchery, South Dakota Game, Fish, and Parks, 19619 Trout Loop, Spearfish,
South Dakota 57783, USA
Downloaded by [Department Of Fisheries] at 21:53 28 February 2013

ity. However, the experimental inoculation of Rainbow Trout

Abstract
Bacteria were isolated from the eggs of landlocked fall Chinook
Salmon Oncorhynchus tshawytscha prior to initial placement in
vertical-flow incubations trays 4 h after spawning and also 27 d
later at the early eyed egg stage of development. Bacterial densi-
ties on the eggs after iodophor disinfection and just before being
placed in trays were very low, and most isolates were gram-positive,
nonfermenting cocci. In contrast, bacterial densities on eggs at in-
cubation day 27 exceeded 1 × 107 CFU/egg and were dominated
by slow-growing, gram-negative coccobacilli. Pseudomonas spp.,
as well as Flavobacterium (closely related to F. johnsoniae) were
also present. Most Pseudomonas and Flavobacterium isolates pro-
duced extracellular proteases, making them candidates for further
investigation into their possible contribution to egg mortality.
Eggs from landlocked fall Chinook Salmon Oncorhynchus
tshawytscha obtained from Lake Oahe, South Dakota, exhibit
relatively poor survival during hatchery rearing (Barnes et al.
2000). Although this mortality is probably due to a number
of factors, bacteria isolated during egg incubation have been
implicated in these deaths in several studies (Barnes et al. 1997,
2003, 2005; Stephenson et al. 2003). Other researchers have
also suggested that bacteria may negatively impact egg survival
(Sauter et al. 1987; Barker et al. 1989; Holcomb et al. 2005),
although this relationship is far from certain (Barker et al. 1990;
Omnes et al. 1993). Soft egg disease, possibly caused by bacteria
(Wood 1979; Erdahl 1993), has also been observed in Lake Oahe
Chinook Salmon (Barnes et al. 2003).
Flavobacterium columnare, a bacterium that secretes extra-
cellular proteases and is pathogenic to fish, was isolated from
eggs of Lake Oahe Chinook Salmon by Barnes et al. (2005),
suggesting that this species may be implicated in egg mortal-
O. mykiss eggs with F. columnare did not result in increased
egg mortality (Barnes et al. 2009). Hence, it is likely that bac-
teria other than F. columnare may be involved in prehatching
mortality of salmonid eggs.
The objective of this study was to conduct a brief survey
of culturable bacteria associated with the hatchery-reared eggs
of Chinook Salmon in order to identify the most common
genera and species present. Species of bacteria known to be
fish pathogens, or which produced extracellular proteases, were
noted.
METHODS
Eggs were obtained from fall Chinook Salmon spawned
on October 18, 2010, at Whitlock Spawning and Imprint Sta-
tion, Lake Oahe, South Dakota. After fertilization and water
hardening, eggs from several spawns were pooled and trans-
ported in lake water for 4 h to the McNenny State Fish Hatch-
ery, Spearfish, South Dakota. Immediately upon arrival at the
McNenny hatchery, the eggs were disinfected in a 100-mg/L,
buffered, free-iodine solution for 10 min and rinsed thoroughly
in well water (total hardness as CaCO3 , 360 mg/L; alkalinity as
CaCO3 , 210 mg/L; pH, 7.6; total dissolved solids, 390 mg/L).
Three samples, each containing 10 recently disinfected eggs,
were placed in 50-mL conical tubes with 30 mL of sterile 0.8%
NaCl and vortexed vigorously for 5 min. Three 0.25-mL aliquots
of each dilution of the resulting suspension of bacteria were
plated out on agar media with 40 µg/mL cycloheximide: either
R2A agar (Reasoner and Geldreich 1985) or Cytophaga agar
(Daskalov et al. 1999) amended with 0.5 g of D( + )-glucose

*Corresponding author: mike.barnes@state.sd.us

Received July 8, 2012; accepted November 29, 2012

159
 160 BERGMANN ET AL.

TABLE 1. Mean (SE) CFU per egg of landlocked fall Chinook Salmon sampled either on the day of spawning or after 27 d of incubation and plated on one of
two types of agar with 40 µg/mL cycloheximide. For the 27-d samples, CFU of small (<2.0 mm), yellow, slow-growing colonies and all other colony morphotypes
are listed separately.

Days postspawn Agar Colony type Mean (SE) CFU/egg

0 R2A All 7.5 × 101 (4.43 × 101)
0 Cytophaga All 2.5 × 101 (8.66 × 100)
27 R2A Small, round, yellow 1.51 × 107 (4.82 × 106)
27 R2A All larger 5.73 × 105 (7.81 × 104)
27 Cytophaga Small, round, yellow 1.70 × 107 (5.07 × 106)
27 Cytophaga All larger 5.10 × 105 (1.46 × 105)

and 0.5 g skim milk. Plates were incubated for 6 d, and 30
randomly selected bacterial colonies were transferred to slants
of R2A agar and characterized according to colony morphol-
ogy, gram-staining (Koneman et al. 1997), cell morphology, the
Downloaded by [Department Of Fisheries] at 21:53 28 February 2013
DNA extracted, 16S rRNA genes amplified by PCR with uni-
versal bacterial primers, and the PCR products sequenced at
the Center for Conservation of Biological Resources at Black
Hills State University, Spearfish, South Dakota, as described

presence of catalase, presence of cytochrome oxidase, growth
and fermentation of glucose in Hugh and Leifson’s oxidation–
fermentation media, hydrolysis of casein in milk agar, hydrolysis
of starch, and growth, motility, H2 S production, and indole pro-
duction in sulfide-indole- motility (SIM) media (Collins et al.
1995; MacFaddin 2000).
After removal of the initial egg samples from the entire batch
of eggs, the remaining eggs were placed in vertical-flow incuba-
tors (Marisource, Fife, Washington) and reared as described by
Barnes et al. (2005), receiving daily, 15-min, antifungal treat-
ments of 1,667 mg/L formalin. After 27 d, three samples of 10
eggs each were removed and processed as previously described
to remove surface bacteria and were subsequently plated on
R2A and Cytophaga agar.
Fifty round, minute, yellow-colored colonies and 50 larger
colonies were selected from R2A agar plates, and 20 irreg-
ular, low, spreading, yellow colonies were selected from Cy-
tophaga agar plates and transferred to R2A slants. These isolates
were characterized as described previously. Representative iso-
lates of each operational taxonomic unit were chosen, genomic
by Barnes et al. (2010). The 16S rDNA sequences were ana-
lyzed using the Classifier program (Wang et al. 2007) of the
Ribosomal Database Project (http://rdp.cme.msu.edu/) and by
the Basic Local Alignment Search Tool (BLAST) at National
Center for Biotechnology Information (NCBI) (Altschul et al.
1990). The 16S rDNA sequences were deposited in GenBank
under accession numbers JX185730– JX185742. Slant cultures
of 20 Flavobacterium isolates were also sent to Dr. Annemie
Decostere at Ghent University, Ghent, Belgium, and tested for
hybridization to oligonucleotide probes for F. columnare.
RESULTS
Low numbers of colonies (0–20 colonies per plate of media)
were observed when samples of undiluted, vortexed suspensions
of freshly spawned and iodine-disinfected eggs were plated out
on media (Table 1). Numbers of colonies on R2A agar were
somewhat higher than on Cytophaga agar.
Twenty-eight isolates from freshly spawned eggs were char-
acterized further (Table 2). Eight isolates were gram-positive

TABLE 2. Groups of bacterial isolates from freshly spawned disinfected eggs of Chinook Salmon prior to hatchery incubation based on seven phenotypic
characteristics (gram stain, cell shape, glucose oxidation, presence of oxidase and catalase, casein hydrolysis, and siderophores). Positive test responses are
indicated by “ + ”, weakly positive responses by “w,” and negative responses are indicated by “–”. Isolates in each group are listed in the last column. (a) 0D: 1–3,
6–8, 13–17, 20–23, 26, 33–48; (b) 0D: 4, 5, 9–11, 18, 19, 24, 25, 27–32.

Number Casein
Isolate Phenotypic of Gram Glucose hydrolysis, Siderophores, Isolates
series group isolates stain Shape fermentation Catalase Oxidase milk agar milk agar in group
0D I 8 + Coccus − + + − − a
0D II 6 + Coccus − + + w − b
0D III 2 + Coccus + + + w − 8, 22
0D IV 3 − Coccus − + + − − 7, 11, 12
0D V 2 − Rod − + + − − 4, 16
0D VI 1 − Rod + + + + − 10
0D VII 1 + Rod + + + + − 3
0D VIII 1 + Rod − + + − − 6
0D IX 1 − Rod + − − − − 9
 COMMUNICATION 161

TABLE 3. Groups of bacterial isolates from eggs of Chinook Salmon incubated for 27 d, based on 11 phenotypic characteristics (gram stain, cell shape, glucose
oxidation, presence of oxidase and catalase, casein hydrolysis, siderophores, H2 S production on SIM agar, indole production on SIM agar, nitrite production on
nitrate broth, and starch hydrolysis). Positive test responses are indicated by “ + ” and negative responses are indicated by “–”. Isolates in each group are listed in
the last column. (a) 27S 1–3, 6–8, 12–17, 20–23, 26, and 33–48, (b) 4–5, 9–11, 18, 19, 24, 25, and 27–32; (c) 27L: 3, 7, 9, 12–14, 21, 23, 27, 30, 31, 33–37, 40,
45, 47, 49, 50; (d) 27L: 2, 10, 22, 25, 26, 43, 44, 46; (e) 27F: 1, 3–5, 7–13, 15, 16, 18 (ND = not determined). Isolate series 27S and 27L were colonies isolated
from R2A agar, while isolate series 27F were flattened, pale yellow colonies taken from Cytophaga agar.

Casein Nitrite on
Isolate Phenotypic Number Gram Glucose hydrolysis, Siderophores, H2 S, SIM Indole, nitrate Starch Isolates in
series group of isolates stain Shape fermentation Catalase Oxidase milk agar milk agar agar SIM agar broth hydrolysis group

27S I 33 − Ovoid − + + − − − − − ND a
27S II 15 − Ovoid − + + − − − − + ND b
27S III 2 − Coccus − + + + − − − − ND 49, 50
27L I 21 − Rod − + + + + − − − ND c
27L II 11 − Rod − + + + + − − + ND d
27L III 3 − Rod − + + + − − − − ND 41, 48, 50
27L IV 3 − Rod − + + − + − − − ND 29,32, 38
27L V 1 − Rod − + − + − − − + ND 20
27L VI 1 − Rod − + + − − − − − ND 19
27L VII 1 − Rod − + + − − − − − ND 4
Downloaded by [Department Of Fisheries] at 21:53 28 February 2013

27L VIII 1 + Rod − + + − − − − − ND 6

27L IX 1 − Coccus − − + − − − − − ND 11
27L X 1 − Coccus − + + − − − − + ND 24
27L XI 1 − Rod − + + ND ND − − − ND 28
27L XII 1 − Rod − + + − − − − + ND 42
27F I 14 − Rod − + + + − − − + + e
27F II 4 − Rod − − + + − − − + + 2, 6, 17, 19
27F III 1 − Rod − + + + − − − − + 14

cocci, catalase and oxidase positive, did not ferment glucose,
and did not hydrolyze casein. Six isolates had the same char-
acteristics but showed weak casein hydrolysis. Three isolates
were gram-negative cocci, catalase and oxidase positive, and
did not ferment glucose or hydrolyze casein. Two isolates were
gram-positive cocci, oxidase and catalase positive, fermented
glucose, and had weak hydrolysis of casein.
Samples of 27-d-old eggs had over 1.5 × 107 CFU /egg of
bacteria, and over 90% of colonies consisted of tiny (1 mm or
less), round, yellow, slow-growing forms that became visible
after six or more days of growth. Because even the plates from
the most diluted samples had in excess of 200 colonies, estimates
of CFU per egg may be too low. Much lower numbers of other,
larger, faster-growing colony morphotypes were observed (over
5 × 105 CFU/egg). Numbers of colonies on R2A and Cytophaga
agar plates were similar. Fifty, tiny, yellow-colored colonies
from R2A and 50 larger colonies from R2A agar plates were
chosen for further characterization (Table 3). Some colonies on
Cytophaga agar were yellow, low, spreading, and irregular in
outline. Twenty of these were presumed to be Flavobacterium
and were further characterized.
Fifty of the numerous, round, yellow, slow-growing colonies
were characterized, of which 33 had ovoid, motile, gram-
negative cells, did not ferment glucose, did not hydrolyze casein
or produce siderophores on milk agar, produced neither indole
nor H2 S on SIM agar, and did not produce nitrite after growth
in nitrate broth. Two of these isolates (27S1 and 27S20) were
identified as Sphingopyxis on the basis of 16S rDNA sequences
(99% identity to NR 024631, S. chilensis sp. S01). Of the 50
colonies, 14 had the same characteristics as the aforementioned
Sphingopyxis but produced nitrite after growth in nitrate broth.
One of these 16 isolates (27S10) was also identified as Sphin-
gopyxis on the basis of 16S rDNA sequences (also 99% identity
to S. chilensis sp. S01), while a second (27S30) was identified as
Acidovorax on the basis of 16S rDNA sequences (99% identity
to FM955883.1, Acidovorax sp. Asd MW-A3).
The majority (32 of 50) of large, rapidly growing colonies
isolated from 27-d eggs were gram-negative, motile rods that
did not ferment glucose, produced no H2 S or indole on SIM
agar, and exhibited casein hydrolysis and yellow siderophore
secretion on skim milk agar. Twenty-one of these isolates did
not produce nitrite in nitrate broth; one of these isolates (27L34)
was classified as Pseudomonas on the basis of its 16S rDNA se-
quence. Of the 32 isolates, 11 produced nitrite on nitrate broth;
two of these isolates (27L2 and 27L5) were also classified as
Pseudomonas on the basis of 16S rDNA sequence (99% identity
to NR 024902, P. mandelii CIP 105273, and nine other Pseu-
domonas isolates in GenBank). Three of the 50 large colonies
from 27-d eggs were gram-negative, motile rods that did not
ferment glucose, produced no H2 S or indole on SIM agar, and
exhibited casein hydrolysis but no siderophore secretion on skim
milk agar; one of these isolates (27L50) was classified as Pseu-
domonas on the basis of its 16S rDNA sequence (99% identity
to NR 024902.1, P. mandelii CIP 105273, and two other Pseu-
domonas isolates). A fourth isolate (27L1), identified as Pseu-
domonas on the basis of 16S rDNA sequence (99% identity
to NR 042451, P. peli R-2085 and to NR042607.1, P. guineae
LMG 24016), was a gram-negative, motile rod that did not
 162 BERGMANN ET AL.
ferment glucose, produced no H2 S or indole on SIM agar, and
exhibited neither casein hydrolysis nor siderophore secretion
on skim milk agar. One isolate (27L20) of the 50 large colonies
had gram-negative, motile rods that were oxidase negative and
catalase positive, did not ferment glucose, produced neither H2 S
nor indole on SIM agar, produced nitrite from nitrate broth, and
hydrolyzed casein, but produced no siderophores. This isolate
was identified as Flavobacterium on the basis of its 16S rDNA
sequence (97% identity to NR 042497.1, F. saccharophilum,
and four other isolates).
Twenty flattened, yellow, spreading, irregular-shaped
colonies were recovered from Cytophaga agar and character-
ized further. These isolates were all gram-negative rods, nonfer-
menting, oxidase positive, hydrolyzed both casein and starch,
and produced neither H2 S nor indole on SIM agar (Table 3). All
isolates grew on Cytophaga agar with 1.0 µg/mL tobramycin.
Downloaded by [Department Of Fisheries] at 21:53 28 February 2013
Five of the isolates were catalase negative, and the rest were
catalase positive. Eighteen produced nitrite from nitrate broth,
while two did not. The 16S rRNA gene was amplified by PCR
from 19 isolates. AluI digestion of the PCR products and agarose
gel electrophoresis showed the same pattern of restriction frag-
ments in all isolates. The PCR products from isolates 8, 15,
and 18 were sequenced and identified as Flavobacterium by the
Classifier program and showed 97% identity to NR 042470.1,
F. aquidurense WB 1.1 and NR 041057, F. frigidimaris KUC-
1). None hybridized to F. columnare probes.
DISCUSSION
Low numbers of bacteria were recovered from the iodine-
disinfected eggs of freshly spawned Chinook Salmon. The
majority of the isolates were gram-positive cocci, which
were nonfermenting and oxidase and catalase positive. The
prevalence of bacteria with these characteristics was also
noted in the ovarian fluid of Chinook Salmon by Barnes et al.
(2010). However, the bacteria in ovarian fluid were primarily
Micrococcus and Arthrobacter (Barnes et al. 2010).
Much higher densities of bacteria were found on the eyed
eggs of Chinook Salmon, as was expected given the results
reported by Trust (1972) and Barnes et al. (1999, 2000) of
increasing bacterial numbers as egg incubation progressed. In
addition to the difference in bacterial numbers between the sam-
pling dates, different groups of oxidase- and catalase-positive,
gram-negative bacteria predominated on the eggs later in devel-
opment in comparison with the bacteria groups observed imme-
diately after disinfection. The isolation of Sphingopyxis later in
development, bacteria common in aquatic and moist environ-
ments (Koskinen et al. 2000; Kämpfer et al. 2002; Yoon and Oh
2005), is not surprising. Neither is the isolation of Pseudomonas,
another very common freshwater bacterial group (Allen et al.
1983), which has been identified previously on stream-incubated
salmonid eggs (Bell et al. 1971). Unlike the results of Barnes
et al. (2005), the Flavobacterium on the 27-d-old eggs, al-
though capable of hydrolyzing casein, were not F. columnare,
which can be easily misidentified using molecular techniques
(Darwish et al. 2004), and may represent an undescribed species
(the closest GenBank matches exhibited only 97% identity, of-
ten considered as the level of identity representing separate
bacterial species). Flavobacterium johnsoniae is pathogenic to
cultured fish, producing skin lesions similar to those made by
F. columnare (Carson et al. 1993; Soltani et al. 1994). Its possi-
ble role as a salmonid egg pathogen is unknown.
The few bacterial species isolated during this study cannot
be presumed to be exhaustive of all of the microbial species
present on the eggs. In a more comprehensive survey of an entire
hatchery operation, Allen et al. (1983) described 600 different
bacterial species. In addition, there are probably nonculturable
bacteria on the egg surface as well (Barnes et al. 2005).
The composition of the bacterial flora of mature Chinook
Salmon eggs is clearly different from that of ovarian fluid
(Barnes et al. 2010) or freshly spawned eggs. Whether this
is related to the colonization of the eggs by bacteria present in
hatchery water or selection for certain taxa of bacteria as a result
of formalin treatment is not clear. The presence of extracellular
proteases, as evidenced by hydrolysis of casein in agar media
in some Pseuodomonas and Flavobacterium isolates, indicates
that it is possible that some of these bacteria may be involved in
the weakening of the egg chorionic membrane, but this must be
tested experimentally.
ACKNOWLEDGMENTS
We thank the South Dakota Biomedical Research Infrastruc-
ture Network for financial support, Carolyn Ferrell for DNA
sequencing, and Annemie Decostere for assistance with identi-
fication of Flavobacterium isolates.
REFERENCES
Allen, D. A., B. Austin, and R. R. Colwell. 1983. Numerical taxonomy of
bacterial isolates associated with a freshwater fishery. Journal of General
Microbiology 129:2043–2062.
Altschul, S. F., W. Gish, W. Miller, E. W. Myers, and D. J. Lipman. 1990. Basic
local alignment search tool. Journal of Molecular Biology 215:403–410.
Barker, G. A., S. N. Smith, and N. R. Bromage. 1989. The bacterial flora
of Rainbow Trout, Salmo gairdneri Richardson, and Brown Trout, Salmo
trutta L., eggs and its relationship to developmental success. Journal of Fish
Diseases 12:281–293.
Barker, G. A., S. N. Smith, and N. R. Bromage. 1990. Effect of oxolinic acid
on bacterial flora and hatching success rate of Rainbow Trout, Oncorhynchus
mykiss, eggs. Aquaculture 91:205–222.
Barnes, M. E., D. Bergmann, J. Jacobs, and M. Gabel. 2009. Effect of Flavobac-
terium columnare inoculation, antibiotic treatments, and resident bacteria on
Rainbow Trout Oncorhynchus mykiss eyed egg survival and external mem-
brane structure. Journal of Fish Biology 74:576–590.
Barnes, M. E., D. Bergmann, R. L. Kelley, R. J. Cordes, P. A. Nero, and
D. J. Durben. 2010. A survey of bacteria in the ovarian fluid of landlocked
fall Chinook Salmon and their relationship with egg survival. North American
Journal of Aquaculture 72:314–320.
Barnes, M. E., D. Bergmann, H. Stephenson, M. Gabel, and R. J. Cordes. 2005.
Bacterial numbers from landlocked fall Chinook Salmon eyed eggs subjected
to various formalin treatments as determined by scanning electron microscopy
 COMMUNICATION
and bacteriological culture methods. North American Journal of Aquaculture
67:23–33.
Barnes, M. E., R. J. Cordes, and W. A. Sayler. 1997. Use of formalin during
incubation of eyed eggs of inland fall Chinook Salmon. Progressive Fish-
Culturist 59:303–306.
Barnes, M. E., R. J. Cordes, W. A. Sayler, and R. P. Hanten. 2003. Soft-
egg disease in landlocked fall Chinook Salmon eggs: possible causes and
therapeutic treatments. North American Journal of Aquaculture 65:126–133.
Barnes, M. E., A. C. Gabel, and R. J. Cordes. 1999. Bacterial populations during
inland fall Chinook Salmon egg culture in vertical-flow tray incubators. North
American Journal of Aquaculture 61:252–257.
Barnes, M. E., A. C. Gabel, and R. J. Cordes. 2000. Bacterial populations during
Rainbow Trout egg culture in vertical-flow tray incubators. North American
Journal of Aquaculture 62:48–53.
Bell, G. R., G. E. Hoskins, and W. Hodgkiss. 1971. Aspects of the character-
ization, identification, and ecology of the bacterial flora associated with the
surface of stream-incubating Pacific salmon (Oncorhynchus) eggs. Journal of
the Fisheries Research Board of Canada 28:1511–1525.
Carson, J., L. M. Schmidtke, and B. L. Munday. 1993. Cytophaga johnsonae:
Downloaded by [Department Of Fisheries] at 21:53 28 February 2013
a putative skin pathogen of juvenile farmed Barramundi, Lates calcarifer
Bloch. Journal of Fish Diseases 16:209–218.
Collins, C. H., P. M. Lyne, and J. M. Grange. 1995. Collins and Lyne’s micro-
biological methods, 7th edition. Butterworth-Heinemann, Oxford, UK.
Darwish, A. M., A. A. Ismaiel, J. C. Newton, and J. Tang. 2004. Identification of
Flavobacterium columnare by a species-specific polymerase chain reaction
and renaming of ATCC43622 strain to Flavobacterium johnsoniae. Molecular
and Cellular Probes 18:421–427.
Daskalov, H., D. A. Austin, and B. Austin. 1999. An improved growth medium
for Flavobacterium psychrophilum. Letters in Applied Microbiology 28:297–
299.
Erdahl, D. A. 1993. Inland salmonid broodstock management handbook. U.S.
Fish and Wildlife Service, Report 712 FW 1, Washington, D.C.
Holcomb, M., J. G. Cloud, and R. L. Ingermann. 2005. Impact of bacteria on
short-term storage of salmonid eggs. Aquaculture Research 36:1555–1561.
Kämpfer, P., R. Witzenberger, E. B. Denner, H. J. Busse, and A. Neef. 2002.
Sphingopyxis witflariensis sp. nov., isolated from activated sludge. Interna-
tional Journal of Systematic and Evolutionary Microbiology 52:2029–2034.
163
Koneman, E. W., S. D. Allen, W. M. Janda, P. C. Schreckenberger, and
W. C. Winn Jr. 1997. Color atlas and textbook of diagnostic microbiology,
5th edition. Lippincott, Philadelphia.
Koskinen, R., T. Ali-Vehmas, P. Kämpfer, M. Laurikkala, I. Tsitko, E. Kostyal, F.
Atroshi, and M. Salkinoja-Salonen. 2000. Characterization of Sphingomonas
isolates from Finnish and Swedish drinking water distribution systems. Jour-
nal of Applied Microbiology 89:687–696.
MacFaddin, J. F. 2000. Biochemical tests for identification of medical bacteria,
3rd edition. Williams and Wilkins, Baltimore, Maryland.
Omnes, M. H., Y. Temey, J. Y. Daniel, and J. L. Nicolas. 1993. Bacterial envi-
ronment of gametes and eggs of turbot (Scophthalmus maximus). European
Aquaculture Society, Special Publication 19, Oostende, Belgium.
Reasoner, D. J., and E. E. Geldreich. 1985. A new medium for the enumeration
and subculture of bacteria from potable water. Applied and Environmental
Microbiology 49:1–7.
Sauter, R. W., C. Williams, E. A. Meyer, B. Celnik, J. L. Banks, and D. A. Leith.
1987. A study of bacteria present within unfertilized salmon eggs at the time
of spawning and their possible relation to early lifestage disease. Journal of
Fish Diseases 10:193–203.
Soltani, M., B. Munday, and J. Carson. 1994. Susceptibility of some freshwater
species of fish to infection by Cytophaga johnsonae. Bulletin of the European
Association of Fish Pathologists 14:133–135.
Stephenson, H., M. Gabel, and M. E. Barnes. 2003. Microbial inhibition in
response to treatments of hydrogen peroxide and formalin on landlocked fall
Chinook Salmon eyed eggs, as determined by scanning electron microscopy.
North American Journal of Aquaculture 65:324–329.
Trust, T. J. 1972. The bacterial population in vertical flow tray hatcheries during
incubation of salmonid eggs. Journal of the Fisheries Research Board of
Canada 29:567–571.
Wang, Q., G. M. Garrity, J. M. Tiedje, and J. R. Cole. 2007. Naı̈ve Bayesian
classifier for rapid assignment of rRNA sequences into the new bacterial
taxonomy. Applied and Environmental Microbiology 73:5261–5267.
Wood, J. W. 1979. Diseases of Pacific salmon: their prevention and treatment,
3rd edition. State of Washington, Department of Fisheries, Olympia.
Yoon, J. H., and T. K. Oh. 2005. Sphingopyxis flavimaris sp. nov., isolated from
sea water of the Yellow Sea in Korea. International Journal of Systematic and
Evolutionary Microbiology 55:369–373.
This article was downloaded by: [Department Of Fisheries]
On: 28 February 2013, At: 19:51
Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,
37-41 Mortimer Street, London W1T 3JH, UK

North American Journal of Aquaculture

Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/unaj20

Heat Shock Protein 70 (HSP70) Responses in Tissues

of White Sturgeon and Green Sturgeon Exposed to
Different Stressors
a b c c
Weifang Wang , Dong-Fang Deng , Nicola De Riu , Giuseppe Moniello & Silas S. O. Hung
d

a
Qingdao Key Laboratory for Marine Fish Breeding and Biotechnology, Yellow Sea Fisheries
Research Institute, Chinese Academy of Fishery Sciences, Qingdao, 266071, China
b
Aquatic Feeds and Nutrition Department, Oceanic Institute, 41-202 Kalanianaole Highway,
Waimanalo, Hawaii, 96795, USA
c
Department of Veterinary Medicine, University of Sassari, Via Vienna 2, Sassari, 07100,
Italy
d
Department of Animal Science, University of California, One Shields Avenue, Davis,
California, 95616-8521, USA
Version of record first published: 08 Feb 2013.

To cite this article: Weifang Wang , Dong-Fang Deng , Nicola De Riu , Giuseppe Moniello & Silas S. O. Hung (2013): Heat Shock
Protein 70 (HSP70) Responses in Tissues of White Sturgeon and Green Sturgeon Exposed to Different Stressors, North American
Journal of Aquaculture, 75:2, 164-169

To link to this article: http://dx.doi.org/10.1080/15222055.2012.747457

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions

This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to
anyone is expressly forbidden.

The publisher does not give any warranty express or implied or make any representation that the contents
will be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses should
be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims,
proceedings, demand, or costs or damages whatsoever or howsoever caused arising directly or indirectly in
connection with or arising out of the use of this material.
 North American Journal of Aquaculture 75:164–169, 2013

C American Fisheries Society 2013

ISSN: 1522-2055 print / 1548-8454 online

DOI: 10.1080/15222055.2012.747457

ARTICLE

Heat Shock Protein 70 (HSP70) Responses in Tissues of White

Sturgeon and Green Sturgeon Exposed to Different Stressors
Weifang Wang
Qingdao Key Laboratory for Marine Fish Breeding and Biotechnology,
Yellow Sea Fisheries Research Institute, Chinese Academy of Fishery Sciences, Qingdao 266071, China

Dong-Fang Deng
Aquatic Feeds and Nutrition Department, Oceanic Institute, 41-202 Kalanianaole Highway, Waimanalo,
Hawaii 96795, USA
Downloaded by [Department Of Fisheries] at 19:51 28 February 2013

Nicola De Riu and Giuseppe Moniello

Department of Veterinary Medicine, University of Sassari, Via Vienna 2, Sassari 07100, Italy

Silas S. O. Hung*
Department of Animal Science, University of California, One Shields Avenue, Davis,
California 95616-8521, USA

Abstract
A factorial experiment was conducted to compare the responses of heat shock protein 70 (Hsp70) in seven different
tissues of White Sturgeon Acipenser transmontanus and Green Sturgeon A. medirostris after they were exposed to four
different stressors. Three White Sturgeon (2.3 ± 0.1 kg [mean ± SE]) and three Green Sturgeon (2.3 ± 0.1 kg [mean
± SE]) were each subjected to one of four different stressors, after which the Hsp70 levels in seven different tissues
were measured using Western blot. The four stressors were heat shock, cold shock, air exposure, and food deprivation;
and the seven tissues sampled were mucus, heart, liver, gastrointestinal tract, gill, spleen, and white muscle. We also
sampled tissues of three White Sturgeon and three Green Sturgeon without any stressor, and measured their Hsp70
levels as a control. We compared Hsp70 responses of the stressed sturgeon with those of the control, which was set at
100%, and found that Hsp70 responses were significantly (P < 0.05) affected by the different stressors and also varied
significantly among the tissues. For both species of sturgeon, heat shock was shown to be the most effective stressor
inducing Hsp70 responses and mucus was the most responsive tissue. Under heat shock stress, Hsp70 responses in all
tissues except liver were significantly higher in the White Sturgeon than in the Green Sturgeon. When both species of
fish were exposed to heat shock, cold shock, or food deprivation, White Sturgeon showed significantly higher Hsp70
responses in mucus than did Green Sturgeon. In summary, heat shock elicited the highest Hsp70 responses and mucus
was the most sensitive tissue. Based on the tissue Hsp70 responses, White Sturgeon were predicted to have a better
defense mechanism against heat shock than Green Sturgeon.

Heat shock proteins (Hsp) are a family of highly conserved
cellular proteins expressed in response to a variety of stressors,
and they have thus been proposed as biomarkers of cellular stress
and environmental insult in animals, including aquatic organ-
isms (Sanders 1993). Heat shock proteins have been shown to
play an important role in maintaining the integrity and enhanc-
ing the resistance of cells to the deleterious effects of stress-
ful conditions (Feige et al. 1996; Yengkokpam et al. 2008).
Based on molecular weight, the Hsp are classified as Hsp100,
Hsp90, Hsp70, Hsp60, and small Hsp (molecular weight

*Corresponding author: sshung@ucdavis.edu

Received August 2, 2012; accepted November 3, 2012

164
 HSP70 RESPONSES TO STRESSORS IN STURGEON
< 40 kilodaltons). Of these, Hsp70 is the most conserved and
widely studied.
Heat shock protein was first discovered in fruit flies exposed
to heat shock (Ritossa 1962); however, other environmental or
biological factors also affect Hsp expression (Yengkokpam et al.
2008). Fish in intensive aquaculture are routinely subjected to
different stressors such as fluctuations of temperature, hypoxia,
feed availability, and handling. Disturbance of physiological
homeostasis due to these stressors could result in depression
of growth, inhibition of reproduction, and reduced resistance to
disease (Palmisano et al. 2000). Furthermore, the response of
Hsp70 could vary among different species of fish (Basu et al.
2001) and among different tissues of a fish exposed to the same
stress (Currie et al. 2000; Fowler et al. 2009).
White Sturgeon Acipenser transmontanus and Green Stur-
geon A. medirostris are both ancient ray-finned fish endemic
Downloaded by [Department Of Fisheries] at 19:51 28 February 2013
to the Pacific coast of North America. White Sturgeon is a
promising aquaculture species in California, but it is also listed
as a species of concern due to declines in the natural popula-
tion (Moyle 2002). Green Sturgeon is an endangered species,
but it is now a potential aquaculture species in the USA be-
cause spawning and rearing techniques are well established (Van
Eenennaam et al. 2008). Under laboratory conditions, Green
Sturgeon exhibited higher growth rates than White Sturgeon
(Deng et al. 2002), but they also showed greater deleterious ef-
fects than White Sturgeon when exposed to high levels of dietary
methylmercury (Lee et al. 2011) and selenomethionine (S. S. O.
Hung, unpublished data). Therefore, we hypothesize that Green
Sturgeon may be more vulnerable to stress than White Sturgeon.
The objective of the present study was to investigate whether
the Hsp70 responses to heat shock, cold shock, air exposure,
and food deprivation differed among the mucus, heart, liver,
gastrointestinal tract, gill, spleen, and white muscle between the
White Sturgeon and Green Sturgeon.
METHODS
Fish supply and maintenance.—White Sturgeon (2.3 ±
0.1 kg [mean ± SE]) and Green Sturgeon (2.3 ± 0.1 kg
[mean ± SE]) were obtained from the Center of Aquatic Bi-
ology and Aquaculture at the University of California, Davis
(UCD). Prior to the experiment, the sturgeon were fed a com-
mercial feed (Silver Cup; Nelson & Sons, Tooele, Utah) and
maintained in circular fiberglass tanks (height = 90 cm; radius
= 183 cm) supplied with flow-through and aerated well wa-
ter (18–19◦ C) at a flow rate of 50 L/min. To minimize stress
of the sturgeon before the experimental treatment, the stock-
ing densities of the White Sturgeon and Green Sturgeon were
very low—1.1 kg/100 L and 0.8 kg/100 L, respectively. The
current experiment followed protocols (15094) approved by the
Campus Animal Care and Use Committee at UCD.
Experimental treatments.—There were four stress
treatments—heat shock, cold shock, air exposure, and
food deprivation—and groups of three White Sturgeon or three
165
Green Sturgeon were subjected to one of the four stressors.
An additional three White Sturgeon and three Green Sturgeon
were used as controls. Only one White Sturgeon and one Green
Sturgeon were kept in a flow-through fiberglass tank (height =
30 cm; radius = 60 cm; flow rate = 10 L/min) with aeration,
and three tanks for each treatment were used to minimize stress
effect from overstocking.
Heat shock.—Sturgeon were subjected to heat shock when
the water temperature was increased from 18◦ C to 26◦ C at a rate
of 4◦ C/h by adjusting the ratio of hot to cold incoming aerated
well water. The water temperature was then maintained at 26◦ C
for 4 h in the flow-through fiberglass tank (height = 30 cm; ra-
dius = 60 cm). Dissolved oxygen levels were monitored hourly
during the heat shock period and decreased from 7.0 mg/L at
the beginning to 5.3 mg/L at the end of the period.
Cold shock.—Sturgeon were abruptly transferred from 18◦ C
to 10◦ C aerated well water in the flow-through fiberglass tank
and kept at 10◦ C for 2 h. Dissolved oxygen levels were moni-
tored hourly during exposure and ranged from 8.5 to 9.5 mg/L.
Air exposure.—Sturgeon were lifted out of the water with a
net for 5 min before they were killed for tissue sampling.
Food deprivation.—The sturgeon were held in the fiberglass
tank with aerated well water (18◦ C) at a flow rate of 10 L/min
and deprived of food for 7 d. Dissolved oxygen levels were
monitored three times daily over this period and ranged from
6.9 to 7.4 mg/L.
Tissue sampling.—At the end of each stress treatment, stur-
geon were stunned with a blunt object and killed by cervical
dislocation, cutting the vertebrate posterior to the head. Tis-
sue samples included mucus (approximately 1 g by scraping
with a razor blade on the skin), white muscle (a cubical section
2 cm in length at the midpoint of the body), gill (the outside
filaments), heart, liver, spleen, and gastrointestinal tract (GIT).
Tissue samples were rinsed in double-distilled water, frozen in
liquid nitrogen, and kept at −80◦ C until they were used for
Hsp70 determination.
Heat shock protein 70 analysis.—Levels of Hsp70 were mea-
sured using Western blot techniques as described by Deng et al.
(2009). An equal amount of protein (25 µg) from each tissue
sample was loaded into gels and separated by SDS-PAGE, then
electroblotted onto a total polyvinylidene fluoride transfer mem-
brane (Millipore, Bedford, Massachusetts). The Hsp70 standard
(SPP-758; Assay Designs, Ann Arbor, Michigan) was loaded
with samples in each gel in order to calculate the relative band
densities of the tissue samples. The primary polyclonal Hsp70
antibody was purchased from Assay Designs, and peroxidase-
labeled anti-rabbit IgG (Amersham Biosciences, Piscataway,
Pennsylvania) was used as the secondary antibody to detect the
Hsp70 probe. The antibody was detected by enhanced chemi-
luminescence reagents (Amersham Biosciences) for 1 min and
then exposed to hyperfilm (Amersham Biosciences). The band
density on the film was quantified using a GS-710 calibrated
imaging densitometer (Bio-Rad, Hercules, California). The rel-
ative band density of each tissue sample was calculated by
 166 WANG ET AL.

TABLE 1. The Hsp70 responses in tissues of White Sturgeon exposed to different stressors. Values represent the response over the control, which was set at
100%. Data are presented as the mean ± SE (n = 3). The lowercase z, y, x, and w indicate significant (P < 0.05) differences among different tissues; the uppercase
Z and Y indicate significant differences among different stressors as determined by Fisher LSD’s posthoc tests.

Tissues Heat shock (%) Cold shock (%) Air exposure (%) Food deprivation (%)
Mucus 3,983 ± 280 vY 576 ± 121 xZ 546 ± 58 yZ 602 ± 117 xZ
Heart 1,100 ± 203 yxY 252 ± 52 yZ 219 ± 12 zZ 197 ± 23 yZ
Liver 1,436 ± 432 xwY 190 ± 34 zyZ 110 ± 6 zZ 44 ± 7 zZ
GIT 1,327 ± 184 yxY 143 ± 6 zyZ 106 ± 12 zZ 103 ± 5 zyZ
Gill 1,386 ± 463 yxwY 99 ± 26 zZ 102 ± 21 zZ 93 ± 17 zZ
Spleen 775 ± 40 zyY 90 ± 3 zZ 108 ± 21 zZ 45 ± 2 zZ
White muscle 346 ± 96 zY 75 ± 13 zZ 82 ± 9 zZ 45 ± 16 zZ
ANOVA: P-value
Tissue (T) <0.001
Stressor (S) <0.001
T×S 0.001
Downloaded by [Department Of Fisheries] at 19:51 28 February 2013

comparison to the Hsp70 standard on the same film. The Hsp70
responses of the stressed sturgeon were compared with the con-
trols of the same tissues, which were set to 100%.
Statistical analysis.—Data are presented as the mean ± SE
and were analyzed with factorial ANOVA using Statistica 6.0
(StatSoft, Tulsa, Oklahoma) after the data were checked for
any violations of the assumptions. The level of significance was
P < 0.05, and significant differences were detected by Fisher’s
least-significant-difference (LSD) posthoc test.
RESULTS
Heat shock protein 70 responses in White Sturgeon and
Green Sturgeon varied significantly among tissues and were
affected differently by the different stressors (Tables 1, 2). The
Hsp70 responses in the mucus of White Sturgeon were signif-
icantly higher than those in the other tissues after the sturgeon
were stressed by any one of the four stressors (Table 1). The
Hsp70 responses in the mucus of Green Sturgeon were signifi-
cantly higher than the other tissues after the sturgeon were cold
shocked, air exposed, and food deprived. However, in the heat
shock group, the Hsp70 responses in the liver were not signif-
icantly different from that in the mucus (Table 2). There is no
clear difference or discernible pattern among the other tissues
except that white muscle tended to have the lowest responses in
both sturgeon species after they were stressed. Among the four
stressors, heat shock is the stressor that stimulated the highest re-
sponse of Hsp70 in all the tissues evaluated in this study. There
was no difference in the Hsp70 response when both species
of sturgeon were exposed to cold shock, air exposure, or food
deprivation.
The Hsp70 responses were significantly different between the
two sturgeon species, and the responses varied among different
stressors (Figure 1). The Hsp70 responses in all tissues except

TABLE 2. The Hsp70 responses in tissues of Green Sturgeon exposed to different stressors. Values represent the response over the control, which was set at
100%. Data are presented as the mean ± SE (n = 3). The lowercase z, y, x, and w indicate significant (P < 0.05) differences among different tissues; the uppercase
Z and Y indicate significant differences among different stressors as determined by Fisher’s LSD posthoc tests.

Tissues Heat shock (%) Cold shock (%) Air exposure (%) Food deprivation (%)
Mucus 2,260 ± 563 xY 431 ± 63 xZ 672 ± 90 wZ 435 ± 107 xZ
Heart 533 ± 137 zY 224 ± 27 yZ 255 ± 42 yZ 228 ± 20 yZ
Liver 2,155 ± 536 xY 187 ± 62 zyZ 420 ± 134 xZ 132 ± 37 zyZ
GIT 843 ± 235 zyY 135 ± 17 zyZ 166 ± 25 zyZ 107 ± 17 zyZ
Gill 731 ± 196 zyY 91 ± 7 zZ 94 ± 12 zZ 96 ± 5 zZ
Spleen 561 ± 107 zyY 95 ± 18 zyZ 136 ± 20 zyZ 74 ± 16 zZ
White muscle 181 ± 49 zY 70 ± 9 zZ 88 ± 13 zZ 23 ± 7 zZ
ANOVA: P-value
Tissue (T) <0.001
Stressor (S) <0.001
T×S 0.002
 HSP70 RESPONSES TO STRESSORS IN STURGEON
Downloaded by [Department Of Fisheries] at 19:51 28 February 2013
FIGURE 1. Comparison of Hsp70 responses in tissues of White Sturgeon and Green Sturgeon exposed to stressors. Different letters indicate significant differences
(P < 0.05) between the two species. Absence of a letter indicates no significant difference between species. Values are presented as the mean + SE (n = 3).
for liver were significantly higher in the White Sturgeon than
the Green Sturgeon after heat shock. The cold shock or food
deprivation did not cause significant differences in the Hsp70
response between the two species of sturgeon except in mucus,
which had a higher Hsp70 response in White Sturgeon than in
Green Sturgeon after they were cold shocked or food deprived.
Air exposure stimulated a higher Hsp70 response in the liver
in Green Sturgeon than in White Sturgeon, and no differences
were observed in the other tissues between the two species of
sturgeon.
DISCUSSION
The results of the present study show that the Hsp70 response
depended on the tissue, stressor, and sturgeon species. Mucus
was the tissue with the greatest increase in Hsp70, while white
muscle was less responsive in both species of sturgeon. Among
the four stressors, heat shock stimulated the highest response of
Hsp70 in both species of sturgeon. Compared with Green Stur-
geon, White Sturgeon exposed to heat shock exhibited a higher
Hsp70 response in all tissues, suggesting that White Sturgeon
may have a better defense mechanism against heat stress.
The present study showed that the heat shock response varied
among the different tissues in sturgeon. Similar results have
been observed in Rainbow Trout Oncorhynchus mykiss, where
heat shock resulted in Hsp70 mRNA expression levels in the
following tissues order: blood > red muscle > heart > brain >
liver > white muscle (Currie et al. 2000). Similarly, Fowler et al.
(2009) found tissue specific Hsp70 expression in heat-shocked
167
Rainbow Trout in decreasing order: heart > liver > gill > muscle
(red and white). Tissue specificity of the Hsp70 response is
also reported for Fathead Minnow Pimephales promelas (Dyer
et al. 1991), Mummichog Fundulus heteroclitus (Koban et al.
1991), Sea Lampreys Petromyzon marinus (Wood et al. 1999),
and Common Carp Cyprinus carpio (Wang et al. 2007). These
different responses among tissues support the hypothesis that
the thermal tolerance of an organism is governed by one tissue
more than others (Dyer et al. 1991).
Fish mucus is known to contain mainly macromolecular gly-
coproteins, antibodies, as well as secondary bioactive metabo-
lites (Ellis 2001) and thus has multiple functions such as me-
chanical protection, hydrodynamic lubrication, and active an-
tiparasitic and antibacterial action (Shephard 1994). Horne and
Sims (1998) reported thinning and shedding mucus in stressed
or diseased fish, indicating that the quality or quantity of mucus
responds to stress and could thus be a good biological indicator
for stress management. However, there is no information on the
cellular stress response of such an important and easily obtain-
able tissue. On the other hand, white muscle was shown to be
a less responsive tissue to stress in both sturgeon species, and
similar results were found in Rainbow Trout (Currie et al. 2000;
Fowler et al. 2009). This suggests that white muscle, a metabol-
ically less active tissue with a slow turnover, is not a suitable
tissue to study the Hsp70 response to the short-term stressors
used in the present experiment.
Exposure of the two sturgeon species to a variety of
husbandry-related stressors either increased or decreased the
 168 WANG ET AL.
responses of Hsp70 among different tissues. Heat shock is the
most studied stressor affecting Hsp expression in animals, and
Hsp70 responds well to heat shock. This was further confirmed
by the present study in which heat stress dramatically increased
the Hsp70 response in the different tissues of the two sturgeon
species compared with the other stressors. The temperature in-
crease caused the fluidization of cell membranes, which can
lead to disintegration of the lipid bilayer. Membrane fluidity de-
creases with a decrease in temperature (Los and Murata 2004),
which may explain why heat shock stimulates the greatest re-
sponse in the present experiment. The different experimental
conditions (such as species of fish, feeding protocol, and stressed
condition) may contribute to the different responses observed
among different studies.
It has been reported that food deprivation induces Hsp70 ex-
pression in larval Gilthead Seabream Sparus auratus and Rain-
Downloaded by [Department Of Fisheries] at 19:51 28 February 2013
bow Trout (Cara et al. 2005), and the quantity of Hsp70 in the
liver of Rohu Labeo rohita fingerlings increased linearly with
the duration of starvation (Yengkokpam et al. 2008). However,
after 1 week of food deprivation, the liver Hsp70 response de-
creased to 44% in the White Sturgeon and increased to 132% in
the Green Sturgeon when compared with the control groups. The
different responses of Hsp70 to food deprivation among differ-
ent species of fish suggest that different strategies are employed
to cope with nutrient deficiency.
The response of Hsp70 varies among different species of
fish (Basu et al. 2001), and this is further confirmed by the
present study. During 1 week of food deprivation, Green Stur-
geon showed a higher body weight loss (9.7%) than White
Sturgeon (7.5%). The increased Hsp70 responses in the most
sensitive tissue, the mucus, were lower in the Green Sturgeon
than White Sturgeon under all stressors except for the air expo-
sure. Heat shock also resulted in higher Hsp70 responses in the
other tissues of the White Sturgeon compared with the Green
Sturgeon, except in the liver. The higher Hsp70 response in
White Sturgeon exposed to the same stressor indicates better
protection for the fish, and thus the current results suggest that
White Sturgeon would have better heat tolerance than Green
Sturgeon under these experimental conditions. Previous studies
(Lee et al. 2011) have shown that Green Sturgeon have lower
body fat and thus lower energy reserves than White Sturgeon
of similar size. Furthermore, the different histopathological re-
sponses between these two sturgeon species suggest that they
may be functionally and structurally different. Silvestre et al.
(2010) showed different protein expression in Green Sturgeon
and White Sturgeon larvae, and higher mortality of Green Stur-
geon larvae exposed to heat stress and selenium. These differ-
ences between the two species of sturgeon could make the Green
Sturgeon more vulnerable to stressors.
ACKNOWLEDGMENTS
This study was partially funded by the Ecological Restora-
tion Program (grant E1183017), California Department of Fish
and Game. Weifang Wang was supported by the China Scholar-
ship Council and the Ocean University of China. Nicola De Riu
was supported by the Fondazione Banco di Sardegna, Italy. We
would like to thank the Center of Aquatic Biology and Aqua-
culture for the use of the culture facilities and the infrastructure
support of the Department of Animal Science and the College
of Agriculture and Environmental Sciences, UCD.
REFERENCES
Basu, N., T. Nakano, E. G. Grau, and G. K. Iwama. 2001. The effects of cortisol
on heat shock protein 70 levels in two fish species. General and Comparative
Endocrinology 124:97–105.
Cara, J. B., N. Aluru, F. J. Moyano, and M. M. Vijayan. 2005. Food-deprivation
induces HSP70 and HSP90 protein expression in larval Gilthead Sea Bream
and Rainbow Trout. Comparative Biochemistry and Physiology 142B:426–
431.
Currie, S., C. D. Moyes, and B. L. Tufts. 2000. The effects of heat shock and
acclimation temperature on hsp70 and hsp30 mRNA expression in Rainbow
Trout: in vivo and in vitro comparisons. Journal of Fish Biology 56:398–408.
Deng, D. F., C. Wang, S. Lee, S. Bai, and S. S. O. Hung. 2009. Feeding rates
affect heat shock protein levels in liver of larval White Sturgeon (Acipenser
transmontanus). Aquaculture 287:223–226.
Deng, X., J. P. Van Eenennaam, and S. I. Doroshov. 2002. Comparison of early
life stages and growth of Green and White Sturgeon. Pages 237–248 in W.
Van Winkle, P. J. Anders, D. H. Secor, and D. A. Dixon, editors. Biology,
management, and protection of North American sturgeon. American Fisheries
Society, Symposium 28, Bethesda, Maryland.
Dyer, S. D., K. L. Dickson, E. G. Zimmerman, and B. M. Sanders. 1991. Tissue-
specific patterns of synthesis of heat-shock proteins and thermal tolerance of
the Fathead Minnow (Pimephales promelas). Canadian Journal of Zoology
69:2021–2027.
Ellis, A. E. 2001. Innate host defense mechanisms of fish against viruses and
bacteria. Developmental and Comparative Immunology 25:827–839.
Feige, U., R. I. Morimoto, I. Yahara, and B. S. Polla, editors. 1996. Stress-
inducible cellular responses. Birkhäuser Verlag, Basel, Switzerland.
Fowler, S. L., D. Hamilton, and S. Currie. 2009. A comparison of the heat
shock response in juvenile and adult Rainbow Trout (Oncorhynchus mykiss):
implications for increased thermal sensitivity with age. Canadian Journal of
Fisheries and Aquatic Sciences 66:91–100.
Horne, M. M., and D. E. Sims. 1998. Preliminary ultrastructural studies of the
surface mucus of Atlantic Salmon. Bulletin of the Aquaculture Association
of Canada 98:85–86.
Koban, M., A. A. Yup, L. B. Agellon, and D. A. Powers. 1991. Molecular adap-
tation to environmental temperature: heat-shock response of the eurythermal
teleost Fundulus heteroclitus. Molecular Marine Biology and Biotechnology
1:1–17.
Lee, J. W., N. De Riu, S. Lee, S. C. Bai, G. Moniello, and S. S.O. Hung.
2011. Effects of dietary methylmercury on growth performance and tissue
burden in juvenile Green (Acipenser medirostris) and White Sturgeon (A.
transmontanus). Aquatic Toxicology 105:227–234.
Los, D. A., and N. Murata. 2004. Membrane fluidity and its roles in the percep-
tion of environmental signals. Biochimica et Biophysica Acta 1666:142–157.
Moyle, P. B. 2002. Inland fishes of California. University of California Press,
Berkeley.
Palmisano, A. N., J. R. Winton, and W. W. Dickhoff. 2000. Tissue-specific
induction of hsp90 mRNA and plasma cortisol response in Chinook Salmon
following heat shock, seawater challenge, and handling challenge. Marine
Biotechnology 2:329–338.
Ritossa, F. M. 1962. A new puffing pattern induced by temperature shock and
DNP in Drosophila. Experientia 18:571–573.
Sanders, B. M. 1993. Stress proteins in aquatic organisms: an environmental
perspective. Critical Reviews in Toxicology 23:49–75.
 HSP70 RESPONSES TO STRESSORS IN STURGEON
Shephard, K. L. 1994. Functions for fish mucus. Reviews in Fish Biology and
Fisheries 4:401–429.
Silvestre, F., J. Linares-Casenave, S. I. Doroshov, and D. Kültz. 2010. A pro-
teomic analysis of Green and White Sturgeon larvae exposed to heat stress
and selenium. Science of the Total Environment 408:3176–3188.
Van Eenennaam, J. P., J. Linares-Casenave, J. B. Muguet, and S. I. Doroshov.
2008. Induced spawning, artificial fertilization, and egg incubation techniques
for Green Sturgeon. North American Journal of Aquaculture 70:434–445.
Wang, Y., J. Xu, L. Sheng, and Y. Zheng. 2007. Field and laboratory investiga-
tions of the thermal influence on tissue-specific hsp70 levels in Common Carp
Downloaded by [Department Of Fisheries] at 19:51 28 February 2013
169
(Cyprinus carpio). Comparative Biochemistry and Physiology 148A:821–
827.
Wood, L. A., I. R. Brown, and J. H. Youson. 1999. Tissue and developmental
variations in the heat shock response of Sea Lampreys (Petromyzon marinus):
effects of an increase in acclimation temperature. Comparative Biochemistry
and Physiology 123A:35–42.
Yengkokpam, S., A. K. Pal, N. P. Sahu, K. K. Jain, R. Dalvi, S. Misra, and
D. Debnath. 2008. Metabolic modulation in Labeo rohita fingerlings during
starvation: hsp70 expression and oxygen consumption. Aquaculture 285:
234–237.
This article was downloaded by: [Department Of Fisheries]
On: 28 February 2013, At: 19:52
Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,
37-41 Mortimer Street, London W1T 3JH, UK

North American Journal of Aquaculture

Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/unaj20

Refining Ammonia Treatments for Control of

Prymnesium parvum in Striped Bass Fingerling
Production Ponds
a b b
Thomas A. Wyatt , Aaron Barkoh & J. Warren Schlechte
a
Texas Parks and Wildlife Department, Inland Fisheries Division, Dundee State Fish
Hatchery, Route 1 Box 123A, Electra, Texas, 76360, USA
b
Texas Parks and Wildlife Department, Inland Fisheries Division, Heart of the Hills Fisheries
Science Center, 5103 Junction Highway, Mountain Home, Texas, 78058, USA
Version of record first published: 08 Feb 2013.

To cite this article: Thomas A. Wyatt , Aaron Barkoh & J. Warren Schlechte (2013): Refining Ammonia Treatments for Control
of Prymnesium parvum in Striped Bass Fingerling Production Ponds, North American Journal of Aquaculture, 75:2, 170-177

To link to this article: http://dx.doi.org/10.1080/15222055.2012.751943

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions

This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to
anyone is expressly forbidden.

The publisher does not give any warranty express or implied or make any representation that the contents
will be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses should
be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims,
proceedings, demand, or costs or damages whatsoever or howsoever caused arising directly or indirectly in
connection with or arising out of the use of this material.
 North American Journal of Aquaculture 75:170–177, 2013

C American Fisheries Society 2013

ISSN: 1522-2055 print / 1548-8454 online

DOI: 10.1080/15222055.2012.751943

ARTICLE

Refining Ammonia Treatments for Control of Prymnesium

parvum in Striped Bass Fingerling Production Ponds
Thomas A. Wyatt
Texas Parks and Wildlife Department, Inland Fisheries Division,
Dundee State Fish Hatchery, Route 1 Box 123A, Electra, Texas 76360, USA

Aaron Barkoh* and J. Warren Schlechte

Texas Parks and Wildlife Department, Inland Fisheries Division,
Heart of the Hills Fisheries Science Center, 5103 Junction Highway, Mountain Home, Texas 78058, USA
Downloaded by [Department Of Fisheries] at 19:52 28 February 2013

Abstract
Texas state fish hatcheries use un-ionized ammonia nitrogen (NH3 -N) treatments of 0.14–0.25 mg/L to control
Prymnesium parvum in production ponds for phase-1 Striped Bass Morone saxatilis based on a published ammonia
tolerance for 4- to 6-d-old Sunshine Bass (female White Bass M. chrysops × male Striped Bass M. saxatilis). Because
fingerling production has been inconsistent and the treatments require frequent reapplications to maintain control
of P. parvum, we conducted this study to answer two questions: (1) are the treatments developed for Sunshine Bass
appropriate for Striped Bass culture, and (2) can the maximum treatments be increased as the fish grow. Striped Bass
(4, 10, 20, and 28 d old) were exposed to NH3 -N concentrations up to 1.2 mg/L for 96 h. Ammonia concentrations and
fish mortalities were monitored at 24-h intervals, their relationship modeled using logistic regression, and ammonia
concentrations that 90% of the fish survived during the various exposure periods (maximum treatments) were
estimated. The 4- to 6-d-old Striped Bass survived ammonia concentrations similar to those of Sunshine Bass of the
same age-group. The 20-d-old fish were the least tolerant of NH3 -N, followed by 4-d-old fish and then 10-d-old fish.
The 28-d-old fish was the most tolerant of ammonia toxicity. The maximum treatment varied with fish age. Thus, the
current practice of treating for concentrations between 0.14 and 0.25 mg NH3 -N/L throughout production of phase-1
fish is inappropriate. Treating with 0.14- to 0.25-mg NH3 -N/L can be suitable for culture of Striped Bass fry and
fingerlings, except for fish in the 22–24-d age range. The upper limit of concentration may be increased up to 0.37
and 0.40 mg NH3 -N/L for 10- to 12-d-old and 28- to 30-d-old Striped Bass, respectively. However, the concentration
must be lowered to 0.1 mg NH3 -N/L for 22- to 24-d-old Striped Bass.

The haptophyte Prymnesium parvum forms toxic blooms
that cause massive and extensive mortalities in cultured fishes
in many parts of the world, including the USA (Guo et al.
1996). Since 2001–2002, this microalga has become a persis-
tent threat to fish production at two Texas Parks and Wildlife
Department (TPWD) fish hatcheries. At these hatcheries, P.
parvum causes massive mortalities among all species and life
stages of fish unless strategies are used to control cell den-
sities, toxicity, or both (Barkoh et al. 2010). Currently, these
facilities use potassium permanganate to mitigate ichthyotoxi-
city and un-ionized ammonia nitrogen (NH3 -N) to control cell
densities. Because NH3 -N is also harmful to fish, conserva-
tive concentrations of 0.14–0.25 mg/L (treatments) are used to
control P. parvum cell densities in production ponds for Mo-
rone spp. fingerlings (Barkoh et al. 2004, 2010). Ammonium
sulfate is used as the source of NH3 -N because it is the most
appropriate chemical option available for control of P. parvum,
where primary and secondary production are necessary for suc-
cessful production of fingerlings (Sarig 1971; Guo et al. 1996;
Barkoh et al. 2003). Ammonium sulfate kills P. parvum without

*Corresponding author: aaron.barkoh@tpwd.state.tx.us

Received August 20, 2012; accepted November 18, 2012

170
 CONTROL OF PRYMNESIUM PARVUM IN STRIPED BASS FINGERLINGS
negatively affecting other life forms, including desirable algae
and zooplankton (Sarig 1971; Shilo 1971). In addition, its rela-
tively low cost, ease of dispersion, high solubility, and fertilizing
effect make ammonium sulfate a preferred chemical for control-
ling P. parvum (Sarig 1971; Guo et al. 1996).
Although NH3 -N can be effective against P. parvum in pro-
duction ponds for phase-1 (38–45-mm TL) Morone spp., there
are occasional fish kills in some ponds. Further, the narrow range
of treatments requires frequent (two to three times weekly)
reapplications of ammonium sulfate to maintain control of P.
parvum. Treatments are based on water temperature, pH, total
dissolved solids (TDS), and total ammonia nitrogen (TAN) lev-
els, and require frequent monitoring in production ponds. Also,
cell density and ichthyotoxicity data for P. parvum are col-
lected as part of the treatment decision-making process. These
requirements make the current control protocol labor intensive
Downloaded by [Department Of Fisheries] at 19:52 28 February 2013
and, with a small margin for error due to the narrow range of the
treatments, missteps can result in fish kills. Consequently, issues
were raised that required investigation to attempt to improve the
current TPWD treatment protocol.
The first issue concerns the appropriateness of the treatments
(0.14–0.25 mg NH3 -N/L) for other moronids (e.g., Striped Bass
Morone saxatilis and Palmetto Bass [female Striped Bass ×
male White Bass M. chrysops) when these treatments were de-
veloped for Sunshine Bass (female White Bass × male Striped
Bass; Barkoh et al. 2004). Because responses to environmental
stressors differ among Striped Bass and their hybrids (Harrell
1997; Kelly and Kohler 1999; Myers and Kohler 2000), it was
thought that these fish at similar ages may have different sen-
sitivities to ammonia toxicity. Previous studies have reported
that sensitivity to ammonia toxicity varies by fish species (Ball
1967; Boudreaux et al. 2007). Thus, if ammonia sensitivity
varies among Striped Bass and their hybrids then occasional
fish kills or low survival in production ponds could be explained
when the other relevant variables (e.g., water quality variables
such as dissolved oxygen [DO] and pH) were suitable for culture
of Morone spp.
Based on reports that sensitivity to ammonia toxicity varies
with fish size or life stage (Rice and Stokes 1975; Rice and
Bailey 1980; Solbé and Shurben 1989), the second issue con-
cerns the appropriateness of the treatments for all sizes during
phase-1 Striped Bass production when the treatments were de-
veloped for 4- to 6-d-old fish (Barkoh et al. 2004). The rationale
was if older fish (≥10 mm) tolerate NH3 -N levels greater than
the current 0.25 mg NH3 -N/L maximum treatment, then higher
rates of ammonium sulfate can be applied as the fish grow and
thereby reduce the frequency of reapplication. This thinking
assumes that ammonia decline rates are similar for all concen-
trations within a defined treatment range.
The 0.14-mg NH3 -N/L treatment is the minimum concentra-
tion that kills approximately 90% of P. parvum in 48 h, whereas
0.25 mg NH3 -N/L is the concentration in which approximately
90% of 4-d-old Sunshine Bass survive in 48 h (48-h maximum
tolerable concentration [MTC]; Barkoh et al. 2003, 2004). Thus,
171
to determine whether current treatments are suitable for Striped
Bass culture, we estimated the 48-h MTC of NH3 -N for 4-d-old
Striped Bass for comparison with that reported for Sunshine
Bass (Barkoh et al. 2004, 2010). Further, we estimated the 48-,
72-, and 96-h MTCs of NH3 -N for age 10-, 20-, and 28-d-old
Striped Bass to determine whether the treatments are appropri-
ate for Striped Bass older than 6 d old, and, if so, whether the
0.25-mg NH3 -N/L treatment can be increased as the fish grow.
We used these life stages of Striped Bass because Palmetto Bass
of similar ages were found to exhibit different sensitivities to
ammonia and pH (Bergerhouse 1993). Lastly, the 24-h MTC
and no-effect levels (NOEL, the concentration a fish survives
in a given period) of NH3 -N for Striped Bass were estimated to
increase flexibility and effectiveness of the NH3 -N treatments
to control P. parvum while minimizing the risk of ammonia
toxicity to the fish.
METHODS
Experimental conditions.—This study was conducted in-
doors at the Dundee State Fish Hatchery (DSFH) near Wichita
Falls, Texas, using water from Lake Diversion, the water source
for the hatchery. This water had calcium, sodium, magnesium,
chlorides, and TDS concentrations of 169; 568; 55; 1,407; and
2,190 mg/L, respectively; total alkalinity and total hardness of 99
and 649 mg/L as CaCO3 , respectively; salinity of 4‰; and con-
ductivity of 4,575 µS/cm. The experimental water was treated
with ultraviolet radiation and ozone to remove P. parvum cells
and toxins (hereafter, “water”; Smith 2005; Barkoh et al. 2010).
The water temperature was maintained close to the historical
mid-April average outdoor pond water temperature of 20◦ C by
using a room air conditioning system. Lighting was adjusted to
simulate daylight between 0700 and 2000 hours. Twenty 6-L
McDonald egg hatching jars were cleaned with soap and water
and thoroughly rinsed with water; four jars were randomly as-
signed to each of four treatments and a control, and randomly
arranged on an egg incubation rack. Each jar was filled with
water (5.4 L) just before each trial began. Before the jars were
filled, microscopic examination and bioassay of the water veri-
fied absence of P. parvum cells and ichthyotoxicity (Ulitzur and
Shilo 1964; Sarig 1971; Green et al. 1982; Larsen 1999).
Experimental fish.—Striped Bass of ages 4, 10, 20, and 28
d were tested; 1 d was the day eggs hatched. Eggs for these
fish were from one spawn, hatched in McDonald jars, and the
fry reared in 75-L plastic vats to the target ages. Vats had flow-
through water at rates of 3–4 L/min, and each was aerated with
compressed air through an aeration ring around the base of a
standpipe. Aeration maintained DO levels of 6 mg/L or greater
(Brown and Gratzek 1980; Piper et al. 1982; Nicholson et al.
1990) in each vat. The 10-d-old and older fish were reared from
4-d-old fish at a stocking rate of 2,000 fish per vat using rou-
tine culture practices. Twice each day, these fish were offered
zooplankton and a commercial diet (Nelson and Sons, Murray,
Utah) consisting of a salmon ration starter, followed by number
 172 WYATT ET AL.
1 granules (50% protein) and number 4 crumbles (32% protein)
as the fish grew. When each target age was attained, the fish were
collected with fine-mesh dip nets and counted into the McDon-
ald jars at 20 fish per jar. Mean total lengths were 7, 12, 19, and
26 mm for the 4-, 10-, 20-, and 28-d-old fish, respectively.
Ammonia treatments.—We selected the ammonia treatment
concentrations based on known tolerance values for Morone
spp. (Barkoh at al. 2004). Treatment concentrations were 0 (i.e.,
ambient; control), 0.2, 0.4, 0.6, and 0.8 mg NH3 -N/L for the
4- and 10-d-old fish; 0, 0.4, 0.6, 0.8, and 1.0 mg NH3 -N/L for
the 20-d-old fish; and 0, 0.6, 0.8, 1.0, and 1.2 mg NH3 -N/L
for the 28-d-old fish. Ammonium sulfate was used as source of
NH3 -N because it is routinely used to control P. parvum in fish
culture ponds at Texas state fish hatcheries (Kurten et al. 2007;
Barkoh et al. 2010). The quantity of ammonium sulfate required
to generate each treatment NH3 -N concentration in 5.5 L of
Downloaded by [Department Of Fisheries] at 19:52 28 February 2013
water was estimated graphically. Various amounts of ammonium
sulfate were weighed on a Mettler electronic balance (Mettler-
Toledo, Columbus, Ohio) to the nearest milligram; each was
dissolved in 100 mL of water and thoroughly mixed with 5.4
L of water in a McDonald jar. The TAN, pH, and temperature
in each jar were measured (see water quality below) and used
along with the TDS of the water to calculate the corresponding
NH3 -N concentration (Colt 2001). A graph of the quantities
of ammonium sulfate verses the corresponding NH3 -N levels
was used to back-estimate the required quantity of ammonium
sulfate for each NH3 -N treatment level of the experiment. We
assumed a linear relationship between ammonium sulfate and
NH3 -N, and no significant changes in pH, temperature, and TDS
of the water.
For each trial, ammonium sulfate treatment solutions
(100 mL each) were prepared individually for the treatment
replicates. Test fish were counted into the McDonald jars, each
with 5.4 L of water, before ammonia treatments to simulate the
application of ammonium sulfate to fish production ponds to
control P. parvum. Treatment solutions were transferred drop-
wise, over a 30- to 40-min period, via small-bulb pipettes into
the appropriate McDonald jars. The fish were exposed to the
treatment and control solutions for 96 h. We used 96-h exposure
treatments to complement earlier studies on moronids (e.g., Op-
penborn and Goudie 1993; Harcke and Daniels 1999). Jars were
examined for dead fish at 24, 48, 72, and 96 h after application
of treatments. Dead fish were removed from jars and counted.
At the end of 96 h, the remaining dead and live fish in each jar
also were counted. Fish were considered dead if they displayed
opaque body coloration or were unresponsive to touch. The test
was conducted separately for each age fish. All tests were static
exposures of treatment and control solutions to the fish. Jars
were not aerated, and test fish were not fed during the ammonia
exposure treatments.
Water quality.—Ammonia, temperature, pH, and DO in each
jar were measured within 2 h of fish transfer and treatment ap-
plications to ensure treatment concentrations were attained and
relevant water quality variables (temperature, pH, and DO) were
within ranges suitable for Striped Bass culture. Temperature,
pH, and DO were measured with a YSI 650 MDS handheld me-
ter fitted with a YSI 600 XL multiprobe sensor (Yellow Springs
Instruments, Yellow Springs, Ohio). These same variables also
were measured when water samples (about 120 mL each) were
taken for ammonia (via TAN) measurements at 24, 48, 72,
and 96 h after the treatments were initiated. Total ammonia
nitrogen was measured with a Denver Instruments Model
250 meter equipped with an Accumet ammonia ion-selective
electrode (Denver Instruments, Denver, Colorado). Un-ionized
ammonia concentrations were calculated from TAN values
with an equation that compensates for temperature, pH, and
TDS (Colt 2001).
Data analysis.—Because some of the fish in the controls
died, we adjusted the percent mortality data for the treatments
by the control mortality values (Schneider-Orelli 1947) before
statistical analysis. We conducted exploratory analysis using a
locally weighted regression scatter plot smoothing procedure
(Neter et al. 1996) to better understand the shape of the dose–
response curve. To estimate MTCs, the relationship between
fish mortality and NH3 -N (final values for each 24-h interval)
was modeled using logistic regression (Hosmer and Lemeshow
1989; Allison 1995; SAS 2008). The NOEL values were esti-
mated using an ANOVA approach that modeled mortality as a
function of NH3 -N treatment level. Because the mortality data
were not normally distributed, we transformed the data into
ranks and used a one-way ANOVA on the ranks, followed by a
Dunnett’s test to determine which treatment differed from the
control (Zar 1984). We defined the NOEL as the mean NH3 -N
concentration of the highest treatment level, where fish mortality
did not significantly differ from that of the control. Differences
in water quality variables or fish mortality among treatments and
control were determined by ANOVA, followed by Tukey’s test
(SAS 2008). Effects or differences were considered significant
at P ≤ 0.05.
RESULTS
Water Quality
Concentrations of TDS were 2,212; 2,220; 2,180; and
2,190 mg/L for the 4-, 10-, 20-, and 28-d-old fish tests, re-
spectively. Temperature, DO, and pH did not significantly differ
among treatments anytime during the experiment (Table 1), and
all were within ranges deemed suitable for the culture of Striped
Bass (Harrell et al. 1990). Only NH3 -N concentration signifi-
cantly differed among treatments. The concentrations of NH3 -N
declined during each 4-d experiment, and the percent decline
increased with treatment concentration (Table 1). For the same
treatment levels percent declines were extremely higher for the
28-d-old fish than for the others, and the reason is unclear.
4-d-Old Fish
Mortalities were low in the controls throughout the 4-d ex-
periment, less than 15% for the 0.2-mg NH3 -N/L treatment,
 CONTROL OF PRYMNESIUM PARVUM IN STRIPED BASS FINGERLINGS 173

TABLE 1. Mean ± SD values of temperature (◦ C), dissolved oxygen (mg/L), pH, un-ionized ammonia nitrogen (NH3 -N; mg/L), and percent NH3 -N declines
over 4-d periods in jars for testing ammonia tolerance of Striped Bass. Only NH3 -N was significantly (P < 0.05) different among treatments; NA = not applicable.

Un-ionized ammonia nitrogen treatments (mg/L)

Water quality
variable Control 0 0.2 0.4 0.6 0.8 1.0 1.2

Fish age: 4 d
Temperature 20.9 ± 0.3 20.9 ± 0.3 20.9 ± 0.3 20.9 ± 0.3 20.9 ± 0.3 NA NA
Dissolved oxygen 8.6 ± 0.3 8.6 ± 0.3 8.6 ± 0.4 8.6 ± 0.3 8.5 ± 0.4 NA NA
pH 8.06 ± 0.06 8.02 ± 0.04 7.98 ± 0.02 7.94 ± 0.02 7.89 ± 0.02 NA NA
NH3 -N 0.003 ± 0.005 0.157 ± 0.031 0.298 ± 0.054 0.478 ± 0.093 0.588 ± 0.125 NA NA
NH3 -N (96-h% 26.98 25.98 33.92 36.9
decline)
Fish age: 10 d
Temperature 20.9 ± 0.6 20.9 ± 0.6 20.9 ± 0.5 21.0 ± 0.6 20.9 ± 0.5 NA NA
Dissolved oxygen 7.9 ± 0.1 7.9 ± 0.1 7.9 ± 0.2 7.9 ± 0.1 7.9 ± 0.1 NA NA
pH 8.06 ± 0.01 8.04 ± 0.02 8.04 ± 0.04 8.0 ± 0.04 8.0 ± 0.04 NA NA
Downloaded by [Department Of Fisheries] at 19:52 28 February 2013

NH3 -N 0.0 0.161 ± 0.027 0.309 ± 0.051 0.433 ± 0.083 0.528 ± 0.056 NA NA
NH3 -N (96-h% 26.07 22.39 31.57 32.97
decline)
Fish age: 20 d
Temperature 21.2 ± 0.1 NA 21.2 ± 0.2 21.2 ± 0.1 21.1 ± 0.1 21.1 ± 0.1 NA
Dissolved oxygen 7.0 ± 0.2 NA 6.6 ± 0.3 6.9 ± 0.2 6.8 ± 0.2 6.5 ± 0.2 NA
pH 8.01 ± 0.05 NA 7.93 ± 0.07 7.90 ± 0.08 7.90 ± 0.08 7.88 ± 0.09 NA
NH3 -N 0.007 ± 0.004 NA 0.258 ± 0.055 0.373 ± 0.091 0.495 ± 0.091 0.601 ± 0.152 NA
NH3 -N (96-h% 35.85 41.47 30.88 40.38
decline)
Fish age: 28 d
Temperature 20.9 ± 0.1 NA NA 20.9 ± 0.1 20.9 ± 0.1 20.9 ± 0.1 20.9 ± 0.1
Dissolved oxygen 6.4 ± 0.9 NA NA 6.2 ± 1.2 6.4 ± 1.1 6.3 ± 1.0 6.2 ± 1.0
pH 7.75 ± 0.20 NA NA 7.71 ± 0.22 7.70 ± 0.26 7.73 ± 0.21 7.73 ± 0.21
NH3 -N 0.005 ± 0.004 NA NA 0.274 ± 0.175 0.367 ± 0.252 0.434 ± 0.276 0.536 ± 0.33
NH3 -N (96-h% 70.85 71.16 70.82 69.35
decline)

and higher than 50% for the three highest treatments (Table 2).
We observed complete mortality in all but one replicate of the
two highest treatments. All fish but two died by 96 h in the
0.6-mg NH3 -N/L treatment, whereas all fish died by 72 h in
the 0.8-mg NH3 -N/L treatment. Mortality increased in each
treatment over the course of the study. Based on the final NH3 -
N concentrations, the mean MTCs were 0.53, 0.21, 0.13, and
0.14 mg NH3 -N/L for 24, 48, 72, and 96 h, respectively (Ta-
ble 3). The NOEL was 0.54 mg NH3 -N/L for 24 h. We could
not estimate NOEL values after 24 h because all treatments
caused significantly higher mortalities than were observed in the
controls.
10-d-Old Fish
Mortalities increased in all treatments over the course of the
study. By 96 h, mortality remained below 10% in the control but
reached approximately 21–100% in the treatments, increasing
with treatment concentration (Table 2). Based on the final NH3 -
N concentrations, mean MTC was at least 0.8 mg NH3 -N/L for
24 h and decreased to 0.20 mg NH3 -N/L by 96 h (Table 3). The
NOEL decreased from 0.5 mg NH3 -N/L by 24 h to 0.25 mg
NH3 -N/L by 72 h. No NOEL was estimated for 96 h because
all treatments caused significantly more mortalities than the
controls by that time.
20-d-Old Fish
Mortalities increased over time in all treatments. At least
half of the fish were dead within 24 h in the three highest
NH3 -N concentrations, and more than 70% were dead in all
treatment concentrations within 72 h (Table 2). At the highest
three concentrations, we observed complete mortality in at least
one replicate, and all fish died by 96 h in treatments 0.8 and
1.0 mg NH3 -N/L. The mean MTC was 0.2 mg NH3 -N/L at
24 h and 0.1 mg NH3 -N/L for 48–96 h based on final ammonia
concentrations (Table 3). The NOEL could not be estimated for
any time step of the study because of significant mortalities in
all treatments.
 174 WYATT ET AL.

TABLE 2. Mean cumulative percent mortalities of Striped Bass exposed to ammonia treatments for 96 h. Data for 28-d-old fish are not included because none
died during the 4-d experiment. Treatment mortalities were adjusted for control mortalities using the Schneider-Orelli’s (1947) formula; NA = not applicable.

Un-ionized ammonia nitrogen treatments (mg/L)

Exposure
time (h) Control 0 0.2 0.4 0.6 0.8 1.0
Fish age: 4 d
24 1.3 ± 2.5 2.5 ± 5.0 1.3 ± 2.5 5.0 ± 0.0 50.0 ± 16.8 NA
48 1.3 ± 2.5 10.0 ± 4.1 16.3 ± 11.1 45.0 ± 7.1 88.8 ± 8.5 NA
72 3.8 ± 4.8 13.8 ± 4.8 45.0 ± 4.1 85.0 ± 10.8 99.5 ± 0.0 NA
96 5.0 ± 5.8 13.8 ± 4.8 55.0 ± 13.5 97.5 ± 5.0 99.5 ± 0.0 NA
Fish age: 10 d
24 2.5 ± 2.9 2.5 ± 5.0 1.3 ± 2.5 0.0 1.3 ± 2.5 NA
48 7.5 ± 2.9 12.5 ± 5.0 3.8 ± 2.5 21.3 ± 21.4 28.8 ± 9.5 NA
72 8.8 ± 2.5 20.0 ± 5.8 17.5 ± 11.9 46.3 ± 25.0 53.8 ± 13.8 NA
± ± 31.3 ± 8.5 66.3 ± 25.0 ±
Downloaded by [Department Of Fisheries] at 19:52 28 February 2013

96 8.8 2.5 21.3 4.8 100.0 0.0 NA

Fish age: 20 d
24 1.7 ± 3.3 NA 26.7 ± 7.7 51.7 ± 16.7 81.5 ± 12.3 98.3 ± 3.3
48 5.0 ± 3.3 NA 46.3 ± 9.5 80.0 ± 5.4 92.1 ± 11.8 100.0 ± 0.0
72 7.9 ± 3.7 NA 74.6 ± 14.4 91.7 ± 8.4 97.5 ± 5.0 100.0 ± 0.0
96 9.6 ± 4.4 NA 81.3 ± 14.6 95.0 ± 6.4 100.0 ± 0.0 100.0 ± 0.0

28-d-Old Fish mined. However, the NOEL (or MTC) was considered to be
No fish died in 96 h regardless of the NH3 -N treatment con- much higher than those of the younger age-groups since no
centration. Thus, the MTC and NOEL values were undeter- fish died in this age-group at the tested higher concentrations.
Based on the final NH3 -N concentrations for the highest ammo-
TABLE 3. Means and CIs of MTC and NOEL of un-ionized ammonia nitro- nia treatment level (1.2 mg NH3 -N/L), we suggest a NOEL (or
gen for different-age Striped Bass exposed to ammonia treatments for 4 d. Data MTC) of at least 0.40 for 48 h and 0.36 mg NH3 -N/L for up to
for 28-d-old fish are not reported because none died during the 4-d experiment.
96 h.
Age of fish (d)
Variablea 4 10 20 DISCUSSION
Exposure time: 24 h Ammonia had a cumulative toxicity effect on Striped Bass
MTC 0.53 at least 0.8 0.20 as indicated by the consistent increases in mortalities over the
0.50–0.56 0.15–0.25 course of the study, and the relationship between ammonia tox-
NOEL 0.54 0.50 icity and Striped Bass age was nonlinear. We found 20-d-old
fish to be the least tolerant of NH3 -N, followed by 4- and 10-d-
Exposure time: 48 h
old fish, in that order. The 28-d-old fish was the most tolerant
MTC 0.21 0.37 0.10
of ammonia toxicity since none of these fish died. Our results,
0.15–0.25 0.30–0.45 0.05–0.15
in some ways, are similar to those of Bergerhouse (1993) who
NOEL 0.40
found 14-d-old Palmetto Bass more tolerant of ammonia than
Exposure time: 72 h 5- and 20-d-old fish, an indication that the relationship between
MTC 0.13 0.30 0.10 ammonia toxicity and Palmetto Bass age is nonlinear. The mea-
0.10–0.20 0.25–0.34 0.01–0.15 sured water quality variables—temperature, DO, and pH—did
NOEL 0.25 not affect ammonia toxicity in this study.
Exposure time: 96 h The present results cannot explain the observed different
MTC 0.14 0.20 0.10 responses to ammonia toxicity of the fish tested, and we are un-
0.13–0.20 0.15–0.25 0.01–0.15 aware of a direct explanation from the literature. In fish, acute
NOEL ammonia toxicosis results from net influx of un-ionized ammo-
a nia across gill membranes and the central nervous system (CNS)
Values in mg NH3 -N/L; MTC is the concentration 90% of the fish survives; NOEL
is the concentration all fish survive; MTC or NOEL for 28- to 32-d-old Striped Bass is appears to be the target organ of ammonia poisoning (Boyd
suggested to be at least 0.36 mg NH3 -N/L. and Tucker 1998). Thus, the ammonia gradient across the gill
 CONTROL OF PRYMNESIUM PARVUM IN STRIPED BASS FINGERLINGS
membrane, permeability of the gill membrane to ammonia dif-
fusion, and sensitivity of the CNS to ammonia are factors that
separately or in combination may explain the different responses
to ammonia toxicity. Because water quality was similar among
the tested fish, it is reasonable to assume that the un-ionized
ammonia concentrations of the external microenvironments of
the gills were similar. Information on the internal microenvi-
ronments of the gills, conditions of the gills, and conditions
of the CNS among the tested fish is unknown. Comparative
histological and physiological studies of the gills and CNS of
the different ages of Striped Bass as well as knowledge of the
internal microenvironments of their gills are needed before we
can offer factual explanation of the observed different responses
of the fish to ammonia toxicity. Meanwhile, we speculate that
the observed extreme sensitivity of the 20-d-old Striped Bass
to ammonia toxicity was due to a steeper, across-gill ammonia
Downloaded by [Department Of Fisheries] at 19:52 28 February 2013
gradient that resulted in a greater net influx of ammonia or due
to oversensitivity of the CNS to ammonia, or both. These fish
were probably near or undergoing metamorphosis from postlar-
vae to juveniles, a process characterized by significant morpho-
logical and physiological changes (Webster 2002), and conse-
quently increased vulnerability of fish to environmental factors
(Grizzle and Mauldin II 1994). Conversely, the 28-d-old fish
may have completed metamorphosis (Grizzle and Mauldin II
1994) and been better equipped to survive the ammonia treat-
ments; hence, none died. The 4- and 10-d fish probably had
conditions intermediate of the two extremes.
This study did not test the ammonia sensitivities of 15- to 19-
d-old and 25- to 27-d-old Striped Bass; thus, our understanding
of ammonia toxicity to Striped Bass of ages 4–32 d remains
incomplete. Because the 20- to 24-d-old fish were the most
vulnerable to ammonia toxicity among the ages investigated in
the present study, we recommend additional studies to better
demarcate the age-related transition in ammonia tolerance of
Striped Bass. Longer-term (e.g., 28–30-d) exposures of fish to
ammonia would help better understand the chronic effects of
ammonia on Striped Bass. This information would help with
effective mitigation of P. parvum blooms without a significant
adverse effect on Striped Bass survival.
The first question this study intended to answer was whether
the 0.14- to -0.25-mg NH3 -N/L treatments are appropriate for
Striped Bass—in other words, can 90% of 4-d-old Striped Bass
survive 0.25 mg NH3 -N/L for 48 h. Barkoh et al. (2004) reported
the 48-h MTC for Sunshine Bass as 0.25 mg NH3 -N/L, the
same as the upper confidence limit of the 48-h MTC for Striped
Bass in this study. Thus, we conclude that the current ammonia
treatments are also good for controlling P. parvum in Striped
Bass fingerling production ponds. However, because the mean
MTC was 0.21 mg/L, we suggest using this as the maximum
treatment for 4- to 6-d-old Striped Bass.
We found the treatments inappropriate for certain sizes of
Striped Bass during phase-1 fingerling production. The mean
MTC for 22- to 24-d-old fish (0.1 mg/L) was lower than the
minimum treatment; thus, these fish cannot survive the treat-
175
FIGURE 1. Maximum concentrations of un-ionized ammonia nitrogen that
90% of Striped Bass of different ages survive in 48 h (48-h MTC).
ments. Similarly, the 7- to 8-d-old fish, with mean MTCs of
0.13–0.14 mg/L, cannot survive most of the treatments. Be-
cause the relationship between Striped Bass age and ammonia
concentration was not a monotonic trend, there seems to be no
easy answer to the question of whether the 0.25 mg NH3 -N/L
treatment can be increased as the fish grow. The 48-h MTCs
suggest that NH3 -N cannot be increased in a linear or mono-
tonic fashion with age of the fish but rather must be seesawed
to suit the different age-groups of Striped Bass (Figure 1). Am-
monia treatments may be increased up to 0.37 and 0.40 mg
NH3 -N/L for 10- to 12-d-old and 28- to 30-d-old Striped Bass,
respectively, but must be lowered to 0.1 mg NH3 -N/L for 20–
22-d Striped Bass. For age 6- to 8-d-old and 12- to 14-d-old fish,
maximum treatments of 0.13–0.14 and 0.20–0.30 mg NH3 -N/L,
respectively, may be appropriate, whereas the maximum treat-
ment must be lowered to 0.1 mg NH3 -N/L for 22–24-d old fish.
Though reducing ammonia concentration to 0.1 mg NH3 -N/L is
feasible by flowing freshwater through ponds, these ammonia
concentrations are ineffective in controlling P. parvum (Barkoh
et al. 2003, 2010). Thus, an alternative treatment should be used
to treat P. parvum, if needed, during these ages when the fish is
most vulnerable to ammonia toxicity.
The ammonia treatments (0.14–0.25 mg NH3 -N/L) were de-
veloped for 4- to 6-d old Sunshine Bass (Barkoh et al. 2004) but
have been used to control P. parvum in ponds with Striped Bass
older than 6 d old with mixed phase-1 fingerling production
results. To avoid ammonia-related toxicity and achieve consis-
tent Striped Bass production results, we make the following
recommendations. Treatments that are effective for the control
of P. parvum (Barkoh et al. 2003) should be selected in con-
cert with concentrations the fish can tolerate, using Table 3 as
a guide. For fish 28 d and older, the maximum concentration
should be 0.36 mg NH3 -N/L. Future studies should address the
issue of whether or not higher ammonia treatments would allow
a reduction in the frequency of ammonia reapplications while
maintaining control of P. parvum in Striped Bass fingerling
 176 WYATT ET AL.
production ponds. Our experience with the 0.14- to 0.25-mg
NH3 -N/L treatments is that applied ammonia declines to lev-
els below 0.14 mg NH3 -N/L in 5–7 d, depending on the algal
biomass of the pond.
The ammonia treatments (MTCs) discussed are conservative
in that we used the ammonia concentrations measured at the end
of each time step of the study rather than the initial exposure con-
centrations or the mean values over the various time intervals or
over the course of the study. Nonetheless, these ammonia treat-
ments must be used with caution for the control of P. parvum
because ammonia is also toxic to Morone spp. (Bergerhouse
1993; Oppenborn and Goudie 1993; Weirich et al. 1993; Ashe
et al. 1996; Harcke and Daniels 1999). Ammonia toxicity levels
or effects on fish are influenced by several physicochemical vari-
ables, including temperature, DO, pH, alkalinity, salinity, and
TDS (Emerson et al. 1975; Thurston et al. 1981; Meade 1985;
Downloaded by [Department Of Fisheries] at 19:52 28 February 2013
Bergerhouse 1993; Colt 2001). For waters with lower salinities,
alkalinities, hardness, TDS, or DO (or a combination of these
variables) than those of the water used in this study, the ammonia
treatments would be inappropriate because toxicity levels would
increase and result in higher fish mortalities. Similarly, under
conditions of higher pH and temperatures than those of our
study, the toxicity levels of the treatments would increase and
cause higher fish mortalities. Successful use of these ammonia
treatments to control P. parvum without risking production of
Morone spp. fingerlings depends on adequate knowledge of the
water quality of ponds, especially pH and temperature, which
fluctuate daily.
ACKNOWLEDGMENTS
We thank the DSFH staff for help with conducting this study
and Loraine Fries, Gerald Kurten, and Bob Betsill for reviewing
an earlier draft of the manuscript. Funding was provided in
part by Federal Aid in Sportfish Restoration grants F-95-D and
F-231-R to the TPWD.
REFERENCES
Allison, P. D. 1995. Survival analysis using SAS R
: a practical guide. SAS
Institute, Cary, North Carolina.
Ashe, D., J. R. Tomasso, A. G. Eversole, and T. I. J. Smith. 1996. Toxicity of
ammonia to fingerling White Bass: effect of selected environments on uptake
dynamics. Progressive Fish-Culturist 58:277–280.
Ball, I. R. 1967. The relative susceptibilities of some species of fresh-water fish
to poisons–I. ammonia. Water Research 1:767–775.
Barkoh, A., D. G. Smith, and J. W. Schlechte. 2003. An effective minimum
concentration of un-ionized ammonia nitrogen for controlling Prymnesium
parvum. North American Journal of Aquaculture 65:220–225.
Barkoh, A., D. G. Smith, J. W. Schlechte, and J. M. Paret. 2004. Ammonia
tolerance by Sunshine Bass fry: implication for use of ammonium sulfate
to control Prymnesium parvum. North American Journal of Aquaculture
66:305–311.
Barkoh, A., D. G. Smith, and G. M. Southard. 2010. Prymnesium parvum control
treatments for fish hatcheries. Journal of the American Water Resources
Association 46:161–169.
Bergerhouse, D. L. 1993. Lethal effects of elevated pH and ammonia on early
life stages of hybrid Striped Bass. Journal of Applied Aquaculture 2(3–4):81–
100.
Boudreaux, P. J., A. M. Ferrara, and Q. C. Fontenot. 2007. Acute toxicity of
ammonia to Spotted Gar, Lepisosteus oculatus, Alligator Gar, Atractosteus
spatula, and Paddlefish, Polyodon spathula. Journal of the World Aquaculture
Society 38:322–325.
Boyd, C. E., and C. S. Tucker. 1998. Pond aquaculture water quality manage-
ment. Kluwer Academic Publishers, Norwell, Massachusetts.
Brown, E. E., and J. B. Gratzek. 1980. Fish farming handbook: food, bait,
tropicals, and Goldfish. AVI, Westport, Connecticut.
Colt, J. 2001. Fish hatchery management appendices—table 9: ammonia cal-
culator (freshwater). Online supplement to G. A. Wedemeyer, editor. Fish
hatchery management, 2nd edition. American Fisheries Society, Bethesda,
Maryland. Available: http//www.fisheries.org/hatchery. (November 2006).
Emerson, K., R. C. Russo, R. E. Lund, and R. V. Thurston. 1975. Aqueous
ammonia equilibrium calculations: effect of pH and temperature. Journal of
the Fisheries Research Board of Canada 32:2379–2383.
Green, J. C., D. J. Hibberd, and R. N. Pienaar. 1982. The taxonomy of Prym-
nesium (Prymnesiophyceae) including a description of a new cosmopolitan
species, P. patellifera sp nov., and further observations on P. parvum N. Carter.
British Phycological Journal 17:363–382.
Grizzle, J. M., and A. C. Mauldin II. 1994. Age-related changes in survival
of larval and juvenile Striped Bass in different concentrations of calcium
and sodium. Transactions of the American Fisheries Society 123:1002–
1005.
Guo, M., P. J. Harrison, and F. J. R. Taylor. 1996. Fish kills related to Prymnesium
parvum N. Carter (Haptophyta) in the People’s Republic of China. Journal of
Applied Phycology 8:111–117.
Harcke, J. E., and H. V. Daniels. 1999. Acute toxicity of ammonia and nitrite to
reciprocal cross hybrid Striped Bass Morone chrysops × M. saxatilis eggs
and larvae. Journal of the World Aquaculture Society 30:496–500.
Harrell, R. M. 1997. Hybridization and genetics. Pages 217–234 in R. M. Harrell,
editor. Striped Bass and other Morone culture. Elsevier, Amsterdam.
Harrell, R. M., J. H. Kerby, and R. V. Minton, editors. 1990. Culture and
propagation of Striped Bass and its hybrids. American Fisheries Society,
Southern Division, Striped Bass Committee, Bethesda, Maryland.
Hosmer, D. W., and S. Lemeshow. 1989. Applied logistic regression. Wiley,
New York.
Kelly, A. M., and C. C. Kohler. 1999. Cold tolerance and fatty acid composition
of Striped Bass, White Bass, and their hybrids. North American Journal of
Aquaculture 61:278–285.
Kurten, G. L., A. Barkoh, L. T. Fries, and D. C. Begley. 2007. Combined nitrogen
and phosphorus fertilization for controlling the toxigenic alga Prymnesium
parvum. North American Journal of Aquaculture 69:214–222.
Larsen, A. 1999. Prymnesium parvum and P. patelliferum (Haptophyta)—one
species. Phycologia 38:541–543.
Meade, J. W. 1985. Allowable ammonia for fish culture. Progressive Fish-
Culturist 47:135–145.
Myers, J. J., and C. C. Kohler. 2000. Acute responses to salinity for Sunshine
Bass and Palmetto Bass. North American Journal of Aquaculture 62:195–202.
Neter, J., M. H. Kutner, C. J. Nachtsheim, and W. Wasserman. 1996. Applied
linear statistical models. McGraw-Hill, Boston.
Nicholson, L. C., L. C. Woods III, and J. G. Woiwode. 1990. Intensive culture
techniques for the Striped Bass and its hybrids. Pages 141–157 in R. M.
Harrell, J. H. Kerby, and R. V. Minton, editors. Culture and propagation of
Striped Bass and its hybrids. American Fisheries Society, Southern Division,
Striped Bass Committee, Bethesda, Maryland.
Oppenborn, J. B., and C. A. Goudie. 1993. Acute and sublethal effects of
ammonia on Striped Bass and hybrid Striped Bass. Journal of the World
Aquaculture Society 24:90–101.
Piper, R. G., I. B. McElwain, L. E. Orme, J. P. McCraren, L. G. Fowler, and J. R.
Leonard. 1982. Fish hatchery management. U.S. Fish and Wildlife Service,
Washington, D.C.
Rice, S. D., and J. E. Bailey. 1980. Survival, size, and emergence of Pink Salmon,
Oncorhynchus gorbuscha, alevins after short- and long-term exposures to
ammonia. U.S. National Marine Fisheries Service Fishery Bulletin 78:641–
648.
 CONTROL OF PRYMNESIUM PARVUM IN STRIPED BASS FINGERLINGS
Rice, S. D., and R. M. Stokes. 1975. Acute toxicity of ammonia to several devel-
opmental stages of Rainbow Trout, Salmo gairdneri. U.S. National Marine
Fisheries Service Fishery Bulletin 73:207–211.
Sarig, S. 1971. Toxin-producing algae: Prymnesium parvum Carter. Pages 17–
43 in S. F. Snieszko and H. R. Axelrod, editors. Diseases of fishes, book
3: the prevention and treatment of diseases of warmwater fishes under sub-
tropical conditions, with special emphasis on intensive fish farming. T. F. H.
Publications, Neptune City, New Jersey.
SAS (Statistical Analysis Systems). 2008. SAS user’s guide: statistics. SAS
Institute, Cary, North Carolina.
Schneider-Orelli, O. 1947. Entomologisches Praktikum; Einführung in die land-
und forstwirtschaftliche Insektenkunde. Sauerländer, Aarau, Switzerland.
(In German.)
Shilo, M. 1971. Toxins of Chrysophyceae. Pages 67–103 in S. Kadis, A.
Ciegler, and S. J. Ajl, editors. Microbial toxins, volume 7. Academic Press,
New York.
Smith, D. G. 2005. Efficacy of ultraviolet radiation to control Prymnesium
parvum cells and toxicity. Pages 66–70 in A. Barkoh and L. T. Fries, editors.
Management of Prymnesium parvum at Texas state fish hatcheries. Texas
Downloaded by [Department Of Fisheries] at 19:52 28 February 2013
177
Parks and Wildlife Department, Management Data Series 236, PWD RP
T3200-1138, Austin.
Solbé, J. F. de L. G., and D. G. Shurben 1989. Toxicity of ammonia to early life
stages of Rainbow Trout (Salmo gairdneri). Water Research 23:127–129.
Thurston, R. V., G. R. Phillips, R. C. Russo, and S. M. Hinkins. 1981. Increased
toxicity of ammonia to Rainbow Trout (Salmo gairdneri) resulting from
reduced concentrations of dissolved oxygen. Canadian Journal of Fisheries
and Aquatic Sciences 38:983–988.
Ulitzur, S., and M. Shilo. 1964. A sensitive assay system for determination of
the ichthyotoxicity of Prymnesium parvum. Journal of General Microbiology
36:161–169.
Webster, C. D. 2002. Hybrid Striped Bass. Pages 327–343 in C. D. Webster
and C. E. Lim, editors. Nutrient requirements and feeding of finfish for
aquaculture. CAB International, New York.
Weirich, C. R., J. R. Tomasso, and T. I. J. Smith. 1993. Toxicity of ammonia and
nitrite to Sunshine Bass in selected environments. Journal of Aquatic Animal
Health 5:64–72.
Zar, J. H. 1984. Biostatistical analysis, 2nd edition. Prentice-Hall, Englewood
Cliffs, New Jersey.
This article was downloaded by: [Department Of Fisheries]
On: 28 February 2013, At: 19:52
Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,
37-41 Mortimer Street, London W1T 3JH, UK

North American Journal of Aquaculture

Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/unaj20

Evaluation of Commercial Marine Fish Feeds for

Production of Juvenile Cobia in Recirculating
Aquaculture Systems
a b c a b d
Paul S. Wills , Charles R. Weirich , Richard M. Baptiste & Marty A. Riche
a
Center for Aquaculture and Stock Enhancement, Harbor Branch Oceanographic Institute at
Florida Atlantic University, 5600 US 1 North, Ft. Pierce, Florida, 34946, USA
b
U.S. Department of Agriculture, Agricultural Research Service, 5600 US 1 North, Ft. Pierce,
Florida, 34946, USA
c
Rancheros del Mar Sociedad Anónima de Capital Variable, Carretera a Pichilingue Kilometro
2.5 L-13, Marina Palmira, La Paz, Baja California Sur, 23000, Mexico
d
U.S. Department of Agriculture, Agricultural Research Service, Stuttgart National
Aquaculture Research Center, 2955 Highway 130 East, Stuttgart, Arkansas, 72160, USA
Version of record first published: 19 Feb 2013.

To cite this article: Paul S. Wills , Charles R. Weirich , Richard M. Baptiste & Marty A. Riche (2013): Evaluation of Commercial
Marine Fish Feeds for Production of Juvenile Cobia in Recirculating Aquaculture Systems, North American Journal of
Aquaculture, 75:2, 178-185

To link to this article: http://dx.doi.org/10.1080/15222055.2012.750635

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions

This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to
anyone is expressly forbidden.

The publisher does not give any warranty express or implied or make any representation that the contents
will be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses should
be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims,
proceedings, demand, or costs or damages whatsoever or howsoever caused arising directly or indirectly in
connection with or arising out of the use of this material.
 North American Journal of Aquaculture 75:178–185, 2013

C American Fisheries Society 2013

ISSN: 1522-2055 print / 1548-8454 online

DOI: 10.1080/15222055.2012.750635

ARTICLE

Evaluation of Commercial Marine Fish Feeds for Production

of Juvenile Cobia in Recirculating Aquaculture Systems
Paul S. Wills*
Center for Aquaculture and Stock Enhancement,
Harbor Branch Oceanographic Institute at Florida Atlantic University,
5600 US 1 North, Ft. Pierce, Florida 34946, USA

Charles R. Weirich1
U.S. Department of Agriculture, Agricultural Research Service, 5600 US 1 North, Ft. Pierce,
Downloaded by [Department Of Fisheries] at 19:52 28 February 2013

Florida 34946, USA

Richard M. Baptiste
Center for Aquaculture and Stock Enhancement,
Harbor Branch Oceanographic Institute at Florida Atlantic University,
5600 US 1 North, Ft. Pierce, Florida 34946, USA

Marty A. Riche2
U.S. Department of Agriculture, Agricultural Research Service, 5600 US 1 North, Ft. Pierce,
Florida 34946, USA

Abstract
The effect of different commercially available marine fish diets on production characteristics and body composition
of juvenile Cobia Rachycentron canadum reared in production-scale recirculating aquaculture systems was evaluated
in a 57-d growth trial. Juvenile Cobia (mean weight ± SE, 26.7 ± 0.9 g) were stocked at an initial density of 1.2 kg/m3.
After stocking, fish were fed one of three closed-formula diets formulated for carnivorous marine finfish (coded diet
A, 50% crude protein : 22% crude lipid; diet B, 49% crude protein : 17% crude lipid; and diet C, 48% crude protein :
17% crude lipid), all at a targeted feed rate of 3–5% body weight per day. At 2-week intervals, 10% of the population
of each tank was sampled to determine mean weight, weight gain, specific growth rate, feed conversion efficiency,
and biomass. At the termination of the trial, the entire population of each tank was harvested to determine the same
characteristics and survival. In addition, fish were sampled to determine relative changes in whole body composition,
energy retention, protein efficiency ratio, and protein productive value. Final weight (203.3 g), specific growth rate
(3.6%/d), feed conversion efficiency (92.2%), biomass (7.3 kg/m3), and protein productive value (25.2%) of fish fed
the high-lipid diet A were significantly higher than those of fish fed the other two diets. No differences in whole body
composition were observed among fish fed the three diets with the exception of dry matter composition. Contrary to
previous reports, the results of the current study indicate that juvenile Cobia reared in production-scale recirculating
aquaculture systems fed high-lipid diets exhibit protein sparing and better growth.

*Corresponding author: pwills2@hboi.fau.edu

1
Present address: Rancheros del Mar Sociedad Anónima de Capital Variable, Carretera a Pichilingue Kilometro 2.5 L-13, Marina Palmira,
La Paz, Baja California Sur 23000, Mexico.
2
Present address: U.S. Department of Agriculture, Agricultural Research Service, Stuttgart National Aquaculture Research Center, 2955
Highway 130 East, Stuttgart, Arkansas 72160, USA.
Received September 19, 2012; accepted November 9, 2012

178
 COMMERCIAL DIETS FOR PRODUCTION OF JUVENILE COBIA
Commercial cultivation of Cobia Rachycentron canadum,
a large migratory marine pelagic fish (Smith 1986), is gain-
ing considerable interest in the United States. Cobia posses
many characteristics that make them ideal for culture, including
well-established captive spawning and larval rearing techniques
that allow mass production of juveniles for grow-out operations
(Dodd 2001; Franks et al. 2001; Liao et al. 2001; Arnold et al.
2002; Kilduff et al. 2002; Weirich et al. 2006; Holt et al. 2007;
Benetti et al. 2008). Cobia exhibit very high growth rates and
feed conversion efficiencies under various culture conditions in
open-ocean net-pens, ponds, and recirculating aquaculture sys-
tems (RASs) and can attain weights up to 4–6 kg in a year (Chou
et al. 2001, 2004; Liao et al. 2004; Weirich et al. 2004, 2010;
Lunger et al. 2006). Cobia adapt well to culture in many types
of systems including tank culture (Schwarz et al. 2004; Weirich
et al. 2004, 2010), and the product is well accepted by con-
Downloaded by [Department Of Fisheries] at 19:52 28 February 2013
sumers (McClane 1974; Oesterling 2001; Rickards 2001; Shiau
2007).
Dietary evaluations with Cobia are ongoing, particularly per-
taining to substitutes for fish meal and fish oil in the formulations
to reduce cost and environmental impacts (Chou et al. 2004;
Zhou et al. 2005, 2011; Lunger et al. 2006, 2007a, b; Fraser and
Davies 2009). Despite studies evaluating ingredient digestibil-
ity (Zhou et al. 2004), select essential amino acid requirements
(Zhou et al. 2006; Zhou et al. 2007), and optimal crude protein
(CP) and crude lipid (CL) inclusion (Chou et al. 2001; Wang
et al. 2005; Craig et al. 2006), knowledge regarding optimal Co-
bia nutrition is limited. Until sufficient nutritional information
is established to formulate precision diets for Cobia, it is impor-
tant to identify commercially available feeds that perform well
for this species under different culture conditions (Fraser and
Davies 2009). These types of studies are difficult to perform
by commercial producers due to the relatively small number
of treatments that typical farms are capable of testing within
a given production cycle. To this end, Weirich et al. (2010)
fed three different commercial diets, with varying levels of CP
and CL content, only one of which was specifically formulated
for marine fish, to juvenile Cobia to determine effects on pro-
duction characteristics and body composition when reared in
RASs. Production characteristics were good for all three formu-
lations tested, with the higher protein marine fish diet yielding
significantly faster growth. The present study was conducted
to evaluate three closed-formula commercial marine fish diets,
each with similar CP and CL levels reported on their labels, for
their effect on production characteristics and body composition
of juvenile Cobia reared in RASs.
METHODS
Experimental systems.—Studies were conducted in four
replicate RASs located within the U.S. Department of Agri-
culture’s Sustainable Tank Aquaculture Recirculating Research
Facility on the campus of Harbor Branch Oceanographic In-
stitute at Florida Atlantic University, Ft. Pierce, Florida. Each
179
45-m3 RAS consisted of four 8-m3 round fiberglass culture tanks
on a shared filtration system. The complete system design and
operational parameters have been described in a prior study
conducted with juvenile Cobia (Weirich et al. 2010).
Experimental design, husbandry, and sampling protocol.—
To initiate a 57-d rearing trial, juvenile Cobia, reared from re-
cently metamorphosed post larvae obtained from the Rosen-
stiel School of Marine and Atmospheric Science, University
of Miami, Miami, Florida were stocked into three of the four
tanks within each of the four RASs at an initial tank density
of 1.2 kg/m3 (350 fish/tank). The mean ± SE weight of fish
at stocking was 26.1 ± 0.9 g (n = 100). Dietary treatments
were assigned to tanks using a randomized complete block de-
sign, with blocking on the system resulting in four replicates per
treatment.
The three diets evaluated were EXTR 450 Sink (Rangen,
Buhl, Idaho), Silvercup Steelhead (Nelson and Sons, Murray,
Utah), and Marine Grower (Zeigler Bros., Gardners, Pennsylva-
nia). These formulations were sold as marine fish diets suitable
for grow out of juvenile Cobia. Tanks of fish were assigned and
fed one of the three slow-sinking diets arbitrarily coded as diet
A, diet B, and diet C. Feed tag specifications were as follows: not
less than 50% CP:15% CL, 45% CP:16% CL, and 44% CP:15%
CL for diets A, B, and C, respectively (we did not link the diet
codes with the manufacturer names). Actual proximate compo-
sition, amino acid analysis, and fatty acid analysis (Tables 1,
2) were performed by a commercial feed laboratory (Midwest
Laboratories, Omaha, Nebraska).
Fish were allowed 30 min to consume their allotted rations
at each feeding. The size of pellets fed from day 1 through day
40 was 3.0 mm; thereafter, fish were fed 5.0-mm pellets until
the conclusion of the trial. The daily amount of feed offered
ranged from 3% to 5% body weight per day, decreasing with
increasing fish size, and was achieved by increasing total feed
offered to each tank by 4% daily. Feed was offered twice per
day up to one half of the daily calculated ration at each feeding
or until satiation, which ever occurred first. Total daily rations
were recorded. Waste feed not automatically removed by the
filtration systems was removed from the tanks at the end of
the day, although very little was observed due to the satiation
feeding method.
At stocking, and at 2-week intervals thereafter, 10% of the
fish population of each tank was sampled and individual weights
recorded to the nearest 1.0 g and lengths to the nearest millime-
ter. Production characteristics evaluated included mean weight,
weight gain, total length, specific growth rate (SGR), feed con-
version efficiency (FCE), condition factor (K), and biomass.
Specific growth rate was calculated using the formula SGR =
[log10 (Wt ) − log10 (Wi )]/t × 100, where Wt = the mean weight
at the end of a sampling period, Wi = the initial mean weight,
and t = the length of time of the sampling period in days.
Feed conversion efficiency was calculated using the following
formula: (g wet weight gain/g dry weight feed fed) × 100. Con-
dition factor, K, was calculated using the following formula:
 180 WILLS ET AL.

TABLE 1. Analyzed composition (g/100 × g dry diet) of commercially avail- TABLE 2. Analyzed fatty acid composition (g/100 g of diet) of commercially
able marine fish diets fed to juvenile Cobia during the 57-d experimental rearing available marine fish diets fed to juvenile Cobia during the 57-d experimental
trial. rearing trial.

Variable Diet A Diet B Diet C Fatty acid composition Diet A Diet B Diet C
Gross energy (cal/g) 5,688 5,323 5,260 Saturated
Dry matter 93.5 95.5 96.0 Lauric (C12:0) 0.05 0.03 0.03
Ash 8.8 8.1 8.9 Myristic (C14:0) 2.80 1.04 0.84
Crude protein 50.6 49.4 47.8 Pentadecanoic (C15:0) 0.13 0.06 0.08
Crude lipid 21.9 16.7 16.5 Palmitic (C16:0) 3.60 3.80 3.57
Indispensable amino acids Heptadecanoic (C17:0) 0.05 0.07 0.07
Arginine 2.48 3.18 4.04 Stearic (C18:0) 0.50 0.99 0.92
Histidine 1.26 1.33 1.53 Arachidic (C20:0) 0.05 0.04 0.04
Isoleucine 1.77 1.33 1.46 Behenic (C22:0) 0.03 0.02 0.03
Leucine 3.71 2.98 3.09 Monounsaturated
Total lysine 3.24 2.88 2.95 Myristoledic (C14:1 Trans) 0.09 0.02 0.02
Methionine 0.94 0.91 0.95 Myristoleic (C14:1 Cis) 0.03 0.03 0.03
Downloaded by [Department Of Fisheries] at 19:52 28 February 2013

Phenylalanine 2.20 1.82 1.83 Palmitelaidic (C16:1 Trans) 0.04 0.07 0.06
Threonine 1.80 1.73 1.89 Palmitoleic (C16:1 Cis) 1.69 1.08 0.84
Tryptophan 0.33 0.29 0.31 Oleic (C18:1 Cis) 2.73 4.43 3.91
Valine 3.02 2.30 1.96 11-Eicosenoic (C20:1) 3.64 0.12 0.21
Dispensable amino acids Erucic (C22:1) 0.33 0.02 0.03
Alanine 4.22 3.68 3.37 Polyunsaturated (n-6)
Aspartic acid 3.71 3.63 3.77 Linoleic (C18:2 Cis) 1.42 2.17 2.52
Cystine 0.44 0.61 0.61 Gamma Linolenic (C18:3 gamma) 0.03 0.03 0.03
Glutamic acid 6.12 5.72 6.74 11–14 Eicosadienoic (C20:2) 0.05 0.04 0.04
Glycine 3.42 2.32 2.11 Homo-gamma Linolenic (C20:3) 0.02 0.03 0.02
Proline 2.44 2.39 2.57 Arachidonic (C20:4) 0.11 0.12 0.10
Serine 1.94 2.08 2.13 Polyunsaturated (n-3)
Tyrosine 1.49 1.27 1.45 Alpha Linolenic (18:3 alpha) 0.33 0.17 0.28
Eicosapentaenoic (C20:5) 1.60 1.19 0.95
Docosapentaenoic (C22:5) 0.27 0.17 0.19
Docosahexaenoic (C22:6) 2.05 0.79 1.53
(weight in g)/(total length in mm)3 × 105. Mortalities were re- Total saturated fats 7.25 6.08 5.64
moved and recorded daily. Mean biomass of fish from each Total polyunsaturated fats 5.91 4.72 5.68
treatment tank, corrected for growth and mortalities, was used Total monounsaturated fats 8.64 5.75 5.12
to recalculate appropriate feed rations biweekly. At the termi- Total trans fatty acids 0.13 0.13 0.11
nation of the rearing trial the entire population of each tank was Total n-3 fatty acids 4.27 2.32 2.96
harvested, individuals counted, and fish batch weighed to deter- Total n-6 fatty acids 1.61 2.37 2.69
mine production characteristics and survival. In addition, five Total n-9 fatty acids 3.06 4.45 3.94
fish were collected from each tank to determine the hepatoso- n-3:n-6 2.65 0.98 1.1
matic index (HSI = liver weight/body weight × 100), and 15
fish were collected and stored at −20◦ C for subsequent analysis
of whole body composition. Insufficient intraperitoneal fat in all Joseph, Michigan), and CP was calculated as N × 6.25. Ash was
treatments precluded a reliable measurement of this parameter. determined following incineration at 600◦ C for 2 h. Crude lipid
Whole body composition and nutritional indices.—Standard was determined gravimetrically following tissue extraction with
procedures were used for determining proximate components of diethyl ether (Ankom XT-15, Ankom Technology, Macedonia,
fish and diets (AOAC 2003). Whole body samples were pooled New York). Gross energy was determined by adiabatic bomb
by tank, minced in a meat grinder, and further homogenized us- calorimetry (Parr 1266, Parr Instruments, Moline, Illinois). In
ing a mortar and pestle. Two samples from each homogenized addition to proximate components, energy retention (ER), pro-
pool were taken for moisture analysis and dried at 105◦ C for tein efficiency ratio (PER), and protein productive value (PPV)
24 h. Tissues remaining in the pooled samples were analyzed for were determined. Energy retention was calculated using the fol-
CP, CL, ash, and gross energy (GE). Nitrogen was determined lowing formula: (Wt × Et − Wi × Ei )/FI × E, where Wt = the
following combustion (TruSpec N-elemental analyzer, Leco, St. weight at the termination of the trial, Wi = the initial weight,
 COMMERCIAL DIETS FOR PRODUCTION OF JUVENILE COBIA 181

TABLE 3. Mean ± SE final weight, specific growth rate (SGR), feed conversion efficiency (FCE), feed intake, condition factor (K), and survival of juvenile
Cobia (initial weight 26.7 g) fed commercially available marine fish diets in a 57-d rearing trial. Values within rows sharing a letter are not significantly different
(ANOVA-Tukey; P > 0.05).

Production variable N Diet A Diet B Diet C

Final weight (g) 4 203.3 ± 3.2 x 170.5 ± 7.8 y 110.9 ± 6.5 z
Total length (mm) 140 275 ± 5x 270 ± 6x 242 ± 3y
Biomass (kg/m3) 4 7.3 ± 0.6 x 5.6 ± 0.5 y 2.8 ± 0.2 z
SGR (%/d) 4 3.6 ± 0.03 x 3.2 ± 0.10 y 2.5 ± 0.10 z
FCE (%) 4 92.2 ± 0.8 x 80.3 ± 2.8y 57.7 ± 2.34 z
Feed intake (g/fish) 4 220.4 ± 4.4 x 212.0 ± 3.6 y 192.4 ± 10.5 z
Condition (K) 140 0.94 ± 0.03 x 0.88 ± 0.08 x 0.71 ± 0.02 y
Survival (%) 4 80.0 ± 3.3 x 72.2 ± 3.4 x 55.9 ± 3.2 y

FI = the feed intake, and E, Et , and Ei = the energy content age data were arcsine–square root transformed prior to analysis.
Downloaded by [Department Of Fisheries] at 19:52 28 February 2013

of the diet, fish at termination, and fish at initiation of the rear-
ing trial, respectively. Protein efficiency ratio was calculated
using the following formula: g wet weight gain/g protein in-
take. Protein productive value was calculated using the formula
(Wt × Pt − Wi × Pi )/FI × P, where P, Pt , and Pi = the protein
content of the diet, fish at termination, and fish at initiation of
the rearing trial, respectively. Although samples were collected
at the initiation of the trial, samples were lost due to a freezer
malfunction. Therefore, initial whole body protein and energy
values from a group of 29.2-g juvenile Cobia from a previous
study (Weirich et al. 2010) grown using the same brand of feed,
formulation, and feeding protocols as used prior to initiation
of the current study were used as surrogate initial whole body
values. As a result, reported values represent relative differences
due to treatment, as the initial compositional values used in the
current study were the same for all treatments.
Water quality.—Temperature, salinity, and pH of water ob-
tained from each RAS sump and dissolved oxygen of each cul-
ture tank were measured twice daily at 0900 and 1700 hours dur-
ing the rearing trial. The total ammonia nitrogen (TAN), nitrite-
nitrogen (NO2 -N), and alkalinity of each RAS were measured
daily at 1000–1200 hours. Temperature and dissolved oxygen
were measured using a LDO HQ20 meter (Hach, Loveland, Col-
orado) and salinity and pH were measured using a YSI 556 m
(YSI, Yellow Springs, Ohio). The TAN and NO2 -N were deter-
mined via Hach colorimetric assays (methods 8155 and 8153,
respectively) using a D/R 2500 spectrophotometer. Alkalinity
was determined using Hach digital titration method 8203. Wa-
ter quality within the RAS remained within acceptable levels
for Cobia culture in each study tank during the entire study.
Water temperature and salinity ranged from 25.6◦ C to 29.0◦ C
and from 23.0 to 30.0 g/L, respectively. Postfeeding dissolved
oxygen levels were maintained >80% saturation. The TAN and
NO2 -N were each maintained at levels below 0.35 mg/L. Al-
kalinity was maintained between 200 and 250 mg/L as CaCO3 .
The pH ranged from 7.2 to 7.6. Sodium bicarbonate was added
daily to the systems to maintain alkalinity levels.
Statistical analyses.—Statistical analyses were performed
using SAS 9.2 (SAS Institute, Cary, North Carolina). Percent-
Values of SGR were transformed prior to analysis. All data were
subjected to analysis of variance (ANOVA) using a randomized
complete block model and means separation was achieved us-
ing Tukey’s studentized range test at a significance level of α =
0.05.
RESULTS
Treatment means for final weight, SGR, and FCE of juvenile
Cobia were significantly different from each other with diet A
> diet B > diet C (Table 3). Total length and condition factor
of fish fed diet A were not significantly different from those fed
diet B; however, both had a greater total length than those fed
diet C (Table 3). Mean final tank biomass of fish fed diets A,
B, and C were significantly different from each other (Table 3).
Mean survival of fish fed diets A and B was significantly higher
than fish fed diet C (Table 3).
Analyzed proximate and amino acid (AA) compositions of
the diets are reported in Table 1. Dietary CP was similar among
all the diets, but CL and GE were higher in diet A relative to
the others. Analyzed fatty acid (FA) composition of the diets is
reported in Table 2. Among the saturated FAs, 14:0 was higher
and 18:0 lower in diet A than in diets B or C. Relative to diets B
and C, diet A was also higher in 20:1 and 22:1 but lower in 16:1.
The n-6 polyunsaturated FAs were similar with the exception of
higher 18:2 in diets B and C than in diet A. Conversely, all of the
n-3 polyunsaturated series were higher in diet A relative to the
other diets. Unlike the other dietary FAs, docosahexaenoic acid
(DHA; 22:6n-3) was higher and eicosapentaenoic acid (EPA;
22:5n-3) was lower in diet B than in diet C.
Whole body composition of juvenile Cobia sampled from
each dietary treatment at the conclusion of the rearing trial
showed no differences in GE, ash, CP, or CL (Table 4). Dry
matter content of fish fed diet A and diet C were not signifi-
cantly different from each other, and fish fed diet B were not
significantly different from fish fed diet C; however, dry matter
content of fish fed diet A was significantly lower than that of
fish fed diet B.
 182 WILLS ET AL.

TABLE 4. Whole body composition (g/100 g dry matter) of juvenile Cobia fed commercially available marine fish diets in a 57-d rearing trial. Values
(means ± SE; n = 4) within rows sharing a letter are not significantly different (P > 0.05).

Variable Diet A Diet B Diet C

Gross energy (cal/g) 6,030 ± 23 z 6,159 ± 41 z 6,153 ± 33 z
Dry matter 28.8 ± 0.3 y 30.6 ± 0.4 z 29.6 ± 0.4 y,z
Ash 9.0 ± 0.3 z 9.0 ± 0.5 z 9.8 ± 0.2 z
Crude protein 53.7 ± 0.4 z 52.0 ± 0.5 z 53.0 ± 1.0 z
Crude lipid 35.4 ± 1.0 z 38.5 ± 0.9 z 37.1 ± 0.7 z

Significant differences existed among the three diets for ER,
PER, and PPV (Table 5). Energy retention of fish fed diet A and
diet B were not significantly different from each other but both
were significantly higher than diet C. The PER and PPV among
the treatments were significantly different in the rank order of
Downloaded by [Department Of Fisheries] at 19:52 28 February 2013
generic commercial marine fish diets (Fraser and Davies 2009;
NRC 2011).
Fish fed diet A exhibited higher ER (8.9% increase) and PPV
(10.7% increase) relative to fish fed diet C. The better ER and
PPV suggest a better balance of AA or protein to digestible

diet A > diet B > diet C. There were no significant differences
in HSI.
DISCUSSION
Despite recent advances, current understanding of Cobia nu-
trition remains limited. Until nutritional requirements are estab-
lished, it will be necessary to identify commercial feeds that
perform well given the applied culture conditions (Fraser and
Davies 2009). The benefits of such an approach and its appli-
cation have been previously demonstrated using diets from a
single manufacturer (Weirich et al. 2010). The current study
evaluated production characteristics of juvenile Cobia reared in
production-scale RASs using commercial diets formulated for
carnivorous marine fish. The differences in performance of Co-
bia fed the three generic marine fish diets underscore the need
for further development of Cobia-specific diets.
Throughout the trial, feed was offered daily up to a calcu-
lated ration or until the fish lost interest in feeding, which ever
occurred first. Cobia are notably aggressive feeders so any re-
duction in FI is cause for attention. The lower FI observed in
fish fed diet C is often indicative of palatability issues in Co-
bia (Lunger et al. 2006; Schock et al. 2012). Although FI of
diet C was lower, feeding behavior was not directly quantified
and so the link between reduced intake and palatability remains
anecdotal. Nevertheless, given our current understanding of Co-
bia nutrition, poor palatability could manifest itself when using
energy ratio for Cobia in diet A. Insufficient or poorly balanced
AA limits protein deposition and increases catabolism of un-
used AA, resulting in increased ammonia production (Mach
and Nortvedt 2011) and increasing the burden of its removal
from the RAS (Timmons and Ebeling 2007). Alternatively, the
higher lipid and dietary energy of diet A may have spared pro-
tein for deposition (Halver and Hardy 2002). The CL of diet
A was 31–33% higher than the other diets, with significantly
better growth, FCE, and protein retention (PER and PPV). As
CL and GE were similar in diets B and C, protein and energy
sources in diet C likely had lower digestibility resulting in the
lower FCE and reduced performance relative to fish fed diet B.
Cobia digest a wide variety of ingredients well; however,
certain protein sources (e.g., rapeseed or meat and bone meal)
show lower digestibility (Zhou et al. 2004). A curvilinear rela-
tionship between PER and poultry by-product meal substitution
for fish meal was observed in juvenile Cobia by Zhou et al.
(2011). Maximum performance was obtained with a substitu-
tion of 30–40% poultry by-product meal (Zhou et al. 2011) but
has been reported as high as 60% (Saadiah et al. 2011). How-
ever, Cobia fed increasing levels of soybean meal and poultry
meal substituted for fish meal resulted in reduced growth and FI
with corollaries in the blood metabolome (Schock et al. 2012).
It was unclear if the reduced growth was due to reduced FI
or effects on energy metabolism. Altering key dietary com-
ponents, as in least-cost feed formulation, could compromise
energy metabolism (Schock et al. 2012). The importance of

TABLE 5. Selected dietary efficiency parameters and hepatosomatic index (HSI) of juvenile Cobia fed commercially available marine fish diets in a 57-d rearing
trial. Values (means ± SE) within rows sharing a letter are not significantly different (P > 0.05).

Parameter N Diet A Diet B Diet C

Energy retention (%) 4 24.9 ± 0.3 y 25.0 ± 1.5 y 16.0 ± 1.0 z
Protein efficiency ratio 4 1.64 ± 0.01 x 1.44 ± 0.06 y 0.93 ± 0.05 z
Protein productive value (%) 4 25.2 ± 0.5 x 22.9 ± 1.2 y 14.5 ± 0.8 z
HSI (%) 20 6.3 ± 0.3 z 5.7 ± 0.3 z 5.3 ± 0.3 z
 COMMERCIAL DIETS FOR PRODUCTION OF JUVENILE COBIA
understanding the effects of varying dietary components, be-
yond balancing nutrients, cannot be understated (de Almeida
Bicudo et al. 2010).
The dietary CP and AA profiles of the diets were similar, with
overall AA profiles of diets B and C nearly identical. Growth
and efficiency were better in fish fed diet A than the other two
diets despite 39% lower Arginine than in diet C. That the 2.5%
Arginine in diet A was not limiting is consistent with the results
of Saadiah et al. (2011) and other marine fish (NRC 2011).
Lysine and methionine are often cited as first limiting AAs in
fish feeds that utilize plant proteins (Gatlin et al. 2007). Although
methionine in all the diets was lower than the 1.19% dietary
requirement for Cobia (Zhou et al. 2006), the levels were the
same across all the diets. While marginal methionine levels
may have reduced overall growth, dietary levels of methionine
were not responsible for growth differences between the diets.
Downloaded by [Department Of Fisheries] at 19:52 28 February 2013
Availability of methionine is generally high in Cobia (Zhou et al.
2004); however, the effect of minor differences in availability
between diets cannot be discounted when dietary levels are at
or marginally below the requirement.
Although there were minor differences, lysine across all the
diets was in excess of the 2.33% dietary requirement established
for Cobia (Zhou et al. 2007). Unless lysine was highly unavail-
able to fish fed diet C, it is unlikely lysine was responsible for
the poor performance of fish fed this diet. The remaining essen-
tial AA levels were similar among all diets except the branch
chain AAs (BCAA), which were higher in diet A. The BCAA
requirements for Cobia remain unknown; however, BCAA lev-
els in all three diets were in excess of requirements for other
fish species (NRC 2011). However, the higher intake of leucine
from diet A may be responsible for the induction of a signaling
pathway resulting in greater protein synthesis (Luo et al. 2010)
and higher PPV in fish fed that diet.
High dietary energy can reduce FI. Typical Cobia feeds con-
tain 15–16% CL (Liao et al. 2004), similar to the levels in diets
B and C but lower than in diet A. The response of Cobia to
dietary CL may be size dependent. Cobia (49 g) exhibited re-
duced growth and FI when fed 18% CL relative to 6% or 12%
CL, whereas differences were not observed in 7-g fish (Craig
et al. 2006). However, FI was not reported and it appears the
reported differences may have been an artifact of the statisti-
cal analysis and interpretation. In 41-g Cobia fed a 40% CP
diet, there was no benefit to greater than 6% dietary lipid (Chou
et al. 2001). However, the low CP in that study likely limited
the genetic growth potential and obscured evidence for protein
sparing. Cobia fed diets with 25% CL demonstrated lower FI
and growth, but FE was significantly better (10%), which may
have positive practical implications (Wang et al. 2005). How-
ever, the duration of the trial was too short (6 weeks) to discern
any benefit.
Less certain were the qualitative effects of the dietary lipid on
performance in the current study. Fish oil (FO) has traditionally
been the primary source of lipid in aquaculture feeds; however,
limits to increasing FO production have intensified efforts to
183
identify suitable alternatives (Glencross and Turchini 2011).
Notwithstanding, it is recognized that rapidly growing marine
fish have a critical requirement for DHA and EPA, which are not
found in terrestrial oil sources. Based on the FA compositions of
the three diets it is likely they contained some terrestrial-plant-
derived oils (Gunstone 2011).
As with AAs, the FA profiles between diets B and C were
similar. However, diet C had higher linoleic acid (18:2n-6),
DHA, and total polyunsaturated FAs, but lower EPA, than diet
B. Relative to diets B and C, A had a higher n-3:n-6 ratio
resulting from higher total n-3 and lower total n-6. Lower total
n-6 was a function of lower 18:2n-6, whereas higher n-3 was a
function of higher EPA and DHA. Although DHA, and to a lesser
degree EPA, are regarded as essential to marine fish, quantitative
requirements for these FAs in Cobia remain unknown. However,
there is some evidence Cobia have a dietary requirement for
DHA, but EPA may only be conditionally required (Trushenski
et al. 2012). Whether DHA in the diets used in the current study
were below those levels reported by Trushenski et al. (2012)
and this resulted in reduced growth is difficult to discern due to
differences in reporting methods.
Nevertheless, the higher DHA, EPA, and 16:1n-7 in diet A
suggests it contained a greater level of FO than the other diets.
No differences in production were observed in Cobia grown
from 60 to 160 g following dietary substitution of soy oil +
DHA for FO (Trushenski et al. 2011). However, the investiga-
tors suggested that diets with higher levels of FO were con-
sumed better by Cobia, which may account for the higher FI of
diet A in the current study. Ensuring palatability of formulated
diets is critical, particularly as new ingredients are employed
(Hertrampf and Piedad-Pascual 2000).
The cost of feed is an important component in the cost of
production of carnivorous fishes, and using an appropriately
formulated ration is critical in maximizing profit. In the cur-
rent trial, the cost of feed per kilogram of fish produced was
US$1.55, $1.43, and $2.30 for diets A, B, and C, respectively.
Higher quality feeds are often associated with higher feed costs,
which was not evident from the results of this study. It is clear
that there were differences in performance between the three
generic commercial marine fish diets tested in this study. This
underscores the need to ensure that diet formulations control
for quality. A quality feed should maximize FI and efficiency
and also be appropriate for the applied culture conditions. De-
spite mixed results with high-lipid diets fed to Cobia (Chou
et al. 2001; Wang et al. 2005; Craig et al. 2006; Weirich et al.
2010) the results of the current study clearly indicate juvenile
Cobia reared in production-scale RASs perform better when fed
high-lipid diets.
ACKNOWLEDGMENTS
Thanks are extended to Jay Adams, Meghan Anderson,
Daniel Benetti, Terri Breeden, Megan Davis, Don Freeman,
Bryan Garr, David Haley, Susan Laramore, Rolland Laramore,
 184 WILLS ET AL.
Todd Lenger, Gary Luisi, Tim Pfeiffer, Bruno Sardenburg,
Peter Stock, and Patrick Tracy for technical support. This study
was funded by the U.S. Department of Agriculture’s Agricul-
tural Research Service under Project No. 6225-63000-007-00D.
Mention of trade names or commercial products in this article is
solely for the purpose of providing specific information and does
not imply endorsement by the U.S. Department of Agriculture.
REFERENCES
AOAC (Association of Analytical Communities). 2003. Official methods of
analysis of AOAC international, 17th edition, 2nd revision. AOAC Interna-
tional, Gaithersburg, Maryland.
Arnold, C. R., J. B. Kaiser, and G. J. Holt. 2002. Spawning of Cobia Rachy-
centron canadum in captivity. Journal of the World Aquaculture Society
33:205–208.
Benetti, D. D., B. Sardenberg, A. Welch, R. Hoenig, M. R. Orhun, and I.
Zink. 2008. Intensive larval husbandry and fingerling production of Cobia
Downloaded by [Department Of Fisheries] at 19:52 28 February 2013
Rachycentron canadum. Aquaculture 281:22–27.
Chou, R. L., B. Y. Her, M. S. Su, G. Hwang, Y. H. Wu, and H. Y. Chen.
2004. Substituting fish meal with soybean meal in diets of juvenile Cobia
Rachycentron canadum. Aquaculture 229:325–333.
Chou, R. L., M. S. Su, and H. Y. Chen. 2001. Optimal dietary protein and lipid
levels for juvenile Cobia (Rachycentron canadum). Aquaculture 193:81–89.
Craig, S. R., M. H. Schwarz, and E. McLean. 2006. Juvenile Cobia (Rachycen-
tron canadum) can utilize a wide range of protein and lipid levels without
impacts on production characteristics. Aquaculture 261:384–391.
de Almeida Bicudo, A. J., L. F. Batista Pinto, and J. E. Possebon Cyrino.
2010. Clustering of ingredients with amino acid composition similar to the
nutritional requirement of Nile Tilapia. Scientia Agricola 67:517–523.
Dodd, Q. 2001. US Cobia culture meets early success. Hatchery International
(July/August):14–17.
Franks, J. S., J. T. Ogle, J. M. Lotz, L. C. Nicholson, D. N. Barnes, and K.
M. Larsen. 2001. Spontaneous spawning of Cobia, Rachycentron canadum,
induced by human chorionic gonadotropin (HCG), with comments on fer-
tilization, hatching, and larval development. Proceedings of the Gulf and
Caribbean Fisheries Institute 52:598–609.
Fraser, T. W. K., and S. J. Davies. 2009. Nutritional requirements of Co-
bia, Rachycentron canadum (Linnaeus): a review. Aquaculture Research
40:1219–1234.
Gatlin, D. M., III, F. T. Barrows, P. Brown, K. Dabrowski, T. G. Gaylord, R.
W. Hardy, E. Herman, G. Hu, Å. Krogdahl, R. Nelson, K. Overturf, M. Rust,
W. Sealey, D. Skonberg, E. J. Souza, D. Stone, R. Wilson, and E. Wurtele.
2007. Expanding the utilization of sustainable plant products in aquafeeds: a
review. Aquaculture Research 38:551–579.
Glencross, B. D., and G. M. Turchini. 2011. Fish oil replacement in starter,
grow-out, and finishing feeds for farmed aquatic animals. Pages 373–404 in
G. M. Turchini, W. K. Ng, and D. R. Tocher, editors. Fish oil replacement
and alternative lipid sources in aquaculture feeds. CRC Press, Boca Raton,
Florida.
Gunstone, F. D. 2011. The world’s oils and fats. Pages 61–98 in G. M. Turchini,
W. K. Ng, and D. R. Tocher, editors. Fish oil replacement and alternative lipid
sources in aquaculture feeds. CRC Press, Boca Raton, Florida.
Halver, J. E., and R. W. Hardy, editors. 2002. Fish nutrition, 3rd edition. Aca-
demic Press, San Diego, California.
Hertrampf, J. W., and F. Piedad-Pascual. 2000. Handbook on ingredients for
aquaculture feeds. Kluwer Academic Publishers, Dordrecht, The Netherlands.
Holt, G. J., C. K. Faulk, and M. H. Schwarz. 2007. A review of the larviculture
of Cobia Rachycentron canadum, a warm water marine fish. Aquaculture
268:181–187.
Kilduff, P., W. DuPaul, M. Oesterling, J. Olney Jr., and J. Tellock. 2002. Induced
tank spawning of Cobia, Rachycentron canadum, and early larval husbandry.
World Aquaculture Magazine 33(2):35–38.
Liao, I. C., T. S. Huang, W. S. Tsai, C. M. Hsueh, S. L. Chang, and E. M. Leaño.
2004. Cobia culture in Taiwan: current status and problems. Aquaculture
237:155–165.
Liao, I. C., H. M. Su, and E. Y. Chang. 2001. Techniques in finfish larviculture
in Taiwan. Aquaculture 200:1–31.
Lunger, A. N., S. R. Craig, and E. McLean. 2006. Replacement of fish meal in
Cobia (Rachycentron canadum) diets using an organically certified protein.
Aquaculture 257:393–399.
Lunger, A. N., E. McLean, and S. R. Craig. 2007a. The effects of organic protein
supplementation upon growth, feed conversion and texture quality parameters
of juvenile Cobia (Rachycentron canadum). Aquaculture 264:342–352.
Lunger, A. N., E. McLean, T. G. Gaylord, D. Kuhn, and S. R. Craig. 2007b.
Taurine supplementation to alternative dietary proteins used in fish meal
replacement enhances growth of juvenile Cobia (Rachycentron canadum).
Aquaculture 271:401–410.
Luo, J., D. Chen, and B. Yu. 2010. Effects of different dietary protein sources
on expression of genes related to protein metabolism in growing rats. British
Journal of Nutrition 104:1421–1428.
Mach, D. T. N., and R. Nortvedt. 2011. Free amino acid distribution in plasma
and liver of juvenile Cobia (Rachycentron canadum) fed increased levels of
lizardfish silage. Aquaculture Nutrition 17:e644–e656.
McClane, A. J. 1974. McClane’s new standard fishing encyclopedia and inter-
national angling guide. Holt, Rinehart and Winston, New York.
NRC (National Research Council). 2011. Nutrient requirements of fish and
shrimp. National Academies Press, Washington, D.C.
Oesterling, M. 2001. Cultured Cobia satisfy tastebuds. Virginia Marine Resource
Bulletin 33(2):23.
Rickards, W. L. 2001. Sustainable Cobia culture and fisheries. University of
Virginia, Virginia Sea Grant College Program, Publication VSG-01-07, Char-
lottesville.
Saadiah, I., A. M. Abol-Munafi, and C. M. Che Utama. 2011. Replacement of
fishmeal in Cobia (Rachycentron canadum) diets using poultry by-product
meal. Aquaculture International 19:637–648.
Schock, T. B., S. Newton, K. Brenkert, J. Leffler, and D. W. Bearden. 2012. An
NMR-based metabolomic assessment of cultured Cobia health in response to
dietary manipulation. Food Chemistry 133:90–101.
Schwarz, M., S. R. Craig, E. McLean, and D. Mowry. 2004. Status of Cobia
research and production. Pages 115–116 in G. J. Flick, T. Rakestraw, and
L. Douglas, editors. Proceedings of the 5th international conference on re-
circulating aquaculture. Virginia Polytechnic Institute and State University,
Roanoke.
Shiau, C. Y. 2007. Biochemical composition and utilization of cultured Cobia
(Rachycentron canadum). Pages 147–156 in I. C. Liao and E. M. Leaño, ed-
itors. Cobia aquaculture: research, development and commercial production.
Asian Fisheries Society, Manila.
Smith, M. M. 1986. Rachycentridae. Pages 661–662 in M. M. Smith and P. C.
Heemstra, editors. Smith’s sea fishes. Springer-Verlag, Berlin.
Timmons, M. B., and J. M. Ebeling. 2007. Recirculating aquaculture. Northeast
Regional Aquaculture Center, Publication 10-007, Cayuga Aqua Ventures,
Ithaca, New York.
Trushenski, J., M. Schwarz, A. Bergman, A. Rombenso, and B. Delbos. 2012.
DHA is essential, EPA appears largely expendable, in meeting the n-3 long-
chain polyunsaturated fatty acid requirements of juvenile Cobia Rachycentron
canadum. Aquaculture 326–329:81–89.
Trushenski, J., M. Schwarz, H. Lewis, J. Laporte, B. Delbos, R. Takeuchi, and
L. A. Sampaio. 2011. Effect of replacing dietary fish oil with soybean oil on
production performance and fillet lipid and fatty acid composition of juvenile
Cobia Rachycentron canadum. Aquaculture Nutrition 17:e437–e447.
Wang, J. T., Y. J. Liu, L. X. Tian, K. S. Mai, Z. Y. Du, Y. Wang, and H. J. Yang.
2005. Effect of dietary lipid level on growth performance, lipid deposition,
hepatic lipogenesis in juvenile Cobia (Rachycentron canadum). Aquaculture
249:439–447.
Weirich, C. R., T. I. J. Smith, M. R. Denson, A. D. Stokes, and W. E. Jenkins.
2004. Pond culture of larval and juvenile Cobia, Rachycentron canadum,
 COMMERCIAL DIETS FOR PRODUCTION OF JUVENILE COBIA
in the southeastern United States: initial observations. Journal of Applied
Aquaculture 16(1–2):27–44.
Weirich, C. R., A. D. Stokes, T. I. J. Smith, W. E. Jenkins, and M. R. Denson.
2006. Outdoor tank and pond pawning of Cobia, Rachycentron canadum, in
coastal South Carolina. Journal of Applied Aquaculture 18(3):1–16.
Weirich, C. R., P. S. Wills, R. M. Baptiste, and M. A. Riche. 2010. Production
characteristics and body composition of juvenile Cobia fed three different
commercial diets in recirculating aquaculture systems. North American Jour-
nal of Aquaculture 72:43–49.
Zhou, Q. C., K. S. Mai, B. P. Tan, and Y. J. Liu. 2005. Partial replacement of fish-
meal by soybean meal in diets for juvenile Cobia (Rachycentron canadum).
Aquaculture Nutrition 11:175–182.
Downloaded by [Department Of Fisheries] at 19:52 28 February 2013
185
Zhou, Q. C., B. P. Tan, K. S. Mai, and Y. J. Liu. 2004. Apparent digestibil-
ity of selected feed ingredients for juvenile Cobia Rachycentron canadum.
Aquaculture 241:441–451.
Zhou, Q. C., Z. H. Wu, S. Y. Chi, and Q. H. Yang. 2007. Dietary lysine re-
quirement of juvenile Cobia (Rachycentron canadum). Aquaculture 273:
634–640.
Zhou, Q. C., Z. H. Wu, B. P. Tan, S. Y. Chi, and Q. H. Yang. 2006. Optimal
dietary methionine requirement for juvenile Cobia (Rachycentron canadum).
Aquaculture 258:551–557.
Zhou, Q. C., J. Zhao, P. Li, H. L. Wang, and L. G. Wang. 2011. Evaluation of
poultry by-product meal in commercial diets for juvenile Cobia (Rachycen-
tron canadum). Aquaculture 322–323:122–127.
This article was downloaded by: [Department Of Fisheries]
On: 28 February 2013, At: 19:53
Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,
37-41 Mortimer Street, London W1T 3JH, UK

North American Journal of Aquaculture

Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/unaj20

Aquaculture Methods for a Genetically Managed

Population of Endangered Delta Smelt
a a a a
Joan C. Lindberg , Galen Tigan , Luke Ellison , Theresa Rettinghouse , Meredith M.
a b
Nagel & Kathleen M. Fisch
a
Fish Conservation and Culture Laboratory, Biological and Agricultural Engineering
Department, University of California–Davis, 1 Shields Avenue, Davis, California, 95616, USA
b
Genomic Variation Laboratory, Department of Animal Science, University of
California–Davis, 1 Shields Avenue, Davis, California, 95616, USA
Version of record first published: 21 Feb 2013.

To cite this article: Joan C. Lindberg , Galen Tigan , Luke Ellison , Theresa Rettinghouse , Meredith M. Nagel & Kathleen M.
Fisch (2013): Aquaculture Methods for a Genetically Managed Population of Endangered Delta Smelt, North American Journal
of Aquaculture, 75:2, 186-196

To link to this article: http://dx.doi.org/10.1080/15222055.2012.751942

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions

This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to
anyone is expressly forbidden.

The publisher does not give any warranty express or implied or make any representation that the contents
will be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses should
be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims,
proceedings, demand, or costs or damages whatsoever or howsoever caused arising directly or indirectly in
connection with or arising out of the use of this material.
 North American Journal of Aquaculture 75:186–196, 2013

C American Fisheries Society 2013

ISSN: 1522-2055 print / 1548-8454 online

DOI: 10.1080/15222055.2012.751942

ARTICLE

Aquaculture Methods for a Genetically Managed Population

of Endangered Delta Smelt
Joan C. Lindberg,* Galen Tigan, Luke Ellison, Theresa Rettinghouse,
and Meredith M. Nagel
Fish Conservation and Culture Laboratory, Biological and Agricultural Engineering Department,
University of California–Davis, 1 Shields Avenue, Davis, California 95616, USA

Kathleen M. Fisch
Genomic Variation Laboratory, Department of Animal Science, University of California–Davis,
Downloaded by [Department Of Fisheries] at 19:53 28 February 2013

1 Shields Avenue, Davis, California 95616, USA

Abstract
In response to Federal listing of the Delta Smelt Hypomesus transpacificus as a threatened species in 1993, intensive
fish culture techniques were developed to provide a supply of fish for research activities. The Delta Smelt was listed
as endangered by the state of California in 2009, and several agencies worked quickly to develop a captive refuge
population under genetic management. Captive 2-year-old wild-origin Delta Smelt served as the founding population
in 2008. Each year, 250 genetically selected, single pair crosses are made in vitro, and the resultant full-sibling families
are combined to rear in multifamily groups. Typically, eight families are reared together from egg to adult stage, with
80% or more of the initial families represented at the adult stage. Multifamily rearing provides an efficient way of
achieving a breeding population of 500 in a smaller facility. Juvenile survival increased from 18% in 2009 to 39%
in 2010, as facilities and methodologies improved. Growth rate also increased significantly from 2009 to 2010 (from
0.19 to 0.25 mm/d). Subdermal alphanumeric tags identified individuals and allowed spawning of select individuals
to preserve genetic diversity in the refuge population. Group marking, by adipose fin clip, provided efficiencies in
time and space. Tagging and genetic analyses enabled in vitro spawning of recommended pair crosses each year. At
present, we recommend completing the majority of spawning from February to mid-May and continuing to augment
the refuge population with wild fish each year. The refuge population provides one type of safeguard against species
extinction and provides an example for endangered fish culture.

Managers of species at risk of extinction are often con-
fronted with little time and few options for recovery (Millen-
nium Ecosystem Assessment 2005; Jackson 2008). Although
bringing fish into aquaculture settings is not ideal, for many
fish species propagation may play an important role in prevent-
ing extinction (Nickum et al. 2004). In the current study, the
cooperative efforts of several agencies resulted in the develop-
ment of a safe place in captivity, a refuge for the endangered
smelt, and by extension the population is termed a refuge pop-
ulation. Use of the term conservation hatchery does not quite
suit the program, as it often implies restocking to the wild. A
captive population implies capture of wild animals and rearing
them under artificial circumstances, whereas a refuge popu-
lation includes preservation of the evolutionary potential of a
species for many generations – in the captive refuge setting.
The refuge population provides one level of protection against
species extinction, allowing more time for habitat restoration
and improved management. The refuge population of Delta
Smelt Hypomesus transpacificus was founded in 2007–2008,
and fish culture efforts have developed rapidly in conjunction
with genetic management (Fisch et al. 2013) of the refuge pop-
ulation.

*Corresponding author: lindberg@steeper.us

Received March 16, 2012; accepted November 18, 2012

186
 DELTA SMELT AQUACULTURE
Delta Smelt are small, silvery fish endemic to the upper San
Francisco Estuary (SF Estuary) in northern California, USA.
Predominantly an annual fish, they spend the majority of their
life cycle in low saline water of the upper SF Estuary and Suisun
Bay (Moyle 2002). They have gained notoriety over the past
decade, as their principal habitat is caught in a battle between
protecting natural aquatic resources and providing Californians
with ample water (Bennett 2005; Sommer et al. 2007; Moyle
2008). Estimates of population abundance declined in 1982
and remained low (Sweetnam 1999; Bennett 2005; Newman
2008). Delta Smelt were federally listed as threatened in 1993
(U.S. Office of the Federal Register 1993) and as endangered
under the California Endangered Species Act in 2009 (CFGC
2009). Three additional pelagic fish in the SF Estuary (Striped
Bass Morone saxatilis, Threadfin Shad Dorosoma petenense,
and Longfin Smelt Spirinchus thaleichthys) also show signs
Downloaded by [Department Of Fisheries] at 19:53 28 February 2013
of population decline since 2002, suggesting the SF Estuary
problems are widespread (Feyrer et al. 2007; Sommer et al.
2007). Many of the possible causes are anthropogenic in
origin (Baxter et al. 2008; Moyle 2008), and returning habitat
complexity and the seasonal and interannual variability in salt
and freshwater flow conditions to the system may aid native
species recovery (Lund et al. 2010; Moyle et al. 2010).
As the risk of extinction for the species increased (Moyle
2002, 2008; Bennett 2005), actions were taken to bring a por-
tion of the wild population into captivity, a refuge, for conserva-
tion management of the Delta Smelt. A breeding program was
initiated to genetically manage and monitor the refuge popu-
lation in collaboration with the Genomic Variation Laboratory,
University of California–Davis (UC Davis; Fisch et al. 2009b).
Fish culture techniques had been previously developed for Delta
Smelt, over the past decade, by the Fish Conservation and Cul-
ture Laboratory (FCCL), UC Davis, to provide a reliable supply
of fish for about 15 research programs annually. The previous
culture methods relied on capturing immature wild stock each
fall to provide the first filial generation (F1 generation; all life
stages) of fish for research the following year. This method of
culture continued until 2007. In 2007, the State of California
Department of Fish and Game further restricted collection of
wild Delta Smelt due to mounting concern over the low species
abundance indices (Sommer et al. 2007). The FCCL proposed
to state and federal agencies that the captive population of adult
wild fish, spawned in 2007, be reared for another year to serve
as the founding population (F0 ) of a new refuge population in
2008. With their support, FCCL facilities were expanded and
new procedures were developed to establish and maintain a
refuge population.
The refuge population was initiated in 2008 with wild-caught
2-year-old Delta Smelt (held in captivity for 1 year). The 328
2-year-old Delta Smelt (birth year 2006, of natural origin) pro-
duced 164 full-sibling families to found the refuge population.
In each subsequent year, we have selected about 450 1-year-old
fish from the refuge population and about 50 wild fish to make
a total of 250 single pair crosses (mating of a single male and
187
female). Using the microsatellite genotypes of individuals, a
pedigree is reconstructed and pairwise kinship values are calcu-
lated (Ballou and Lacy 1995; but see Fisch et al. 2013). Breeding
pairs are then selected with the aim of minimizing average co-
ancestry and inbreeding, and to maintain equal representation
of founder alleles in the refuge population (Fisch et al. 2013).
Delta Smelt are reared well in excess of the annual target pop-
ulation of 500 breeding individuals, starting with over 200,000
eggs, in order to ensure an adequate pool of adult fish (>6,000)
from which to select the broodfish for the refuge population
and to continue to provide fish for research each year. Using
these methods to determine preferred pair crosses, the refuge
population has progressed to spawn the F3 generation in 2011.
The captive refuge population could serve as source material to
replenish a depleted natural population, if necessary; however,
there are no current plans to supplement the wild population
with cultured Delta Smelt.
The overall goal of the Delta Smelt refuge population pro-
gram is to create a captive population that maintains genetic
diversity and is representative of the wild population in suc-
cessive generations. Three primary entities helped initiate the
program: (1) the FCCL of UC Davis, with proven delta smelt
culture techniques and facilities; (2) the Genomic Variation Lab
of UC Davis, which developed the microsatellite markers (Fisch
et al. 2009a) and provides the genetic management component
of the breeding program annually (Fisch et al. 2009b, 2010,
2012, 2013); and (3) the U.S. Fish and Wildlife Service, which
supported the initial genetics work and maintains a smaller
population of Delta Smelt at Livingston Stone National Fish
Hatchery (LSNFH), Shasta Lake, California, to protect against
catastrophic losses at either facility.
There are two main components to developing a successful
Delta Smelt refuge population: appropriate fish husbandry tech-
niques to support a genetic breeding program and to rear all life
stages of the fish, as discussed in this paper, and the genetic man-
agement and monitoring of the population which was described
in brief above and in detail elsewhere (Fisch et al. 2013). The
fish husbandry techniques are described in three main sections:
(1) facility description and rearing, (2) founding population and
progeny, F0 –F1 subadult stage; and (3) new aquaculture tech-
niques in support of the genetically managed population, F1 –F3.
The successful development of a refuge population for Delta
Smelt may serve as an example for culture of other endangered
fishes.
METHODS
Facilities description and general rearing techniques.—The
capacity of the FCCL research facility has increased over the
years to about 20,000 adult Delta Smelt. Initially, the refuge
population was reared solely in the research facility until the
refuge facility became operational. Most of the fish culture tech-
niques developed for Delta Smelt are described in report form
 188 LINDBERG ET AL.
(Baskerville-Bridges et al. 2005); general culture methods are
described here, but in less detail.
Research facilities.—The water source for the research
facility is raw surface water derived from a man-made reservoir,
Clifton Court Forebay, in Contra Costa County, California,
and the FCCL is adjacent to the forebay. Water is pumped to
three settling tanks (715 × 238 × 75 cm deep; 12,760 L) for
removal of larger particles and is then passed through a drum
filter (50-µm mesh; PR Aqua Nanaimo, British Columbia; ca.
340 L/min) to remove smaller particles (>50 µm). The water
is then treated with ozone (65.1 g/h output unit; Pacific Ozone
Technology, Benicia, California) and foam-fractionated before
distribution to fish-rearing systems.
Most of the fish-rearing systems are recirculating and biofil-
tered. Both recirculating and flow-through systems are tempera-
ture controlled. Water is circulated by 0.5–1.0-hp pumps through
Downloaded by [Department Of Fisheries] at 19:53 28 February 2013
biofilters, UV filters, and a particle filter in recirculating systems.
Approximately 10% of system capacity is renewed daily through
tank cleaning and water flushing. Water is aerated by central air
blowers, or airstones. The research facility has two independent
larval-fish rearing systems of 10 tanks (130 L, 68-cm diameter
black polyethylene) and a capacity of 1,770 L each. The late-
larval stage fish are reared in a recirculating system of 20 tanks
(400 L, 68-cm diameter black polyethylene) and a total capacity
of 10,460 L. The juvenile-stage through the adult morphology
fish are reared in recirculating or flow-through systems with
larger tanks (1,100 L; 152-cm diameter; black-interior insulated
fiberglass). An indoor recirculating system of 12 tanks (system
also includes three larger tanks [1,930 L; 183-cm diameter] used
to rear research juveniles, for a total system capacity of 19,570 L)
is used to rear the younger juvenile stages (Table 1). Light con-
ditions in the indoor facility where generally late-larval to juve-
nile fish are reared are approximately 1–2 µmol/m2/s. The older
juvenile to adult stages are reared in an outdoor flow-through
system of 13 tanks with temperature control (Table 1), where the
TABLE 1. Delta Smelt life stages for use in aquaculture rearing and transitions between systems. Categories are general guidelines. Length measurements are
TL for larval stage and FL for all other stages. Delta Smelt life stages are defined in more detail in Mager et al. (2004).
Days Average Tank
posthatch length volume
Life stage (dph) (mm) (L)
Larval 0–40 5–17 130
Late larval 41–80 18–23 400
Subjuvenile to 81–199 24–49 1,100
juvenile
Subadult 200–249 50–54 1,100
Adult >250 >55 1,100
ozonated source water circulates between one or two 10-hp heat
pumps and a 3,820-L storage tank before passing to the rearing
systems. Outdoor tanks have shade-cloth (mesh-fabric) covers.
Refuge facilities.—The new refuge facility was mod-
eled after the FCCL research facility; the main building is
12.2 × 18.3 m. Water (750–950 L/min) from the reservoir
(same as for research facility) is sand-filtered (Model SM48-2,
80 PSI max; Everfilt, Mira Loma, California) and is UV treated
to supply the egg incubation system, live-prey culture units,
and fish-rearing units. For 1 year, 2009, larvae and late larvae
were reared in a recirculating system containing both larval and
late-larval tanks with the same lighting conditions. However,
because the fish appear to have life stage-dependent sensitivities
to light, the life stages were separated and light levels were
adjusted in 2010 to accommodate the larval and late-larval
stages (4–5 and 1–2 µmol/m2/s, respectively; Table 2). Particle
filters (Aquadyne Hartwell, Georgia) were added to each system
to remove excess algae and other particulates. The adult rearing
systems consist of two recirculating systems of 10 adult tanks
per system (as described for the research facility), and these
systems are under an awning adjacent to the main building.
An effort is made to maintain similar light conditions in the
juvenile- and adult-rearing system at both refuge and research
facilities. The light conditions in the adult-rearing system are
approximately 1–3 µmol/m2/s. System capacities are similar to
those of the research facility, and currently the refuge population
is distributed about equally between the two facilities.
Spawning and egg incubation.—Cultured Delta Smelt may
spawn from December to August but more typically from Jan-
uary to June in a temperature-controlled environment. The
FCCL has opted to manually express gametes and fertilize in
vitro rather than allowing mature adults to spawn in their holding
tanks. This is performed because the demersal adhesive eggs are
difficult to collect from the tanks and to separate from food and
feces, and egg release is inhibited in tanks. Manual expression
Rearing system
Recirculating rearing system, black interior tanks
Recirculating rearing system, black interior tanks
Recirculating or flow-through rearing system, indoor black
interior tanks, 80–120 dph; recirculating or flow-through
rearing system, outdoor black interior tanks, with tank covers,
awning and peripheral shade cloth 121–199 dph
Recirculating or flow-through rearing system, outdoor black
interior tanks, clear water conditions, shade cloth on tanks;
preferably under awning
Recirculating or flow-through rearing system, outdoor black
interior tanks, or under awning
 DELTA SMELT AQUACULTURE 189

TABLE 2. Initiation of Delta Smelt refuge population F0 –F3 ; description of major changes in fish husbandry over the first 3 years is provided.

Rearing location of life stage

Research Major differences between
Generation Birth year facility Refuge facility year-classes Result Data
F0 2006 Wild adults Wild 2-year-old fish, random High fecundity Figure 2
mating
F1 2008 Larvae to Late subadult Reared primarily in research Lower fecundity in Figure 2
subadult and adult facility, cultured 1-year-old, 1-year-olds
managed mating
F2 2009 All life stages All life Larval and late larval stages Low juvenile survival Figure 5
stages reared in one system at refuge
with higher light; rotifers
grown with Nannochloropsis
intiated weaning to prepared
diet early
Downloaded by [Department Of Fisheries] at 19:53 28 February 2013

F3 2010 All life stages All life Larval and late larval separated Increased juvenile Figures 5, 6
stages at refuge, with low light growth and survival
levels for late larval stage;
rotifers grown with enriched
supplement; weaning to
prepared diet delayed

of eggs results in higher quality and number of eggs, and allows
for select pair crosses to be made. A single clutch of eggs (fish
can produce several egg clutches per season) is fertilized in a
290–500-mL plastic bowl; larger bowls are used for 2-year-old
fish. Water is added to activate sperm, and eggs disperse and ad-
here to the bottom by means of an adhesive stalk (Mager et al.
2004). Water is replaced, and the bowls of developing embryos
are floated in water baths at 16◦ C for 3 d, which allows staff
to monitor and remove dead eggs. After 3 d, eggs are gently
freed with fingertips and rinsed with a clay mixture (Bentonite,
16 g/L; Sigma-Aldrich, St. Louis, Missouri) to minimize cohe-
sion. The total number of fertile eggs from each pair cross is
volumetrically estimated and recorded along with weights and
lengths of parents.
A column-style incubator consists of a vertical clear plastic
tube (5 × 42-cm-long Plexiglas) with a 250-µm mesh screen
in the bottom to hold a 200-mL mix of coarse (number 7) and
fine sand (number 60). At the top of the incubator, a 1.3-cm
diameter clear tube extends down to a 9.5- or 19.0-L black bucket
with screened standpipe. The incubators receive a recirculating,
upwelling supply of filtered water that creates a fluidized sand
bed in the columns to keep the eggs moving just above the
surface of the sand. At hatch, the larvae swim up, aided by the
upwelling water current, and out of the incubator into the bucket.
Rearing, feeding, and fish tank transfers.—From newly
hatched larval to adult stage, fish are transferred five or more
times between fish-rearing systems to accommodate life stages
and breeding program (Table 1). The five development stages
(for details see Mager et al. 2004) important to Delta Smelt
culture at FCCL are: (1) larval stage (0–40 days posthatch
[dph]; 5–17-mm TL; small, transparent, and elongate larvae);
(2) late-larval stage (41–80 dph; 18–23-mm FL; elongate form,
swim bladder development); (3) juvenile stage (81–200 dph;
25–50-mm FL; metamorphosing into adult fusiform morphol-
ogy and coloration, and increasing in size); (4) immature stage
“subadults” (200–249 dph; 50–54-mm FL; fish have gained
weight and are heartier and less sensitive to sunlight); and (5)
adults (>250 dph; >55-mm FL; fish begin to develop mature
gametes at about 55 mm, and cultured fish can reach 90 mm in
the first year). The first three transfers are made water to water
to reduce stress.
As Delta Smelt grow, they transition from live prey to a com-
mercial diet. Live prey cultures include brackish water rotifers
Brachionus plicatus (Reed Mariculture, Campbell, California)
and brine shrimp nauplii Artemia franciscana (dry cysts avail-
able from Artemia International, Fairview, Texas). Live prey are
fed to fish 6 times/d at a target density of 10 rotifers/mL from 3
to 40 dph. Artemia are fed at a target density of 1–3 nauplii/mL
from 10 to 120 dph until weaning to a commercial feed. An
algal concentrate Nannochloropsis (Reed Mariculture) is added
to the larval- and late larval-stage rearing systems to increase the
turbidity of the water to 9 nephelometric turbidity units (NTU)
and promote feeding (Baskerville-Bridges et al. 2004). After
fish transition to the outdoor tanks, algae are used for several
weeks to help reduce stress by reducing visibility.
To wean fish from live prey, Artemia are supplemented
with a dry feed mixture at a 2:1 ration of Cyclop-eeze (Argent
Laboratories, Redmond, Washington) and Lancy 2/4 (INVE
 190 LINDBERG ET AL.
Aquaculture) 2–4 times/d at 120 dph. After the fish are weaned,
they are fed a 2:1 ration of 4/6 NRD (INVE Aquaculture)
and 370 Hikari (By-Rite Pet Supply, Hayward, California)
15 times/d at a 1–3% body weight via vibratory feeders. Juve-
nile fish were transferred to the larger outdoor tanks at about
120 dph and 30–35-mm FL to rear to maturity. To monitor and
assess growth, 10 fish per tank in 10–12 tanks were measured
biweekly through 140–160 dph. To assess survival, fish were
counted when transferred from the “late-larval” tanks at
1,500–3,000 fish per tank and ≥20 mm-FL to the “adult” tanks,
and subsequently when stocking density was adjusted. The total
population was 48,000–96,000 fish. Analysis of variance was
used to compare survival between years, and linear regression
analysis was used to compare growth in juvenile fish.
The founding population and progeny, F0 to F1 subadult
stage.—The founding population of wild Delta Smelt (resident
Downloaded by [Department Of Fisheries] at 19:53 28 February 2013
on the FCCL site 1 year) was randomly bred to represent as
many wild fish in the new refuge population as possible; 328
fish were mated. The 2-year-old broodfish had an average clutch
size of over 4,500 eggs, or three to four times as many eggs as
1-year-old fish, used in subsequent years. Egg contribution was
limited and equalized at 1,000 eggs per full-sibling group (FSG);
multifamily rearing groups (MFG) were made by combining
three to six full-sibling families per MFG in order to rear all
families in the limited space available.
To accommodate the 164 families, both the normal larval
tanks (130 L) and a system of smaller tanks (70 L, recirculating
system) were used; the 70-L tank system was used in this year
only. The F1 generation larval stocking density was adjusted
to 3,000 larvae per tank in the 70-L tanks, versus 5,000–6,000
larvae in the larger 130-L tanks (which were used in subsequent
years). Following the early larval-rearing phase, the fish were
transferred at 40–50 dph to the 400-L tank system, usually with
1,500–3,000 fish per tank, to rear to 80 dph, for the late-larval
stage. These MFGs were then transferred into the larger indoor
adult tanks. Subadults were combined into 18 of the 1,100-L
outdoor adult tanks containing 2–19 full-sibling families per
tank in preparation for spawning.
New aquaculture techniques in support of the geneti-
cally managed population, F1 –F3 : subadult transfer to off-site
location and consolidation using adipose fin clip.—Once Delta
Smelt reach the subadult to adult stage (50–60 mm), fish are
thinned to 250 fish per each of the 32 MFG in the late fall or
winter. A subpopulation of each MFG, 50 fish per multifamily
group, are transferred in oxygenated tanks, with 5 g/L of salt, to
LSNFH as a safeguard against catastrophic loss. In 2010, and
thereafter, two MFGs each with 200 fish were consolidated at the
FCCL by use of an adipose fin clip to mark one of the two groups.
Tagging, fin-clipping, spawning, and family recovery assess-
ment.—Broodfish required both an individual identification tag
and a fin clip to implement the genetically managed breeding
plan. The tagging and fin-clipping process was usually initiated
in January or February, prior to the spawning season, starting
with 20 fish per MFG. Fish were tagged, anesthetized (100 mg/L
tricaine methanesulfonate; Argent Laboratories), allowed to re-
cover, and combined to adult tanks (<400 per tank). Small
plastic visible implant alphanumeric tags (Northwest Marine
Technologies, Olympia, Washington) were inserted under the
skin near the dorsal fin (Figure 1), and two small samples of fin
were removed and stored in a 95% solution of ethanol for DNA
processing and archiving. Attempts were made to tag an equal
number of males and females over the season. As fish mature,
males and females are sorted to separate tanks.
Fish from each MFG are also subsampled for weight and
length comparisons prior to the spawning season in January and
again in April to monitor growth over the spawning season;
data are presented for 2010. Wild fish are processed separately
during spawning operations.
Managed breeding, family recovery, and wild fish incorpora-
tion.—During the spawning season (ca. February through May)
the female tank is sorted for ripe females twice a week, tag
numbers are recorded and sent to the geneticist, males are rec-
ommended (and recovered from the tank housing the tagged
males), and spawns are completed within hours. Gender is ob-
served as running milt in males and distended belly and egg
development observed at vent in females. In maturing females,
eggs can be extracted by mild pressure applied to the abdomen.
Assessment of family recovery of the tagged broodfish pool
begins after the initial tagging and fin-clipping operations are
complete, but tagging and assessment continue until all spawns
have been completed for the season. Recovery is defined as
successful parentage assignment of individuals from each full-
sibling family in the tagged pool of broodfish. At the end of
the spawning season, the number of families recovered in the
tagged pool of broodfish are tallied and the percent recovered
is calculated based on the number of families initiating each
generation. Full-family recovery includes the number of families
successfully spawned as a percentage of the initial families for
each generation.
In late fall of each year, subadult wild fish (usually greater
than 50 mm, some with initial development of gametes) are col-
lected from the lower Sacramento River, where they congregate
prior to migrating into fresher waters for the spring spawning
season (Moyle 2002). A target population of about 50 wild
subadult fish is currently collected (under permit) to become
part of the refuge population on an annual basis.
Rearing changes between 2009 and 2010.—Eggs were
combined with equal representation of full-sibling families
(FSG) by combining 750 eggs from each of eight adult pairs,
within 10 d of each other, to make an MFG. Each MFG had
6,000 eggs, and resulting larvae (anticipating 5–10% embryo
mortality) are incubated together, as described earlier.
In 2010, several feeding and rearing changes were imple-
mented to help promote growth and survival of the F3 gener-
ation (birth year [BY] 2009) based on observations made in
2009, and to compensate the higher larval stocking density of
6,000 versus 5,000 larvae per tank in previous years (prior to
the refuge population; Table 2). Changes include the following:
 DELTA SMELT AQUACULTURE
Downloaded by [Department Of Fisheries] at 19:53 28 February 2013
FIGURE 1. A cultured Delta Smelt with an alphanumeric tag inserted below the dorsal fin. [Figure available in color online.]
(1) rotifers cultured with RotiGrow Plus (omega fatty acid preen-
riched microalgal blend), considered an enrichment diet for the
rotifers and larvae instead of the microalgae Nannochlorop-
sis, used prior to 2010 (both from Reed Mariculture), and ro-
tifers fed at higher density in 2010 (17 rotifers/mL/larval tank
in 2010 versus 10/mL prior); (2) delayed weaning of fish to a
prepared diet mixture (Cyclop-eeze from Argent Laboratories,
and EPAC/NRD 4/6 from INVE Aquaculture, 1:2 mixture) from
70 to 120 dph, based on unpublished FCCL data of improved
performance; and (3) rearing late-larval fish under low light
(1.4 µmol/m2/s) by separating the larval and late-larval rear-
ing systems, observed to promote better swimming and feeding
behaviors from past experience. In addition, light levels of the
outdoor adult tanks were reduced at the refuge facility by dou-
bling the shade cloth covers and adding a shade cloth drape
to the overhead awning perimeter (1–3 µmol/m2/s) to help the
juvenile fish transition to the brighter outdoor environment.
RESULTS
The Founding Population and Offspring, F0 –F1
Subadult Stage
Random mating of the 2-year-old wild fish produced 164 F1
families. Mean fecundity of the F0 generation was 4,569 eggs
191
per clutch, and mean length of the female was 98.3-mm FL in
2008 (Figure 2).
New Aquaculture Techniques in Support
of the Genetically Managed Population, F1 –F3
Subadult transfer to off-site location.—Fish transfer to
LSNFH was successful, with less than 1% mortality.
Adipose fin clip and tagging results.—The new adipose fin
clip procedure resulted in labor and space efficiencies. Adipose
fins did not regenerate, and the mark was effective in distinguish-
ing between two groups of adults housed together throughout
the 5–6-month spawning period. Combining two MFGs in each
broodfish tank resulted in significant production efficiencies.
The retention of alpha tags for individual fish identification
was generally good (40–100%), but there was a marked effect
of fish size on tag retention (Figure 3). All broodfish grew over
the spawning season, but fish in the last five MFGs were still
significantly shorter in early April than the first five MFGs, av-
eraging 61.9- versus 74.3-mm FL (ANOVA: P < 0.0002; Figure
2A). The older and larger F2 broodfish (hatched earlier in the
previous season) exhibited better tag retention, 92% retaining
the tag in MFGs 1–20 versus 72% in MFGs 21 and higher (the
youngest fish of the season; Figure 3). Wild fish lost 59% of
their tags when receiving tags on January 14, 2010; however,
 192 LINDBERG ET AL.
Downloaded by [Department Of Fisheries] at 19:53 28 February 2013

FIGURE 2. Data are illustrated for 2-year-old founding population (2008, black circle) and 1-year-old F1 and F2 broodfish in 2009 (gray square) and 2010 (black
cross), respectively, in terms of (A) fork length and (B) date (December through June). Data are for the selected spawns of the tagged broodfish pool in the refuge
population and do not capture the full-season reproductive potential of the smelt in terms of timing and frequency of spawns or fecundity of females, as only one
egg clutch per female is depicted.

FIGURE 3. Variation in broodfish size, tag retention, and recovery of families
within each MFG for F2 delta smelt in 2010. (A) Length of fish (average FL
and SE; n = 20 fish per MFG) measured on January 26–27, 2010 (open circles)
and again on April 7–9, 2010 (filled circles); (B) tag retention, represented by
number of live fish with tag of total fish tagged (%; n = 20 tagged fish per
MFG 2 months after tagging [squares]); recovery of full-sibling families from
each MFG also shown (% recovered of eight inititial families stocked per tank
[diamonds]).
fish survival was high (89.5%), and fish retagged February 11,
2010, were observed to have good tag retention.
Managed breeding, family recovery, and wild fish incorpora-
tion.—The F1 generation (BY2008, spawn 2009) was initiated
with 164 families in 2008 (Table 3). At the subadult or adult
stage, most fish were combined with two or more multifamilies
per tank, resulting in 2–19 full-sibling families per tank in
18 tanks. More than 1,400 broodfish were individually tagged,
and parentage was analyzed based on DNA from fin clips. Of the
initial 164 full-sibling families, 153 (93%) were recovered in the
F1 adult population in 2009, and 145 of the recovered families
(88%) were spawned to create the F2 generation (Table 3). A
total of 508 fish were selected from the tagged F1 broodfish
pool and wild broodfish tanks (53 wild fish contributing) to
make 254 pair crosses designed to minimize mean kinship
(Fisch et al. 2013; Table 3). The F1 refuge population was
spawned mid-February through May in 2009 (Figure 1). The
mean fecundity of the 1-year-old F1 broodfish of 1,579 eggs per
clutch and average female length was 71.9-mm FL (Figure 2).
The F2 generation (BY2009, spawn 2010) was initiated with
254 families. One tank of larvae, eight families (MFG 27), was
lost to a technical problem. The F2 broodfish were consolidated
into 18 adult tanks at the refuge facility. The F2 generation
(BY2009, spawn 2010) broodfish pool consisted of more than
1,800 tagged fish; 219 of the initial 254 full-sibling families were
recovered (86%), and 206 of those families (81%) were success-
fully spawned. From the tagged F2 broodfish pool, 432 cultured
fish and 34 wild broodfish contributed to the refuge population
to make 233 total pair crosses (Table 3); average egg clutch size
was 1,471, and average female length was 72.0-mm FL. Fish
 DELTA SMELT AQUACULTURE 193

TABLE 3. Delta Smelt refuge population management at the UC Davis FCCL. Families recovered in broodfish life stage (750 eggs per FSG) after rearing in
MFGs (eight FSGs per MFG) are determined through parentage analysis (see Fisch et al. 2012).

Delta Smelt generation Founder F1 F2 F3

Birth year (spawn year) 2006 (2008) 2008 (2009) 2009 (2010) 2010 (2011)
Number of tagged broodfish from a population >1,400 >1,800 >1,700
of ca. 6,400
Number of FSGs initiating the generation 164 254 233
Number of FSGs recovered in tagged pool 153 (93%) 219 (86%) 197 (85%)
of broodfish (as % of initial FSGs)
Number of FSGs included in successful spawns 145 (88%) 206 (81%) 187 (80%)
(as % of initial FSGs)
Number of wild fish supplementing refuge 53 34 64
population
Number of select pair crosses made to initiate 164 254 233 256
the next generation
Downloaded by [Department Of Fisheries] at 19:53 28 February 2013

spawned early in the previous season (prior to May 4) were
larger and spawned earlier—by the end of April (188 of the 250
pair crosses, or 75%); the remaining 25% were smaller, hatched
late in the previous season, and required longer to mature and
spawn (Figure 4). F3 progeny of this later-spawning group had
lower survival than progeny of early spawners (Figure 4).
The F3 generation (BY2010, spawn 2011) broodfish pool
consisted of more than 1,700 tagged fish; 197 of the initial 233
full-sibling families were recovered (85%), and 187 of those
families (80%) were successfully spawned. From the tagged F3
FIGURE 4. Recovery of the F2 generation and survival of the F3 genera-
tion Delta Smelt juveniles as a function of MFG. Total number of F2 parents
spawned per MFG in 2010 (solid black line) declines with spawn date, which
coincides with increasing MFG number as spawning season progresses (from
mid-February to through mid-June). The F2 parents that hatched late in the
previous year, 2009 (those with high MFG numbers), tended to spawn late in
2010 season, e.g., not until after May 4 (dashed black line). Survival of F3
juveniles (at 80 dph; dashed line–triangle marker, secondary y-axis) reflects this
pattern, as F3 fish hatched later in the spawning season, 2010 (with higher MFG
numbers), had poor survival. The low survival of juvenile fish from MFGs 7–10
is attributed to a temporary disease problem in these groups. Lines are included
representing the best linear fit of the data.
broodfish pool, 448 cultured fish and 64 wild fish were selected
to make 256 pair crosses to produce the F4 generation.
Rearing changes between 2009 and 2010, comparison of
growth and survival of the F2 and F3 generations.—Juvenile
fish survival improved significantly from the F2 generation to the
F3 generation, averaging 18% and 36%, respectively (ANOVA:
P < 0.0001; Figure 5).
Growth rate was also significantly higher in the F3 generation
versus the F2 generation fish (0.248 versus 0.188 mm/d; 5.2-mm
intercept; P < 0.0001; Figure 6). Average water temperature
varied within 1.3◦ C across systems (15.4–16.7◦ C) and between
years. Temperature averaged 0.5◦ C higher for larval and late-
larval stages, but 0.8◦ C lower for juvenile to adult stages in 2010
versus 2009.
In 2009, larval and late-larval stages were reared in one sys-
tem with the same lighting conditions (4–5 µmol/m2/s) for both.
Under these conditions the larvae were observed to be swimming
and feeding actively in the upper water column, whereas indi-
viduals in the late-larval stage were lower in the water column
and appeared to be more stressed. In 2010 the rearing systems
were separated by life stage, reducing incident light levels to
25% of the 2009 levels in the late-larval fish-rearing system,
and said fish demonstrated more active and normal behavior at
the lower light level (1–2 µmol/m2/s).
Juvenile fish are also sensitive to light. Reducing the inci-
dent light to the outdoor tanks appears to have contributed to
decreased juvenile mortality from 970 in 2009 to 219 in 2010
during 3 d of transitioning to outdoor tanks.
DISCUSSION
Successful fish husbandry methods are described that support
a genetically managed breeding program for the Delta Smelt
refuge population, and no significant loss of genetic diversity
has been observed to date (Fisch et al. 2013). Wild fish are
 194 LINDBERG ET AL.
Downloaded by [Department Of Fisheries] at 19:53 28 February 2013

FIGURE 5. Survival of cultured juvenile Delta Smelt spawned in 2009 and 2010. Data represent average survival for all MFGs of juveniles spawned in month
indicated at transfer to final adult tanks (ca. 100–120 d posthatch; 2010 data: black line, square marker; 2009 data: dashed line, triangle marker; SE of the average
included). Survival was higher in 2010 than 2009 (P < 0.0001) and also for the 3 months March–May (P < 0.009).

supplemented, as founders, each year to help maintain genetic
diversity and minimize genetic drift.
Facility expansion and modifications in rearing techniques
contributed to improved fish rearing success over the 3-year
period of this study. Improved juvenile survival and growth
in the F3 generation (Figure 6) are thought to be attributable
to methodological changes, but several changes were made
simultaneously and so weighting importance of each change is
not possible. Differential lighting of life stage rearing systems
appeared to benefit the late-larval life stage as these fish tended
to be more active, feeding, and swimming higher in the water
column when the light levels were lowered. An increase in
feed quality and quantity, and a delay in weaning to a dry diet
FIGURE 6. Comparison of cultured Delta Smelt growth in length over the first
5 months of life for 2009 and 2010. Daily growth rate is 0.188 and 0.248 mm/d
for 2009 and 2010, respectively, and significantly higher in 2010 (P < 0.0001);
intercept is fixed at 5.20 mm based on average size at hatch (authors’ unpublished
data).
may also have contributed to the improved growth and survival
observed in 2010; age at weaning may be species specific but is
also influenced by diet quality and co-feeding of live and inert
diets (Person-Le Ruyet et al. 1993; Chen et al. 2006; Engrola
et al. 2009). In addition, Delta Smelt transitioning from the
late-larval to the juvenile stage appear to be sensitive to light and
crowding, so improved growth and survival may also be due to
transitioning the fish into larger tanks earlier and keeping the fish
in a darkened indoor environment longer, before moving them
to well-shaded outdoor tanks. Slightly warmer temperatures
(0.5◦ C higher mean temperature) for larval to late larval-stage
rearing may also have contributed to faster growth and higher
survival.
Adopting the practice of rearing the smelt in multifamily
groups (usually eight families per group) has permitted good
family retention while reducing labor and facility costs. Loss
of families has been less than 20% of the approximate 250
initial families each year. Retention of a high proportion of
the families to the adult stage is likely due, at least in part, to
equalizing family size at the egg stage (750 eggs per full-sibling
group). Equalizing family size has been recommended as a
strategy to maximize effective population size and reduce
domestication selection for traits more suitable in a nonnatural
environment in captive populations (Allendorf 1993; Frankham
2008). However, family representation is worse in those MFGs
spawned early or late in the season, as reflected in juvenile
survival (Figure 3). Modifying the MFG structure for the larval
rearing phase(s) in these early and late MFGs may improve
family retention for the most compromised groups. Potential
gains may be had by combining half the number of families
 DELTA SMELT AQUACULTURE
(i.e., four versus eight families), increasing representation of
each family in the MFG (1,000–1,200 eggs per MFG versus 750
eggs per MFG typically used), and reducing stocking density
from 6,000–4,000 larvae per tank. Smaller rearing groups can
later be combined to the standard eight families per MFG after
the period of high larval mortality has passed.
Removal of adipose fin clip and alphanumeric tagging tech-
niques were both useful in combining families and identifying
individuals, contributing to the efficiency and the ability to im-
plement the genetic breeding program. The best fish tagging
and fin clip sampling schedule for accomplishing both the fish
breeding and the genetic and pedigree analysis may be a mix
of tagging 600–700 fish in late January to early February prior
to the spawning season, and then supplementing the pool of
tagged fish by tagging more fish throughout the season as they
become mature. In 2009, fish were tagged and fin-clipped for
Downloaded by [Department Of Fisheries] at 19:53 28 February 2013
DNA and parentage analysis throughout the busy spawning sea-
son. In 2010 and 2011, a large representative group of fish was
tagged and fin-clipped in January. The latter method proved
useful in creating a larger pool of fish from which to select
breeding pairs, especially early in the season, and also reduced
the tagging effort during the busy spawning season. However,
the youngest broodfish require more time to reach an adequate
size for tagging (Figure 3). Additionally, care should be taken
to include in the pool of tagged fish those that mature in mid-
to-late season to help minimize imposed seasonal spawning
bias.
Spawning.—The wild population of Delta Smelt is thought
to spawn primarily from early April to mid-May (Moyle 2002),
but the spawning period varies year to year and probably extends
from February to June in some years (Wang 1986). Therefore,
an effort was made to spawn fish during the full spawning sea-
son in the refuge population. However, fish hatched late in the
spawning season, from mid-May to mid-June, appear to have
limited utility to the refuge population overall. These younger
fish are significantly smaller (P < 0.0002) and appear to spawn
later in the season (F2 adults, MFG ≥ 21; Figure 4), and their
progeny do not survive as well as those spawned midseason (F3
juveniles; Figure 4). Taken together, these factors contribute to
a cycle of diminishing family recovery for fish hatched late in
the season.
The captive founding population produced 164 families
through random gamete fertilization. In the F1 –F3 generations,
close to 250 families were produced each year with 80–88%
recovery of one or more individuals from each family (Table 3).
Documenting the family recovery is an important component
of a genetically managed population, especially over multiple
generations, with the intention to monitor and to minimize the
potential loss of families (and within-population diversity) in
each generation (Williamson 2001). As more fish species be-
come imperiled, developing managed breeding programs will
become a more common fish management tool. A well-managed
population constitutes a genetic bank and provides one level of
security against species extinction.
195
CONCLUSIONS
The refuge Delta Smelt population has been maintained
through the F3 generation as of 2011. The value of the refuge
population lies in the safeguard it provides against extinction and
in the additional time it allows for the improved management of
its natural habitat.
ACKNOWLEDGMENTS
We are deeply indebted to Bradd Baskerville-Bridges for his
enthusiastic fish culture efforts on behalf of the Delta Smelt. We
acknowledge all of the FCCL staff for their dedication. Devel-
opment of the Delta Smelt culture program and of the refuge
population program was made possible with the support of the
CALFED Bay–Delta Program, the California State Department
of Water Resources, the Interagency Ecological Program, and
the U.S. Bureau of Reclamation (contract R10AC20014). We
also acknowledge the Biological and Agricultural Engineering
Department, UC Davis, for its support of the project.
REFERENCES
Allendorf, F. W. 1993. Delay of adaptation to captive breeding by equalizing
family size. Conservation Biology 7:416–419.
Ballou, J. D., and R. C. Lacy. 1995. Identifying genetically important individuals
for management of genetic variation in pedigreed populations. Pages 76–111
in J. D. Ballou, M. Gilpin, and T. J. Foose, editors. Population management for
survival and recovery: analytical methods and strategies in small population
conservation. Columbia University Press, New York.
Baskerville-Bridges, B., J. C. Lindberg, and S. I. Doroshov. 2004. The effect of
light intensity, alga concentration, and prey density on the feeding behavior of
Delta Smelt larvae. Pages 219–227 in F. Feyrer, L. R. Brown, R. L. Brown, and
J. J. Orsi, editors. Early life history of fishes in the San Francisco Estuary and
watershed. American Fisheries Society, Symposium 39, Bethesda, Maryland.
Baskerville-Bridges, B., J. C. Lindberg, and S. I. Doroshov. 2005. Manual for
the intensive culture of Delta Smelt (Hypomesus transpacificus). University
of California–Davis, Report to CALFED Bay–Delta Program, ERP-02-P31,
Sacramento.
Baxter, R., R. Breuer, L. Brown, M. Chotkowski, F. Feyrer, M. Gingras, B.
Herbold, A. Mueller-Solger, M. Nobriga, T. Sommer, and K. Souza. 2008.
Pelagic organism decline progress report: 2007 synthesis of results. Inter-
agency Ecological Program for the San Francisco Estuary, Technical Re-
port 227, Sacramento, California. Available: www.science.calwater.ca.gov/
pdf/workshops/POD/2007 IEP-POD synthesis report 031408.pdf. (March
2012).
Bennett, W. A. 2005. Critical assessment of the Delta Smelt population in
the San Francisco Estuary, California. San Francisco Estuary and Watershed
Science [online serial] 3(2):article 1.
CFGC (California Fish and Game Commision). 2009. Uplisting the Delta
Smelt to endangered species status. Amend Title 14, CCR, Section 670.5 of
California Endangered Species Act.
Chen, B. N., J. G. Qin, M. S. Kumar, W. G. Hutchinson, and S. M. Clarke.
2006. Ontogenetic development of digestive enzymes in Yellowtail Kingfish
Seriola lalandi larvae. Aquaculture 260:264–271.
Engrola, S., L. Figueira, L. E. C. Conceição, P. J. Gavaia, L. Ribeiro, and M. T.
Dinis. 2009. Co-feeding in Senegalese sole larvae with inert diet from mouth
opening promotes growth at weaning. Aquaculture 288:264–272.
Feyrer, F., M. L. Nobriga, and T. R. Sommer. 2007. Multidecadal trends for three
declining fish species: habitat patterns and mechanisms in the San Francisco
Estuary, California, USA. Canadian Journal of Fisheries and Aquatic Sciences
64:723–734.
 196 LINDBERG ET AL.
Fisch, K. M., J. A. Ivy, R. S. Burton, and B. May. 2013. Evaluating the per-
formance of captive breeding techniques for conservation hatcheries: a case
study of the Delta Smelt captive breeding program. Journal of Heredity 104:
92–104.
Fisch, K. M., B. Mahardja, T. Rettinghouse, L. Ellison, G. Tigan, J. Lindberg,
and B. May. 2010. Captive breeding plan for the endangered Delta Smelt:
genetic management and fish rearing modifications for 2010. Interagency
Ecological Program Newsletter 23(3):13–20.
Fisch, K. M., B. Mahardja, T. Rettinghouse, L. Ellison, G. Tigan, J. Lindberg,
and B. May. 2012. Delta Smelt captive refugial population—2011 season
summary. Interagency Ecological Program Newsletter 25(1):9–10.
Fisch, K. M., J. L. Petersen, M. R. Baerwald, J. K. Pedroia, and B. May.
2009a. Characterization of 24 microsatellite loci in Delta Smelt, Hypomesus
transpacificus, and their cross-species amplification in two other smelt species
of the Osmeridae family. Molecular Ecology Resources 9:405–408.
Fisch, K. M., T. Rettinghouse, L. Ellison, G. Tigan, J. Lindberg, and
B May. 2009b. Delta Smelt refugial population development and ge-
netic management—2009 season summary. Interagency Ecological Program
Newsletter 22(3):3–9.
Downloaded by [Department Of Fisheries] at 19:53 28 February 2013
Frankham, R. 2008. Genetic adaptation to captivity in species conservation
programs. Molecular Ecology 17:325–333.
Jackson, J. B. C. 2008. Ecological extinction and evolution in the brave
new ocean. Proceedings of the National Academy of Sciences of the USA
105(Supplement 1):11458–11465.
Lund, J. R., E. Hanak, W. E. Fleenor, W. A. Bennett, R. E. Howitt, J. F. Mount,
and P. B. Moyle. 2010. Comparing futures for the Sacramento–San Joaquin
Delta. University of California Press, Berkeley.
Mager, R. C., S. I. Doroshov, J. P. Van Eenennaam, and R. L. Brown. 2004.
Early life stages of Delta Smelt. Pages 169–180 in F. Feyrer, L. R. Brown,
R. L. Brown, and J. J. Orsi, editors. Early life history of fishes in the San
Francisco Estuary and watershed. American Fisheries Society, Symposium
39, Bethesda, Maryland.
Millennium Ecosystem Assessment. 2005. Ecosystems and human well-being:
biodiversity synthesis. World Resources Institute, Washington, D.C.
Moyle, P. B. 2002. Inland fishes of California. University of California Press,
Berkeley.
Moyle, P. B. 2008. The future of fish in response to large-scale change in
the San Francisco Estuary, California. Pages 357–374 in K. D. McLaughlin,
editor. Mitigating impacts of natural hazards on fishery ecosystems. American
Fisheries Society, Symposium 64, Bethesda, Maryland.
Moyle, P. B., W. A. Bennett, W. E. Fleenor, and J. R. Lund. 2010. Habitat
variability and complexity in the upper San Francisco Estuary. San Francisco
Estuary and Watershed Science [online serial] 8(3):article 1.
Newman, K. B. 2008. Sample design-based methodology for estimating Delta
Smelt abundance. San Francisco Estuary and Watershed Science [online se-
rial] 6(3):article 3.
Nickum, M. J., P. M. Mazik, J. G. Nickum, and D. D. MacKinlay, editors.
2004. Propagated fish in resource management. American Fisheries Society,
Symposium 44, Bethesda, Maryland.
Person-Le Ruyet, J., J. C. Alexandre, L. Thébaud, and C. Mugnier. 1993. Ma-
rine fish larvae feeding: formulated diets or live prey? Journal of the World
Aquaculture Society 24:211–224.
Sommer, T., C. Armor, R. Baxter, R. Breuer, L. Brown, M. Chotkowski, S.
Culberson, F. Feyrer, M. Gingras, B. Herbold, W. Kimmerer, A. Mueller-
Solger, M. Nobriga, and K. Souza. 2007. The collapse of pelagic fishes in the
upper San Francisco Estuary. Fisheries 32:270–277.
Sweetnam, D. A. 1999. Status of Delta Smelt in the Sacramento–San Joaquin
Estuary. California Fish and Game 85(1):22–27.
U.S. Office of the Federal Register. 1993. Endangered and threatened wildlife
and plants: determination of threatened status for the Delta Smelt. Federal
Register 58:42(5 March 1993):12854–12864.
Wang, J. C. S. 1986. Fishes of the Sacramento–San Joaquin Estuary and adjacent
waters, California: a guide to the early life histories. U.S. Department of the
Interior, Bureau of Reclamation, Mid-Pacific Region, Interagency Ecological
Program Technical Report 9, Byron, California.
Williamson, J. H. 2001. Broodstock management for imperiled and other fishes.
Pages 397–482 in G. A. Wedemeyer, editor. Fish hatchery management, 2nd
edition. American Fisheries Society, Bethesda, Maryland.
This article was downloaded by: [Department Of Fisheries]
On: 28 February 2013, At: 19:53
Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,
37-41 Mortimer Street, London W1T 3JH, UK

North American Journal of Aquaculture

Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/unaj20

Comparison of Traditional and Fermented Soybean

Meals as Alternatives to Fish Meal in Hybrid Striped
Bass Feeds
a a a
Artur Rombenso , Curtis Crouse & Jesse Trushenski
a
Fisheries and Illinois Aquaculture Center and Department of Zoology, Southern Illinois
University Carbondale, 1125 Lincoln Drive, Carbondale, Illinois, 62901, USA
Version of record first published: 26 Feb 2013.

To cite this article: Artur Rombenso , Curtis Crouse & Jesse Trushenski (2013): Comparison of Traditional and Fermented
Soybean Meals as Alternatives to Fish Meal in Hybrid Striped Bass Feeds, North American Journal of Aquaculture, 75:2,
197-204

To link to this article: http://dx.doi.org/10.1080/15222055.2012.756440

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions

This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to
anyone is expressly forbidden.

The publisher does not give any warranty express or implied or make any representation that the contents
will be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses should
be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims,
proceedings, demand, or costs or damages whatsoever or howsoever caused arising directly or indirectly in
connection with or arising out of the use of this material.
 North American Journal of Aquaculture 75:197–204, 2013

C American Fisheries Society 2013

ISSN: 1522-2055 print / 1548-8454 online

DOI: 10.1080/15222055.2012.756440

ARTICLE

Comparison of Traditional and Fermented Soybean Meals

as Alternatives to Fish Meal in Hybrid Striped Bass Feeds
Artur Rombenso, Curtis Crouse, and Jesse Trushenski*
Fisheries and Illinois Aquaculture Center and Department of Zoology,
Southern Illinois University Carbondale, 1125 Lincoln Drive, Carbondale, Illinois 62901, USA
Downloaded by [Department Of Fisheries] at 19:53 28 February 2013

Abstract
Soybean meal is one of the most common fish meal alternatives used in aquafeeds because of its high protein density,
favorable amino acid profile, comparatively low price, and widespread availability. However, palatability issues and
antinutritional factors limit soybean meal inclusion, particularly in feeds for carnivorous fishes. Various processing
strategies, including fermentation, may offer some advantage in terms of reducing or eliminating antinutritional
factors while enhancing protein content, improving protein absorption rate, and feed intake. Accordingly, we assessed
production performance of juvenile hybrid Striped Bass (White Bass Morone chrysops × Striped Bass M. saxatilis)
fed diets containing 30% menhaden fish meal, or reduced or fish-meal-free feeds (0, 5, or 10% fish meal) containing
traditional soybean meal (47.5% crude protein) or PepSoyGen fermented soybean meal (52.0% crude protein) as
the principal sources of dietary protein. Each dietary treatment was fed to quadruplicate tanks of fish (10 fish/tank;
average individual weight = 18.1 ± 0.2 g [mean ± SE], n = 4) in a recirculation system for 8 weeks. Production
performance tended to decline among fish fed diets containing less than 10% fish meal; however, this effect was
less overt among those fed the fermented soybean meal-based feeds. At each level of fish meal replacement, fish
fed the fermented soybean meal treatment outgrew those fed the corresponding traditional soybean meal feed,
though differences at the 5% fish meal level were not statistically significant. Fermentation appears to increase the
acceptability and biological value of soybean meal in hybrid Striped Bass feeds, and PepSoyGen shows promise as a
fish meal alternative for aquafeeds for carnivorous fishes.

Fish meal (∼60–70% crude protein) has been widely used as
a protein source in aquafeeds due to its price and favorable amino
acid profile (Trushenski et al. 2006). As a result of increasing
demand for fish meal and harvest restrictions, which maintain
essentially fixed annual production (6 million metric tons/year,
FAO 2010), the price of fish meal has risen—quite dramatically
in some years—from a preceding 5-year average of US$412/ton
in 2000 to $1,536/ton in 2012 (FAO 2012). The rising cost of
fish meal is a major issue in aquaculture, and identification of
alternative protein sources is considered an ongoing priority for
the industry.
Numerous alternative protein sources have been explored in
aquafeeds (Gatlin et al. 2007), with a strong emphasis on plant
products because of their availability and competitive pricing.
Research and development in this area continue and have diver-
sified to include novel crops and coproducts as well as a range of
processing strategies. However, soybean meal (∼46–48% crude
protein) remains one of the most common fish meal alternatives
used in aquafeeds because of its high protein density, favorable
amino acid profile, comparatively low price, and widespread
availability (Trushenski et al. 2006; Gatlin et al. 2007). Feeds
based on soybean meal are well accepted by omnivorous fish
species such as Channel Catfish Ictalurus punctatus (Robinson
and Li 1994) and Nile Tilapia Oreochromis niloticus (El-Saidy
and Gaber 2002), and soybean meal can facilitate complete fish
meal replacement in feeds for these taxa. Fish-meal sparing
with soybean meal is less successful in feeds for carnivorous
fishes, and, depending on the extent of fish meal replacement,
implementation of soy-derived protein (as soybean meal and
in other forms) can impair the production performance (Brown

*Corresponding author: saluski@siu.edu

Received October 19, 2012; accepted December 1, 2012

197
 198 ROMBENSO ET AL.
et al. 1997; Refstie et al. 1998; Chou et al. 2004; Gaylord
et al. 2004; Takagi et al. 2008; Blaufuss and Trushenski 2012)
and physiological competence (Laporte and Trushenski 2012;
Trushenski et al. 2013) of both marine and freshwater taxa.
Accordingly, a considerable amount of research effort has
been focused on solving problems associated with palatability,
digestion, and utilization of soy-based feeds, including attempts
to neutralize or remove antinutritional factors (ANF) thought
to be the cause of these negative attributes (Watanabe and
Pngmaneerat 1993; Keembiyehtty and Gatlin 1997; Mambrini
et al. 1999; Zhou et al. 2005). Antinutritional factors found in
soy, including protease inhibitors, phytate, saponins, lectins,
and oligosaccharides, can reduce digestion of polypeptides,
reduce mineral bioavailability, and decrease feed intake and
digestibility, ultimately leading to production performance
losses (Bureau et al. 1998; Refstie et al. 1998; Robaina et al.
Downloaded by [Department Of Fisheries] at 19:53 28 February 2013
1998; Francis et al. 2001; Lin et al. 2006; Blaufuss and
Trushenski 2012). A variety of strategies have been used to
address the ANF content of soy, including processing strategies
such as extrusion, purification, defatting, heat-treating, and fer-
mentation (Hancock et al. 1989, 1990; Jones et al. 1989; Mital
and Garg 1990; Reddy and Pierson 1994; Burnham et al. 2000).
Soybean fermentation processes, which have increased di-
gestibility, calcium, and the levels of some vitamins in soy-based
foods, have been used for many years in East Asia for human
food preparation (Lee 1998; Kim et al. 1999). PepSoyGen fer-
mented soybean meal (∼52% crude protein) is a protein feed-
stuff manufactured by the solid-state fermentation of soybean
meal by dehydrated Aspergillus oryzae and Bacillus subtilis.
This product has been used successfully in young animal cul-
ture, such as turkey poults and weanling pigs (Jones et al. 2008;
Yoo et al. 2009), and appears to offer some advantage in terms of
reducing or eliminating ANF while enhancing protein content
(Mital and Garg 1990; Reddy and Pierson 1994) and improving
protein absorption rate in young animals due to its higher pro-
portion of small peptides (Hong et al. 2004; Min et al. 2009).
Because of its modified composition and reduced ANF con-
tent, PepSoyGen has proven more suitable than soybean meal
in feeds for sensitive life stages of terrestrial livestock. Although
this product would seem to have great potential in aquafeeds, it
has not been widely evaluated in feeds for carnivorous fishes.
Hybrid Striped Bass (White Bass Morone chrysops ×
Striped Bass M. saxatilis) exhibit many favorable characteristics
for aquaculture such as their tolerance of intensive culture
conditions in a variety of culture systems, relatively simple
husbandry, excellent growth performance, disease resistance,
and high market value as a live or fresh product (Kohler 2004).
These attributes helped to make hybrid Striped Bass one of the
fastest growing sectors of U.S. aquaculture, with production
expanding rapidly through the 1990s to reach approximately 12
million pounds per year in the mid-2000s (Turano 2012). Since
reaching this peak, the hybrid Striped Bass industry has slowly
contracted, and current U.S. production is approximately 8–9
million pounds per year. Although there are a number of
issues that have contributed to this shrinkage, feedback from
members of the Striped Bass Growers’ Association indicates
that rising feed cost, driven primarily by the price of fish meal,
is consistently the number one challenge facing the industry
(Turano 2012). Nutritionists have assessed a variety of fish meal
alternatives in hybrid Striped Bass feeds, including soy protein
(Gallagher 1994; Brown et al. 1997; Blaufuss and Trushenski
2012; Laporte and Trushenski 2012); however, full replace-
ment of fish meal with soybean meal typically results in the
performance impairments discussed above. Thus, the use of
PepSoyGen may be particularly advantageous in feeds for
hybrid Striped Bass. Accordingly, we assessed the production
performance of juvenile hybrid Striped Bass fed diets containing
menhaden fish meal as the primary protein source, or graded lev-
els of soybean meal or PepSoyGen as alternative protein sources.
METHODS
Feed preparation and analyses.—A practical feed, including
300 g/kg menhaden fish meal (FM; 64.2% crude protein;
Special Select, Omega Protein, Houston, Texas) and 40 g/kg
menhaden fish oil (Virginia Gold, Omega Protein) and similar to
formulations previously verified in hybrid Striped Bass culture
(Blaufuss and Trushenski 2012; Laporte and Trushenski 2012),
served as the positive control feed in the present work (30% FM;
Table 1). Experimental feeds were derived from the 30% FM
formulation using traditional soybean meal (SBM; 47.5% crude
protein; Siemer Enterprises, Teutopolis, Illinois) or PepSoyGen
fermented soybean meal (PSG; 52.0% crude protein; Nutra-Flo
Protein and Biotech Products, Sioux City, Iowa) as alternatives
to fish meal. The resultant formulations contained 0, 5, or 10%
fish meal with soybean meal or PepSoyGen replacing the bulk
of fish meal-derived protein (i.e., 0% FM SBM, 0% FM PSG,
5% FM SBM, 5% FM PSG, 10% FM SBM, and 10% FM PSG;
Table 1). All feeds, including the 30% FM and the PepSoyGen-
based feeds, contained at least 255 g/kg soybean meal. This level
of dietary soybean meal is common in hybrid Striped Bass feeds
and does not present any of the issues that have been associated
with higher soybean meal inclusion rates. Thus, to maintain a
degree of practicality and facilitate direct comparisons between
soybean meal and PepSoyGen in sparing fish meal, minimum
soybean meal inclusion rates were set at 255 g/kg. All feeds were
formulated to meet or exceed the known nutrient requirements
of hybrid Striped Bass (Gatlin 1997; NRC 2011). Particular at-
tention was paid to dietary methionine levels, as this amino acid
is much less abundant in soybean meal and PepSoyGen than in
menhaden fish meal (Table 2). A crystalline methionine supple-
ment was used to ensure adequate dietary levels of this amino
acid; however, the maximum inclusion rates were set at 10 g/kg.
Thus, any further limitations regarding dietary methionine
levels had to be addressed via modifications to other feedstuff
inclusion rates. Feeds were prepared at the Fisheries and Illinois
Aquaculture Center (FIAC; Carbondale, Illinois) according to
standard FIAC feed manufacturing protocols: feedstuffs were
 PEPSOYGEN IN HYBRID STRIPED BASS FEEDS 199

TABLE 1. Formulation and proximate composition of test diets fed to hybrid Striped Bass. FM = fish mean, SBM = soybean meal, PSG = PepSoyGen.

Test diet
Diet component 30% FM 10% FM SBM 10% FM PSG 5% FM SBM 5% FM PSG 0% FM SBM 0% FM PSG
Ingredient (g/kg, as-fed basis)
Menhaden fish meala 300.0 100.0 100.0 50.0 50.0 0.0 0.0
Soybean mealb 260.4 561.9 255.0 639.7 255.0 717.5 255.0
PepSoyGenc 0.0 0.0 303.4 0.0 381.0 0.0 462.5
Wheat bran 290.0 181.6 182.2 148.4 147.3 115.2 102.6
Menhaden fish oila 40.0 40.0 40.0 40.0 40.0 40.0 40.0
Soybean oil 46.5 64.5 66.9 69.3 72.4 74.1 78.4
Carboxymethyl cellulose 20.0 20.0 20.0 20.0 20.0 20.0 20.0
Sodium phosphate 32.9 21.5 12.3 21.5 14.1 21.5 21.4
Choline 6.0 6.0 6.0 6.0 6.0 6.0 6.0
Stay-Cd 2.0 2.0 2.0 2.0 2.0 2.0 2.0
Vitamin premixe 1.2 1.2 1.2 1.2 1.2 1.2 1.2
Downloaded by [Department Of Fisheries] at 19:53 28 February 2013

Mineral premixf 1.0 1.0 1.0 1.0 1.0 1.0 1.0

Methionineg 0.0 0.3 10.0 0.9 10.0 1.5 10.0
Proximate compositionh (g/kg, dry matter basis, except Dry matter)
Dry matter 909 ± 1 960 ± 1 958 ± 1 960 ± 1 970 ± 1 946 ± 1 965 ± 1
Protein 373 ± 8 370 ± 8 387 ± 8 380 ± 8 395 ± 8 382 ± 8 389 ± 8
Lipid 153 ± 4 148 ± 4 148 ± 4 139 ± 4 148 ± 4 129 ± 4 137 ± 4
Ash 128 ± 3 87 ± 3 81 ± 3 78 ± 3 73 ± 3 76 ± 3 71 ± 3
a
Omega Protein, Houston, Texas.
b
47.5% crude protein; Siemer Enterprises, Teutopolis, Illinois. 52.0% crude protein; Nutra-Flo Protein and Biotech Products, Sioux City, Iowa.
c
Nutra-Flo Protein and Biotech Products, Sioux City, Iowa.
d
Argent Laboratories, Redmond, Washington.
e
Formulated to contain 25.000% L-ascorbyl-2-polyphosphate, 14.000% RRR-alpha tocopheryl acetate, 13.160% vitamin K, 12.500% inositol, 12.500% nicotinic acid, 7.500%
riboflavin, 6.250% calcium pantothenate, 2.500% pyridoxine hydrochloride, 1.250% thiamine mononitrate, 1.000% vitamin A palmitate, 0.500% cyanocobalamin, 0.450% folic acid,
0.125% biotin, and 0.010% cholecalciferol in a cellulose base.
f
Formulated to contain 24.897% zinc oxide, 14.933% ferrous sulfate, 3.470% manganese oxide, 0.967% cupric carbonate, 0.262% potassium iodide, 0.060% sodium selenate, and
0.030% cobalt carbonate in a cellulose base.
g
Maximum dietary inclusion rates for crystalline amino acid supplements were set at 10 g/kg.
h
Values represent least-square means ± pooled SE of triplicate samples.

mixed with a cutter–mixer (Model CM450, Hobart Corporation,
Troy, Ohio), pelleted using a commercial-grade food grinder
(1.5 hp electric grinder, Cabela’s, Sidney, Nebraska), and
dried in a commercial-grade food dryer (100◦ C; Harvest Saver
R-5A, Commercial Dehydrator Systems, Eugene, Oregon) to
approximately 950 g/kg dry matter. All feeds were maintained
frozen (−20◦ C) throughout the trial. Proximate analysis of
the feeds was conducted in triplicate in order to confirm feed
composition (Table 1). Briefly, the samples were lyophilized
(Freezone 6, Labconco Corporation, Kansas City, Missouri)
to determine moisture content by the loss of mass and then
pulverized for further analysis. Protein was determined using a
nitrogen–protein analyzer (LECO FP-528, LECO Corporation,
St Joseph, Michigan), ash was determined gravimetrically after
incineration (650◦ C for 4 h), and lipid content was determined
gravimetrically after chloroform–methanol extraction using a
method adapted from Folch et al. (1957). Minor differences
in dietary lipid content probably occurred due to small
inaccuracies in feedstuff weighing, incomplete mixing during
feed manufacturing or sample collection, or both.
Experimental design and feeding trial.—Juvenile hybrid
Striped Bass were sourced from a commercial vendor (Keo Fish
Farm, Keo, Arkansas) and stocked into a recirculation system at
the FIAC comprising biological and mechanical filters, twenty-
eight 110-L, round, fiberglass tanks, and supplied with con-
stant aeration by air diffusers (10 fish/tank; average individual
weight = 18.1 ± 0.2 g [mean ± SE]). Each dietary treatment
was assigned randomly to four replicate tanks (n = 4). All fish
were fed once a day to apparent satiation.
Water temperature and dissolved oxygen (YSI-85 series
dissolved oxygen meter, Yellow Springs Instruments, Yellow
Springs, Ohio) were measured daily. Total ammonia-, nitrite-
and nitrate-nitrogen were determined weekly by spectrophoto-
metric analysis (Hach, Loveland, Colorado) and alkalinity was
measured weekly by the bromocresol green–methyl red titration
method (Hach). Throughout the trial, photoperiod was kept at
24 h light to maximize growth performance, and all water qual-
ity conditions were maintained within the following acceptable
ranges for hybrid Striped Bass culture (Kohler 2000): temper-
ature range = 16.4–26.3◦ C (22.4 ± 0.3◦ C [mean ± SE]),
 200 ROMBENSO ET AL.

TABLE 2. Percent amino acid composition of the primary protein ingredients Feed Intake (% body weight/day) = 100
of test diets fed to hybrid Striped Bass. average individual dry matter feed intake
×
(average intial individual weight × average final individual weight)0.5 /
Protein source days of feeding
Menhaden Soybean
Amino acid (%) fish meala PepSoyGenb mealc Three fish per tank were then euthanized by an overdose of tri-
caine methanesulfonate (MS-222, Argent Chemical Laborato-
Arginine 4.0 3.6 3.6 ries, Redmond, Washington; fish were exposed to a ∼200-mg/L
Histidine 1.8 1.6 1.3 solution in culture water until cessation of opercular movement,
Isoleucine 2.7 2.2 2.6 ∼5 min) and then weighed and dissected to determine hepato-
Leucine 4.6 5.4 3.8 somatic index (HSI) and liposomatic index (LSI) as follows:
Lysine 5.0 3.2 2.2 HSI = 100 × (liver weight/whole body weight), and
Methionine 2.2 0.7 0.7 LSI = 100 × (intraperitoneal fat weight/whole body
Cystine 0.6 1.0 0.7 weight).
Phenylalanine 2.5 2.6 2.7 Statistical analysis.—Although data comprising multiple
Tyrosine 2.1 1.7 1.2 fish were collected for each tank, replicate tanks served as the
Threonine 2.7 2.2 2.0 experimental units for all statistical analyses (n = 4). All data
Downloaded by [Department Of Fisheries] at 19:53 28 February 2013

Tryptophan 0.5 0.6 0.7 were analyzed by one-way ANOVA (PROC GLIMMIX; Statis-
Valine 3.2 2.3 2.7 tical Analysis Software version 9.1, SAS Institute, Cary, North
Alanine 4.2 2.3 NR Carolina). For parameters exhibiting significant treatment ef-
Aspartic acid 5.8 6.0 NR fects, Tukey’s honestly significant difference (HSD) tests were
Glutamic acid 8.6 9.9 NR used for pairwise comparison of means. To further assess the
Glycine 4.8 2.4 NR main and interactive effects of fish meal sparing level and soy-
Proline 3.3 2.6 NR bean meal type, data from the experimental groups (excluding
Serine 2.5 2.8 NR the 30% FM control) were also subjected to a two-way ANOVA
Taurine 0.5 NRd NR (PROC GLIMMIX). In all cases, differences were considered
Hydroxyproline 1.4 NR NR significant at P < 0.05.
a
Omega Protein, Houston, Texas.
b
c
Nutra-Flo Protein and Biotech Products, Sioux City, Iowa. RESULTS
NRC (2011).
d
Not reported. In comparison with the 30% FM control group, production
performance tended to decline among fish fed diets containing
less than 10% FM; however, this effect was less overt among
dissolved oxygen range = 5.9–9.3 mg/L (7.6 ± 0.1 mg/L), to- those fed the PepSoyGen-based feeds (Table 3). In terms of
tal ammonia nitrogen range = 0.09–0.14 mg/L (0.12 ± 0.01), weight gain and SGR, at each level of fish meal replacement,
nitrite-nitrogen range = 0.01–0.06 mg/L (0.04 ± 0.01 mg/L), fish fed the PepSoyGen treatment outgrew those fed the cor-
nitrate-nitrogen range = 1.9–7.1 mg/L (5.7 ± 0.1 mg/L), and responding soybean meal feed, though differences at the 5%
alkalinity range = 60–324 mg/L CaCO3 (162 ± 50 mg/L FM level were not statistically significant for either parameter.
CaCO3 ). These results were underscored by significant effects of soy-
Growth performance, sample collection, and analysis.— bean meal type on weight gain and SGR (P < 0.001 for both
After 8 weeks, fish were group-weighed by tank and counted in parameters), indicating that PepSoyGen-based feeds generally
order to calculate standard metrics of production performance yielded greater growth performance than soybean meal-based
as follows.
feeds. Weight gain and SGR were also significantly affected by
fish meal sparing level (P < 0.001 for both parameters), indicat-
Weight Gain (%) = 100
ing reduced performance in the 0% FM series compared with
(average individual final weight − average individual intial weight)
× the 5% FM and 10% FM series across soybean meal types. A
average individual intial weight
significant interaction between fish meal sparing level and soy-
bean meal type was observed for SGR (P = 0.014) but not for
Feed Conversion Ratio (FCR) weight gain (P = 0.067).
average individual dry matter feed intake Differences observed in weight gain were the result of cor-
=
average individual weight gain responding differences in feed intake and FCR (Table 3). Com-
pared with the 30% FM control, all of the experimental feeds
Specific Growth Rate (SGR, % body weight/day) = 100 yielded equivalent FCRs and only fish fed the 0% FM feeds
log average individual final weight − loge average individual intial weight exhibited significantly reduced feed intake. However, the two-
× e
days of feeding way ANOVA revealed more subtle trends in these parameters.
 PEPSOYGEN IN HYBRID STRIPED BASS FEEDS 201

TABLE 3. Production performance of hybrid Striped Bass by dietary treatment. Values represent least-squares means; pooled SE and P-values resulting from
one-way ANOVA tests are also provided. Mean values with common letter labels are not significantly different (P > 0.05). FM = fish mean, SBM = soybean
meal, PSG = PepSoyGen, HSI = hepatosomatic index, LSI = liposomatic index.

Test diet
30% 10% FM 10% FM 5% FM 5% FM 0% FM 0% FM Pooled
Performance measure FM SBM PSG SBM PSG SBM PSG SE P-value
Survival (%) 100 98 100 100 100 100 100 1 0.451
Initial weight (g) 18.1 18.3 18.3 17.7 18.3 17.8 18.4 0.7 0.933
Final weight (g) 76.9 yz 63.5 xy 82.6 z 67.1 xyz 74.3 xyz 39.7 w 58.6 x 4.8 <0.001
Weight gain (%) 324.7 z 247.3 xy 351.9 z 278.7 xyz 305.7 yz 122.1 w 219.1 x 23.6 <0.001
FCR 1.00 yz 1.16 yz 1.04 yz 0.98 y 0.97 y 1.28 z 1.05 yz 0.09 0.029
SGR 2.58 z 2.22 xy 2.68 z 2.37 xyz 2.50 yz 1.42 w 2.06 x 0.1 <0.001
(% body weight/d)
Feed intake 2.82 yz 2.75 xyz 3.04 z 2.50 xy 2.63 xyz 1.85 w 2.29 w 0.1 <0.001
(% body weight/d)
Downloaded by [Department Of Fisheries] at 19:53 28 February 2013

HSI 1.0 0.8 1.0 1.0 1.2 0.9 1.4 0.4 0.377
LSI 2.0 ± 0.3 1.8 ± 0.3 2.5 ± 0.3 2.4 ± 0.3 2.7 ± 0.3 2.2 ± 0.3 2.7 ± 0.3 0.2 0.346

Specifically, feed intake was affected by fish meal sparing level
(10% FM > 5% FM > 0% FM; P < 0.001) and soybean meal
type (PSG > SBM; P = 0.003), though no interaction effect
was observed (P = 0.408). The FCR was similarly affected by
soybean meal type (SBM > PSG; P = 0.049) and fish meal
sparing level (0% FM > 5% FM, 10% FM = 0% FM and 5%
FM; P = 0.038), but not by an interaction of these effects (P =
0.304).
No significant differences (using one-way or two-way
ANOVA) were observed in HSI, LSI, or survival. Only a single
mortality was observed during the trial, and this was a fish that
jumped from its culture tank.
DISCUSSION
In the present work, increasing amounts of soybean meal
in the diet formulation resulted in declining production perfor-
mance among fish fed diets containing less than 10% fish meal.
Soybean meal has been successfully used to spare fish meal
in hybrid Striped Bass diets (Gallagher 1994; Keembiyehetty
and Gatlin 1997), but aggressive sparing or complete replace-
ment negatively affects production performance (Brown et al.
1997) and potentially physiological competence (Laporte and
Trushenski 2012). In the present work production performance
expectedly declined as soybean meal inclusion was increased,
but this effect was less overt among fish fed the PepSoyGen-
based feeds. At each level of fish meal replacement, PepSoyGen-
fed fish outperformed those fed the corresponding soybean meal
feed (albeit nonsignificantly, in the case of the 5% FM treat-
ments), indicating that fermented soybean meal may be supe-
rior to traditional soybean meal in fish diets. Although hybrid
Striped Bass appear able to tolerate high inclusion rates of soy-
bean meal in feeds based on apparent digestibility coefficients
(Rawles and Gatlin 2000; Thompson et al. 2008), high amounts
of soybean meal in the diet negatively affect growth, more so
than PepSoyGen.
Brown et al. (1997) tested several types of soybean products
(solvent-extracted soybean meal, roasted soybeans, and raw-
unprocessed soybeans) in juvenile hybrid Striped Bass and dis-
covered that raw soybean meal has limited potential in feeds for
this taxon, probably due to the presence of unattenuated ANF.
Those authors concluded that unprocessed soybeans have lit-
tle potential as an alternative protein source for fish meal, but
utilization can be improved by treatment. For example, hybrid
Striped Bass fed roasted soybean feeds outgrew fish fed the
raw soybean-based feeds. Similarly, solvent extraction proved
to be an effective treatment to yield a good protein alternative
in feeds for hybrid Striped Bass, allowing higher levels of re-
placement while maintaining production performance (Brown
et al. 1997). Others have similarly reported positive effects of
extrusion (Burnham et al. 2000), purification (Hancock et al.
1989), and defatting (Jones et al. 1989) of soybeans. Other re-
fined soy protein products have been developed, including soy
protein concentrate and soy protein isolates. Soy protein concen-
trate (SPC) is prepared by aqueous alcohol extraction of defat-
ted soybeans, and is almost completely free of phytate, lectins,
saponins, and oligosaccharides (Anderson and Wolf 1995). Soy
protein isolate (SPI) is prepared by continuous aqueous extrac-
tions conducted at different pH levels (Dersjant-Li 2002; Lin
et al. 2006). Although SPI contains higher saponin levels than
SPC, total ANF in this product may be equal or lower than in SPI
(Blaufuss and Trushenski 2012). Both SPC and SPI have been
successful in aquaculture feeding trials with partial replacement
of fish meal. Levels of 50% replacement with SPC have been
achieved in some carnivorous species including Rainbow Trout
Oncorhynchus mykiss and Atlantic Salmon Salmo salar (Médale
 202 ROMBENSO ET AL.
et al. 1998; Mambrini et al. 1999; Refstie et al. 2001) and 40%
with SPI in Rainbow Trout feeds (Glencross et al. 2005). How-
ever, both can reduce feed intake and growth rates depending
on the species and their inclusion rates. They can also be used
in combination with soybean meal to spare fish meal (Blaufuss
and Trushenski 2012), though those authors concluded that uti-
lization of these products in aquafeeds will depend on their
acceptability to carnivorous fishes (particularly juveniles) and
their commercial cost. As these examples illustrate, none of
the aforementioned processing strategies or protein refinements
completely neutralize ANF, and the utilization of these ingredi-
ents in carnivore aquafeeds is still constrained by biological and
economic limitations (Gatlin et al. 2007).
Fermented soy protein products such as PepSoyGen repre-
sent a next step in processing to reduce ANF. Fermented soy
products offer an ANF-reduced or ANF-free feed ingredient
Downloaded by [Department Of Fisheries] at 19:53 28 February 2013
with increased protein content (Mital and Garg 1990; Reddy
and Pierson 1994) and a higher proportion of small peptides
(Hong et al. 2004; Min et al. 2009). PepSoyGen has been used
successfully in diets for young terrestrial livestock (Jones et al.
2008; Yoo et al. 2009). Guandalini and Rubino (1982) deter-
mined that the absorption rate of glycine peptide in PepSoyGen
was greater than free glycine in rabbit intestines, probably due
to the presence of dipeptide transport mechanisms present in the
gut (Adibi and Phillips 1968; Li et al. 1999). Most PepSoyGen-
based diets used in weanling pig studies achieved higher growth
performance and nitrogen digestibility than did traditional soy-
bean meal diets (Min et al. 2004). PepSoyGen is also effective
in sparing higher cost refined soy proteins; Min et al. (2009)
had success in using PepSoyGen to replace SPC in weanling
and weaned pig diets, achieving adequate growth performance
and amino acid digestibility. Yoo et al. (2009) analyzed the ef-
fects of fermented soy protein by different organisms, including
PepSoyGen, on nitrogen balance, apparent fecal and ileal di-
gestibility in weaned pigs. Of those evaluated, PepSoyGen-
based diets improved feed utilization, probably due to the ex-
tensive hydrolysis of protein during fermentation that produces
readily available free amino acids and peptides (Steinkraus
1996; Sarkar et al. 1997; Yoo et al. 2009). In addition, the
PepSoyGen group had a lower level of urine-excretion nitrogen
than the other groups. Aspergillus oryzae fermentation can de-
crease trypsin inhibitors in soybean meal by 84% (Hong et al.
2004), and Bacillus spp. fermentation results in major biochem-
ical changes in the substrate leading to an increase in solubility
and in vitro digestibility (Kiers et al. 2000). Thus, PepSoyGen
can achieve the same results of reduced ANF in young pig di-
ets as those from the more costly soy protein concentrates and
isolates.
The benefits of fermented soy protein observed in the live-
stock studies mentioned above appear to translate to fish as well.
In the present study, PepSoyGen was able to maintain compa-
rable growth at the 5% FM level and at each level of fish meal
replacement outperformed fish fed the corresponding soybean
meal. Differences observed in weight gain were the result of
corresponding differences in feed intake and FCR. Compared
with the 30% FM control, all of the experimental feeds yielded
equivalent FCR, and only fish fed the 0% FM feeds exhibited
significantly reduced feed intake. Specifically, feed intake was
affected by fish meal sparing level (10% FM > 5% FM > 0%
FM; P < 0.001) and soybean meal type (PSG > SBM; P =
0.003), highlighting that at each level of fish meal replacement
the PepSoyGen-based feeds had higher feed intake compared
with soybean meal feeds. This agrees with previous research and
is probably related to the elimination of intake-suppressing ANF,
specifically oligosaccharides, which are almost completely re-
moved by the fermentation process (Jones et al. 2008; Min et al.
2009). While complete fish meal replacement was not possible
using our PepSoyGen-based formulations, PepSoyGen appears
to be more readily accepted by juvenile hybrid Striped Bass
than soybean meal. Importantly, PepSoyGen is also a practical
ingredient for aquafeeds as it is less costly than other refined soy
proteins and can be incorporated into feeds for extrusion with-
out causing negative effects on the pellets (Fallahi et al. 2012).
Further investigation with blends of alternative proteins includ-
ing PepSoyGen may lead to diet formulations that successfully
eliminate the need for fish meal as a primary protein source in
feeds for hybrid Striped Bass and other carnivorous fishes.
ACKNOWLEDGMENTS
We thank Bonnie Mulligan, Brian Gause, Franklin Woitel,
John Bowzer, and Matthew Young for their assistance with
data collection and analysis. We also thank Terry Waugh and
Jason Sewell for technical support and Nutraferma for financial
support and in-kind contributions of feedstuffs.
REFERENCES
Adibi, S. A., and E. Phillips. 1968. Evidence for greater adsorption of amino
acids from peptide than from free form in human intestine. Clinical Research
16:446.
Anderson, R. L., and W. J. Wolf. 1995. Compositional changes in trypsin in-
hibitors, phytic acid, saponins and isoflavones related to soybean processing.
Journal of Nutrition 125(3 Supplement):581–588.
Blaufuss, P., and J. Trushenski. 2012. Exploring soy-derived alternatives to fish
meal: using soy protein concentrate and soy protein isolate in hybrid Striped
Bass feeds. North American Journal of Aquaculture 74:8–19.
Brown, P. B., B. J. Brown, S. Hart, J. Curry, and A. Hittle-Hutson. 2008.
Comparison of soybean-based practical diets containing 32, 36, or 40% crude
protein fed to hybrid Striped Bass in earthen culture ponds. North American
Journal of Aquaculture 70:128–131.
Brown, P. B., R. Twibell, Y. Jonker, and K. A. Wilson. 1997. Evaluation of three
soybean products in diets fed to juvenile hybrid Striped Bass Morone saxatilis
× M. chrysops. Journal of the World Aquaculture Society 28:215–223.
Bureau, D. P., A. M. Harris, and C. Y. Cho. 1998. The effects of purified alcohol
extracts from soy products on feed intake and growth of Chinook Salmon
(Oncorhynchus tshawytscha) and Rainbow Trout (Oncorhynchus mykiss).
Aquaculture 161:27–43.
Burnham, L. L., I. H. Kim, J. O. Kang, H. W. Rhee, and J. D. Hancock. 2000.
Effects of sodium sulfite and extrusion on the nutritional value of soybean
products for nursery pigs. Asian-Australasian Journal of Animal Sciences
13:1584–1592.
 PEPSOYGEN IN HYBRID STRIPED BASS FEEDS
Chou, R. L., B. Y. Her, M. S. Su, G. Hwang, Y. H. Wu, and H. Y. Chen.
2004. Substituting fish meal with soybean meal in diets of juvenile Cobia
Rachycentron canadum. Aquaculture 229:325–333.
Dersjant-Li, Y. 2002. The use of soy protein in aquafeeds. Pages 541–558 in L.
E. Cruz-Suárez, D. Ricque-Marie, M. Tapia-Salazar, M. G. Gaxiola-Cortés,
and N. Simoes, editors. Avances en nutrición acuı́cola VI: memorias del VI
simposium internacional de nutrición acuı́cola, Cancun, Mexico.
El-Saidy, D. M. S. D., and M. M. A. Gaber. 2002. Complete replacement of fish
meal by soybean meal with dietary L-lysine supplementation for Nile Tilapia
Oreochromis niloticus (L.) fingerlings. Journal of the World Aquaculture
Society 33:297–306.
Fallahi, P., K. Muthukumarappan, K. A. Rosentrater, and M. L. Brown. 2012.
Twin-screw extrusion processing of vegetable-based protein feeds for Yellow
Perch (Perca flavescens) containing distillers dried grains, soy protein con-
centrate, and fermented high protein soybean meal. Journal of Food Research
1:230–246.
FAO (Food and Agriculture Organization of the United Nations). 2010. Fish-
meal market report—March 2010. FAO, Fisheries Department, Globe-
fish, Rome. Available: www.globefish.org/fishmeal-market-report-march-
Downloaded by [Department Of Fisheries] at 19:53 28 February 2013
2010.html. (September 2010).
FAO (Food and Agriculture Organization of the United Nations). 2012. The state
of world fisheries and aquaculture (SOFIA). FAO, Fisheries and Aquaculture
Department, Rome.
Folch, J., M. Lees, and G. H. Sloane-Stanley. 1957. A simple method for the
isolation and purification of total lipides from animal tissues. Journal of
Biological Chemistry 226:497–509.
Francis, G., H. P. S. Makkar, and K. Becker. 2001. Antinutritional factors
present in plant-derived alternate fish feed ingredients and their effects in
fish. Aquaculture 199:197–227.
Gallagher, M. L. 1994. The use of soybean meal as a replacement for fish meal in
diets for hybrid Striped Bass (Morone saxatilis × M. chrysops). Aquaculture
126:119–127.
Gatlin, D. M., III. 1997. Nutrition and feeding of Striped Bass and hybrid Striped
Bass. Pages 235–251 in R. M. Harrell, editor. Striped Bass and other morone
culture. Elsevier, Amsterdam.
Gatlin, D. M., III, F. T. Barrows, P. Brown, K. Dabrowski, T. G. Gaylord, R.
W. Hardy, E. Herman, G. Hu, Å. Krogdahl, R. Nelson, K. Overturf, M. Rust,
W. Sealey, D. Skonberg, E. J. Souza, D. Stone, R. Wilson, and E. Wurtele.
2007. Expanding the utilization of sustainable plant products in aquafeeds: a
review. Aquaculture Research 38:551–579.
Gaylord, T. G., S. D. Rawles, and D. M. Gatlin III. 2004. Amino acid
availability from animal, blended, and plant feedstuffs for hybrid Striped
Bass (Morone chrysops × M. saxatilis). Aquaculture Nutrition 10:345–
352.
Glencross, B., D. Evans, K. Dods, P. McCafferty, W. Hawkins, R. Maas, and S.
Sipsas. 2005. Evaluation of the digestible value of lupin and soybean protein
concentrates and isolates when fed to Rainbow Trout, Oncorhynchus mykiss,
using either stripping or settlement faecal collection methods. Aquaculture
245:211–220.
Guandalini, S., and A. Rubino. 1982. Digestion and adsorption of protein. Pages
88–91 in T. Brody, editor. Nutritional biochemistry, 2nd edition. Academic
Press, San Diego, California.
Hancock, J. D., A. J. Lewis, and E. R. Peo Jr. 1989. Effects of ethanol extrac-
tion on the utilization of soybean protein by growing rats. Nutrition Reports
International 39:813–821.
Hancock, J. D., E. R. Peo Jr., A. J. Lewis, and J. D. Crenshaw. 1990. Effects
of ethanol extraction and duration of heat treatment of soybean flakes on the
utilization of soybean protein by growing rats and pigs. Journal of Animal
Science 68:3233–3243.
Hong, K. J., C. H. Lee, and S. W. Kim. 2004. Aspergillus oryzae GB-107
fermentation improves nutritional quality of food soybeans and feed soybean
meals. Journal of Medicinal Food 7:430–435.
Jones, C. K., J. M. DeRouchey, J. L. Nelssen, M. D. Tokach, S. S. Dritz, and
R. D. Goodband. 2008. Effects of pepsoygen and dried porcine solubles 50
203
in nursery pig diets. Pages 52–61 in Kansas swine day 2008 proceedings.
Kansas State University Agricultural Experiment Station and Cooperative
Extension Service, 2009-10-06, Manhattan.
Jones, D. B., J. D. Hancock, J. L. Nelsson, and D. F. Li. 1989. Effect of
replacing milk products with a soybean protein isolate milk replacer in diets
for nursery pigs. Pages 54–60 in Kansas swine day 1989 proceedings. Kansas
State University Agricultural Experiment Station and Cooperative Extension
Service, Manhattan.
Keembiyehetty, C. N., and D. M. Gatlin III. 1997. Performance of Sunshine
Bass fed soybean-meal-based diets supplemented with different methionine
compounds. Progressive Fish-Culturist 59:25–30.
Kiers, J. L., A. E. A. Van laeken, F. M. Rombouts, and M. J. R. Nout. 2000. In
vitro digestibility of Bacillus fermented soya bean. International Journal of
Food Microbiology 60:163–169.
Kim, S. H., J. L. Yang, and Y. S. Song. 1999. Physiological functions of
chungkukjang. Food Industry and Nutrition 4:40–46.
Kohler, C. C. 2000. Striped Bass and hybrid Striped Bass culture. Pages 898–907
in R. R. Stickney, editor. Encyclopedia of aquaculture. Wiley, New York.
Kohler, C. C. 2004. A white paper on the status and needs of hybrid Striped
Bass aquaculture in the north central region. North Central Regional
Aquaculture Center, Iowa State University, Ames. Available: www.ncrac.
org/content/status-and-needs-hybrid-striped-bass-aquaculture-north-central-
region. (November 2011).
Laporte, J., and J. Trushenski. 2012. Production performance, stress toler-
ance and intestinal integrity of Sunshine Bass fed increasing levels of soy-
bean meal. Journal of Animal Physiology and Animal Nutrition 96:513–
526.
Lee, H. J. 1998. Health functional peptides from soybean foods. Korea Soybean
Digest 15:16–22.
Li, D. F., X. H. Zhao, T. B. Yang, E. W. Johnson, and P. A. Thacker. 1999.
A comparison of the intestinal absorption of amino acids in piglets when
provided in free form or as a dipeptide. Asian-Australasian Journal of Animal
Science 12:939–943.
Lin, J., P. G. Krishnan, and C. Wang. 2006. Retention of isoflavones and saponins
during the processing of soy protein isolates. Journal of the American Oil
Chemists’ Society 83:59–63.
Mambrini, M., A. J. Roem, J. P. Carvèdi, J. P. Lallès, and S. J. Kaushik. 1999.
Effects of replacing fish meal with soy protein concentrate and of DL-
methionine supplementation in high-energy, extruded diets on the growth
and nutrient utilization of Rainbow Trout, Oncorhynchus mykiss. Journal of
Animal Science 77:2990–2999.
Médale, F., T. Boujard, F. Vallée, D. Blanc, M. Mambrini, A. J. Roem, and S.
J. Kaushik. 1998. Voluntary feed intake, nitrogen and phosphorus losses in
Rainbow Trout (Oncorhynchus mykiss) fed increasing dietary levels of soy
protein concentrate. Aquatic Living Resources 11:239–246.
Min, B. J., J. H. Cho, Y. J. Chen, H. J. Kim, J. S. Yoo, Q. Wang, I. H. Kim,
W. T. Cho, and S. S. Lee. 2009. Effects of replacing soy protein concentrate
with fermented soy protein in starter diet on growth performance and ileal
amino acid digestibility in weaned pigs. Asian-Australasian Journal of Animal
Science 22:99–106.
Min, B. J., J. W. Hong, O. S. Kwon, W. B. Lee, Y. C. Kim, I. H. Kim, W.
T. Cho, and J. H. Kim. 2004. The effect of feeding processed soy protein
on the growth performance and apparent ileal digestibility in weanling pigs.
Asian-Australasian Journal of Animal Science 17:1271–1276.
Mital, B. K., and S. K. Garg. 1990. Tempeh—technology and food value. Food
Reviews International 6:213–224.
NRC (National Research Council). 2011. Nutrient requirements of fish and
shrimp. National Academies Press, Washington, D.C.
Rawles, S. D., and D. M. Gatlin III. 2000. Nutrient digestibility of common
feedstuffs in extruded diets for Sunshine Bass Morone chrysops♀ × M.
saxatilis♂. Journal of the World Aquaculture Society 31:570–579.
Reddy, N. R., and M. D. Pierson. 1994. Reduction in antinutritional and toxic
components in plant foods by fermentation. Food Research International
27:281–290.
 204 ROMBENSO ET AL.
Refstie, S., T. Storebakken, G. Baeverfjord, and A. J. Roem. 2001. Long-term
protein and lipid growth of Atlantic Salmon (Salmo salar) fed diets with
partial replacement of fish meal by soy protein products at medium or high
lipid level. Aquaculture 193:91–106.
Refstie, S., T. Storebakken, and A. J. Roem. 1998. Feed consumption and
conversion in Atlantic Salmon (Salmo salar) fed diets with fish meal, ex-
tracted soybean meal or soybean meal with reduced content of oligosaccha-
rides, trypsin inhibitors, lectins and soya antigens. Aquaculture 162:301–
312.
Robaina, L., M. S. Izquierdo, F. J. Moyano, J. Socorro, J. M. Vergara, and D.
Montero. 1998. Increase of the dietary n–3 / n–6 fatty acid ratio and addition
of phosphorus improves liver histological alterations induced by feeding diets
containing soybean meal to Gilthead Seabream, Sparus aurata. Aquaculture
161:281–293.
Robinson, E. H., and M. H. Li. 1994. Use of plant proteins in catfish feeds:
replacement of soybean meal with cottonseed meal and replacement of fish
meal with soybean meal and cottonseed meal. Journal of the World Aquacul-
ture Society 25:271–276.
Sarkar, P. K., L. J. Jones, G. S. Craven, S. M. Somerset, and C. Palmer. 1997
Downloaded by [Department Of Fisheries] at 19:53 28 February 2013
Amino acid profiles of kinema, a soybean-fermented food. Food Chemistry
59:69–75.
Steinkraus, K. H., editor. 1996. Handbook of indigenous fermented foods.
Marcel Dekker, New York.
Takagi, S., H. Murata, T. Goto, M. Endo, H. Yamashita, and M. Ukawa. 2008.
Taurine is an essential nutrient for Yellowtail Seriola quinqueradiata fed
non-fish meal diets based on soy protein concentrate. Aquaculture 280:198–
205.
Thompson, K. R., S. D. Rawles, L. S. Metts, R. G. Smith, A. Wimsatt, A.
L. Gannam, R. G. Twibell, R. B. Johnson, Y. J. Brady, and C. D. Webster.
2008. Digestibility of dry matter, protein, lipid, and organic matter of two
fish meals, two poultry by-product meals, soybean meal, and distiller’s dried
grains with solubles in practical diets for Sunshine Bass, Morone chrysops
× M. saxatilis. Journal of the World Aquaculture Society 39:352–363.
Trushenski, J. T., C. S. Kasper, and C. C. Kohler. 2006. Challenges and oppor-
tunities in finfish nutrition. North American Journal of Aquaculture 68:122–
140.
Trushenski, J. T., M. Schwarz, W. V. N. Pessoa, B. Mulligan, C. Crouse, B.
Gause, F. Yamamoto, and B. Delbos. 2013. Amending reduced fish-meal
feeds with marine lecithin, but not soy lecithin, improves the growth of
juvenile Cobia and may attenuate heightened responses to stress challenge.
Journal of Animal Physiology and Animal Nutrition 97:170–180.
Turano, M. 2012. 2012 U.S. hybrid Striped Bass industry update. Presentation
at Aquaculture America 2012, World Aquaculture Society, Louisiana State
University, Baton Rouge.
Watanabe, T., and J. Pongmaneerat. 1993. Potential of soybean meal as a pro-
tein source in extruded pellets for Rainbow Trout. Nippon Suisan Gakkaishi
59:1415–1423.
Yoo, J. S., H. D. Jang, J. H. Cho, J. H. Lee, and I. H. Kim. 2009. Effects
of fermented soy protein on nitrogen balance and apparent fecal and ileal
digestibility in weaned pigs. Asian-Australasian Journal of Animal Science
22:1167–1173.
Zhou, Q. C., K. S. Mai, B. P. Tan, and Y. J. Liu. 2005. Partial replacement of fish-
meal by soybean meal in diets for juvenile Cobia (Rachycentron canadum).
Aquaculture Nutrition 11:175–182.
This article was downloaded by: [Department Of Fisheries]
On: 28 February 2013, At: 19:54
Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,
37-41 Mortimer Street, London W1T 3JH, UK

North American Journal of Aquaculture

Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/unaj20

Pretreating Channel Catfish with Copper Sulfate Affects

Susceptibility to Columnaris Disease
a a a a
Bradley D. Farmer , Benjamin H. Beck , Andrew J. Mitchell & David L. Straus
a
U.S. Department of Agriculture, Agricultural Research Service, Harry K. Dupree Stuttgart
National Aquaculture Research Center, Post Office Box 1050, Stuttgart, Arkansas, 72160, USA
Version of record first published: 26 Feb 2013.

To cite this article: Bradley D. Farmer , Benjamin H. Beck , Andrew J. Mitchell & David L. Straus (2013): Pretreating Channel
Catfish with Copper Sulfate Affects Susceptibility to Columnaris Disease, North American Journal of Aquaculture, 75:2,
205-211

To link to this article: http://dx.doi.org/10.1080/15222055.2012.758210

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions

This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to
anyone is expressly forbidden.

The publisher does not give any warranty express or implied or make any representation that the contents
will be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses should
be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims,
proceedings, demand, or costs or damages whatsoever or howsoever caused arising directly or indirectly in
connection with or arising out of the use of this material.
 North American Journal of Aquaculture 75:205–211, 2013
American Fisheries Society 2013
ISSN: 1522-2055 print / 1548-8454 online
DOI: 10.1080/15222055.2012.758210

ARTICLE

Pretreating Channel Catfish with Copper Sulfate Affects

Susceptibility to Columnaris Disease
Bradley D. Farmer,* Benjamin H. Beck, Andrew J. Mitchell, and David L. Straus
U.S. Department of Agriculture, Agricultural Research Service,
Harry K. Dupree Stuttgart National Aquaculture Research Center, Post Office Box 1050,
Stuttgart, Arkansas 72160, USA
Downloaded by [Department Of Fisheries] at 19:54 28 February 2013

Abstract
Columnaris disease is one of the most important bacterial diseases affecting Channel Catfish Ictalurus punctatus
commercially grown in the USA. This disease can greatly diminish the profitability of aquaculture operations by
large-scale mortality events, particularly in the fingerling production phase. Three experiments were conducted to
evaluate the susceptibility of Channel Catfish fingerlings to columnaris disease when they were preexposed to copper
sulfate (CuSO4 ). In experiment 1, fish were exposed for 24 h to 0, 1, 2, or 4 mg/L CuSO4 and were challenged
immediately with Flavobacterium columnare, the etiological agent of columnaris disease. The resulting survival data
indicated that fish preexposed to CuSO4 and then challenged exhibited significantly lower survival than did fish not
exposed and then challenged. Experiment 2 was designed as above, except after the 24-h static exposure to CuSO4 ,
fish were subjected to an additional 24 h in flow-through water prior to the challenge with F. columnare. In contrast
to experiment 1, fish preexposed to CuSO4 and allowed an additional 24 h in flow-through water had a significantly
higher survival rate than fish not exposed and then challenged. Experiment 3 evaluated the longevity of resistance to
columnaris disease afforded by preexposure to CuSO4 ; in this experiment the remaining fish from experiment 2 were
challenged 9 d after exposure to CuSO4 . The increased survival rate of fish preexposed to CuSO4 was still significantly
different, indicating the incurred resistance to F. columnare lasts for at least a week after exposure to CuSO4 .

Copper sulfate (CuSO4 ) has a long and storied history of use
in aquaculture including use as an algicide, an anti-biofouling
agent, and as a therapeutant for disease applications (Tucker and
Robinson 1990). A recent development for an additional use of
CuSO4 arises from both anecdotal and scientific evidence as-
serting that the preexposure of fish to CuSO4 can affect the
fish’s susceptibility to disease. Recent reports by fish farmers in
North Carolina indicated that CuSO4 reduced the incidence rate
of disease and possibly made fish more resistant when exposed
to infectious disease (Steve Gabel, North Carolina State Uni-
versity Extension, personal communication). However, many
questions remain, especially the mechanism by which copper
seems to protect fish from pathogens, which pathogens it pro-
tects against, the duration of protection, and the doses required.
The preexposure of Channel Catfish Ictalurus punctatus to
CuSO4 can increase resistance to enteric septicemia of catfish
(ESC; Griffin and Mitchell 2007). MacFarlane et al. (1986)
reported that preexposure of Striped Bass Morone saxatilis to
CuSO4 protected against Flavobacterium columnare challenge.
Conversely, other studies have concluded that CuSO4 exposure
can result in immunosuppression and provokes a higher suscep-
tibility to disease (Nemcsók and Boross 1982; Zelikoff 1993).
Increased susceptibility to bacterial, viral, and fungal infections
after exposure has been reported in Steelhead Onchorhynchus
mykiss. Steelhead were more susceptible to enteric redmouth
infection (Knittel 1981), infectious hematopoietic necrosis
(Hetrick et al. 1979), and saprolegniasis (Carballo et al. 1995)
after sublethal exposure to copper.
The mechanism by which CuSO4 exerts its beneficial ef-
fects is probably multifaceted. Low concentrations of copper
can have a stimulatory effect on the immune system of fish
(Muhvich et al. 1995; Dautremepuits et al. 2004). Copper

*Corresponding author: bradley.farmer@ars.usda.gov

Received August 29, 2012; accepted December 7, 2012

205
 206 FARMER ET AL.
sulfate can also have a direct inhibitory effect on bacteria
through the displacement of essential metals from their na-
tive binding sites or through ligand interactions, resulting in
conformational changes in the nucleic acids and proteins that
alter the oxidative phosphorylation cascade and osmotic bal-
ance (Borkow and Gabbay 2005). The current series of exper-
iments investigated the utility of preexposing Channel Catfish
to CuSO4 and its effects on their susceptibility to columnaris
disease, one of the most important bacterial diseases of Channel
Catfish commercially grown in the USA (USDA 2010).
METHODS
Fish husbandry.—Channel Catfish fingerlings were cultured
indoors at the Harry K. Dupree Stuttgart National Aquaculture
Research Center (SNARC), Stuttgart, Arkansas, in six 250-L
Downloaded by [Department Of Fisheries] at 19:54 28 February 2013
fiberglass stock tanks and were fed and monitored daily. The
turnover rate of the flow-through well water was once every 2 h.
Average weight of fingerlings was 5.5 ± 1.5 g (mean ± SD)
and average length was 9.3 ± 0.7 cm.
Water temperatures ranged from 26.2◦ C to 27.5◦ C. Tem-
perature and dissolved oxygen were measured daily with a
YSI Pro20 meter (Yellow Springs Instruments, Yellow Springs,
Ohio). The mean dissolved oxygen concentration was 5.6 ±
1.2 mg/L. Total ammonia nitrogen (TAN) concentrations were
determined in each tank with a Hach DR/4000V spectropho-
tometer using the Nessler method 8038 (Hach, Loveland, Col-
orado). An Orion 720A pH meter (Fisher Scientific, Waltham,
Massachusetts) was used to measure pH at the beginning of the
study; pH range was 7.5–8.2. Standard titration methods (APHA
2005) were used to measure total alkalinity (213 mg/L) and total
hardness (112 mg/L).
Fish were observed daily for clinical signs associated with
the disease. Fish unable to maintain neutral buoyancy were con-
sidered moribund and removed for sampling. Fish were not fed
during CuSO4 exposure or during the first day after challenge,
but were offered food on day 2 and throughout the rest of the
study. Animal care and experimental protocols were approved
by the SNARC Institutional Animal Care and Use Committee
and conformed to U.S. Department of Agriculture, Agricultural
Research Service Policies and Procedures 130.4 and 635.1.
Experimental design.—Three experiments were conducted
to evaluate the effects of preexposure to CuSO4 . In each exper-
iment there were six treatments and three replications of each
treatment (n = 3) for a total of 18 tanks. Treatments were as fol-
lows: (1) fish not exposed to CuSO4 and then challenged with
F. columnare (challenge control), (2) fish exposed to 1 mg/L
CuSO4 and then challenged with F. columnare, (3) fish exposed
to 2 mg/L CuSO4 and then challenged with F. columnare, (4)
fish exposed to 4 mg/L CuSO4 and then challenged with F.
columnare, (5) fish exposed to 4 mg/L CuSO4 but not chal-
lenged with F. columnare, (6) fish not exposed to CuSO4 nor
challenged with F. columnare (negative control). Experiments 1
and 2 were conducted under similar conditions.
Fish for all experiments had the same preexposure; however,
in experiment 1, the fish were challenged immediately after the
24-h static exposure to CuSO4 . In experiment 2, an additional
24 h in flow-through water was allowed before fish were stocked
out and then challenged with F. columnare. Experiment 3 was
conducted on the remaining fish from experiment 2; fish were
challenged either 9 d after exposure to CuSO4 or 1 week since
the challenge in experiment 2.
Copper sulfate exposure.—Exposure to CuSO4 was con-
ducted in six 250-L stock tanks; each contained 2 kg (approxi-
mately 360 fish) of Channel Catfish fingerlings to ensure that all
fish representing a treatment group were exposed equally. The
fish were acclimated in these tanks for 2 weeks. The CuSO4
concentrations chosen were based on a previous recommenda-
tion of 1% of the alkalinity, which was approximately 200 mg/L
(as CaCO3 ) in our water. This resulted in a range of 0.5 × , 1 × ,
and 2 × for the 1, 2, and 4 mg/L CuSO4 exposures, respectively.
The exposure was 24 h static with supplemental aeration. The
chemical was applied from a stock solution of CuSO4 prepared
at a concentration of 10 g/L. An aliquot of each stock solution
was added to the respective tank. After 5 min unfiltered wa-
ter samples were collected and at 24 h filtered and unfiltered
samples were collected to determine total and dissolved copper
concentrations. Copper concentrations were verified by stan-
dard methods (APHA 2005) using inductively coupled plasma
(ICP) spectroscopy.
Challenge with F. columnare.—Fish exposed to CuSO4 and
nonexposed fish were moved to aquaria containing 10 L of
aerated flow-through well water using the “Ultra Low-Flow
System” described by Mitchell and Farmer (2010). The flow
rate was set to 29 ± 1 mL/min and monitored daily; this rate
allowed a natural progression of the disease in a flow-through
environment.
Fish were allocated from the stock tanks to the challenge
aquaria at a rate of 0.5 kg (90 fish) per tank in experiments 1 and
2. In experiment 3, fish were stocked at 20 fish/tank because
of the limited number of fish remaining in the second group.
Tanks receiving the bacterial challenge were then exposed to
F. columnare isolate LV-359-01, an isolate previously shown
to produce consistent mortality in the low-flow system (Farmer
et al. 2011). The isolate was retrieved from the −80◦ C freezer
and streaked on Ordals medium (Anacker and Ordal 1959). After
48 h the isolate was dislodged from the agar using a sterile cotton
swab and inoculated into 5 mL of F. columnare growth medium
(FCGM; Farmer 2004). This suspension was incubated at 28◦ C
for 24 h, and then the 5-mL starter culture was used to inoculate
1 L of FCGM. The inoculated broth was incubated in a flask
for 24 h at 28◦ C in an orbital shaker incubator at 200 rpm.
When the bacterial growth reached an absorbance of 0.68 at
550 nm (approximately 4.0 × 1010 bacteria/mL), the flask was
removed and placed on a stir plate at room temperature. A 10-mL
sample was removed for serial dilution and colony-forming-
unit (CFU) enumeration. In experiment 1, 50 mL of bacterial
suspension was added to tanks receiving bacterial challenges. In
 COPPER SULFATE PRETREATMENT AND RESISTANCE TO COLUMNARIS
TABLE 1. Total copper (Cu) concentrations at 5 min and 24 h and dissolved copper concentrations at 24 h (percent dissolved of total in parentheses) from each
experiment (Exp.) in the study on the pretreatment of Channel Catfish with CuSO4 to increase resistance to columnaris disease.
Total Cu at 5 min
CuSO4 (mg/L) Exp. 1 Exp. 2
1.0 1.03 1.03
2.0 2.09 2.21
4.0 4.24 4.42
0 <0.02 <0.02
experiments 2 and 3, the challenge dose was increased to 100 mL
to maximize the potential for finding a significant difference in
survival.
Bacteriology.—Dead and moribund fish were removed from
Downloaded by [Department Of Fisheries] at 19:54 28 February 2013
the tanks daily and samples were taken from the caudal fin, gills,
and liver. The samples were cultured on selective Cytophaga
agar containing 5 µg/mL neomycin sulfate and 200 U/mL
polymyxin B to limit nontargeted bacterial growth (Hawke and
Thune 1992). If present, a maximum of three moribund or dead
fish were sampled from each tank per day. Cultures were incu-
bated at 28◦ C for 48 h then scored as being positive or negative
for growth based on colony morphology (i.e., flat, yellow rhizoid
colonies).
Clinical observation and histology.—Fish were observed
twice daily for signs of disease and observations recorded. Mor-
talities were collected twice daily and analyzed as a function
of time. Gill samples from five fish from each treatment group
were collected and immediately fixed in Davidson’s solution
at the conclusion of the 24-h CuSO4 exposure from the stock
tanks in experiments 1 and 2. In experiment 2, an additional
five gill samples were collected 24 h after the end of the CuSO4
exposure and just before the bacterial challenge. In experiment
3, gills were sampled 9 d after the conclusion of the CuSO4
exposure and just before challenge again from the stock tanks.
Sampled tissues were transferred to 70% isopropanol 24–48 h
after fixation and stored until routine paraffin embedding, at
which time tissues were sectioned to 5–6 µm, mounted on glass
slides, and stained with hematoxylin and eosin for pathological
evaluation.
Statistical analysis.—Survival data were analyzed with
SigmaPlot 11 (Systat Software, San Jose, California) using
Kaplan–Meier log rank survival analysis, and all pairwise mul-
tiple comparisons used the Holm–Sidak method with type III
adjusted P-values. Each experiment was analyzed separately,
and tank was the fixed effect and replicates were the random ef-
fect. Treatment effects were considered significant at P ≤ 0.05.
RESULTS
Copper Analysis
The background level of copper in the unexposed tanks was
below detection limit (<0.02 mg/L) in all experiments. There
207
Total Cu at 24 h Dissolved Cu at 24 h (% of total)
Exp. 1 Exp. 2 Exp. 1 Exp. 2
0.77 0.86 0.54 (70.8) 0.65 (75.6)
1.64 1.47 1.10 (67.1) 1.21 (82.3)
3.65 3.67 2.22 (60.8) 2.87 (78.2)
<0.02 <0.02 <0.02 <0.02
was not a significant difference in the copper measured in each
exposure level between the experiments (Table 1).
Clinical Signs
Moribund Channel Catfish in challenged tanks displayed
signs consistent with an F. columnare infection; fish were lethar-
gic, and minimal feeding activity was noted on day 4. This ac-
tivity increased in fish in all tanks over the rest of the study
period. Gross pathologies were also typical; the skin of mori-
bund fish initially had discrete depigmented areas that became
multifocal to diffuse as the infection progressed. Dermal ulcer-
ation and cutaneous sloughing exposed the underlying muscles,
and severely frayed fins were also observed. The gills had fo-
cal to multifocal branchial necrosis with yellowish pigment. No
internal gross changes or lesions were observed.
Bacteriology and Histology
Notably, F. columnare was cultured from at least one tissue
from all fish sampled. The culture data included 85 total necrop-
sies: 30 from experiment 1 and 55 total from experiments 2 and
3. Growth was considered positive if at least one colony matched
the colony morphology of F. columnare.
Histological analysis of gills sampled immediately after
CuSO4 exposure indicated some level of pathology in all CuSO4
treated fish (Figure 1). The most extensive gill damage was ob-
served in fish exposed to 4 mg/L CuSO4 for 24 h with no
recovery time. The most common lesions in this treatment were
moderate hyperplasia of the epithelium and moderate to severe
edema that lifted the epithelia (Figure 1, panel B). Gills of fish
exposed to lower levels of CuSO4 showed less damage, and mild
hyperplasia was commonly observed (not shown). Figure 1 also
illustrates the stages of recovery 24 h (panel C) and 9 d (panel
D) after the conclusion of the CuSO4 exposure experiments.
There was obvious evidence of repair by 24 h, and a complete
recovery (lack of pathologic changes) at 9 d.
Survival
Experiment 1.—Channel Catfish survival was significantly
affected by preexposure to CuSO4 in all experiments. In experi-
ment 1, survival was significantly lower in all groups exposed to
CuSO4 . The mean ( ± SD) percent survival for the nonexposed
challenged control was 50.3 ± 13.3%. Fish exposed to 1, 2,
 208
FIGURE 1. Representative photomicrographs of gills of Channel Catfish pre-
treated with CuSO4 to increase resistance to columnaris disease. (A) Untreated
control, (B) treated with 4 mg/L CuSO4 for 24 h, (C) treated with 4 mg/L CuSO4
for 24 h followed by a 24-h recovery time, (D) treated with 4 mg/L CuSO4 for
24 h followed by a 9-d recovery time. Scale bar = 100 µm.
Downloaded by [Department Of Fisheries] at 19:54 28 February 2013
and 4 mg/L CuSO4 and then challenged with F. columnare had
mean survival rates of 25.4 ± 7.9%, 33.0 ± 17.5%, and 38.9 ±
13.0%, respectively. Fish in all tanks exposed to CuSO4 had sig-
nificantly lower rates of survival than did the nonexposed groups
(P ≤ 0.05), but survival rates were not significantly different
from each other. One fish died in the nonexposed and nonchal-
lenged groups (negative control) of unknown causes. No fish
died in the fish exposed to 4 mg/L CuSO4 but not challenged
with F. columnare. Mortality persisted for 5 d with the majority
of fish deaths occurring by day 3. The study was concluded on
day 7 after 2 d of no mortalities. Survival curves are depicted in
Figure 2.
Experiment 2.—Fish survival in experiment 2 was also sig-
nificantly affected by CuSO4 exposure prior to challenge. How-
ever, in this experiment, an additional 24 h was allowed after the
conclusion of the CuSO4 exposure and before the challenge. All
fish exposed to CuSO4 had significantly higher survival than did
the unexposed challenged fish. Survival rates for the 1-, 2-, and
4-mg/L CuSO4 exposures were 46.6 ± 22.8%, 65.9 ± 19.3%,
and 85.9 ± 2.7%, respectively, compared with 8.7 ± 13.8%
survival in the challenge control. Mortality attributed to canni-
balism did occur in the negative control and 4-mg/L CuSO4 and
nonchallenged tanks at a rate of 1.75% and 0.4%, respectively.
Again, the disease progressed for 5 d with most mortalities
occurring on day 3. Experiment 2 was also concluded on day
7 after 2 d of no mortalities. Mortalities were collected twice
daily and analyzed as a function of time; results are depicted in
Figure 3.
Experiment 3.—Experiment 3 investigated the longevity of
the improved resistance of Channel Catfish to columnaris dis-
ease described in experiment 2. The fish remaining in the stock
tank that had been exposed to CuSO4 , as well as unexposed
fish, were maintained for 9 d after the challenge in experiment
2. Twenty fish were stocked into each aquarium and challenged
as in experiment 2. Signs of columnaris were observed and mor-
talities began on day 2 and persisted to day 6. Fish exposed to
CuSO4 (all doses) 9 d prior to challenge still had significantly
FARMER ET AL.
FIGURE 2. Kaplan–Meier survival curve of Channel Catfish exposed to
CuSO4 for 24 h under static conditions and then challenged immediately with
Flavobacterium columnare. Treatments were: (1) fish not exposed to CuSO4
and then challenged with F. columnare (challenge control; CC), (2) fish exposed
to 1 mg/L CuSO4 and then challenged with F. columnare (1C), (3) fish exposed
to 2 mg/L CuSO4 and then challenged with F. columnare (2C), (4) fish ex-
posed to 4 mg/L CuSO4 and then challenged with F. columnare (4C), (5) fish
exposed to 4 mg/L CuSO4 but not challenged with F. columnare (4NC), (6) fish
not exposed to CuSO4 nor challenged with F. columnare (negative control, NC).
higher survival than nonexposed fish (P ≤ 0.05) (Figure 4). This
indicates that the diminished susceptibility persists for at least
a week after exposure to CuSO4 . Survival rates for fish exposed
to 1, 2, and 4 mg/L CuSO4 were 46.7 ± 25.7%, 43.3 ± 5.8%,
and 70.0 ± 20.0%, respectively, compared with 23.3 ± 10.4%
survival in the challenge control. No mortality was observed
in the negative control treatment. Resulting survival curves are
shown in Figure 4. For comparison, the mean percent survival
(±SD), difference in mean survival relative to the challenged
control, and statistical significance within an experiment for all
experiments are given in Table 2.
DISCUSSION
The cumulative results of the three experiments clearly
demonstrate that preexposure to CuSO4 can significantly affect
the susceptibility of Channel Catfish fingerlings to F. columnare.
Intriguingly, if the pathogen is introduced immediately after
CuSO4 exposure, then susceptibility to columnaris disease is
increased significantly compared with nonexposed fish. This in-
crease in susceptibility can be attributed, at least partially, to the
pathological lesions induced by the CuSO4 exposure as shown
in Figure 1. However, if a day is granted to recover from the
CuSO4 exposure, then a significant reduction in susceptibility
 COPPER SULFATE PRETREATMENT AND RESISTANCE TO COLUMNARIS 209
Downloaded by [Department Of Fisheries] at 19:54 28 February 2013

FIGURE 3. Kaplan–Meier survival curve of Channel Catfish exposed to
CuSO4 for 24 h under static conditions, given an additional 24 h in flow-through
well water, and then challenged with Flavobacterium columnare. Treatments
were: (1) fish not exposed to CuSO4 and then challenged with F. columnare
(challenge control; CC), (2) fish exposed to 1 mg/L CuSO4 and then challenged
with F. columnare (1C), (3) fish exposed to 2 mg/L CuSO4 and then challenged
with F. columnare (2C), (4) fish exposed to 4 mg/L CuSO4 and then chal-
lenged with F. columnare (4C), (5) fish exposed to 4 mg/L CuSO4 but not
challenged with F. columnare (4NC), (6) fish not exposed to CuSO4 nor chal-
lenged with F. columnare (negative control, NC).
FIGURE 4. Kaplan–Meier survival curve of Channel Catfish exposed to
CuSO4 for 24 h under static conditions, maintained for 9 d in flow-through
well water, and then challenged with Flavobacterium columnare. Treatments
were: (1) fish not exposed to CuSO4 and then challenged with F. columnare
(challenge control; CC), (2) fish exposed to 1 mg/L CuSO4 and then challenged
with F. columnare (1C), (3) fish exposed to 2 mg/L CuSO4 and then challenged
with F. columnare (2C), (4) fish exposed to 4 mg/L CuSO4 and then chal-
lenged with F. columnare (4C), (5) fish exposed to 4 mg/L CuSO4 but not
challenged with F. columnare (4NC), (6) fish not exposed to CuSO4 nor chal-
lenged with F. columnare (negative control, NC).

is achieved that persists for at least 1 week. This can possibly
be explained by the partial recovery of the lesions 24 h
postexposure and full recovery by 9 d postexposure (Figure 1).
Findings reported here add support to the conjecture that
CuSO4 exposure can reduce the susceptibility of fish to F.
columnare and Edwardsiella ictaluri (MacFarlane et al. 1986;
Griffin and Mitchell 2007). Furthermore, additional evidence
is presented that shows this benefit can persist for a minimum
of 9 d. However, there are conflicting studies that report that
CuSO4 exposure can result in immunosuppresson and higher

TABLE 2. Mean ( ± SD) percent survival on the pretreatment of Channel Catfish with CuSO4 to increase resistance to columnaris disease, difference in mean
survival relative to the challenged control ( CC), and statistical significance within the experiment. Treatments were: (1) fish not exposed to CuSO4 and then
challenged with Flavobacterium columnare (challenge control; CC), (2) fish exposed to 1 mg/L CuSO4 and then challenged with F. columnare (1C), (3) fish
exposed to 2 mg/L CuSO4 and then challenged with F. columnare (2C), (4) fish exposed to 4 mg/L CuSO4 and then challenged with F. columnare (4C), (5) fish
exposed to 4 mg/L CuSO4 but not challenged with F. columnare (4NC), (6) fish not exposed to CuSO4 nor challenged with F. columnare (negative control, NC).
Means within a column with different letters are significantly different (P ≤ 0.05).

Experiment 1 Experiment 2 Experiment 3

Treatment % Survival  CC % Survival  CC % Survival  CC
NC 99.7 x 98.3 v 100 w
4NC 100 x 99.6 v 100 w
CC 50.3 y (13.3) 8.7 w (13.8) 23.3 x (10.4)
1C 25.4 z (7.9) (−) 20.1 46.6 x (22.8) ( + ) 37.9 46.7 y (25.7) ( + ) 23.4
2C 33.0 z (17.5) (−) 17.3 65.9 y (19.3) ( + ) 57.2 43.3 y (5.8) ( + ) 20.0
4C 38.9 z (13.0) (−) 11.4 85.9 z (2.7) ( + ) 77.2 70.0 z (20.0) ( + ) 46.7
 210 FARMER ET AL.
susceptibility to disease (Nemcsók and Boross 1982; Zelikoff
1993). Rougier et al. (1996) noted that low doses of copper,
before and after challenge with Listeria monocytogenes, were
protective to Zebrafish Brachydanio rerio, while high doses of
copper were counter protective. Farmer et al. 2012 showed that
CuSO4 reducesdthe number of F. columnare on fish and in water,
and improved survival in infected fish. However, in the present
study, no copper was added to the aquaria in which the fish were
challenged with F. columnare, eliminating the possibility for
direct bactericidal action on the bacteria. This suggests that the
resulting difference in susceptibility must be from physiological
changes in the fish. Ongoing studies in our laboratory are ex-
amining the mechanisms by which CuSO4 directly affects fish
physiology and results in tangible effects on fish susceptibility
to disease.
The negative results from experiment 1 are possibly ex-
Downloaded by [Department Of Fisheries] at 19:54 28 February 2013
plained by the pathologic changes caused by the copper ex-
posure. Epithelial hyperplasia and edema were observed in fish
exposed to the highest level of CuSO4 . However, resolution of
these changes at later time points could have reduced bacte-
rial attachment and subsequent infection. An alternative theory
is that CuSO4 exposure stimulates a nonspecific immune re-
sponse or modulates how that response is mediated. Low con-
centrations of copper can have a stimulatory effect on the im-
mune system of fish (Muhvich et al. 1995; Dautremepuits et al.
2004).
Copper sulfate is inexpensive, and the protocol we have pre-
sented could potentially be used as a management strategy to
avert losses from columnaris and possibly reduce the use of
more expensive antibiotics. Particularly noteworthy are the re-
sults from experiment 3, which indicated resistance was main-
tained for a considerable period of time. These findings suggest
copper application could be beneficial to a number of situations
in catfish aquaculture. Specifically, the harvest of a fingerling
pond for grading and stocking can be stressful to the fish and
conducive to outbreaks of columnaris disease. Additionally, the
seasonal pattern of spring and fall disease outbreaks could lend
themselves to prophylactic application of CuSO4 .
Dissolved copper ions are the active component of CuSO4 in
water (Boyd 2000). Copper ions bind rapidly to organic and in-
organic materials, and further testing is warranted to determine
whether laboratory results are repeatable under typical pond
conditions. Also of concern is the potential for copper toxicity,
which is closely tied to water chemistry conditions, specifi-
cally hardness and alkalinity, and varies between regional water
sources (Boyd 2000). A toxic level was not achieved, even at
twice the normal dose or 4 mg/L, and higher ranges could be
investigated.
These data represent an important step towards understand-
ing the positive attributes of CuSO4 exposure as a possible
tool for the management of diseases in aquaculture. Additional
studies should investigate the mechanism by which resistance
occurs and also evaluate the effectiveness of preexposure to
CuSO4 in the prevention of diseases caused by other aquatic
pathogens.
ACKNOWLEDGMENTS
The authors thank Matt Barnett and Cindy Ledbetter for
their technical assistance throughout the course of the study.
We are grateful for constructive comments and review of early
manuscripts. This study was funded by the U.S. Department
of Agriculture (USDA), Agricultural Research Service under
Project No. 6225-32000-004-00D. Mention of trade names or
commercial products in this article is solely for the purpose of
providing specific information and does not imply recommen-
dation or endorsement by the USDA. The USDA is an equal
opportunity provider and employer.
REFERENCES
Anacker, R. L., and E. J. Ordal. 1959. Studies on the myxobacterium Chon-
drococcus columnaris: I. serological typing. Journal of Bacteriology 78:25–
32.
APHA (American Public Health Association). 2005. Standard methods for the
examination of water and wastewater, 21st edition. APHA, in conjunction
with the American Water Works Association and the Water Environment
Federation, Washington, D.C.
Borkow, G., and J. Gabbay. 2005. Copper as a biocidal tool. Current Medicinal
Chemistry 12:2163–2175.
Boyd, C. E. 2000. Water quality in warmwater fish ponds. Auburn University,
Agricultural Experiment Station, Auburn, Alabama.
Carballo, M., M. J. Munoz, M. Cuellar, and J. V. Tarazona. 1995. Effects
of waterborne copper, cyanide, ammonia, and nitrite on stress parameters
and changes in susceptibility to Saprolegniosis in Rainbow Trout (On-
corhynchus mykiss). Applied and Environmental Microbiology 61:2108–
2112.
Dautremepuits, C., S. Betoulle, S. Paris-Palacios, and G. Vernet. 2004.
Immunology-related perturbations induced by copper and chitosan in carp
(Cyprinus carpio L.). Archives of Environmental Contamination and Toxi-
cology 47:370–378.
Farmer, B. D. 2004. Improved methods for the isolation and characterization
of Flavobacterium columnare. Master’s thesis. Louisiana State University,
Baton Rouge.
Farmer, B. D., B. H. Beck, and D. L. Straus. 2012. Effectiveness of cop-
per sulfate and potassium permanganate on Channel Catfish infected with
Flavobacterium columnare. North American Journal of Aquaculture 74:
320–329.
Farmer, B. D., A. J. Mitchell, and D. L. Straus. 2011. The effect of high total
ammonia concentration on the survival of Channel Catfish experimentally
infected with Flavobacterium columnare. Journal of Aquatic Animal Health
23:162–168.
Griffin, B. R., and A. J. Mitchell. 2007. Susceptibility of Channel Cat-
fish, Ictalurus punctatus (Rafinesque), to Edwardsiella ictaluri challenge
following copper sulphate exposure. Journal of Fish Diseases 30:581–
585.
Hawke, J. P., and R. L. Thune. 1992. Systemic isolation and antimicrobial
susceptibility of Cytophaga columnaris from commercially reared Channel
Catfish. Journal of Aquatic Animal Health 4:109–113.
Hetrick, F. M., M. D. Knittel, and J. L. Fryer. 1979. Increased suscep-
tibility of Rainbow Trout to infectious hematopoietic necrosis virus af-
ter exposure to copper. Applied and Environmental Microbiology 37:198–
201.
 COPPER SULFATE PRETREATMENT AND RESISTANCE TO COLUMNARIS
Knittel, M. D. 1981. Susceptibility of Steelhead Trout Salmo gairdneri Richard-
son to redmouth infection Yersinia ruckeri following exposure to copper.
Journal of Fish Diseases 4:33–40.
MacFarlane, R. D., G. L. Bullock, and J. J. A. McLaughlin. 1986. Ef-
fects of five metals on susceptibility of Striped Bass to Flexibacter
columnaris. Transactions of the American Fisheries Society 115:227–
231.
Mitchell, A., and B. Farmer. 2010. Evaluation of an ultra-low-flow water deliv-
ery system for small experimental tanks. North American Journal of Aqua-
culture 72:195–200.
Muhvich, A. G., R. T. Jones, A. S. Kane, R. S. Anderson, and R. Reimscheussel.
1995. Effects of chronic copper exposure on the macrophage chemilumines-
cent response and gill histology in Goldfish (Carassius auratus L.). Fish and
Shellfish Immunology 5:251–264.
Downloaded by [Department Of Fisheries] at 19:54 28 February 2013
211
Nemcsók, J., and L. Boross. 1982. Comparative studies on the sensitivity of dif-
ferent fish species to metal pollution. Acta Biologica Academiae Scientiarum
Hungaricae 33:23–27.
Rougier, F., A. Menudier, C. Bosgiraud, and J. A. Nicolas. 1996. Copper and
zinc exposure of Zebrafish, Brachydanio rerio (Hamilton-Buchanan): effects
in experimental listeria infection. Ecotoxicology and Environmental Safety
34:134–140.
Tucker, C. S., and E. H. Robinson. 1990. Channel Catfish farming handbook.
Van Nostrand Reinhold, New York.
USDA (U.S. Department of Agriculture). 2010. Part 1: reference of fingerling
catfish health and production practices in the United States, 2010. USDA
National Animal Health Monitoring System, Fort Collins, Colorado.
Zelikoff, J. T. 1993. Metal pollution-induced immunomodulation in fish. Annual
Review of Fish Diseases 3:305–325.
This article was downloaded by: [Department Of Fisheries]
On: 19 March 2013, At: 23:21
Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,
37-41 Mortimer Street, London W1T 3JH, UK

North American Journal of Aquaculture

Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/unaj20

Safety of Feed Treated with 17α-Methyltestosterone

(17MT) to Larval Nile Tilapia
a b b b
David L. Straus , James D. Bowker , Molly P. Bowman , Daniel G. Carty , Andrew J.
a a a
Mitchell , Bradley D. Farmer & Cynthia K. Ledbetter
a
U.S. Department of Agriculture, Agricultural Research Service, Harry K. Dupree Stuttgart
National Aquaculture Research Center, Post Office Box 1050, Stuttgart, Arkansas, 72160, USA
b
U.S. Fish and Wildlife Service, Aquatic Animal Drug Approval Partnership Program, 4050
Bridger Canyon Road, Bozeman, Montana, 59715, USA
Version of record first published: 05 Mar 2013.

To cite this article: David L. Straus , James D. Bowker , Molly P. Bowman , Daniel G. Carty , Andrew J. Mitchell , Bradley D.
Farmer & Cynthia K. Ledbetter (2013): Safety of Feed Treated with 17α-Methyltestosterone (17MT) to Larval Nile Tilapia,
North American Journal of Aquaculture, 75:2, 212-219

To link to this article: http://dx.doi.org/10.1080/15222055.2012.758211

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions

This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to
anyone is expressly forbidden.

The publisher does not give any warranty express or implied or make any representation that the contents
will be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses should
be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims,
proceedings, demand, or costs or damages whatsoever or howsoever caused arising directly or indirectly in
connection with or arising out of the use of this material.
 North American Journal of Aquaculture 75:212–219, 2013
American Fisheries Society 2013
ISSN: 1522-2055 print / 1548-8454 online
DOI: 10.1080/15222055.2012.758211

ARTICLE

Safety of Feed Treated with 17␣-Methyltestosterone

(17MT) to Larval Nile Tilapia
David L. Straus*
U.S. Department of Agriculture, Agricultural Research Service,
Harry K. Dupree Stuttgart National Aquaculture Research Center,
Post Office Box 1050, Stuttgart, Arkansas 72160, USA

James D. Bowker, Molly P. Bowman, and Daniel G. Carty

U.S. Fish and Wildlife Service, Aquatic Animal Drug Approval Partnership Program,
Downloaded by [Department Of Fisheries] at 23:21 19 March 2013

4050 Bridger Canyon Road, Bozeman, Montana 59715, USA

Andrew J. Mitchell, Bradley D. Farmer, and Cynthia K. Ledbetter

U.S. Department of Agriculture, Agricultural Research Service,
Harry K. Dupree Stuttgart National Aquaculture Research Center,
Post Office Box 1050, Stuttgart, Arkansas 72160, USA

Abstract
As a synthetic androgen, 17α-methyltestosterone (17MT) is frequently used to redirect the course of sex differ-
entiation by exposing the undifferentiated gonad to a sufficient dosage. This hormone has been widely accepted as a
safe and effective treatment for sex reversal in many fish species, and it is administered to larval tilapia (3–12 d old)
for ∼28 consecutive days to produce populations of >90% males. This study assessed the safety of 17MT-treated feed
when administered to larval Nile Tilapia Oreochromis niloticus at one, three, and five times (i.e., 1×, 3×, and 5×)
the proposed dosage of 9 mg 17MT/kg fish daily for 28 consecutive days. Despite elevated total ammonia nitrogen
levels measured during the last 5 d of the study, environmental conditions were acceptable for rearing tilapia. Fish fed
aggressively regardless of the concentration of 17MT in the feed, behavior was considered normal with no dose-related
differences detected, and no mortality was observed in the 3× treatment group. Fish that were treated with five times
the proposed therapeutic dosage had significant pathological changes. Based on the results of this study, the 17MT
margin of safety extends to at least 3× (27 mg 17MT/kg fish daily) the proposed dosage of 9 mg 17MT/kg fish daily
when administered in feed for 28 d to Nile Tilapia.

Larval stages of many teleosts contain both ovarian and
testicular tissues, and sexual differentiation commences shortly
after hatching or initiation of feeding (Yamamoto 1969; Don-
aldson and Hunter 1982; Yamazaki 1983). Various techniques
have been developed to control sexual differentiation in fishes
(Donaldson and Hunter 1982). These techniques have typically
involved androgen or estrogen treatments to override endoge-
nous mechanisms of sex determination in developing larval
stages. Treatment regimens have generally included immersion
of larval fish in water containing a steroid, incorporation of a
steroid in the larval diet, or both. Results have been variable
because timing, dosage, and environmental conditions can in-
fluence treatment efficacy. The steroid 17α-methyltestosterone
(17MT) is frequently used to redirect the course of sex differ-
entiation by exposing the undifferentiated gonad to a sufficient
dosage, and it is a male-specific hormone commonly used to
induce sex reversal in teleost fishes (Hunter and Donaldson
1983).

*Corresponding author: dave.straus@ars.usda.gov

Received October 8, 2012; accepted December 7, 2012

212
 SAFETY OF 17MT TO LARVAL TILAPIA
Gonadal differentiation in tilapia Oreochromis spp. occurs
at 8–25 d posthatch, and these fish begin to reproduce at 3–6
months of age. Early and prolific reproduction is a restriction
on commercial production of these fish. An all-male population
allows for greater feed conversion and larger, more marketable
fish. Oral administration of 17MT to newly hatched tilapia fry
(3–12 d old) for ∼28 consecutive days results in populations
composed of >90% males (Green et al. 1997; Rani and Macin-
tosh 1997; Teichert-Coddington et al. 2000). Excess androgen
introduced into the early life stage of fish overrides endogenous
hormones and directs sexual differentiation towards the for-
mation of testes. Orally administered 17MT is an efficacious,
cost-effective, and efficient way to produce populations of
primarily male tilapia. Therefore, oral administration of 17MT
is ideal for the commercial production of tilapia.
In the United States, public agencies and private industry are
working to obtain U.S. Food and Drug Administration (FDA)
Downloaded by [Department Of Fisheries] at 23:21 19 March 2013
approval of 17MT for use in aquaculture for tilapia. To obtain
this approval, studies must demonstrate that this treatment regi-
men is safe to representative target animals. These target-animal
safety studies are designed to estimate a margin of safety asso-
ciated with overdosing and overexposing healthy test fish (e.g.,
Greenlees 1997; Gaikowski et al. 2003; Straus et al. 2012). In
this study, the margin of safety of 17MT-treated feed was as-
sessed when administered to larval Nile Tilapia O. niloticus at
one, three, and five times (i.e., 1×, 3×, and 5×) the proposed
therapeutic dosage of 9 mg 17MT/kg fish daily for 28 consecu-
tive days.
METHODS
Nile Tilapia fry were obtained from AmeriCulture, Animas,
New Mexico, and shipped to the Harry K. Dupree Stuttgart
National Aquaculture Research Center (SNARC), Stuttgart,
Arkansas. At SNARC, the fish were held for 4 d in two reference-
population aquaria before being moved into test aquaria and
allowed to acclimate for 3 d; fish weight at the start of the study
was 0.032 g ± 0.008 (mean ± SD). Males and females were
assumed to be present in roughly equal proportions. Fish health
evaluations were conducted onsite by a fish pathologist certified
by the American Fisheries Society. Reference population fish
were fed nontreated tilapia starter feed (Tilapia Sex Reversal
Control Feed, Rangen, Buhl, Idaho) at 15% of body weight per
day (BW/d) with an automatic feeder (Fish Mate F14 Aquar-
ium Feeders, Pet Mate, Hersham, Surrey, UK). Feed was not
administered the day fish were transferred from the reference
population to the test aquaria.
Larval Nile Tilapia were randomly stocked into 16 test
aquaria filled with recirculating well water at 50 fish per aquar-
ium. An additional four aquaria were stocked identically and
used as “surrogates” to monitor fish growth during the exposure
period. Diets (treatment groups) were randomly assigned to four
replicate aquaria per treatment. Nontreated control feed was ad-
ministered to the surrogate aquaria, and the fish in these aquaria
213
were weighed weekly to calculate proper feeding quantities in
the test aquaria. Treatment groups were 0×, 1×, 3×, or 5× (0,
9, 27, and 45 mg 17MT/kg fish daily, respectively) the proposed
therapeutic dose of 9 mg 17MT/kg fish daily for 28 consecutive
days.
Static test aquaria (n = 20) were glass (22 L), and each
was filled with 15 L of water and contained a biofilter (Tetra
Whisper 10, Blacksburg, Virginia). The laboratory in which the
study was conducted was equipped with overhead lights set on
a timer to produce a 12 h light : 12 h dark cycle. Test and
surrogate aquaria were cleaned daily. Approximately 20% of
the water in each aquarium was removed daily and replaced to
maintain adequate rearing conditions.
Water temperature and dissolved oxygen (DO) concentra-
tion were measured with a YSI Pro20 dissolved oxygen and
temperature meter (YSI Environmental, Yellow Springs, Ohio),
alkalinity and hardness were measured by titration using stan-
dard methods (APHA 2005), and pH was measured with an
Orion Research 720A meter (Thermo Electron Corporation,
Beverly, Massachusetts). Total ammonia nitrogen (TAN) and
nitrite (NO2 ) concentrations were measured with a HACH col-
orimeter (Hach, Loveland, Colorado). The TAN and NO2 levels
became elevated near the end of the study due to fish growth
and increased feeding, and it was necessary to change 50% of
the water daily during the last few days.
Rangen prepared 5-kg batches of Tilapia Sex Reversal Feed
with the appropriate dose of 17MT (Spectrum Chemical Man-
ufacturing, Gardena, California), and one 20-kg bag of non-
treated tilapia starter feed (0×). Samples were collected from
each batch of feed for dose verification analyses to confirm that
(1) the concentration of 17MT in each batch of treated feed was
within ± 20% of the target concentration, (2) 17MT was mixed
homogenously throughout each batch of treated feed, (3) 17MT
was not present in nontreated feed, and (4) the 17MT concentra-
tion in each batch remained stable throughout the study. Anal-
ysis was performed by Maxxam Analytics (Burnaby, British
Columbia) using HPLC.
During the 28-d treatment period, feed was administered to
fish at a rate of 15% BW/d using automatic feeders. The amount
of feed was adjusted daily based on a projected growth curve
for tilapia previously cultured at SNARC. The fish in the test
tanks were assumed to grow at a similar rate as fish in the
surrogate tanks and adjusted feed amounts were delivered to
test tanks based on surrogate fish size. Separate feed containers
were filled each week with the appropriate amount of feed to be
delivered during the week and stored in a freezer until use.
Prior to distributing fish in test aquaria, 30 fish were sam-
pled from the reference population (15 fish/aquaria) for general
necropsy and 10 of these fish were randomly selected for his-
tological evaluation; due to the small size of these fish, they
were examined under a microscope to assess general health. On
the day after the last treatment, 20 live fish were taken from
each test aquaria, measured for length and weight, and assessed
for general health using standard necropsy procedures. Fish
 214 STRAUS ET AL.
health evaluations included a visual examination of (1) skin for
discoloration and presence and severity of dermal lesions, (2)
gills for pallor, and (3) internal tissues (liver, spleen, and kid-
ney) for signs of gross lesions and abnormalities. Ten of the 20
fish from each test aquarium were randomly selected, placed
in individual containers, and processed for histological evalu-
ation. Over the course of the in-life phase, nine fish died or
became moribund. A fish health evaluation was performed on
each of these fish, and then they were processed for histological
evaluation.
Histology.—Upon completion of fish health evaluations, the
10 fish selected from each test aquarium were placed in David-
son’s fixative solution for 48–72 h, and then transferred into 70%
ethyl alcohol. If fish were small enough, they were processed
whole; if fish were larger, the head section was removed and
they were embedded in two separate Omnisette tissue cassettes
(Fisher Scientific, Fair Lawn, New Jersey) as follows: (1) the
Downloaded by [Department Of Fisheries] at 23:21 19 March 2013
head of larger fish was removed posterior to the gills and cut in
half longitudinally, and both halves were placed into a cassette,
(2) the muscle wall was removed from the left side of the body
and the entire body section was placed into another cassette,
(3) cassettes were infiltrated with paraffin with a Leica ASP
300 advanced smart processor (Leica Microsystems, Nussloch,
Germany), (4) samples were embedded in paraffin blocks with
a Leica EG 1160 tissue embedding system, (5) blocks were sec-
tioned (5 µm thick) on a Leica RM2255 rotary microtome, and
(6) sections were mounted on glass microscope slides, stained
with hematoxylin and eosin (H&E), and protected with a cover
slip.
Of the 10 fish sampled from each aquarium for histology,
eight were designated for evaluation of gill, liver, anterior kid-
ney, and posterior kidney. Tissues evaluated in the remaining
two fish included those four tissues as well as brain, heart,
muscle, skin, spleen, pyloric intestine, and rectal intestine. A
histopathologist evaluated each tissue for cellular changes or
lesions that might provide evidence of 17MT-induced toxicity.
Lesions were scored using a six-point ordinal severity scale
(Straus et al. 2012); tissue lesions in fish from the 0× or 5×
treatment groups were examined and compared first. If lesions
were detected in one or more tissues of the 5× group that were
characterized as marked or severe and not observed in the 0×
group, then all fish from the 3× group were examined for that
specific tissue. If such lesions were detected in one or more
tissues of the 3× group, then all fish from the 1× group were
examined for that specific tissue. Because differences were de-
tected among histological evaluation scores as stated above, fish
from all treatment groups were evaluated histologically.
The histologist used a Zeiss light microscope (Carl Zeiss
International, Jena, Germany) to evaluate fish tissues and doc-
umented features associated with lesions (e.g., inflammation,
edema), whether lesions were scattered or focal, and whether
lesions were considered to be artifacts. For data analysis pur-
poses, a dichotomized histological scoring scheme was devel-
oped in which histology scores of 0, 1, 2, or 3 were considered
nonpathological and given a score of 0. Histology scores of 4 or
5 were considered pathological and were given a score of 1.
Data analysis.—For each of the treatment groups, mortal-
ity and histology data were analyzed using SAS Proc Glimmix
(SAS version 9.1.3, SAS Institute, Cary, North Carolina). Treat-
ment effects were tested at α = 0.10 significance level. All other
data were summarized or statistically analyzed, or both, using
Microsoft Office Excel (2003 version), SYSTAT 12.02 (Systat
Software, Chicago, Illinois), or SigmaPlot 11.2 (Systat Soft-
ware, San Jose, California). Mean fish TL and body weight
for treatment groups at the end of the study were analyzed by
ANOVA. A pairwise multiple comparison test was used if a sig-
nificant difference was detected by ANOVA. Treatment effects
were tested at α = 0.05 significance level.
RESULTS
Mortality was observed in the 1× and 5× treatment groups;
the single mortality in the 1× group was considered incidental
(i.e., cannibalism). Although mean cumulative mortality in the
5× group (9%) was greater than in the other treatment groups
(mean cumulative mortality ranged from 0.0 to 0.5%), differ-
ences were not significant.
Fish behavior was characterized as normal in the 0×, 1×,
and 3× treatment groups. Behavior of fish in the 5× group was
also characterized as normal, but these fish did not consume
feed as aggressively. During the first 24 d of the study, fish
fed actively and consumed virtually all feed offered. However,
in the last 5 d of the study, fish in most aquaria, including
fish in surrogate tanks, consumed less than 100% of the feed
(often consuming only 75% of the feed). In two aquaria of the
5× treatment group, less feed was consumed during the last
5 d possibly because the amount of feed offered to fish was
not adjusted to account for mortality (feed was preweighed in
1-week increments). As a result, fish were fed at rates ranging
from 15.3 to 19.2% BW/d.
Over the course of the in-life phase, mean TL and weight of
fish in all treatment aquaria increased. Mean ( ± SD) TL and
weight of the 30 fish collected from the reference population
before the start of the study was 9.8 ± 0.65 mm and 0.012
± 0.003 g, respectively. A significant difference was detected
in mean TL and weight among the four treatment groups (n =
80 for each group) at the end of the study. Based on results
from a pairwise comparison test, fish from the 1× group (53 ±
5.07 mm and 2.52 ± 0.78 g) were significantly larger than fish
in the 0× (49 ± 5.18 mm and 2.11 ± 0.70 g) and 5× (47 ±
4.67 mm and 1.64 ± 0.50 g) groups. Fish in the 0× and 3×
(50 ± 3.85 mm and 2.15 ± 0.48 g) groups were significantly
larger than fish in the 5× group. Size of fish in the 3× group
was not significantly different from the 0× and 1× groups.
All fish sampled from the reference population showed
mostly normal tissues, except that moderate lifting of gill ep-
ithelium and moderately severe vacuolation of liver hepatocytes
was common. Nearly all moribund fish collected during the
 SAFETY OF 17MT TO LARVAL TILAPIA
Downloaded by [Department Of Fisheries] at 23:21 19 March 2013
FIGURE 1. Proteinaceous material (dark orange–red) in blood vessels of gill tissue, clearly visible in filament blood (just below center from right to left) of the
5× treatment group in the study on the safety of 17MT to larval Nile Tilapia. [Figure available in color online.]
study had gill lesions, and filamentous bacteria were observed
in H&E-stained sections. No bacterial pathogens were observed
on any live fish sampled.
Fish sampled from all treatment groups had mostly normal
brain, skeletal muscle, skin, spleen, and intestine tissues. Le-
sions that were characterized as marked or severe and consid-
ered pathological were detected in the gill, liver, anterior kidney,
posterior kidney, and heart of fish from all groups. The follow-
ing tissue changes were common: (1) gill tissue had moderate to
moderately severe edema, mild proliferation of epithelium, and,
in some fish, low numbers of eosinophilic granular cells, (2)
liver tissue had moderate to moderately severe vacuolation and
mild nuclear pleomorphism of hepatocytes, and (3) kidney tis-
sue had minimal to moderate hyaline droplet degeneration and
necrosis of tubule epithelium, numbers of regenerating (devel-
oping, immature) tubules, and scattered eosinophilic granular
cells.
Histological changes that were specific to the treatment
groups were as follows. A few fish from those sampled from
the 1× group had minimal to mild amounts of proteinaceous
material in the heart atrium (i.e., eosinophilic deposits were not
seen in blood vessels of any fish in this group); hypertrophy
(mild to moderate) of cardiac muscle cells was also confined to
the heart atrium. Fish sampled from the 3× group had intravas-
cular deposits of proteinaceous material, except for fish in one
aquarium; 10 of the 3× group fish showed changes similar to 5×
group fish. Fish sampled from the 5× group had intravascular
deposits of proteinaceous material, which was the most remark-
215
able finding; accumulation of this material was most pronounced
in blood vessels of the gills (Figure 1). Moderate to severe hy-
pertrophy, degeneration, and necrosis of cardiac muscle cells
were seen in heart tissue of fish from the 5× group (Figures 2,
3). Mild to moderately severe accumulations of intracellular
protein in hepatocytes was evident. Skeletal muscle appeared
normal in all fish examined in the 5× group.
Tissues in which lesions were considered pathological are
summarized in Table 1. Statistical results indicated that signif-
icant differences were not detected in the severity of lesions
between the 5× and 0× treatment groups.
Mean concentrations of 17MT in the treated feeds (Table 2)
were within ± 20% of the target dose. All individual feed sam-
ples analyzed from each batch of feed were less than the target
feed concentration by 6.7–11.9%, but were also within ± 20%
of the target dose. No 17MT was detected in the 0× control
feed.
Mean water temperature among test tanks was 28.0◦ C, and
individual measurements ranged from 26.9◦ C to 28.8◦ C. Mean
DO concentrations among test aquaria was 6.7 mg/L, and in-
dividual measurements ranged from 4.0 to 7.6 mg/L. All DO
concentrations measured in test aquaria were in a range con-
sidered acceptable for rearing tilapia. Mean water hardness
(131.1 ± 2.84 mg/L as CaCO3 ), alkalinity (231.3 ± 4.07 mg/L
as CaCO3 ), and pH (8.3 ± 0.24) in test aquaria were also within
the range considered acceptable for rearing tilapia.
Mean concentrations of TAN (<0.2 mg/L) and NO2
(<0.09 mg/L) were considered low during the first half of the
 216 STRAUS ET AL.
Downloaded by [Department Of Fisheries] at 23:21 19 March 2013

FIGURE 2. Hypertrophy and lysis of cardiac muscle cells of the 5× treatment group in the study on the safety of 17MT to larval Nile Tilapia. [Figure available
in color online.]

TABLE 1. Treatment group histology scores of marked (4) or severe (5) fish study. During the last 9 d of the study, mean daily concen-
tissues considered pathological in the study on the safety of 17MT to larval Nile trations of TAN ranged from 0.27 to 6.70 mg/L. The overall
Tilapia.
concentration of TAN in all test aquaria during this 9-d period
Histology score was 2.97 mg/L (n = 7 sample days), and the mean concentra-
tion of TAN in tank water in each of the four treatment groups
Tissue 4 5 ranged from 2.59 (3× group) to 3.23 mg/L (1× group).
Gill Water quality was suitable for holding and rearing healthy
Epithelial lifting 0× None tilapia, despite elevated TAN levels during the last 5 d of the
Edema 0×, 1×, 3×, 5× None study. Therefore, survival of test fish, general and feeding (ap-
Proliferation 1×, 5× 5× petite) behaviors, fish health evaluations, and 17MT-induced
Proteinaceous material 5× None gross and microscopic lesions evident in test fish were used to
Protein in blood 5× None estimate a margin of safety.
Liver
Vacuolation 5× None
DISCUSSION
Protein in hepatocytes 5× None
Posterior Kidney Based on the results of this study, the 17MT margin of safety
Degeneration 0×, 3× None extends to at least 3× (27 mg 17MT/kg fish daily) the proposed
Necrosis 0×, 3× None dosage of 9 mg 17MT/kg fish daily when administered in feed
Protein in blood 5× None for 28 d to Nile Tilapia. Fish fed aggressively regardless of
Anterior Kidney the concentration of 17MT in the feed offered, behavior was
Protein in blood 5× None considered normal with no dose-related differences detected,
Heart and no mortality was observed in the 3× treatment group.
Degeneration 3×, 5× None A reason that the margin of safety was not recommended to
Protein in blood 3×, 5× None extend the standard dosage to 5× was the presence of protein
in blood of the gill, heart, liver, and kidney and proteinaceous
 SAFETY OF 17MT TO LARVAL TILAPIA 217
Downloaded by [Department Of Fisheries] at 23:21 19 March 2013

FIGURE 3. Ventricle (top), bulbous (left), and atrium (bottom) chambers of the heart of the 5× treatment group in the study on the safety of 17MT to larval Nile
Tilapia. [Figure available in color online.]

material in gill tissue. These lesions were observed in some
fish in all treatment groups and were considered 17MT-related,
as this finding was not present in the 0× group (control) and
the severity of this finding increased in a dose-dependent man-
ner. However, the differences among pathological lesions in the
0× and 5× groups were not significant. Some of the observed
lesions (primarily in the gill and liver) were attributed to the
elevated concentrations of TAN and NO2 in the water during
the last 9 d of the study.
In the present study, nontreated control feed (i.e., 0×) was
administered to fish in the surrogate aquaria; the mean weight
of fish in these aquaria was used for the adjustment of feed in
the test aquaria. Based on the significant growth difference of
the fish in the different treatment groups, the use of the mean
weight of fish in the surrogate aquaria for the calculation of
feed for the test aquaria resulted in a slightly different amount
given to test fish than required. Feed amounts were not adjusted
for mortality that occurred during the week for which feed had
been aliquoted, resulting in the 5× group being offered more
feed (based on weight). Based on the mean weight of fish in the
surrogate tanks on the final day of the study (1.69 ± 0.11 g),
the larger fish from the 0×, 1× and 3× groups were offered less
feed; this weight difference could be attributed to the treated
fish not being stressed by handling as were the surrogates.
Based on a mean water temperature of 28◦ C and pH of 8.3,
approximately 12.5% of the total TAN in an aqueous solution is

TABLE 2. Mean concentrations of 17MT in the treated feeds and analytical results for each dose in the study on the safety of 17MT to larval Nile Tilapia.

Treatment Target feed Actual feed Target dose Actual dose

group concentrationa concentrationa Difference to fishb to fishb Difference
1× 60 52.8 −11.9% 9 7.9 −12.2%
3× 180 160.3 −10.9% 27 24.1 −10.7%
5× 300 279.8 −6.7% 45 42.0 −6.7%
a
µg 17MT/g feed.
b
mg 17MT/kg fish daily.
 218 STRAUS ET AL.
un-ionized ammonia (NH3 ; Avault 1996). As a result, during the
last 9 d of the study, mean daily concentrations of NH3 ranged
from 0.04 to 0.84 mg/L. The overall concentration of NH3 in
all test tanks during this period was 0.37 mg/L, and the mean
concentration in each of the four treatment groups ranged from
0.32 (3× group; 2.59 mg/L TAN) to 0.40 mg/L (1× group; 3.23
mg/L TAN) NH3 .
Although the NH3 and NO2 concentrations were elevated,
others have found the concentrations were within a range tol-
erated by tilapia. Redner and Stickney (1979) found that the
48-h LC50 for Blue Tilapia O. aureus was 2.46 mg/L NH3 ,
and in fish acclimated to low concentrations of NH3 , 100% sur-
vival was achieved when fish were exposed to 3.4 mg/L NH3
for 48 h. Daud et al. (1988) reported that the NH3 LC50 val-
ues at 48, 72, and 96 h in Red Tilapia (O. mossambicus × O.
niloticus hybrids) fry were 6.6, 4.1, and 2.9 mg/L, respectively.
Atwood et al. (2001) found that the 96-h LC50 value for small
Downloaded by [Department Of Fisheries] at 23:21 19 March 2013
Nile Tilapia (4.4 g) was 81 mg/L NH3 , but that the 96-h LC50
value for large Nile Tilapia (90.7 g) was 8 mg/L NH3 . However,
El-Shafai et al. (2004) found that concentrations of NH3 rang-
ing from 0.07 to 0.14 mg/L negatively affected growth of Nile
Tilapia fed fresh duckweed. Some of the histological findings
of note (e.g., gill edema and proliferation) were also observed
in fish from the 0× or 1× treatment groups of the present study
and were therefore not considered to be 17MT related.
During the last 9 d of the study, mean daily concentrations
of NO2 ranged from 1.4 to 4.1 mg/L. The overall concentra-
tion of NO2 during this period was 2.88 mg/L, and the mean
concentration in each of the four treatment groups ranged from
2.41 (3× group) to 3.33 mg/L (1× group). In future studies,
procedures should be established to minimize buildup of NH3
and NO2 concentrations, such as transferring fish from smaller
aquaria to larger aquaria, using filtering systems designed to
remove nitrogen compounds, or exchanging aquaria water with
fresh water more frequently.
The heart may be a target organ of 17MT pathologies in-
cluding lysis and hypertrophy of cardiac cells (degeneration
and necrosis), which could lead to decreased cardiac function.
In the present study, lysis of muscle cells suggests that the
heart is a target tissue of 17MT, and damage was most severe
in the atrium of the heart (Figure 1). Such findings would be
consistent with one of the manifestations of methyltestosterone
use or abuse in humans. Given that the heart damage was also
treatment-related and dose-dependent in the present study, it
was suspected that the intravascular eosinophilic material might
have been myoglobin, a muscle breakdown protein (Jeff Wolf,
Pathology Manager and Veterinary Pathologist, Experimental
Pathology Laboratories, Sterling, Virginia, personal communi-
cation). It was conjectured that the material might have been
some sort of lipoprotein, possibly linked to the liver’s metabolic
changes due to the presence of 17MT (Hugh Ferguson, Profes-
sor of Pathology, St. George’s University, Grenada, West Indies,
personal communication). Hornung et al. (2004) reported that
17MT aromatizes to 17α-methylestradiol and contributes to the
estrogenic effects in Fathead Minnow Pimephales promelas af-
ter exposure to the androgen, resulting in significant vitellogenin
production in exposed fish.
Gill epithelial proliferation was observed and is often induced
by exposure to irritants (e.g., ammonia) or is known to be an
artifact of sampling. In this case, elevated TAN and NO2 levels
near the end of the study might have increased susceptibility of
moribund fish to bacterial colonization.
This study is considered to be a valid test of the safety of
9 and 27 mg 17MT/kg fish daily administered in feed to Nile
Tilapia for 28 d. In spite of the elevated TAN levels measured
during the latter stages of the study, the testing was conducted
under environmental conditions that were acceptable for rearing
tilapia. Fish used to establish the reference population were
larval Nile Tilapia that were approximately 10 d old (about
the time when treatment would be started in industry to produce
predominantly male populations of fish), and the 0-mg 17MT/kg
fish daily test fish served as satisfactory controls.
ACKNOWLEDGMENTS
We thank Damon Seawright from AmeriCulture for sharing
his insights on tilapia production. Thanks to Ben Beck, Cather-
ine Childress, and Ken Davis who provided critical reviews of
the manuscript. Mention of trade names or commercial products
in this article is solely for the purpose of providing specific in-
formation and does not imply recommendation or endorsement
by the U.S. Department of Agriculture (USDA) or the U.S Fish
and Wildlife Service (USFWS). The USDA and USFWS are
equal opportunity providers and employers.
REFERENCES
APHA (American Public Health Association). 2005. Standard methods for the
examination of water and wastewater, 21st edition. APHA, in conjunction
with the American Water Works Association and the Water Environment
Federation, Washington, D.C.
Atwood, H. L., Q. C. Fontenot, J. R. Tomasso, and J. J. Isely. 2001. Toxicity of
nitrite to Nile Tilapia: effect of fish size and environmental chloride. North
American Journal of Aquaculture 63:49–51.
Avault, J. W., Jr. 1996. Fundamentals of aquaculture: a step-by-step guide to
commercial aquaculture. AVA Publishing, Baton Rouge, Louisiana.
Daud, S. K., A. Hasbollah, and A. T. Law. 1988. Effect of unionized ammonia
on Red Tilapia (Oreochromis mossambicus × O. niloticus hybrid) fry. Pages
411–414 in R. S. V. Pullin, T. Bhukaswan, K. Tonguthai, and J. L. Maclean,
editors. The second international symposium on tilapia in aquaculture: con-
ference proceedings 15. International Center for Living Aquatic Resources
Managment, Manila.
Donaldson, E. M., and G. A. Hunter. 1982. Sex control in fish with particular
reference to salmonids. Canadian Journal of Fisheries and Aquatic Sciences
39:99–110.
El-Shafai, S. A., F. A. El-Gohary, F. A. Nasr, N. P. van der Steen, and H. J.
Gijzen. 2004. Chronic ammonia toxicity to duckweed-fed tilapia
(Oreochromis niloticus). Aquaculture 232:117–127.
Gaikowski, M. P., J. C. Wolf, R. G. Endris, and W. H. Gingerich. 2003. Safety
of Aquaflor (florfenicol, 50% type A medicated article), administered in feed
to Channel Catfish, Ictalurus punctatus. Toxicologic Pathology 31:689–697.
Green, B. W., K. L. Veverica, and M. S. Fitzpatrick. 1997. Fry and fingerling
production. Pages 215–243 in H. S. Egna and C. E. Boyd, editors. Dynamics
of pond aquaculture. CRC Press, Boca Raton, Florida.
 SAFETY OF 17MT TO LARVAL TILAPIA
Greenlees, K. J. 1997. Laboratory studies for the approval of aquaculture drugs.
Progressive Fish-Culturist 59:141–148.
Hornung, M. W., K. M. Jensen, J. J. Korte, M. D. Kahl, E. J. Durhan, J. S.
Denny, T. R. Henry, and G. T. Ankley. 2004. Mechanistic basis for estrogenic
effects in Fathead Minnow (Pimephales promelas) following exposure to the
androgen 17α-methyltestosterone: conversion of 17α-methyltestosterone to
17α-methylestradiol. Aquatic Toxicology 66:15–23.
Hunter, G. A., and E. M. Donaldson. 1983. Hormonal sex control and its
application to fish culture. Pages 223–303 in W. S. Hoar, D. J. Randall, and
E. M. Donaldson, editors. Fish physiology, volume 9, part B: reproduction—
behavior and fertility control. Academic Press, New York.
Rani, A., and D. J. Macintosh. 1997. An evaluation of the effects of hor-
mone concentration, treatment period, feeding regime, and rearing salinity
on production of all-male Nile Tilapia (Oreochromis niloticus) fry using
17 α-methyltestosterone. Pages 791–802 in K. Fitzsimmons, editor. Pro-
ceedings from the fourth international symposium on tilapia in aquaculture.
Downloaded by [Department Of Fisheries] at 23:21 19 March 2013
219
Northeast Regional Agricultural Engineering Service, NRAES-106, Ithaca,
New York.
Redner, B. D., and R. R. Stickney. 1979. Acclimation to ammonia by Tilapia
aurea. Transactions of the American Fisheries Society 108:383–388.
Straus, D. L., J. D. Bowker, M. P. Bowman, D. Carty, A. J. Mitchell, and B.
D. Farmer. 2012. Safety of Aquaflor-medicated feed to Sunshine Bass. North
American Journal of Aquaculture 74:1–7.
Teichert-Coddington, D., B. Manning, J. Eya, and D. Brock. 2000. Concentra-
tion of 17α-methyltestosterone in hormone treated feed: effects of analytical
technique, fabrication, and storage temperature. Journal of the World Aqua-
culture Society 31:42–50.
Yamamoto, T. O. 1969. Sex differentiation. Pages 117–175 in W. S. Hoar and
D. J. Randall, editors. Fish physiology, volume 3: reproduction and growth
bioluminescence, pigments, and poisons. Academic Press, New York.
Yamazaki, F. 1983. Sex control and manipulation in fish. Aquaculture 33:329–
354.
This article was downloaded by: [Department Of Fisheries]
On: 19 March 2013, At: 23:22
Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,
37-41 Mortimer Street, London W1T 3JH, UK

North American Journal of Aquaculture

Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/unaj20

Ontogenetic Development of the Digestive Tract in

Larvae of American Shad
a b a a a a
X. Y. Hong , X. P. Zhu , K. C. Chen , D. B. Pan & K. B. Li
a
Pearl River Fisheries Research Institute, Chinese Academy of Fishery Sciences, Guangzhou,
510380, China
b
College of Fisheries and Life Science, Shanghai Ocean University, Shanghai, 201306, China
Version of record first published: 05 Mar 2013.

To cite this article: X. Y. Hong , X. P. Zhu , K. C. Chen , D. B. Pan & K. B. Li (2013): Ontogenetic Development of the Digestive
Tract in Larvae of American Shad, North American Journal of Aquaculture, 75:2, 220-227

To link to this article: http://dx.doi.org/10.1080/15222055.2012.761166

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions

This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to
anyone is expressly forbidden.

The publisher does not give any warranty express or implied or make any representation that the contents
will be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses should
be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims,
proceedings, demand, or costs or damages whatsoever or howsoever caused arising directly or indirectly in
connection with or arising out of the use of this material.
 North American Journal of Aquaculture 75:220–227, 2013

C American Fisheries Society 2013

ISSN: 1522-2055 print / 1548-8454 online

DOI: 10.1080/15222055.2012.761166

COMMUNICATION

Ontogenetic Development of the Digestive Tract in Larvae

of American Shad
X. Y. Hong
Pearl River Fisheries Research Institute, Chinese Academy of Fishery Sciences,
Guangzhou 510380, China

X. P. Zhu*
Pearl River Fisheries Research Institute, Chinese Academy of Fishery Sciences, Guangzhou 510380,
China; and College of Fisheries and Life Science, Shanghai Ocean University, Shanghai 201306, China
Downloaded by [Department Of Fisheries] at 23:22 19 March 2013

K. C. Chen, D. B. Pan, and K. B. Li

Pearl River Fisheries Research Institute, Chinese Academy of Fishery Sciences,
Guangzhou 510380, China

we suggest initiating the weaning process at 23–25 DAH in the

Abstract
The American Shad Alosa sapidissima was introduced into
China for aquaculture in 2003. The histological characteristics of
the digestive tract of American Shad larvae were examined from
hatching to 33 d after hatching (DAH) with the aim of cultivat-
ing the larvae efficiently. The fertilized American Shad eggs were
imported from Oregon and incubated in barrels with spring wa-
ter maintained at 20.3–21.9◦ C in Qingyuan, Guangdong Province,
China. The larvae were transferred to cement pools (water tem-
perature at 24.0–25.0◦ C) at 13 DAH. Larvae were fed Artemia
nauplii starting at 3 DAH, Artemia nauplii supplemented with
gradually increasing proportions of commercial formulated feed
after 18 DAH, and then only the formulated feed by 30 DAH.
Based on the source of nutrition and structural changes in the
digestive tract, larval development was divided into three stages:
(1) lecitotrophic (0–2 DAH), (2) lecitoexotrophic (3 DAH), and (3)
exotrophic (4–33 DAH). At 0 DAH of the lecitotrophic stage, the
digestive system was undifferentiated and contained a large aci-
dophilic yolk sac. By 1 DAH, the mouth and anus opened, the liver
appeared, and the buccopharynx and intestine began to differen-
tiate. The lecitoexotrophic stage began at the start of exogenous
feeding, which led to the disappearance of the yolk sac. During
the exotrophic stage, the esophagus could be differentiated into
two different regions by 4 DAH. The intestinal tract developed
quickly and differentiated into an obvious foregut and hindgut
by 11 DAH. The primary stomach formed by 8 DAH and com-
pletely developed by 19 DAH with the presence of a few gas-
tric glands. By 25 DAH, many gastric glands and a number of
pyloric ceca were observed. The results of the study show the
quick development of the digestive system in American Shad and
larvae.
The formation and development of the digestive system in
fish is required for proper ingestion, digestion, and assimilation
of food. Thus, this developmental period is critical during the
early life of a fish and impacts growth and survival. During the
early stages of development, fish larvae eat live prey but can
be transitioned to a formulated diet, as demonstrated in Persian
Sturgeon larvae of Acipenser persicus (Shakourian et al. 2011).
Delayed first feeding causes significant injury to the organiza-
tional structure of the digestive system (Chen et al. 2007), and a
proper transition to a formulated diet is important in fish larval
rearing (Sánchez-Amaya et al. 2007). Histological and biochem-
ical research on the ontogeny of the fish’s digestive system may
provide important information for establishing sound rearing
methods that improve commercial larval rearing (Zaiss et al.
2006). Indeed, there have been many previous histological stud-
ies of the development of the digestive tract in other fish species,
such as the Atlantic Cod Gadus morhua (Kjørsvik et al. 1991),
California Halibut Paralichthys californicus (Gisbert et al.
2004), Yellowtail Kingfish Seriola lalandi (Chen et al. 2006),
Sterlet Acipenser ruthenus (Wegner et al. 2009), and Yellow
Catfish Pelteobagrus fulvidraco (Yang et al. 2010).
The American Shad Alosa sapidissima is the largest anadro-
mous fish of the family Clupeidae (herrings) and is amenable

*Corresponding author: zhuxinping 1964@yahoo.com.cn

Received January 3, 2012; accepted December 17, 2012

220
 COMMUNICATION
to introduction outside its native habitat. The highly revered
Chinese Reeves Shad Tenualosa reevesii is similar to the
American Shad in appearance and taste; however, it is endan-
gered (Wang 1996, 1998). Therefore, American Shad has been
raised in aquaculture in China since 2003. American Shad aqua-
culture was initially unsuccessful due to the low survival rate of
larvae during rearing (Du et al. 2006). In general, larval rearing
is the major bottleneck to effective aquaculture.
Previous studies have documented the effects of ecological
factors, such as salinity (Chittenden 1973; Limburg and Ross
1995) and water temperature (Chittenden 1972), on the growth
and mortality of American Shad larvae. Additionally, the early
development of American Shad embryos and larvae has been
studied (Shardo 1995; Hong et al. 2011a), but the ontogenetic
development of the digestive system has not been described.
Furthermore, a feeding strategy that correlates with morpholog-
ical development has not been developed. Thus, to enhance the
Downloaded by [Department Of Fisheries] at 23:22 19 March 2013
success of American Shad larval rearing, it is necessary to under-
stand the development of its digestive system. In this paper, we
describe the morphological development of the larval digestive
system of American Shad between hatching and metamorpho-
sis. It is our hope that the results presented here will benefit the
American Shad aquaculture practice in China and elsewhere.
METHODS
Rearing of fertilized eggs and larvae.—The fertilized eggs
were collected from the lower Columbia River in Oregon with
the help of local fishers in June 2008. All eggs were incu-
bated in three 3.5-m3, flow-through, cylindrical–conical tanks
at the fish farm (24.61◦ N, 112.58◦ E) supplied with water from a
mountain spring (yearly water temperature range, 18.2–19.0◦ C)
in Qingyuan, Guangdong Province, China. Egg density was
5,000–6,000 eggs/m3. The water was filtered through a 25-µm-
mesh filter, and the water exchange rate was 20–30 L/min during
the incubation period. Water quality conditions were monitored
daily in the incubation tanks and were as follows: dissolved
oxygen, 7.2–8.8 mg/L; pH, 7.7–7.9; water temperature, 20.3–
21.9◦ C. At 13 d after hatching (DAH), the larvae were trans-
ferred from tanks to three cement ponds (3.5 × 3 × 1 m) with
six air stones in each. The water temperature in the cement
ponds was 24.0–25.0◦ C, which was controlled by the incoming
spring water. The density of larvae in ponds was 1,000–1,200
individuals/m3. Eggs and fish were on a 16 h light : 8 h dark pho-
toperiod. The tanks and ponds were cleaned daily by siphoning.
Larvae were fed with Artemia nauplii (O.S.I. brine shrimp
eggs, Ocean Star International, Snowville, Utah) at a rate of 10–
15 individuals/mL, five times per day from 0700 to 2200 hours
beginning at 3 DAH. After 12 DAH, they were fed six times
daily. All the Artemia nauplii were newly hatched in artificial
salt water at a salinity 25–28‰ after 28–32 h of aeration. After
18 DAH, commercial formulated feed (diameter, 100–200 µm;
Shangdong shengsuo Fishery Culture Feed Research Center,
China) was gradually added to their food before feeding Artemia
221
TABLE 1. Gross composition (%) of the compound diet used to feed Ameri-
can Shad.
Nutrient component Portion (%)
Crude protein ≥50%
Crude fat ≥8%
Crude fiber ≤3%
Crude ash ≥16.5%
Calcium ≥5%
Phosphorus ≥1%
Water content ≤12%
Lysine ≥2%
nauplii while the quantity of Artemia nauplii was gradually
reduced. By 30 DAH, they were fed a diet only of formulated
feed (diameter, 150–250 µm) (Table 1).
About 1,000 eggs were sampled to determine the hatch rate.
Hatch rate was expressed as the percentage of visible larvae
hatched from fertilized eggs. Survival rate was expressed as
S33 /(D0−33 + S33 ), where the S33 is the total quantity of survival
larvae at 33 DAH in the three tanks and D0−33 is the cumulative
total quantity of dead larvae during the rearing progress.
Fish sampling and growth measurements.—For morpholog-
ical and histological analysis, American Shad larvae were col-
lected (10 specimens from each tank or pond, randomly) daily
from 0 to 15 DAH and then only on days 17, 19, 21, 23, 25, 27,
30, 33. Sample larvae were euthanized in a solution of tricaine
methanesulfonate (MS-222). The TL (i.e., from the tip of the
snout to the posterior margin of the body) of 30 specimens was
measured to the nearest 0.1 mm using an ocular micrometer with
the aid of a dissecting microscope and Vernier calipers. Growth
in length was assessed by measuring the absolute growth rate
(AGR) as millimeters per day and specific growth rate (SGR) as
percent increase per day (Hopkins 1992). Equations used were
AGR = (TLfinal − TLinitial )/d,
and
SGR = {[loge (TLfinal ) − loge (TLinitial )]/d} × 100,
where the TLfinal and TLinitial are the final and initial fish TL
(millimeters), respectively, and d is the time interval (days)
between samples.
Histological analysis.—Ten specimens were randomly col-
lected and preserved in Bouin’s solution for standard histolog-
ical procedures. Larvae were embedded in paraffin wax and
serially cut into 5-µm sections along the craniocaudal axis us-
ing a rotational microtome (Leica RM 2016). Sections were then
stained with hematoxylin and eosin (H&E) and photographed
under a microscope (Nikon Eclipse E60) equipped with a digital
camera (Nikon DXM 1200).
 222 HONG ET AL.
Downloaded by [Department Of Fisheries] at 23:22 19 March 2013

FIGURE 1. Total length of American Shad from 0 to 33 d after hatching.

Developmental phases and diets are indicated.

RESULTS
Fish Growth
The average length of larval fish increased from 8.56 mm
TL (SD, 0.36) at hatching to 33.10 mm (SD, 1.58) by 33 DAH
(Figure 1). During this period, the AGR and the SGR were
0.76 mm/d (SD, 0.47) and 5.01 %/d (SD, 3.57) respectively,
and the hatch and survival rates were about 81.6% and 88.0%,
respectively.

Buccopharynx
At hatching, the buccopharynx was closed and the digestive
tract consisted of a simple straight tube. Both the mouth and
anus were closed. The anterior half of the body was dorsal to
a large acidophilic yolk sac that contained no oil droplets. At 1
DAH, we observed that the mouth and anus had both opened.
We also noted that the buccopharynx was formed and lined by
a single layer of squamous epithelium (Figures 2a, 3a).
Larvae began to feed on Artemia nauplii at the beginning of
3 DAH. The yolk sac was almost completely absorbed by the
end of 3 DAH. Additionally, the buccopharyngeal cavity had
expanded, the gill arch was differentiated, and a small number
of goblet cells were distinguishable on 3 DAH. By 4 DAH,
the buccopharynx had expanded towards the posterior part of
the body, and an increase in the size and quantity of goblet
cells was observed (Figure 3b). At 8 DAH, the taste buds, the
pseudobranch, four pairs of gill arches, and the lamella were
all formed. Also by 8 DAH, the wall of the buccopharynx in-
cluded a mucosa, a submucosa, a muscular layer, and adventitia
(Figure 3d). Goblet cells were distributed in the buccopharynx
epithelial area at 19 DAH (Figure 2d).
Esophagus
At hatching, the esophagus was a wire-like tube indistin-
guishable from the rest of the digestive tract (Figure 3a). At 3
FIGURE 2. Microsections of the digestive tract during the development of
American Shad larvae. (a) 1 d after hatching (DAH); note the large acidophilic
yolk sac and the relatively smaller body. (b) 4 DAH. (c) 13 DAH; note the
posterior esophagus that has expanded with lots of mucus cells. (d) 19 DAH;
the stomach has formed with few gastric glands, the esophagus is well developed,
and a pseudobranch, gill arch, and lamella are present. (e) 25 DAH; note the
number of gastric glands and the shape of the stomach. BC, buccopharyngeal
cavity; CS, cardiac stomach; DT, digestive tract; E, eye; FS, fundic stomach;
GA, gill arch; GG, gastric glands; H, heart; I, intestine; L, liver; LA, lamella;
M, muscle; MC, mucous cell; N, notochord; OE, esophagus; P, pancreas; PC,
pyloric ceca; PB, pseudobranch; PS, pyloric stomach; SB, swim (gas) bladder;
YS, yolk sac. [Figure available in color online.]
DAH, the esophageal cavity was open, effectively connecting
the cavum pharyngis to the intestine. A pseudostratified cuboidal
epithelium lining the esophagus and goblet cells in the mucosa
were observed (Figure 2a). At 4 DAH, the anterior epithelium
of esophagus was a stratified squamous epithelium densely
covered with goblet cells. Conversely, the posterior region with
no goblet cells was lined by a simple columnar epithelium (Fig-
ure 3b). During development, the anterior stratified squamous
 COMMUNICATION 223
Downloaded by [Department Of Fisheries] at 23:22 19 March 2013

FIGURE 3. Sagittal section of the digestive tract during the development of American Shad larvae. (a) 1 d after hatching (DAH); note the large acidophilic yolk
sac compared with the small body. (b) 4 DAH. (c) 5 DAH; the intestine was lined by a single columnar epithelium. (d) 8 DAH; note the obvious differences
between the esophagus and intestine. (e) 11 DAH; epithelial cells of the hindgut were filled with acidophilic supranuclear bodies (arrow), and the fore intestine
epithelium was filled with a few goblet cells (arrowheads). (f) General view of the intestine in 23 DAH American Shad larvae; both the hindgut and the foregut
intestine epithelium had a few goblet cells (arrowhead). BC, buccopharyngeal cavity; DT, digestive tract; E, eye; FI, foregut intestine; GA, gill arch; H, heart; HG,
hindgut; I, intestine; L, liver; LA, lamella; M, muscle; N, notochord; OE, esophagus; SB, swim (gas) bladder; YS, yolk sac. [Figure available in color online.]

epithelium extended posteriorly (Figure 3d) towards the liver, Intestine

while the posterior esophagus expanded to form the stomach At 1 DAH, the intestine was a tube lined by a simple ciliated
(Figure 2c). By 25 DAH, longitudinal folds formed and goblet columnar epithelium (Figure 2a). By the end of 3 DAH, the in-
cells accumulated between the esophagus and stomach (Fig- testine lengthened and thickened rapidly to reach approximately
ure 2e), and no other distinct histological changes were apparent. two-thirds the length of the yolk sac (Figure 3a).
 224
By 4 DAH, the intestinal valve between the esophagus and
foregut intestine was visible, even though the intestine was only
a simple columnar epithelium with no distinct subsections (Fig-
ure 3c). As the fish grew, the intestinal tract became larger and
had more folds. By 11 DAH, the intestine was clearly divided
into foregut and hindgut by one intestinal valve (Figure 3e). The
foregut intestine was covered by a single layer of columnar ep-
ithelium with a few goblet cells. The first half of the hindgut was
similar to the foregut, but the posterior half showed numerous
large acidophilic supranuclear vacuoles occupying most of the
cytoplasmic space. The enterocytes were surrounded by a sin-
gle layer of columnar epithelium with a few goblet cells but no
supranuclear vacuoles at 23 DAH (Figure 3f). Many pyloric ceca
were observed at the boundary between the pyloric stomach and
intestine at 25 DAH (Figure 2e). During further development,
the pyloric ceca differentiated into a simple columnar epithelium
with goblet cells. However, by 33 DAH, the pyloric epithelium
Downloaded by [Department Of Fisheries] at 23:22 19 March 2013
no longer contained goblet cells, although it had continued to
grow.
Stomach
The stomach was the final organ in the digestive tract to dif-
ferentiate. Before 8 DAH, the stomach was just a spindly tube
FIGURE 4. Sagittal section of the liver and pancreas in American Shad during ontogeny. (a) 3 d after hatching (DAH), (b) 9 DAH, (c) 19 DAH, and (d) 25 DAH.
GB, gall bladder; I, intestine; L, liver; M, muscle; P, pancreas; PI, pancreatic islet. [Figure available in color online.]
HONG ET AL.
that did not have an obvious boundary with the esophagus and
intestine. At 8 DAH, the primary stomach appeared posterior
to the esophagus and was lined by simple columnar epithelium
(Figure 3d). By 19 DAH, the stomach obtained its characteristic
shape but occupied a smaller cavity than the intestine (Fig-
ure 2d). At this time, the cardiac, fundic, and pyloric portions of
the stomach were distinguishable. The majority of gastric glands
were located in the fundic region of the stomach, whereas goblet
cells were restricted to the cardiac portion of the stomach.
At 25 DAH, the gastric cavity was much larger (Figure 2e),
and the stomach took on a “Y” shape. The fundic stomach elon-
gated to form the largest portion of the stomach and contained
many gastric glands cells. As the stomach grew, no goblet cells
were observed in the stomach epithelium at 33 DAH.
Liver
At 1 DAH, the liver was a small cluster of cells adjacent
to the early intestine, behind the yolk sac (Figure 2a). At 3
DAH, we observed the hepatic cells as basophilic polygonal
cells with distinct nucleoli that were arranged around hepatic
sinusoids. At this time, the cytoplasm of hepatic cells was mostly
occupied by lipid vacuoles (Figures 2b, 4a). Blood cells within
the hepatic sinusoids were observed by 3 DAH (Figure 4a). As
 COMMUNICATION
Downloaded by [Department Of Fisheries] at 23:22 19 March 2013
FIGURE 5. Summary of the major histological and morphological developments in American Shad from initial hatching to 33 d after hatching.
hepatic cell numbers increased, so did the lipid vacuoles within
them (Figure 4b, c). However, at 25 DAH the lipid vacuoles
had disappeared and the hepatic cells adopted a regular shape
(Figure 4d).
PANCREAS
The pancreas appeared later than the liver at 1 DAH but at-
tained almost the same cross-sectional area by 3 DAH. It was
located dorsal to the liver and was composed of basophilic cells
with acidophilus zymogen granules (Figures 2b, 4a). The pan-
creas grew, and by 9 DAH the Islets of Langerhans were clearly
visible. The cells that constituted the Langerhans islets stained
slightly, and the abundance of capillaries spread in the islets.
By 25 DAH, the basophilic cells proximal to gland cells and
the acidophilus cells in the upper apex weredistinctly different
(Figure 4d).
The major histological and morphological development of
the digestive system in American Shad is summarized in
Figure 5.
DISCUSSION
The ontogeny of the digestive tract of American Shad
followed the general pattern reported in the current literature
for most fish species. The American Shad larvae just had 1 d
in the lecitoexotrophic stage, and the peculiar developmental
characteristics of American Shad larvae were mainly related
to the development of stomach and intestine. This information
will be helpful in enhancing the success of rearing American
Shad larvae.
Similar to other teleost fishes (Buddington 1985; Bisbal and
Bengtson 1995; Chen et al. 2006), the development of the di-
225
gestive system in larvae of American Shad can be divided into
three stages according to sources of nutrition and developmen-
tal changes in the digestive tract: (1) lecitotrophic (0–2 DAH),
(2) lecitoexotrophic (3 DAH), and (3) exotrophic (4–33 DAH).
The American Shad larvae displayed a short lecitoexotrophic
phase, which is typically a critical period during the larval cul-
ture because fish larvae need mechanisms to obtain exogenous
nutrients (Segner et al. 1993). During this stage, the buccopha-
ryngeal cavity expanded, the esophageal cavity was open, goblet
cells appeared, and the liver and pancreas differentiated quickly.
These notable morphological changes of the digestive system
of American Shad larvae suggest they are ready to absorb and
digest exogenous foods.
The esophagus in American Shad can be subdivided into
two different regions at 4 DAH, similar to other fish species
such as the Yellow Catfish (Yang et al. 2010) and the Sterlet
(Wegner et al. 2009). The anterior part of the esophagus was
a stratified squamous epithelium with densely covered goblet
cells. The posterior region of the esophagus was lined by a
simple columnar epithelium with no goblet cells. Others have
hypothesized that esophageal goblet mucus cells function to lu-
bricate food and aid in digestion prior to the development of
the stomach (Murray 1994; Sire and Vernier 1995; Fiertak and
Kilarski 2002). Furthermore, Scocco et al. (1996) suggested that
mucus secreted by esophageal goblet cells protect the mucosa
from degradation by neutralizing gastric acid in addition to pre-
venting bacterial, physical, and chemical damage (Allen 1989).
Therefore, we conclude that the goblet cells in the anterior part
of the esophagus of American Shad larvae are important during
the early larval stage.
During the lecitotrophic stage, the intestine was a simple tube
that grew to be two-thirds of the body length by 3 DAH. As the
 226 HONG ET AL.
esophagus and stomach developed, the lumen of the intestine
enlarged while its length shrank to about half of the body length.
In other words, the reduction in intestine length relative to body
length reflects the differentiation of the stomach. This reduction
suggests a mechanism whereby we may be able to feed Ameri-
can Shad a formulated food based on the relative length between
intestine and body length. Additionally, American Shad larvae
with slender bodies did not inflate their gas bladder until 19
DAH; instead the larvae wagged their tails and swam in a circle
all of the time.
In most teleost fish, the enterocytes of the foregut intestine
are used for the absorption and storage of lipids (Tanaka 1971;
Kjørsvik et al. 1991; Elbal et al. 2004). However, the foregut
intestine of American Shad had mucous cells, but no lipid vac-
uoles, similar to Large Yellow Croaker Pseudosciaena crocea
(also known as Croceine Croaker Larimichthys crocea) (Mai
et al. 2005), California Halibut (Gisbert et al. 2004), and At-
Downloaded by [Department Of Fisheries] at 23:22 19 March 2013
lantic Halibut Hippoglossus hippoglossus (Luizi et al. 1999).
There are two possible explanations for this observation: (1) the
lipid content of feed did not exceed the fatty acid absorption and
exporting capacities of enterocytes, or (2) the food was passed
through the intestinal tract too rapidly for total absorption of
nutrients to occur (Gisbert et al. 2004). In Yellow Catfish, the
first explanation seemed to be more likely: Yellow Catfish are
able to transport lipids in the foregut intestine after feeding on
live rotifers (Yang et al. 2010).
After 13 DAH, the hindgut epithelium of American Shad
was filled with acidophilic supranuclear vacuoles indicating that
pinocytosis was an important mechanism of intracellular diges-
tion (Mai et al. 2005; Herrera et al. 2010). Supranuclear vacuoles
are indicative of intracellular digestion of proteins before the ap-
pearance of a functional stomach (Tanaka 1971). As the stomach
grew, the number of supranuclear vacuoles decreased. The liver
formed early on day 1 after hatching, and lipid vacuoles in the
liver were observed by 3 DAH. This observation highlights the
importance of lipid nutrition even before the complete develop-
ment of the digestive tract.
The gastric glands of American Shad appeared at 19 DAH,
indicating the beginning of the structural and functional stom-
ach. The digestive enzymes secreted by the gastric glands of the
stomach allow for the digestion of protein within the digestive
tract (Govoni et al. 1986; Cahu and Zambonino Infante 2001;
Chen et al. 2006) and increase digestive efficiency. But their
functionality may still be rudimentary because the presence of
supranuclear vacuoles in the hindgut suggests that the digestive
capabilities of American Shad larvae are limited (Önal et al.
2010). The appearance of a gastric gland and pyloric ceca is
accepted as a histological indicator of the transition from larvae
to juvenile (Baglole et al. 1997). The number of pyloric ceca in
adult American Shad reached 100 (Hong et al. 2011b), which
helps them to adjust to new environments and food sources.
A high level of digestive enzyme activity was been detected
in the pyloric ceca of Trahira Hoplias malabaricus (Alencar
et al. 2003) and Tambaqui Colossoma macropomum (de Souza
Bezerra et al. 2000), indicating that the pyloric ceca had strong
digestive functions.
The gastric gland and pyloric ceca had an important directive
function in larval rearing (Tanaka 1971; Segner et al. 1993). At
19 DAH, both the formulated feed and Artemia nauplii were fed
simultaneously to American Shad with a transition to only the
formulated feed by 30 DAH. Önal et al. (2010) suggested that
formulated feed could be used exclusively after the supranu-
clear vacuoles disappear. However, Howey (1985) believed that
formulated feed should not be fed to American Shad until TL
reaches 25 mm. We observed that the supranuclear vacuoles
disappeared in the hindgut at 23 DAH; the average length of
American shad was 24.56 mm (SD, 0.56) at 25 DAH when
the gastric gland appeared and pyloric ceca were numerous.
Thus, we suggest initiating the weaning process at 23–25 DAH
(24.5 mm TL) in American Shad larvae.
In general, ontogeny of the digestive tract of American Shad
larvae occurred within the first month of development. The time
span of developmental phase 2 was relatively short, and the
weaning time may be between 23 and 25 DAH. To further
refine larval culture protocols, further work should focus on
combining histological techniques with analysis of digestive
enzyme activity to provide more detailed information on the
temporal development of the digestive tract in American Shad.
ACKNOWLEDGMENTS
This work was financed by the Science and Technology Pro-
gram of Guangdong Province (A200901E03).
REFERENCES
Alencar, R. B., M. M. Biondi, P. M. G. Paiva, V. L. A. Vieira, L. B. Carvalho Jr.,
and R. de Souza Bezerra. 2003. Alkaline proteases from the digestive tract of
four tropical fishes. Brazilian Journal of Food Technology 6:279–284.
Allen, A. 1989. Gastrointestinal mucus. Pages 359–382 in S. G. Schultz, J. G.
Forte, and B. B. Rauner, editors. Handbook of physiology, section 6: the
gastrointestinal system—volume 3, salivary, gastric, pancreatic, and hepato-
biliary secretion. American Physiological Society, Bethesda, Maryland.
Baglole, C. J., H. M. Murray, G. P. Goff, and G. M. Wright. 1997. Ontogeny
of the digestive tract during larval development of Yellowtail Flounder: a
light microscopic and mucous histochemical study. Journal of Fish Biology
51:120–134.
Bisbal, G. A., and D. A. Bengtson. 1995. Development of the digestive tract in
larval Summer Flounder. Journal of Fish Biology 47:277–291.
Buddington, R. K. 1985. Digestive secretions of Lake Sturgeon, Acipenser
fulvescens, during early development. Journal of Fish Biology 26:715–723.
Cahu, C., and J. Zambonino Infante. 2001. Substitution of live food by formu-
lated diets in marine fish larvae. Aquaculture 200:161–180.
Chen, B. N., J. G. Qin, J. F. Carragher, S. M. Clarke, M. S. Kumar, and W.
G. Hutchinson. 2007. Deleterious effects of food restrictions in Yellowtail
Kingfish Seriola lalandi during early development. Aquaculture 271:326–
335.
Chen, B. N., J. G. Qin, M. S. Kumar, W. Hutchinson, and S. Clarke. 2006. On-
togenetic development of the digestive system in Yellowtail Kingfish Seriola
lalandi larvae. Aquaculture 256:489–501.
Chittenden, M. E., Jr. 1972. Responses of young American Shad, Alosa sapidis-
sima, to low temperatures. Transactions of the American Fisheries Society
101:680–685.
 COMMUNICATION
Chittenden, M. E., Jr. 1973. Salinity tolerance of young American Shad, Alosa
sapidissima. Chesapeake Science 14:207–210.
de Souza Bezerra, R., J. F. Dos Santos, M. A. da Silva Lino, V. L. A. Vieira,
and L. B. Carvalho Jr. 2000. Characterization of stomach and pyloric caeca
proteinases of Tambaqui (Colossoma macropomum). Journal of Food Bio-
chemistry 24:189–199.
Du, H., Q. Wei, F. Gan, J. Liu, X. Chen, and D. Yang. 2006. Changes in serum
cortisol and blood biochemical parameters after stress in American Shad.
Chinese Journal of Zoology 41:80–84.
Elbal, M. T., M. P. Garcı́a Hernández, M. T. Lozano, and B. Agulleiro. 2004.
Development of the digestive tract of Gilthead Sea Bream (Sparus aurata
L.): light and electron microscopic studies. Aquaculture 234:215–238.
Fiertak, A., and W. M. Kilarski. 2002. Glycoconjugates of the intestinal
goblet cells of four cyprinids. Cellular and Molecular Life Sciences 59:
1724–1733.
Gisbert, E., R. H. Piedrahita, and D. E. Conklin. 2004. Ontogenetic development
of the digestive system in California Halibut (Paralichthys californicus) with
notes on feeding practices. Aquaculture 232:455–470.
Govoni, J. J., G. W. Boehlert, and Y. Watanabe. 1986. The physiology of diges-
tion in fish larvae. Environmental Biology Fishes 16:59–77.
Downloaded by [Department Of Fisheries] at 23:22 19 March 2013
Herrera, M., I. Hachero-Cruzado, A. Naranjo, and J. M. Mancera. 2010. Organo-
genesis and histological development of the Wedge Sole Dicologoglossa
cuneata M. larva with special reference to the digestive system. Reviews in
Fish Biology and Fisheries 20:489–497.
Hong, X., X. Zhu, K. Chen, D. Pan, and K. Li. 2011a. Study on the develop-
ment of the embryo and larva of American Shad, Alosa sapidissima. Acta
Hydrobiologica Sinica 35:153–162.
Hong, X., X. Zhu, K. Chen, D. Pan, and K. Li. 2011b. Morphological character-
istics and karyotype of American Shad Alosa sapidissima. Journal of Dalian
Ocean University 26:180–184.
Hopkins, K. D. 1992. Reporting fish growth: a review of the basics. Journal of
the World Aquaculture Society 23:173–179.
Howey, R. G. 1985. Intensive culture of juvenile American Shad. Progressive
Fish-Culturist 47:203–212.
Jia, Y., Y. Chen, C. A. Goudie, B. A. Simco, and Q. Liu. 2007. Potential invasion
risk of the introduced American Shad Alosa sapidissima to aquatic ecosystem
in China. Acta Zoologica Sinica 53:625–629.
Kjørsvik, E., T. van der Meeren, H. Kryvi, J. Arnfinnson, and P. G. Kvenseth.
1991. Early development of the digestive tract of cod larvae, Gadus morhua
L., during start-feeding and starvation. Journal of Fish Biology 38:1–15.
Limburg, K. E., and R. M. Ross. 1995. Growth and mortality rates of lar-
val American Shad, Alosa sapidissima, at different salinities. Estuaries 18:
335–340.
Luizi, F. S., B. Gara, R. J. Shields, and N. R. Bromage. 1999. Further description
of the development of the digestive organs in Atlantic Halibut (Hippoglossus
hippoglossus) larvae, with notes on differential absorption of copepod and
Artemia prey. Aquaculture 176:101–116.
227
Mai, K., H. Yu, H. Ma, Q. Duan, E. Gisbert, J. L. Zambonino Infante, and
C. L. Cahu. 2005. A histological study on the development of the digestive
system of Pseudosciaena crocea larvae and juveniles. Journal of Fish Biology
67:1094–1106.
Murray, H. M., G. M. Wright, and G. P. Goff. 1994. A study of the posterior
esophagus in the Winter Flounder, Pleuronectes americanus, and the Yellow-
tail Flounder, Pleuronectes ferruginea: morphological evidence for pregastric
digestion? Canadian Journal of Zoology 72:1191–1198.
Önal, U., İ. Çelik, and Ş. Cirik. 2010. Histological development of digestive tract
in discus, Symphysodon spp. larvae. Aquaculture International 18:589–601.
Sánchez-Amaya, M. I., J. B. Ortı́z-Delgado, Á. Garcı́a-López, S. Cárdenas,
and C. Sarasquete. 2007. Larval ontogeny of Redbanded Seabream Pagrus
auriga Valenciennes, 1843, with special reference to the digestive system: a
histological and histochemical approach. Aquaculture 263:259–279.
Scocco, P., P. Ceccarelli, and G. Menghi. 1996. Glycohistochemistry of the
Tilapia spp. stomach. Journal of Fish Biology 49:584–593.
Segner, H., R. Rösch, J. Verreth, and U. Witt. 1993. Larval nutritional physiol-
ogy: studies with Clarias gariepinus, Coregonus lavaretus and Scophthalmus
maximus. Journal of the World Aquaculture Society 24:121–134.
Shakourian, M., M. Pourkazemi, M. A. Y. Sadati, M. H. S. Hassani, H. R.
Pourali, and U. Arshad. 2011. Effects of replacing live food with formulated
diets on growth and survival rates in Persian Sturgeon (Acipenser persicus)
larvae. Journal of Applied Ichthyology 27:771–774.
Shardo, J. D. 1995. Comparative embryology of teleostean fishes: I. development
and staging of the American Shad, Alosa sapidissima (Wilson, 1811). Journal
of Morphology 225:125–167.
Sire, M. F., and J. M. Vernier. 1995. Partial characterization of eosinophilic
granule cells (EGCs) and identification of mast cells of the intestinal lamina
propria in Rainbow Trout (Oncorhynchus mykiss): biochemical and cyto-
chemical study. Biology of the Cell 85:35–41.
Tanaka, M. 1971. Studies on the structure and function of the digestive system
in teleost larvae: III. development of the digestive system during postlarval
stage. Japanese Journal of Ichthyology 18:164–174.
Wang, H. 1996. Status and conservation of Reeves shad resources in China.
Naga, the ICLARM Quarterly 19:20–22.
Wang, S. 1998. Pisces. Pages 24–26 in P. Yue and Y. Chen, editors. China red
data book of endangered animals. Science Press, Beijing.
Wegner, A., T. Ostaszewska, and W. Rożek. 2009. The ontogenetic development
of the digestive tract and accessory glands of Sterlet (Acipenser ruthenus L.)
larvae during endogenous feeding. Reviews in Fish Biology and Fisheries
19:431–444.
Yang, R., C. Xie, Q. Fan, C. Gao, and L. Fang. 2010. Ontogeny of the digestive
tract in Yellow Catfish Pelteobagrus fulvidraco larvae. Aquaculture 302:
112–123.
Zaiss, M. M., I. E. Papadakis, E. Maingote, and P. Divanach. 2006. Ontogeny of
the digestive tract in Shi Drum (Umbrina cirrosa) reared using the mesocosm
larval rearing system. Aquaculture 260:357–368.
This article was downloaded by: [Department Of Fisheries]
On: 19 March 2013, At: 23:23
Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,
37-41 Mortimer Street, London W1T 3JH, UK

North American Journal of Aquaculture

Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/unaj20

Effects of Water Temperature, Photoperiod, Eyestalk

Ablation, and Non-Hormonal Treatments on Spawning
of Ovary-Mature Red Swamp Crayfish
a a a a
Shengli Liu , Shiyuan Gong , Jinmei Li & Wenhu Huang
a
Key Lab of Agricultural Animal Genetics, Breeding, and Reproduction of the Ministry of
Education, and College of Fishery, Huazhong Agricultural University, Wuhan, 430070, China
Version of record first published: 19 Mar 2013.

To cite this article: Shengli Liu , Shiyuan Gong , Jinmei Li & Wenhu Huang (2013): Effects of Water Temperature, Photoperiod,
Eyestalk Ablation, and Non-Hormonal Treatments on Spawning of Ovary-Mature Red Swamp Crayfish, North American Journal
of Aquaculture, 75:2, 228-234

To link to this article: http://dx.doi.org/10.1080/15222055.2012.746247

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions

This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to
anyone is expressly forbidden.

The publisher does not give any warranty express or implied or make any representation that the contents
will be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses should
be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims,
proceedings, demand, or costs or damages whatsoever or howsoever caused arising directly or indirectly in
connection with or arising out of the use of this material.
 North American Journal of Aquaculture 75:228–234, 2013

C American Fisheries Society 2013

ISSN: 1522-2055 print / 1548-8454 online

DOI: 10.1080/15222055.2012.746247

ARTICLE

Effects of Water Temperature, Photoperiod, Eyestalk

Ablation, and Non-Hormonal Treatments on Spawning
of Ovary-Mature Red Swamp Crayfish

Shengli Liu, Shiyuan Gong,* Jinmei Li, and Wenhu Huang

Key Lab of Agricultural Animal Genetics, Breeding, and Reproduction of the Ministry of Education,
and College of Fishery, Huazhong Agricultural University, Wuhan 430070, China
Downloaded by [Department Of Fisheries] at 23:23 19 March 2013

Abstract
This study evaluated the temperature and photoperiod requirements of spawning in ovary-mature red swamp
crayfish Procambarus clarkii and examined the feasibility of inducing spawning in them by inexpensive and practical,
nonhormonal treatments. In a spawning induction experiment lasting 20 d, four water temperature gradients (12–
14◦ C, 16–18◦ C, 20–22◦ C, 24–26◦ C) and four photoperiods (completely light [CL], 16 h light : 8 h dark [16L:8D],
12 h light : 12 h dark [12L:12D], and completely dark [CD]) were evaluated. Spawning in ovary-mature red swamp
crayfish could be observed in a relatively broad temperature range, from 16◦ C to 22◦ C. A period (5–14 d) of low
temperature (16–18◦ C) significantly induced spawning in ovary-mature females (p < 0.01), but hibernation began
to occur and spawning could not be observed when the temperature was below 14◦ C. Spawning occurred in both
CD and CL circumstances, but long-period light-groups showed a higher spawning rate—68.9 ± 10.2% (mean ±
SD) for CL and 69.6 ± 6.4% for 16L:8D—than that in short-period light-groups—53.3 ± 5.7% for 12L:12D and
33.3 ± 8.3% for CD (p < 0.05). In another spawning induction experiment lasting 3 d, eyestalk interventional
stimulating with a high concentration of Na + combined with a low concentration of Ca2 + induced spawning with a
high survival rate (SR = 95.8 ± 7.3% [mean ± SD]) in ovary-mature females. A high concentration of K + combined
with a high concentration of Ca2 + caused no spawning and a high mortality rate (SR = 20.8 ± 19.0% [mean ± SD]).
In this study we also observed that water removal or removing from water can be a stimulating factor of inducing fast
and synchronous spawning in ovary-mature crayfish, inducing 46 (16.0%) of 288 ovary-mature red swamp crayfish
spawned in 10–185 s.

Mass production of stocking seeds in red swamp crayfish
Procambarus clarkii is important for the resource utilization
of this species. The low fecundity, on average 223 young pro-
duced per female (Trimble and Gaudé 1988), makes it neces-
sary to induce maturation in a certain amount of captive females
and prompt them to spawn synchronously. The traditional tech-
nique of accelerating ovary development to induce synchronous
ovarian maturation in decapods is eyestalk ablation (Brown and
Jones 1949; Okumura and Aida 2001; Khazraeenia and Khazrai-
inia 2009; Varalakshmi and Reddy 2010). This is considered
to be one of the determinant factors for mass production of
quality stocking seeds in captive shrimp. As alternatives, in-
ducing ovary maturation via injection with serotonin (5-HT) or
dopamine (DA) antagonist, or a combined treatment with 5-
HT and DA antagonist, were extensively examined and showed
good prospects in practice (Vaca and Alfaro 2000; Chen et al.
2003; Alfaro et al. 2004; Tinikul et al. 2008). However, the
newly developed injection with 5-HT, DA antagonists, or both
is expensive, must be performed repeatedly, and is immature in
extensive application at this time. Eyestalk removal frequently
kills the females either at the time of surgery or after the oper-
ation, and leads to permanent damage as well as loss of seed
quality.
Furthermore, ovarian maturation generally does not cause
immediate spawning (herein, spawning and egg laying are syn-
onymous), although it is a prerequisite to spawning. In Northern

*Corresponding author: gsy@mail.hzau.edu.cn

Received February 6, 2012; accepted October 29, 2012

228
 INDUCING SPAWNING IN OVARY-MATURE RED SWAMP CRAYFISH
Crayfish Orconectes virilis, egg laying was proposed to occur
only if the crayfish were exposed to low temperature (4◦ C)
and complete darkness for 4–5 months followed by exposure
to spring temperature water (at least 12◦ C; Aiken 1969). The
conclusion is that while temperature and photoperiod interact to
control ovarian maturation, water temperature controls spawn-
ing. Similarly, in Spinycheek Crayfish O. limosus, it was found
that once the ovaries matured temperature alone stimulated
egg laying, photoperiod having no effect (Dubé and Portelance
1992). Flood regimes or “hydroperiods” (Gutiérrez-Yurrita
et al. 1999), and light requirements (Wurts and Stickney 1984)
were also proposed to be closely related to trigger egg laying
in past studies. Hence, techniques for inducing spawning in
ovary-mature females deserve to be developed to collect a
certain amount of brood crayfish carrying eggs. However,
in previous studies, little attention was paid to stimulating
synchronous spawning in ovary-mature female crayfish.
Downloaded by [Department Of Fisheries] at 23:23 19 March 2013
In this laboratory, we found that the ovarian maturation of
red swamp crayfish in the Wuhan area of China mostly occurred
during later September to earlier November but that spawning
often lasted to the next summer, which implied that the ovar-
ian maturation of crawfish in the same area was almost syn-
chronous but the spawning could be asynchronous (Gong et al.
2008). Moreover, we observed that the intense burrowing ac-
tivity and extensive sheltering in burrows during the spawning
period make the photoperiod and temperature requirements of
spawning in red swamp crayfish controversial and confusing
(authors, unpublished). We designed an investigation to find out
the photoperiod and temperature requirements of spawning in
ovary-mature crayfish. Meanwhile, we examined the feasibility
of inducing spawning with nonhormonal treatments. We tenta-
tively tested eyestalk interventional stimulation, removing the
crawfish from water, and water removal. For a better under-
standing of the spawning stimulation and for comparison, we
also tested widely accepted ovulation-inducing agents for fish,
using the recommended dosage, and also traditional eyestalk
ablation.
METHODS
Collection of red swamp crayfish.—Adult individuals (4:1,
female : male; 25–35 × g; 75–82 mm; males culled before ex-
perimentation) of red swamp crayfish were collected from an
aqua farm near Wuhan, China, in October and November (the
later period of the peak season of its ovarian maturation in this
area; Gong et al. 2008) of 2009 and 2010.
The ovary-mature red swamp crayfish used in this study
refers to females in a maturation stage of ovarian development
with dark brown oocytes whose diameter is greater than 1.4 mm,
as described before by Li and Zhao (1999), Ando and Makioka
(1998), and Kulkarni et al. (1991). Our sample, before selection,
was made up of females having the following characteristics:
weight, 25–35 × g; length (from the eye to the end of the telson),
75–82 mm; long and forficate oostegites (prepared for incuba-
229
tion); and thicker carapace with coarse yellow or red carapace
granules.
Maintenance of crayfish.—All selected crayfish were trans-
ferred into the laboratory immediately after collection. First,
they were placed into a plastic tank with a water depth of 1.5 cm
for 4 h. The active ones were then selected and maintained in
120 × 60 × 50-cm self-circulation aquaria with temperature
regulating devices under 12 h light : 12 h dark (12L:12D) pho-
toperiods. Crayfish were given surimi (70% in total food weight)
and vegetable (30% in total food weight) feed ad libidum. Every
aquarium had 35 cm of water, 15 artificial caves ( = 5 cm,
L = 18 cm), and two polyethylene climbing nets set on a sub-
stratum of gravel. The density of crayfish was eight females per
aquarium. To acclimatize them to laboratory conditions, these
crayfish were cultured for 8 d at 22–26◦ C before treatment.
Sampling process.—First, females that had completed
spawning and dead ones were removed and counted every day.
Second, at the end of the experiment, ovaries of individuals
that did not spawn were dissected and those without matured
ovaries were not counted as valid samples for analysis. Oocyte
diameter was measured under a Leica MZ7.5 high-performance
stereomicroscope using an ocular micrometer.
Examining temperature and photoperiod requirements of
spawning in ovary-mature red swamp crayfish.—For tem-
perature and photoperiod requirements examination, we used
three parameters: the average spawning rate (SPR % =
100N spawned /[N matured – N died ]), the average mortality (mortality
% = 100N died /N matured ), and the days before spawning (DBS)
in three parallel aquaria for each group. The experiments lasted
20 d. Females were randomly assigned to each aquarium.
Temperature.—Four water temperature gradients (12–14◦ C,
16–18◦ C, 20–22◦ C, and 24–26◦ C) were set with a daily pho-
toperiod (16 h light : 8 h dark [16L:8D]) in this experiment.
Females were randomly assigned to these four temperature gra-
dients (24 females to three aquaria of each temperature gradient)
after acclimatization.
Photoperiod.—Four photoperiods—complete light 24 h/d
(CL), 12L:12D/d, 16L:8D/d, and complete dark 24 h/d (CD)—
were set in this experiment. Each photoperiod had two water
temperature gradients: 16–18◦ C and 24–26◦ C. Females were
randomly assigned to the eight treatments (24 females to three
aquaria of each treatment) after acclimatization.
Tests of inducing spawning.—For the spawning induction
tests, we used two parameters—the average spawning induction
rate (SPR % = 100N spawned /[N matured – N died ]) and the average
SR (SR % = 100N survived /N matured )—as evaluating indicators.
The experiments lasted 3 d.
Induction spawning via eyestalk stimulation.—
Polysaccharide from Bletilla striata (BSP) was used as
the carrier in the interventional stimulating agent preparation.
Bletilla striata is a Chinese herbal medicine (Chinese name,
Bai Ji) used to stop bleeding caused by traumatic injuries,
heal wounds, reduce swelling, and promote regeneration of
tissue (Yeung 1985). In recent years, BSP has also been used
 230
as an “embolizing” agent (Feng et al. 1996; Zheng et al. 1996;
Zheng et al. 1998) and nonviral gene vector (Xia et al. 2008).
In this study, the process of extracting BSP was modified
from the alcohol precipitation method (Xia et al. 2008).
Two combinations of ions were evaluated by being made
into two kinds of interventional stimulating agents with this
carrier. First, the two combinations of ions were made into
the following two solutions: (1) Van Harreveld (1936) crayfish
saline solution adjusted with a high concentration of Na + and
a low concentration of Ca2 + (280 mM Na + , 0.4 mM K + ,
0.1 mM Ca2 + , 16 mM Mg2 + , 312.6 mM Cl−, 2.5 mM Hepes),
and (2) Van’s solution adjusted with high concentrations of K +
and Ca2 + (117 mM Na + , 80 mM K + , 20 mM Ca2 + , 0.5 mM
Mg2 + , 238 mM Cl−, 2.5 mM Hepes). Then, BSP was dissolved
in these solutions respectively to prepare two B. striata gelatins
(BSG) for injection. The BSG with a high concentration of
Na + and a low concentration of Ca2 + was named BSG-1; the
Downloaded by [Department Of Fisheries] at 23:23 19 March 2013
second was named BSG-2.
For this part of the study, females were divided into four
groups (three experimental groups and one control group, 24
animals randomly assigned to three parallel aquaria in each
group). Crayfish in two experimental groups were injected with
BSG-1 and BSG-2, respectively. The eyestalk tissue of each
female was injected with these BSGs (0.2 mL/eyestalk) using a
1-mL insulin syringe with a permanently attached 29G × 1/2-in
BD ultra-fine needle. Crayfish in the third experimental group
had one eyestalk ablated. The experiments lasted 3 d.
Induction spawning by water removal or removing from wa-
ter.—For this part of the study, females were randomly divided
into two groups (24 animals in three parallel aquaria of each
group). Crayfish were cultured 6 d at 18–19◦ C at a photoperiod
of 16L:8 D. The water of one aquarium of the first group was
then siphoned out, and from the other two aquaria of this group
the females were fished out and immediately put into a plastic
container without water; the time for spawning was recorded
one by one. If spawning occurred after treatment in the first
group and not in the second group, we siphoned out the water of
the second group to verify removing water induces spawning.
Induction spawning with hormonal treatments.—In this
study with red swamp crayfish, we tested ovulation-inducing
agents for fish, which included combinations of domperidone
(DOM; a dopamine D2 antagonist) and LHRH-A2 (luteinizing
hormone-releasing hormone analog), using the recommended
TABLE 1. The spawning rate and mortality of red swamp crayfish at four temperature gradients under a 16L:8D photoperiod during a 20-d examination. Letters
(z and y) represent statistical differences; values of mortality and SPR represent mean (SD in parentheses; one-way ANOVA and Holm–Sidak t-test: p < 0.01).
Evaluating indicators 12–14
Mortality (%) 66.7 z (19.0)
SPR (%) 0.0
DBS (d) Hibernate or die
LIU ET AL.
dosage range of about 1.0 µg/g body weight (BW) of DOM and
0.1 µg/g BW of LHRH-A2 (Peter et al. 1988; Peter and Yu 1997;
Lin and Liu 2006). We also tested spawning induction using the
proposed effective dosage of 5-HT, 10−6 mol per female, for
induction ovarian maturation and spawning (Vaca and Alfaro
2000; Alfaro et al. 2004).
Domperidone (Sigma; CAS 57808-66-9) and 5-HT (Sigma;
CAS 153-98-0) were dissolved in a Van Harreveld crayfish
saline solution and made into a 2 mg/mL solution. We made
LHRH-A2 (D-Ala6 analog, CAS 51230-19-4; Chinese Ningbo
Sansheng Pharmaceutical) into a 10 µg/mL solution. In addi-
tion, the prepared DOM and LHRH-A2 solutions were diluted
20 and 10 times, respectively, with the Van Harreveld crayfish
saline solution before injection.
For this part of the study, females were divided into five
groups (four experimental groups and one control group, 24 an-
imals in each group). The first group was injected with 0.25 mL
DOM and 0.25 mL 5-HT solution; the second group was in-
jected with 0.25 mL DOM and 0.25 mL LHRH-A2 solution;
the third group was injected with 0.25 mL DOM (injections
were performed twice, once on the first day and once on the
second day); the fourth group had eyestalks ablated; and the
control group was not treated.
Statistics.—Experimental data were analyzed with a
SigmaPlot 11 statistical program using ANOVA and Holm–
Sidak t-test.
RESULTS
Spawning of Ovary-Mature Red Swamp Crayfish:
Temperature Effects
In the 20 d of experiments with four temperature gradients un-
der photoperiod 16L:8D, spawning was observed in a relatively
broad temperature range—from 16◦ C to 22◦ C—but lower tem-
perature (16–18◦ C) significantly induced spawning (p < 0.01;
Table 1). Specifically, an SPR of 69.6 ± 6.4% (mean ± SD)
was observed in the 16–18◦ C group and an SPR of 18.7 ± 9.7%
(mean ± SD) was observed in the 20–22◦ C group. Figure 1a
shows the daily spawning and mortality observed in the four
temperature gradients of this study. These observations suggest
that spawning in ovary-mature females of red swamp crayfish
requires a period (5–14 d) of lower temperature. However, when
the temperature was below 14◦ C, death and hibernation began
Temperature gradients (◦ C)
16–18 20–22 24–26
12.5 y (12.5) 12.5 y (12.5) 37.5 zy (0.0)
69.6 z (6.4) 18.7 y (9.7) 0.0
9.93 (1.2) ≥14 >20
 INDUCING SPAWNING IN OVARY-MATURE RED SWAMP CRAYFISH
Downloaded by [Department Of Fisheries] at 23:23 19 March 2013
FIGURE 1. Daily spawning and mortality observed in different temperature
gradients and photoperiods. (a) Result observed in four temperature gradi-
ents (12–14◦ C, 16–18◦ C, 20–22◦ C, 24–26◦ C) under a photoperiod of 16L:8D;
(b) and (c) results observed in four photoperiods (Dark, 12L:12D, 16L:8D,
Light) under 16–18◦ C and 24–26◦ C, respectively; Dark = CD, Light = CL.
to occur in the females, and spawning could not be further
observed.
Photoperiod Effects
Spawning occurred in both CD and CL circumstances, but
long-period light-groups (CL and 16L:8D) showed a higher SPR
(68.9 ± 10.2% and 69.6 ± 6.4% [mean ± SD], respectively)
than short-period light-groups (12L:12D and CD [53.3 ± 5.7%
231
and 33.3 ± 8.3% {mean ± SD}, respectively]; p < 0.05;
Table 2). Panels b and c of Figure 1 show the daily spawning
and mortality observed in four photoperiods of two temperature
gradients.
Hormonal Treatment and Eyestalk Ablation Effects
Hormonal treatments DOM + LARH-A2 or DOM injec-
tions alone (which were both effective dose ranges for induction
ovulation and spawning in ovary-mature freshwater fish in
8–24 h) did not induce spawning in ovary-mature females of red
swamp crayfish, even in 3 d. Meanwhile, combined injections
with DOM + 5-HT (which was the effective dose range
for induction ovarian maturation and spawning in economic
decapods) did not induce spawning in ovary-mature crayfish,
but eyestalk ablation did (Table 3).
Long-Acting Eyestalk Interventional Stimulation Effects
In the 3 d of spawning induction tests, eyestalk interventional
stimulating with a high concentration of Na + combined with a
low concentration of Ca2 + induced synchronous spawning with
a high SR in ovary-mature red swamp crayfish. Meanwhile, a
high concentration of K + combined with a high concentration
of Ca2 + caused no spawning and high mortality (SR = 20.8 ±
19.0% [mean ± SD]). Table 3 shows the performance of these
long-acting eyestalk interventional stimulations compared with
eyestalk ablation.
Transient Stimulations
In this study, 46 (16.0%) of 288 ovary-mature red swamp
crayfish spawned in 10–185 s after being fished out or by si-
phoning out the water, suggesting that removing from water or
water removal can be a stimulating factor inducing fast spawn-
ing.
DISCUSSION
The reproductive process consists of oogenesis, vitellogene-
sis, maturation leading to ovulation, and then spawning. Ovar-
ian maturation is a prerequisite to spawning but generally does
not mean spawning. Because of the limited number of eggs a
single red swamp crayfish can produce, it is important to in-
duce a certain amount of captive females to mature and spawn
synchronously for the mass production of stocking seeds or ju-
veniles for use in aquaculture and other purpose. A prior study
investigated the feasibility of using a University of Southwestern
Louisiana crayfish hatchery in propagating red swamp crayfish,
reporting that, with maintenance, 22% of the stocked females
had produced offspring after 7 months, 223 juveniles per female
(Trimble and Gaudé 1988). It is proposed that 205 female red
swamp crayfish would produce sufficient young to stock an acre
of grow-out pond with 10,000 individuals (Trimble and Gaudé
1988). We feel this method used a long reproductive cycle with
low efficiency. In fact, in the natural circumstances of Wuhan,
there are enough matured crawfish already present in the peak
season of its ovarian maturation available for spawning (Gong
 232 LIU ET AL.

TABLE 2. The spawning rate and mortality of red swamp crayfish at four photoperiod gradients under 16–18◦ C during a 20-d examination. Letters (z and y)
represent statistical differences; values of mortality and SPR represent mean (SD in parentheses; one-way ANOVA and Holm–Sidak t-test: p < 0.05).

Photoperiod gradients
Evaluating indicators CL 16L:8D 12L:12D CD
Mortality (%) 29.2 (7.3) 12.5 (12.5) 25.0 (0.0) 29.2 (14.4)
SPR (%) 68.9 z (10.2) 69.6 z (6.4) 53.3 y (5.7) 33.3 y (8.3)
DBS (d) 6–15 5–14 8–16 8–14

et al. 2008). Hence, inducing spawning in ovary-mature females
in a reasonably short period of time has practical importance in
the industry of mass production of stocking seeds in red swamp
crayfish and deserves to be developed into a special technique.
It can be used as the downstream technique of inducing ovar-
Downloaded by [Department Of Fisheries] at 23:23 19 March 2013
reproduction in decapod crustaceans as alternatives to eye-
stalk ablation (Nagaraju 2011). Both increases in the gonado-
somatic index and oocyte diameter were reported to be signif-
icantly induced by spiperone (10−6 mol per crayfish) and 17
α-hydroxyprogesterone (10−7 mol per crayfish) injected twice

ian maturation or as an independent technique to achieve mass a week for 3 weeks in red swamp crayfish females during early
production of ovigerous females (with pleopodal eggs) in red vitellogenesis (Rodrı́guez et al. 2002). However, no evidence
swamp crayfish. was provided that spawning can be induced by hormonal treat-
ments in decapod crustaceans in prior studies, as was shown
in fish. In fish, hormonal agents such as GnRH analogues and
Hormonal and Nonhormonal Treatments in Inducing dopamine antagonists are effective and widely used for induc-
Spawning in Ovary-Mature Red Swamp Crayfish tion ovulation in ovary-mature females, leading to spawning
During the past few decades, significant advancement has in 8–24 h (Peter et al. 1988; Peter and Yu 1997; Lin and Liu
been made in understanding the roles of neurotransmitters– 2006). This study proved that the effective dosage of hormonal
neuromodulators, and peptide and nonpeptide hormones in- agents DOM (25 µg per crayfish, 0.8–1.0 µg/g BW) and DOM
volved in reproduction. Many researchers have focused on en- (0.8–1.0 µg/g BW) + LARH-A2 (2.5 µg per crayfish, 0.08–
docrinological manipulation with those hormonal chemicals 0.10 µg/g BW) for induction ovulation in fish induced no spawn-
(such as serotonin, dopamine, octopamine, and methyl farne- ing in ovary-mature red swamp crayfish, even in 3 d. Meanwhile,
soate) as well as vertebrate-type steroids, prostaglandins, and combined injections of DOM (0.8–1.0 µg/g BW) and 5-HT
mammalian hormones, and other related compounds to induce (0.5 mg per crayfish, 15–20 µg/g BW) were effective dosages
in inducing ovarian maturation and spawning in economic de-
capods (Kulkarni et al. 1992; Vaca and Alfaro 2000; Meeratana
et al. 2006; Wongprasert et al. 2006) but did not induce spawn-
TABLE 3. The performance of eyestalk interventional stimulations and eye-
stalk ablation in induction spawning in ovary-mature females of red swamp ing in ovary-mature red swamp crayfish. Nevertheless, eyestalk
crayfish (3-d examination). Letters (z and y) represent statistical differences; ablation did. These results reveal, when the ovary matured, stim-
SD in parentheses (one-way ANOVA and Holm–Sidak t-test: p < 0.01). ulating the eyestalk region might lead to spawning. In this study
of the long-acting eyestalk interventional stimulation method
Evaluating indicators mean (SD) we developed for inducing spawning, eyestalk interventional
+
Treatments SPR (%) Survival rate (%) injection of BSG with high concentration of Na and low con-
2+
centration of Ca , an SR (43.8 ± 15.8% [mean ± SD]) was
Eyestalk intervention with 43.8 z (15.8) 95.8 z (7.3) achieved that was similar to bilateral eyestalk ablation (SPR =
BSG-1a 50 ± 16.6% [mean ± SD]) but with a much lower mortal-
Eyestalk intervention with 0y 20.8 y (19.0) ity (SR = 95.8 ± 7.3% [mean ± SD]) than eyestalk ablation
BSG-2b (SR = 37.5 ± 12.5% [mean ± SD]). Hence, it is a potential
Eyestalk ablation 50 z (16.6) 37.5 y (12.5) alternative, nonhormonal technique instead of eyestalk ablation
Control 0y 91.7 z (14.4) and deserves further assessment. Moreover, treated females that
a + 2+
BSG-1: gelatin with a high concentration of Na and a low concentration of Ca ,
are not spawned can be sold as food afterwards because the
made of a polysaccharide extracted from BSP and an adjusted crayfish saline (280 mM treatment is nonhormonal, thereby meeting market demands for
Na + , 0.4 mM K + , 0.1 mM Ca2 + , 16 mM Mg2 + , 312.6 mM Cl−, 2.5 mM Hepes) for food safety. Meanwhile, it is much cheaper than all hormonal
injection.
b + 2+
BSG-2: gelatin with high concentrations of K and Ca , made of BSP and an
methods in inducing spawning in red swamp crayfish. A special
adjusted crayfish saline (117 mM Na + , 80 mM K + , 20 mM Ca2 + , 0.5 mM Mg2 + , injector for eyestalk injection should be developed for better
238 mM Cl−, 2.5 mM Hepes) for injection. efficiency of labor.
 INDUCING SPAWNING IN OVARY-MATURE RED SWAMP CRAYFISH
Temperature and Light Requirements of Spawning in Red
Swamp Crayfish as Well as Their Roles in Inducing
Spawning
In natural conditions photoperiod, temperature, and water
are the most critical factors controlling reproduction of aquatic
animals. Among these factors, light requirements for spawn-
ing can vary by species (Carmona-Osalde et al. 2002). In red
swamp crayfish, Dendy (1978) proposed that the spawning of
this species needs reduction of light duration, as found in the
photoperiod of autumn. Another prior study lasting 70 d with
females of initial gonadosomatic index (GSI) of about 0.05% to
0.07% showed that the mean of increase of GSI of females under
a long photoperiod (14 h light : 10 h dark) is dramatically higher
than those under the short period (10 h light : 14 h dark) after 42 d
(Daniels et al. 1994). In this study with ovary-mature red swamp
crayfish, spawning was observed in both CD and CL circum-
stances, but long-period light-groups showed a higher spawning
Downloaded by [Department Of Fisheries] at 23:23 19 March 2013
rate than short-period light-groups. This result does not com-
pletely agree with that of a previous study in red swamp crayfish,
which advocated that constant darkness promoted a high egg
laying percentage (Provenzano and Handwerker 1995). Possi-
bly, photoperiod plays a secondary role in influencing spawning
relative to temperature in ovary-mature females.
Temperature was proposed to be the only critical factor con-
trolling spawning in ovary-mature Northern Crayfish (Aiken
1969) and ovary-mature Spinycheek Crayfish (Dubé and Porte-
lance 1992). Spawning requires lower temperature and temper-
ature fluctuations. In red swamp crayfish, Daniels et al. (1994)
reported that the temperature effect on GSI increases at only 14 d
after stocking. They believed that temperature controls the on-
set of ovarian development, which is accelerated by increasing
photoperiod. The result of our present study with ovary-mature
females of red swamp crayfish also indicates temperature is a
greater factor in a quicker onset of spawning than photoperiod.
As observations made in Northern Crayfish and Spinycheek
Crayfish, a lower temperature (16–18◦ C) also significantly in-
duced spawning (p < 0.01) in ovary-mature red swamp crayfish,
although spawning could be observed in a relatively broad tem-
perature range, from 16◦ C to 22◦ C. However, when the temper-
ature is below 14◦ C, mortality increases and hibernation begins
to occur in the females, and further spawning can seldom be ob-
served. These results explain the wide distribution of red swamp
crayfish around the world and one maturation peak with many
spawning peaks around the year in Wuhan (Gong et al. 2008).
Moreover, a low water temperature below 14◦ C might be useful
in shifting “autumn spawning” females to “spring spawning”
ones, as Dendy (1978) presumed.
Transient Stimulations of Water and Their Roles in
Induction Spawning of Red Swamp Crayfish
In this study, removal of water or removing from water (we
used cool water, 18–20◦ C) can be transient stimulations in in-
duction of fast (10–185 s) and synchronous spawning in 16.0%
(n = 288) of ovary-mature red swamp crayfish, which may have
233
completely matured. This result may be the reason why spawn-
ing in natural conditions has been shown to be directly related
to the hydrologic regime (Carmona-Osalde et al. 2002) of flood
(Gutiérrez-Yurrita et al. 1999) and rain (Rodrı́guez-Serna et al.
2000).
ACKNOWLEDGMENTS
We are grateful to Professor W. Zhao of Shanghai Ocean Uni-
versity, Professor W. Wang, senior engineer M. Wang, and Z.
Luo of Huazhong Agricultural University for their contributions
and assistance. This research was supported by the Transfor-
mation of Agricultural Science and Technology Achievements
program (Project 4002–092062) financed by the Ministry of
Finance, China.
REFERENCES
Aiken, D. E. 1969. Ovarian maturation and egg laying in the crayfish Orconectes
virilis: influence of temperature and photoperiod. Canadian Journal of Zool-
ogy 47:931–935.
Alfaro, J., G. Zúñiga, and J. Komen. 2004. Induction of ovarian maturation and
spawning by combined treatment of serotonin and a dopamine antagonist,
spiperone in Litopenaeus stylirostris and Litopenaeus vannamei. Aquaculture
236:511–522.
Ando, H., and T. Makioka. 1998. Structure of the ovary and mode of oogenesis
in a freshwater crayfish, Procambarus clarkii (Girard). Zoological Science
15:893–901.
Brown, F. A., Jr., and G. M. Jones. 1949. Ovarian inhibition by a sinus–gland
principle in the fiddler crab. Biological Bulletin 96:228–232.
Carmona-Osalde, C., M. Rodrı́guez-Serna, and M. A. Olvera-Novoa. 2002. The
influence of the absence of light on the onset of first maturity and egg laying
in the crayfish Procambarus (Austrocambarus) llamasi (Villalobos, 1955).
Aquaculture 212:289–298.
Chen, Y. N., H. F. Fan, S. L. Hsieh, and C. M. Kuo. 2003. Physiological
involvement of DA in ovarian development of the freshwater giant prawn,
Macrobrachium rosenbergii. Aquaculture 228:383–395.
Daniels, W. H., L. R. D’Abramo, and K. F. Graves. 1994. Ovarian de-
velopment of female red swamp crayfish (Procambarus clarkii) as influ-
enced by temperature and photoperiod. Journal of Crustacean Biology 14:
530–537.
Dendy, J. S. 1978. Preliminary experiment with photoperiod to influence craw-
fish spawning. Aquaculture 15:379–382.
Dubé, P., and B. Portelance. 1992. Temperature and photoperiod effects on
ovarian maturation and egg laying of the crayfish, Orconectes limosus. Aqua-
culture 102:161–168.
Feng, G., B. Kramann, C. Zheng, and R. Zhou. 1996. Comparative study on
the long-term effect of permanent embolization of hepatic artery with Bletilla
striata in patients with primary liver cancer. Journal of Tongji Medical Uni-
versity 16:111–116. (In German.)
Gong, S. L. J., R. Sun, L. Li, and X. He. 2008. The study on reproductive
biology of Procambarus clarkii. Freshwater Fisheries 38:23–25. Available:
mall.cnki.net/magazine/Article/DSYY200806006.htm. (February 2012).
Gutiérrez-Yurrita, P. J., J. M. Martı́nez, M. Ilhéu, M. A. Bravo-Utrera, J. M.
Bernardo, and C. Montes. 1999. The status of crayfish populations in Spain
and Portugal. Pages 161–192 in F. Gherardi and D. M. Holdich, editors.
Crayfish in Europe as alien species: how to make the best of a bad situation?
A. A. Balkema, Rotterdam, The Netherlands.
Khazraeenia, S., and P. Khazraiinia. 2009. Effects of bilateral eyestalk ablation
on gonadal maturity, moulting and biochemical changes in the hemolymph
of female Potamon persicum crabs (Decapoda, Brachyura, Potamidae). Inter-
national Journal of Veterinary Research 3:143–150.
 234 LIU ET AL.
Kulkarni, G. K., L. Glade, and M. Fingerman. 1991. Oogenesis and effects of
neuroendocrine tissues on in vitro synthesis of protein by the ovary of the red
swamp crayfish Procambarus clarkii (Girard). Journal of Crustacean Biology
11:513–522.
Kulkarni, G. K., R. Nagabhushanam, G. Amaldoss, R. G. Jaiswal, and M. Fin-
german. 1992. In vivo stimulation of ovarian development in the red swamp
crayfish, Procambarus clarkii (Girard), by 5-hydroxytryptamine. Invertebrate
Reproduction and Development 21:231–239.
Li, S., and W. Zhao. 1999. Structural changes of mandibular organ during the
ovary developing cycle in crayfish. Journal of Shanghai Fisheries University
8:12–18.
Lin, H. R., and X. C. Liu. 2006. Experimental techniques and methods of fish
physiology. Guangdong Higher Education Press, Guangzhou, China.
Meeratana, P., B. Withyachumnarnkul, P. Damrongphol, K. Wongprasert, A.
Suseangtham, and P. Sobhon. 2006. Serotonin induces ovarian maturation
in giant freshwater prawn broodstock, Macrobrachium rosenbergii de Man.
Aquaculture 260:315–325.
Nagaraju, G. P. C. 2011. Reproductive regulators in decapod crustaceans: an
overview. Journal of Experimental Biology 214:3–16.
Okumura, T., and K. Aida. 2001. Effects of bilateral eyestalk ablation on molting
Downloaded by [Department Of Fisheries] at 23:23 19 March 2013
and ovarian development in the giant freshwater prawn, Macrobrachium
rosenbergii. Fisheries Science 67:1125–1135.
Peter, R. E., H. R. Lin, and G. Van Der Kraak. 1988. Induced ovulation and
spawning of cultured freshwater fish in China: advances in application of
GnRH analogues and dopamine antagonists. Aquaculture 74:1–10.
Peter, R. E., and K. L. Yu. 1997. Neuroendocrine regulation of ovulation in
fishes: basic and applied aspects. Reviews in Fish Biology and Fisheries
7:173–197.
Provenzano, A. J., Jr., and T. S. Handwerker. 1995. Effects of photoperiod on
spawning of red swamp crayfish, Procambarus clarkii, at elevated tempera-
ture. Freshwater Crayfish 8:311–320.
Rodrı́guez, E. M., D. A. Medesani, L. S. López-Greco, and M. Fingerman. 2002.
Effects of some steroids and other compounds on ovarian growth of the red
swamp crayfish, Procambarus clarkii, during early vitellogenesis. Journal of
Experimental Zoology 292:82–87.
Rodrı́guez-Serna, M., C. Carmona-Osalde, M. A. Olvera-Novoa, and J. L.
Arredondo-Figuero. 2000. Fecundity, egg development and growth of ju-
venile crayfish Procambarus (Austrocambarus) llamasi (Villalobos 1955)
under laboratory conditions. Aquaculture Research 31:173–179.
Tinikul, Y., A. Joffre Mercier, N. Soonklang, and P. Sobhon. 2008. Changes in
the levels of serotonin and dopamine in the central nervous system and ovary,
and their possible roles in the ovarian development in the giant freshwater
prawn, Macrobrachium rosenbergii. General and Comparative Endocrinol-
ogy 158:250–258.
Trimble, W. C., and A. P. Gaudé III. 1988. Production of red swamp crawfish
in a low-maintenance hatchery. Progressive Fish-Culturist 50:170–173.
Vaca, A. A., and J. Alfaro. 2000. Ovarian maturation and spawning in the white
shrimp, Penaeus vannamei, by serotonin injection. Aquaculture 182:373–385.
Van Harreveld, A. 1936. A physiological solution for freshwater crustaceans.
Proceedings of the Society for Experimental Biology and Medicine 34:428–
432.
Varalakshmi, K. N., and R. Reddy. 2010. Effects of eyestalk ablations on
growth and ovarian maturation of the freshwater prawn Macrobrachium
lanchesteri (de Man). Turkish Journal of Fisheries and Aquatic Sciences 10:
403–
410.
Wongprasert, K., S. Asuvapongpatana, P. Poltana, M. Tiensuwan, and B. Withy-
achumnarnkul. 2006. Serotonin stimulates ovarian maturation and spawning
in the black tiger shrimp Penaeus monodon. Aquaculture 261:1447–1454.
Wurts, W. A., and R. R. Stickney. 1984. An hypothesis on the light require-
ments for spawning penaeid shrimp, with emphasis on Penaeus setiferus.
Aquaculture 41:93–98.
Xia, X., X. Li, G. Feng, Y. Zhao, and C. Zheng. 2008. Preparation and char-
acterization of a non-viral gene vector from a polysaccharide isolated from
the Chinese medicinal herb Bletilla striata. Chinese Pharmaceutical Journal
43:1369–1373.
Yeung, H. 1985. Handbook of Chinese herbal formulas. Institute of Chinese
Medicine, Los Angeles.
Zheng, C., G. Feng, and H. Liang. 1998. Bletilla striata as a vascular emboliz-
ing agent in interventional treatment of primary hepatic carcinoma. Chinese
Medical Journal 111:1060–1063.
Zheng, C., G. Feng, and R. Zhou. 1996. New use of Bletilla striata as embolizing
agent in the intervention treatment of hepatic carcinoma. Chinese Journal of
Oncology 18:305–307. (In Chinese.)
This article was downloaded by: [Department Of Fisheries]
On: 19 March 2013, At: 23:24
Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,
37-41 Mortimer Street, London W1T 3JH, UK

North American Journal of Aquaculture

Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/unaj20

Channel Catfish Hatchery Production Efficiency Using

a Vertical-Lift Incubator (the See-Saw) at Various Egg
Loading Rates
a a b b
Les Torrans , Brian Ott , Robert “Shorty” Jones & Robert Jones Jr.
a
U.S. Department of Agriculture, Agricultural Research Service, Thad Cochran National
Warmwater Aquaculture Center, Catfish Genetics Research Unit, Post Office Box 38,
Stoneville, Mississippi, 36776, USA
b
Needmore Fisheries, LLC, 1017 Greenfield Road, Glen Allan, Mississippi, 38744, USA
Version of record first published: 19 Mar 2013.

To cite this article: Les Torrans , Brian Ott , Robert “Shorty” Jones & Robert Jones Jr. (2013): Channel Catfish Hatchery
Production Efficiency Using a Vertical-Lift Incubator (the See-Saw) at Various Egg Loading Rates, North American Journal of
Aquaculture, 75:2, 235-243

To link to this article: http://dx.doi.org/10.1080/15222055.2013.763879

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions

This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to
anyone is expressly forbidden.

The publisher does not give any warranty express or implied or make any representation that the contents
will be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses should
be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims,
proceedings, demand, or costs or damages whatsoever or howsoever caused arising directly or indirectly in
connection with or arising out of the use of this material.
 North American Journal of Aquaculture 75:235–243, 2013

C American Fisheries Society 2013

ISSN: 1522-2055 print / 1548-8454 online

DOI: 10.1080/15222055.2013.763879

ARTICLE

Channel Catfish Hatchery Production Efficiency

Using a Vertical-Lift Incubator (the See-Saw)
at Various Egg Loading Rates

Les Torrans* and Brian Ott

U.S. Department of Agriculture, Agricultural Research Service,
Thad Cochran National Warmwater Aquaculture Center, Catfish Genetics Research Unit,
Post Office Box 38, Stoneville, Mississippi 36776, USA

Robert “Shorty” Jones and Robert Jones Jr.

Downloaded by [Department Of Fisheries] at 23:24 19 March 2013

Needmore Fisheries, LLC, 1017 Greenfield Road, Glen Allan, Mississippi 38744, USA

Abstract
Channel Catfish Ictalurus punctatus spawns are typically incubated in 0.25-in-mesh baskets suspended in water
that is agitated with paddles positioned between the baskets. We previously tested a vertical-lift incubator (the “See-
Saw”) using Channel Catfish spawns. When loaded with spawns at rates higher than recommended for traditional
paddle-type incubators, as often occurs during the peak of the spawning season, survival to the swim-up stage was
significantly (2.3 times) higher with the See-Saw than with traditional incubators. This project examined the effect
of spawn loading rate and added oxygen on fry production with the See-Saw. In study 1, See-Saws were loaded
with 15.0 ± 0.1 lb (mean ± SE; 219,825 eggs), 30.1 ± 0.1 lb (446,055 eggs), 45.1 ± 0.1 lb (668,206 eggs), or
60.1 ± 0.1 lb (891,157 eggs) of spawns per trough. Water flow averaged 2.1 gal/min—roughly 40% of the rate
recommended for paddle-type incubators. Swim-up fry production increased in proportion to egg loading density up
to 45 lb/trough, with the 15-, 30-, and 45-lb loading rate treatments producing 132,658, 263,828, and 429,422 swim-up
fry, respectively. However, the 60-lb treatment produced only 417,237 swim-up fry. Survival to the swim-up stage in
the 15-, 30-, and 45-lb treatments averaged 60 ± 9, 59 ± 6, and 64 ± 4%, respectively, but survival to swim-up in
the 60-lb treatment averaged only 46 ± 8%. In study 2, the addition of purified oxygen to See-Saw incubators loaded
with 36 spawns/trough (∼45 lb/trough) increased the DO concentration by 4.0 ppm (to 125% air saturation) but
had no effect on survival to swim-up. The See-Saw incubator can effectively incubate up to 45 lb of spawns (670,000
eggs)—up to three times as many eggs as traditional paddle-type incubators—while using less than half the water;
thus, the See-Saw incubator can provide significant savings in space, water, and energy use.

Channel Catfish Ictalurus punctatus were first spawned by
fish culturists nearly a century ago (Shira 1917); since then,
a variety of artificial incubation methods have been used to
simulate the activity of male Channel Catfish, which in nature
press against the eggs, vigorously wiggle, and fan the egg mass
to provide oxygen to the eggs (Clemens and Sneed 1957).
Channel Catfish spawns or “egg masses” (10,000–20,000 eggs
attached together by a gelatinous matrix) can be chemically
separated and incubated in hatching jars (Ringle et al. 1992),
but this is rarely done commercially. Channel Catfish eggs
can also be successfully incubated by using forced air to
agitate whole spawns (Carmichael et al. 1994). However, while
forced air is typically used to provide supplemental aeration in
hatching and fry rearing troughs (Avery and Steeby 2004), it is
not used commercially to incubate Channel Catfish spawns.
Nearly all modern commercial Channel Catfish hatcheries
use an incubation method that is almost identical to one de-
veloped over eight decades ago (Clapp 1929). In a typical

*Corresponding author: les.torrans@ars.usda.gov

Received September 13, 2012; accepted December 28, 2012

235
 236 TORRANS ET AL.
Downloaded by [Department Of Fisheries] at 23:24 19 March 2013

FIGURE 1. (A) Top view of a paddle-type Channel Catfish egg incubator (empty), showing alternating placement of baskets and paddles. The curved paddles in
this model rotate through 360◦ . Slightly less than half of the surface area in the incubator is used for paddles and is not available for egg masses. (B) A paddle-type
incubator fully loaded with two large spawns in each partitioned basket is shown. The straight paddles in this model oscillate through approximately a 90◦ arc.
Much trough space is wasted, but the loading of additional spawns reduces circulation between and through the spawns, thus affecting hatch rate.

commercial Channel Catfish hatchery, spawns are placed into
six or seven 0.25-in-mesh plastic or hardware cloth baskets sus-
pended in hatching troughs that are 8–10 ft long, 18–24 in wide,
and 10–12 in deep, holding 100–120 gal of water (Figure 1).
Each basket usually has a cross-partition to separate the spawns.
Paddles located between the baskets agitate the spawns, simulat-
ing the fanning provided in nature by the male Channel Catfish
(Tucker and Robinson 1990). The paddles are attached to a shaft
running the length of the hatchery trough and either rotate 360◦
or oscillate through a 45–60◦ arc (Avery and Steeby 2004) at
a rate of approximately 15 cycles/min. A single electric mo-
tor may power paddles in several troughs. Recently, as a safety
measure, high-density polyethylene plastic paddles have been
used to replace metal paddles (Steeby et al. 2004).
Normally, 12 spawns (∼15–20 lb of eggs or 222,000–
296,000 eggs total, depending on egg size) are held in each
hatchery trough (2 spawns/basket; Figure 1B), with a recom-
mended single-pass water flow of 5 gal/min at 78–82◦ F (Avery
and Steeby 2004). This high water requirement has hindered the
development of large commercial hatcheries in some regions
with limited groundwater (e.g., western Alabama and eastern
Mississippi).
Channel Catfish eggs hatch in 6 d at 79◦ F (Small and Bates
2001). Although sac fry can be stocked directly into fingerling
ponds, survival to the fingerling stage is higher if the sac fry are
transferred to fry rearing tanks for development into swim-up
fry before being stocked in ponds (Mischke et al. 2012). Fry
may be stocked upon reaching the swim-up stage or may be fed
for one or more days, depending on the farm (USDA 2010).
Although this incubation system has met the needs of a
growing industry for almost a century, it has limitations. Over-
crowding a trough by placing more than two spawns per bas-
ket (i.e., one spawn per “compartment”) is common during the
peak of the spawning season, when hatchery space is limited.
Overcrowding results in reduced water circulation around and
through individual spawns that at least partially overlap and de-
creases the amount of oxygen delivered to the center of each
spawn (Avery and Steeby 2004). A common early sign of inad-
equate water circulation resulting from overcrowded hatching
troughs is the presence of dead eggs in the center of the spawn
 VERTICAL-LIFT INCUBATOR FOR CHANNEL CATFISH 237

(Tucker 1991). Dead eggs may serve as foci for fungal and bac-
terial infections, which spread rapidly to the rest of the spawn
and adjoining spawns, thereby greatly reducing the hatch rate
from the trough.
Hatchery managers attempt to maximize the delivery of oxy-
gen through the spawns by a combination of (1) increasing the
ambient dissolved oxygen (DO) concentration in the hatching
trough (Torrans and Steeby 2006, 2008), (2) periodically turn-
ing or repositioning spawns that are touching (USDA 2010), and
(3) separating spawns into smaller pieces to increase the surface-
to-volume area. Even with these practices, the optimum egg
loading rate in traditional paddle-type incubators is restricted
(Avery and Steeby 2004; Ott and Torrans 2012).
Torrans et al. (2009) described a new vertical-lift Channel
Catfish incubator (the “See-Saw”) that was developed and
tested at two commercial catfish hatcheries in Arkansas and
Mississippi during 2007 and 2008. The See-Saw incubator was
Downloaded by [Department Of Fisheries] at 23:24 19 March 2013

designed to improve water movement between and through

the spawns. Spawns were held in closed baskets, which were
moved up and down through the water column at a rate of 6
cycles/min (Figure 2). This movement forced water through
the baskets in alternating directions, thus separating and
repositioning the spawns with every cycle. Ott and Torrans
(2012) demonstrated that the See-Saw incubator can incubate
and hatch more eggs with less water exchange than traditional
paddle-type incubators.
In this paper, we present two additional on-farm studies using
the See-Saw incubator. The purpose of the first study (study 1,
conducted in 2010) was to examine the effect of Channel Catfish
egg loading rate on hatch rate and swim-up fry production and
to establish the maximum practical egg loading rate for the See-
Saw incubator. The second study (study 2, conducted in 2011)
was used to confirm the maximum loading rate determined in
the first study and to evaluate the effect of added oxygen on the
Channel Catfish hatch rate at that loading rate.

METHODS
The two studies were conducted at Needmore Fisheries,
LLC (Glen Allan, Mississippi), during May and June 2010
and 2011. This commercial Channel Catfish hatchery had
eight See-Saw incubators or “modules,” each operated by a
separate motor. Torrans et al. (2009) and Ott and Torrans (2012)
described the construction and mechanics of the See-Saw
incubator. Briefly, an angle-iron frame was constructed and
bolted to two pairs of existing hatching troughs, thus creating a
unit (module) consisting of four hatching troughs (Figure 2A).
The troughs were 8 ft × 22 in, with a water depth maintained
at approximately 10.75 in (total water volume = 98.3 gal).
A split-phase, 115-V AC, 60-Hz, 0.25-hp, 5.4-RPM gear-
motor (Dayton Model 6Z399A; Dayton Electric Manufacturing
Company, Niles, Illinois) oscillated a solid steel shaft running
the length of the incubators by means of an eccentric cam and
a vertical push rod. The steel shaft had four crossbars welded
FIGURE 2. (A) A four-trough See-Saw incubator in operation. The baskets
occupy most of the trough area since space is not needed for paddles between
baskets. (B) Each basket in the See-Saw has cross-partitions and can be loaded
with up to 15 lb of Channel Catfish spawns, for a total of 45 lb/trough. The
example shown (with the lid open) was loaded with eight spawns, or approx-
imately 30 lb/trough. Due to the alternating up-and-down movement of the
baskets through the water, spawns are continually repositioned, allowing for
multiple spawns to be placed in each compartment. The bolt near the center of
the basket (indicated by the arrow) attaches the basket to the aluminum rack
and also holds the lid closed. Additional labeled photographs are provided by
Torrans et al. (2009) and Ott and Torrans (2012).
to it that extended equidistant over each of the paired troughs.
The crossbars were connected by a chain to rectangular racks
constructed of 2-in angle-aluminum that supported the hatching
baskets. These racks fit inside the troughs.
 238 TORRANS ET AL.
Hatching baskets were constructed of 0.25-in square-mesh,
PVC-coated hardware cloth; each basket was approximately
24 × 17 × 3 in and contained a cross-partition in the center
to evenly distribute spawns within four sections of the basket
(Figure 2B). The baskets had a hinged lid with a 1-in lip to
contain the eggs. Three baskets were secured to each rack by
using stainless-steel bolts, washers, and wing nuts through the
center of each basket. This bolt system also served to keep the
lid closed, as did 11-in bungee cords that were strapped to both
the basket and the rack.
The movement of the incubator is similar to the movement
of a see-saw in that as the horizontal support bars lift the rack
(containing three baskets) in one paired trough up through the
water, the counterbalanced rack in the adjacent trough is low-
ered. The chain supporting the racks was adjusted to stop the
racks at approximately 1 in above the bottom of each trough to
prevent crushing any sac fry that had hatched and sunk to the
Downloaded by [Department Of Fisheries] at 23:24 19 March 2013
trough bottom.
Previous studies (Torrans et al. 2009; Ott and Torrans 2012)
indicated that raising the baskets completely out of the water
and allowing the spawns to briefly drain of water during each
cycle appeared to provide better oxygenation of the eggs. It was
also believed that the spawns absorbed some oxygen directly
from the atmosphere while they were raised above the water.
Although this may have been true, it was subsequently observed
(our unpublished data) that when baskets were loaded with an
increasing weight of spawns, eggs that were forced against the
screen when the basket was lifted out of the water were damaged.
Therefore, in the studies reported here, the vertical lift was
limited to approximately 1 in below the water surface (measured
from the water surface to the bottom of the basket), allowing
some buoyance to the eggs.
Well water with an alkalinity of 325 ppm and hardness (as
CaCO3 ) of 51 ppm was passed through a propane-fueled boiler
to maintain a constant temperature and then was passed through
an elevated tower to aerate the water and to provide uniform head
pressure for the water supply. Purified oxygen was supplied from
a 50-gal liquid oxygen (LOX) Dewar and was metered into the
main hatchery water supply line, allowing the DO in the water
supplied to the troughs to be maintained at or near air saturation
(8.0 ppm at 80◦ F and 100% air saturation in 2010; 7.7 ppm at
80.6◦ F and 97% air saturation in 2011).
Water was supplied to each trough through a 0.75-in-diameter
PVC hose bibb. Flow regulators were fabricated from 0.75-in-
diameter plastic hose caps, which were drilled to allow a water
flow of approximately 2 gal/min when installed on the hose
bibb and with the valve fully opened. This provided a water
exchange rate of approximately once every 49 min. Water flows
were measured in each trough twice daily during both studies.
A plastic container was placed under the inflow for 15 s, the
water was poured into a graduated cylinder, and the volume was
measured. Flow (gal/min) was calculated as volume (gal) ×
4. Dissolved oxygen concentration (at the inflow and outflow
ends of each trough) and water temperature were measured daily
with a YSI hand-held DO meter (Model 550A; Yellow Springs
Instruments, Inc., Yellow Springs, Ohio).
The eggs used in this study were collected from commercial
spawning ponds that were stocked with unselected commercial
strains of Channel Catfish. Spawning cans in any given pond
were checked every 3 d in rotation, and all spawns found were
removed and taken to the hatchery. The spawns could have
varied in age from 0 to 3 d old, but no attempt was made to sort
them by age when they were placed into the incubators.
Study 1.—During 2010, spawns were collected from spawn-
ing ponds by farm employees in the afternoon and were placed
in traditional incubators overnight. The next morning, spawns
were individually weighed and placed into the test incubators at
the target treatment densities. A 150–300-egg sample was taken
from 10% of the spawns each day. The egg samples were drained
of water, blotted, weighed, and separated with a 1.5% solution
of sodium sulfite (Ringle et al. 1992). The separated eggs were
placed in a petri dish and photographed for later enumeration.
The egg sample counts were used to calculate the total number
of eggs loaded into each incubator, which served as the basis for
calculating hatch rates.
The four hatching troughs comprising a See-Saw incubator
were sequentially loaded with eggs at four different target rates
of approximately 15, 30, 45, and 60 lb/trough. The spawns were
distributed as equally as possible among the four compartments
within each of the three baskets in each trough (Figure 2B).
The four troughs in each See-Saw incubator served as one repli-
cate for each loading rate in the study. Each loading rate was
replicated five times during the study.
Eggs were treated prophylactically by hatchery personnel
once daily with one teaspoon (∼0.4 oz or 30 ppm) of copper
sulfate (CuSO4 ·5H2 O). The copper sulfate was dissolved in a
2-qt pitcher of hatchery water and was poured along the length
of the trough. Neither water flow nor See-Saw operation was
stopped during the prophylactic treatments.
As eggs hatched, the sac fry dropped through the 0.25-in
mesh of the baskets to the bottom of the troughs and were si-
phoned out twice daily at approximately 0800 and 2000 hours.
The sac fry were measured volumetrically in a graduated cylin-
der and then were moved to fry rearing troughs. A 0.2–0.4-oz
sample of sac fry was collected from each trough daily af-
ter hatching began; the sample was measured volumetrically,
placed in a petri dish, and photographed for later enumeration.
These sample counts were used to determine the number of sac
fry produced daily from each trough.
The fry rearing troughs at this hatchery would normally hold
sac fry that hatched from 10–20 lb of eggs. Since more sac fry
were produced in this study from hatching troughs with higher
egg loading rates (30-, 45-, and 60-lb loading rate treatments)
than would normally be reared in a single fry rearing trough, the
sac fry from those hatching troughs were split among several
rearing troughs to maintain similar densities in all fry rearing
troughs. The identity of sac fry from each replicate hatching
trough was maintained by the use of tags that accompanied
 VERTICAL-LIFT INCUBATOR FOR CHANNEL CATFISH
each batch of sac fry from the hatching trough to the rearing
troughs. Thus, several rearing troughs could contain swim-up
fry from one hatching trough; these were sampled individually
and totaled by replicate when moved out of the hatchery.
After 2–3 d in the fry rearing troughs, the sac fry absorbed
their yolk sacs, turned black, and began to swim up. At this time,
sample counts were done on the fry from each trough; the fry
were batch-weighed on a digital scale (60- × 0.02-kg capac-
ity; Model 1GB; Shanghai Ishida Electronic Scales Co., Ltd.,
Pudong, Shanghai, China) by the hatchery personnel and were
stocked into fry rearing ponds. Total fry weight (combined from
all fry rearing troughs with fry produced from the same replicate
hatching trough) and sample counts of swim-up fry from each
rearing trough were used to estimate the total number of swim-
up fry produced from each loading rate replicate in the study.
Study 2.—Only one egg loading rate (36 spawns; ∼45 lb/
trough) was used in 2011. This was the highest loading rate
Downloaded by [Department Of Fisheries] at 23:24 19 March 2013
in study 1 that showed no effect of increasing egg density on
swim-up fry production, and it appeared to be the maximum
recommended egg loading rate for the See-Saw incubator
under these conditions. Study 2 compared the performance of
See-Saw incubators maintained either without added oxygen
(i.e., as was done in study 1; air saturation [control] treatment,
N = 15 replicate troughs) or with the addition of purified
oxygen (added oxygen treatment, N = 17 troughs).
Purified oxygen was distributed through the hatchery from a
50-gal LOX Dewar via a line that supplied individual oxygen
flow meters (FR2000 Model 2A00; Key Instruments, Trevose,
Pennsylvania) at each trough. Oxygen was metered through a
12.1- × 2.4-in MicroBubble diffuser (Model AS303; Point Four
Systems, Inc., Coquitlam, British Columbia, Canada) that was
placed in the approximate center of each trough belonging to the
added oxygen treatment. Oxygen flow rates were monitored and
adjusted daily to maintain an average trough DO concentration
of approximately 3 ppm above air saturation.
Unlike study 1, eggs in study 2 were not held in separate
troughs overnight before being counted and weighed into the
test incubators by the authors. Rather, the farm employees
counted, weighed, and moved the spawns directly from the
transport tank used to collect spawns from the ponds into the
incubators. A waterproof tag containing information on trough
number, date, number of spawns, and weight of spawns was
placed on each trough.
Because all treatments were loaded with the same number
of spawns and because the number of fry produced per pound
of spawn incubated is the main measure of hatchery efficiency
used by most farms, eggs counts were not determined for
individual spawns. Sac fry were siphoned out and moved to fry
rearing troughs by hatchery personnel at multiple times around
the clock (to prevent large amounts of sac fry and debris from
accumulating at the bottom of troughs, which could lead to suf-
focation and mortality); this was in contrast to the twice-daily
siphoning performed in study 1. Produced sac fry were not
measured volumetrically or sampled, but sample counts and
239
total weight of swim-up fry were determined by the hatchery
crew for each fry rearing trough when the fry were moved to the
ponds. From these data, the number of swim-up fry produced
per pound of eggs incubated was determined for study 2.
Statistical analyses were conducted using one-way ANOVA
(SAS 2008). Percentage data were arcsine transformed prior to
analysis. If the differences among loading rates in study 1 were
significant, means were separated by use of Tukey’s studen-
tized range test. In all statistical comparisons, responses were
considered significant at P-values less than 0.05.
RESULTS
Study 1
Spawns were loaded into the See-Saw incubators from May
18 through May 21, 2010, at an average rate of 15.0 ± 0.08,
30.1 ± 0.11, 45.1 ± 0.05, and 60.1 ± 0.11 lb/trough (mean ±
SE; Table 1). Egg numbers per treatment ranged from a low of
219,825 ± 3,144 eggs/trough (mean ± SE) in the 15-lb treat-
ment to a high of 891,157 ± 14,710 eggs/trough in the 60-lb
treatment (Table 1; Figure 3). As planned, the number and total
weight of spawns per trough and the number of eggs per trough
differed significantly among treatments (P < 0.0001). The re-
lationship between egg loading rate (lb/trough) and number of
eggs was linear (r = 0.997, P < 0.0001).
Mean DO at the outlet end of the troughs decreased with
increasing egg loading rate (Table 1; Figure 4). The mean DO
concentrations in the 45- and 60-lb treatments were significantly
lower than the mean DO concentration in the 15-lb treatment
(Table 1; Figure 4); the mean DO concentration in the 30-lb
treatment was intermediate to and not significantly different
from those in the other treatments.
Eggs first began to hatch after 2 d in the hatching troughs
(Figure 4), but the bulk of the eggs (93.9%) hatched after 3–5 d
in the troughs at an estimated age of 6 d postfertilization. As the
eggs developed, metabolic oxygen demand increased and DO
FIGURE 3. Number (+SE) of Channel Catfish eggs loaded, number of sac
fry removed, and number of swim-up fry produced from five replicate See-Saw
incubators that were loaded with 15, 30, 45, or 60 lb of spawn per hatching
trough. For a given life stage (eggs, sac fry, or swim-up fry), bars with the same
letter are not significantly different (P > 0.05).
 240 TORRANS ET AL.

TABLE 1. Performance (mean ± SE) of See-Saw incubators loaded with Channel Catfish eggs at four different target loading rates: 15, 30, 45, and 60 lb of
spawn/trough. Within a given row, means followed by different letters are significantly different (P < 0.05).

Variable 15 lb/trough 30 lb/trough 45 lb/trough 60 lb/trough

Spawns (N) 11.2 ± 0.05 w 22.6 ± 0.51 x 32.6 ± 0.75 y 44.2 ± 0.49 z
Total spawn weight (lb) 15.0 ± 0.08 w 30.1 ± 0.11 x 45.1 ± 0.05 y 60.1 ± 0.11 z
Mean spawn weight (lb) 1.34 ± 0.06 z 1.33 ± 0.03 z 1.38 ± 0.03 z 1.36 ± 0.02 z
Total eggs (N)a 219,825 ± 3,144 w 446,055 ± 6,790 x 668,206 ± 10,143 y 891,157 ± 14,710 z
Total sac fry moved (N)b 178,929 ± 15,209 x 362,045 ± 15,118 y 589,553 ± 34,620 z 654,753 ± 48,619 z
Survival to hatch (%) 81.6 ± 7.4 81.2 ± 3.3 88.3 ± 5.5 73.4 ± 5.2
Total swim-up fry moved (N)c 132,658 ± 19,575 x 263,828 ± 26,438 y 429,442 ± 22,606 z 417,237 ± 85,109 z
Survival, hatch to swim-up (%) 75.6 ± 11.0 73.0 ± 6.8 74.0 ± 6.2 62.9 ± 9.6
Survival, egg to swim-up (%) 60.3 ± 9.0 59.1 ± 6.0 64.4 ± 3.7 46.3 ± 8.5
Swim-up fry per lb of eggs (N) 8,867 ± 1,303 8,877 ± 895 9,530 ± 500 6,934 ± 1,398
Mean DO concentration at outlet (ppm) 7.5 ± 0.17 z 7.0 ± 0.20 zy 6.6 ± 0.13 y 6.6 ± 0.20 y
a
Estimated from sample counts on 10% of the spawns used each day (55 total samples).
Downloaded by [Department Of Fisheries] at 23:24 19 March 2013

b
Estimated based on one sample count from each replicate hatching trough on each day the sac fry were moved.
c
Estimated based on one sample count from each fry trough at the time the fry were moved out to the ponds.

concentrations (measured at the discharge end of the troughs)
decreased (Figure 4). The lowest DO concentrations occurred
on day 3. After that, the increased metabolism of the remaining
embryos was presumably offset by the removal of sac fry, and the
DO concentrations increased. The minimum DO concentration
on all days was above 6.0 ppm.
The See-Saw incubator performed with similar efficiency
at egg loading rates of 15, 30, and 45 lb/trough. There were
significant differences among the 15-, 30-, and 45-lb treatments
with respect to the number of sac fry produced and the number
of swim-up fry produced (Table 1; Figure 3), and there was
a linear relationship between the number of eggs loaded in

FIGURE 4. Mean dissolved oxygen (DO) concentration measured over the 5-d Channel Catfish egg incubation period at the discharge end of five replicate
See-Saw incubators (troughs), each loaded with 15, 30, 45, or 60 lb of eggs. Mean temperature was 80◦ F. The secondary y-axis indicates the mean percentage of
sac fry that were moved from the incubators into fry troughs each day.
 VERTICAL-LIFT INCUBATOR FOR CHANNEL CATFISH 241

those treatments and number of sac fry (r = 0.96, P < 0.0001)

or swim-up fry (r = 0.93, P < 0.0001) produced (Table 1;
Figure 3). However, neither sac fry production nor swim-up
fry production differed significantly between the 45- and 60-lb
treatments (Table 1; Figure 3). In fact, the yield of swim-up fry
from the 60-lb treatment (417,237 fry; 46.3% of eggs loaded)
was actually less than that from the 45-lb treatment (429,422
fry; 64.4% of eggs loaded).
Although the actual hatch from the 60-lb treatment appeared
to be good, many dead sac fry were observed when the fry were
siphoned out of the hatching troughs twice daily. The similar
densities of dead and live sac fry prevented complete separation
of the two groups at the time of siphoning; more dead sac fry
were removed after the sac fry were placed in fry troughs for
maturation to swim-up fry, as the swimming movement of the
live sac fry facilitated removal of the dead individuals.
Downloaded by [Department Of Fisheries] at 23:24 19 March 2013

FIGURE 5. Mean (±SE) calculated number of swim-up fry produced per

Study 2
In 2011, 32 See-Saw hatching troughs were each loaded with
36 Channel Catfish spawns—a total of 1,359 lb of eggs (esti-
mated at 21 million eggs). Fifteen replicate hatching troughs
(air saturation treatment) were operated with the same general
methods as were used for the 45-lb treatment in study 1. Sev-
enteen replicate hatching troughs had purified oxygen (supplied
from a LOX Dewar) added through a ceramic diffuser at an av-
erage rate of 0.00423 ft3/min (0.12 L/min), resulting in a mean
DO concentration of 10.0 ppm at the outlet in comparison with
6.03 ppm for the air saturation treatment (P < 0.05; Table 2).
Aside from the difference in DO concentration, there were no
significant differences in any measured performance variable be-
tween the two DO treatments (Table 2). The number of swim-up
fry produced per pound of spawn (Figure 5) averaged 10,299 ±
890 (mean ± SE; survival from egg stage = 69.6 ± 6.0%)
for incubators in the added oxygen treatment and 10,355 ±
448 (70.0 ± 0.08%) for the incubators with inlet water at air
saturation.
pound of Channel Catfish eggs incubated. Values are means of five replicates
per treatment in 2010 (treatments: 15, 30, 45, or 60 lb of eggs loaded per
hatching trough) and either 15 replicates (air saturation treatment with 45 lb
of eggs/trough) or 17 replicates (added oxygen treatment with purified liquid
oxygen [LOX] and 45 lb of eggs/trough) in 2011.
DISCUSSION
Sac fry production (r = 0.96, P < 0.0001) and swim-up
fry production (r = 0.93, P < 0.0001) were proportional to
spawn loading rates up to 45 lb/trough in study 1, but lower-
than-expected production of both sac fry and swim-up fry was
observed in the 60-lb treatment. Poorer hatch rate or survival to
the swim-up stage can result from a variety of causes, including
maternal broodfish strain, age, or nutrition; male fertility; ge-
netic factors affecting completion of embryonic development;
fungus or other pathogens in the hatchery; and temperature,
hardness, DO concentration, or other environmental factors. In
this study, spawns were distributed randomly among the four

TABLE 2. Performance (mean ± SE) of See-Saw incubators loaded with 36 Channel Catfish spawns per incubator and supplied with water with DO concentration
at air saturation (N = 15 troughs) or with oxygen added at a rate of 0.00423 ft3/min through ceramic diffusers (N = 17 troughs).

Treatment
Variable Added oxygen Air saturation P
Total spawn weight per trough (lb) 44.7 ± 1.88 46.6 ± 1.89 0.47
Mean spawn weight (lb) 1.24 ± 0.05 1.30 ± 0.05 0.47
Total eggs loaded (N)a 660,946 ± 27,829 690,184 ± 27,938 0.47
Total weight of swim-up fry moved (lb) 29.03 ± 2.21 31.98 ± 2.16 0.35
Total swim-up fry moved (N)b 445,857 ± 32,678 478,869 ± 23,696 0.43
Fry per lb of eggs (N) 10,299 ± 890 10,355 ± 448 0.96
Survival to swim-up fry stage (%) 69.6 ± 6.0 70.0 ± 3.0 0.96
Mean DO concentration at outlet (ppm) 10.00 ± 0.14 6.03 ± 0.08 0.0001
a
Estimated from an average egg count of 14,796 eggs/lb for the 2010 spawning season (i.e., study 1) at this hatchery.
b
Estimated based on one sample count from each fry trough at the time the fry were moved out to the ponds.
 242 TORRANS ET AL.
treatments, negating all but treatment and resulting water qual-
ity effects on hatching and survival to the swim-up stage.
Mean DO concentration in troughs of the 60-lb treatment
was similar to that in troughs of the 45-lb treatment (Table 1;
Figure 4), presumably resulting in a comparable hatch per se
(i.e., embryos exiting the chorion and sinking through the screen
basket to the bottom of the trough). However, we believe that
the microenvironment at the bottom of the incubator (a layer
up to perhaps 1 in thick that included sac fry and debris: dead
eggs freed from the spawn by adjacent hatching eggs, egg shells,
and remnants of matrix) resulted in reduced water circulation
(Dhiyebi et al. 2013) and localized anoxia, leading to the death
of many sac fry before they were moved to rearing troughs.
We did not measure ammonia or DO concentration within this
layer, but it is unlikely that ammonia concentrations could have
increased to toxic levels faster than DO decreased.
Conventional paddle-type incubators are less efficient at pro-
Downloaded by [Department Of Fisheries] at 23:24 19 March 2013
viding oxygen through and between the incubating spawns (Ott
and Torrans 2012), thus resulting in reduced hatch at higher
spawn loading rates; however, the rotating paddles do a good
job of creating turbulence on the bottom of the trough, keeping
debris and infertile or dead eggs and egg shell debris suspended
in the trough until they are flushed out the drain, leaving the
heavier sac fry on the bottom of the trough. The higher water
flow and exchange rates normally used in traditional incubators
also facilitate continual removal of debris. The See-Saw, while
providing better oxygen transfer through the incubating spawns
and increased hatch at higher spawn loading rates (Torrans et al.
2009; Ott and Torrans 2012), does not produce a similar turbu-
lence at the bottom of the trough. With the See-Saw, most of the
debris is typically not suspended by turbulence. In addition, the
See-Saw’s water flow of 2 gal/min (as opposed to the 5 gal/min
recommended for traditional incubators) leads to a reduction in
debris removal. Although some of the debris is flushed out of
the trough, much remains in and around the sac fry in a layer on
the bottom of the trough. We siphoned sac fry only twice daily,
and although this was satisfactory when applied to the 15-, 30-,
and 45-lb treatments, it was apparently inadequate for the large
number of eggs hatching from the 60-lb treatment between si-
phoning times. It is likely that more than 45 lb of eggs could be
successfully incubated in the See-Saw if the sac fry are siphoned
more frequently, as would occur with normal hatchery practices.
Dissolved oxygen concentrations measured in the 45-lb treat-
ment See-Saw incubators were 6.6 ppm (81% saturation) dur-
ing study 1 and 6.0 ppm (75% saturation) during study 2. In
both studies, the DO concentration was below the critical level
(88.1% saturation) required for developing Channel Catfish eggs
on the last day of incubation, as determined by Torrans and
Steeby (2008). Higher DO concentrations measured in incuba-
tors with lower egg loading rates in study 1 (15-lb treatment:
7.5 ppm and 94% saturation; 30-lb treatment: 7.0 ppm and 88%
saturation) or with added oxygen in study 2 (10 ppm or 125%
saturation) yielded no improvement in sac fry or swim-up fry
production (Tables 1, 2). Although we did not directly measure
the DO concentration in the interior of spawns as they developed
in the See-Saw incubator, we believe that these data validate the
operational premise of the See-Saw: moving the spawns up and
down through the water column produces improved oxygen de-
livery throughout each spawn, resulting in optimum hatch at a
lower ambient DO concentration than has been predicted from
previous laboratory studies.
During the 2 years of this study, 2,210 lb of Channel Cat-
fish spawns (∼33 million eggs) were incubated, making this
the largest-scale hatchery research study ever reported. Mea-
surements were not made at every critical control point in the
hatchery process, but we believe that this was not necessary to
assess the commercial application of this new Channel Catfish
egg incubator. Overall efficiency in a commercial hatchery may
be best defined as the number of swim-up fry that are transferred
to ponds per pound of spawn brought into the hatchery (A. Jones,
Bear Creek Fisheries, personal communication). When pooling
all treatments in both studies (except for the 60-lb treatment in
study 1), hatchery efficiency from use of the See-Saw incuba-
tors averaged 9,565 swim-up fry/lb of eggs. While comparable
data are not available from industry surveys, personal experi-
ence indicates that this is at least similar to the efficiency of
well-managed commercial hatcheries.
Up to three times as many eggs can be loaded into the See-
Saw as in traditional incubators (45 lb of spawns versus the
15–20 lb recommended for traditional paddle-type incubators)
while using only 40% as much water (2 gal/min compared with
the 5 gal/min recommended for conventional incubators), with
no negative effects on hatch rate or fry production. This water
conservation reduces effluents and pumping costs by 86%.
If well water needs to be heated by only 10◦ F to obtain an op-
timum incubation temperature of 80◦ F (Small and Bates 2001),
then the additional cost savings in reduced heating (propane)
cost alone would justify conversion to the See-Saw incubator.
One See-Saw (i.e., consisting of four troughs) could incubate
180 lb of eggs or 2.66 million eggs at one time. Based on
(1) a water flow rate of 5 gal/min and an egg loading rate of
15 lb/trough for the conventional incubators, (2) corresponding
rates of 2 gal/min and 45 lb/trough for the See-Saw incuba-
tors, (3) a heater with 80% efficiency, and (4) a propane cost of
$2.68 per gallon, one four-trough See-Saw incubator would save
$1,142.00 in propane cost alone for a single hatching cycle (6 d)
compared with the number of conventional paddle-type incuba-
tors that would be needed to incubate the same number of eggs.
Use of the See-Saw incubator also saves labor. Due to the ten-
dency of spawns to touch in the traditional incubators (thereby
negatively impacting water circulation and hatch), hatchery per-
sonnel routinely turn the spawns. In the latest industry survey
(USDA 2010), a majority of hatcheries (53%) turned spawns
at least three times daily. With the See-Saw, spawns are never
touched after being placed into the incubator, and this results in
a large savings in labor. Management is largely limited to the
verification of water flow and mechanical operation, the appli-
cation of prophylactic treatments, and the removal of sac fry. All
of the hatchery management duties in study 2 were conducted
with existing (non-English-speaking) hatchery employees, thus
 VERTICAL-LIFT INCUBATOR FOR CHANNEL CATFISH
confirming the practical operational ease of the See-Saw under
actual commercial conditions.
If a large number (tens of millions) of Channel Catfish fry is
to be produced, as is the case with most commercial Channel
Catfish hatcheries, conversion from traditional paddle-type in-
cubators to the See-Saw is recommended. The savings in floor
space would permit the inclusion of more fry rearing troughs, al-
lowing for greater fry growth before stocking in ponds, or would
allow space for diversification with other species. The reduced
water use is significant, perhaps allowing establishment of com-
mercial hatcheries in areas with limited groundwater, such as
western Alabama. If well water must be heated to optimum in-
cubation temperature (80◦ F), the savings in heating costs would
in itself justify the use of the See-Saw.
If the facility’s Channel Catfish fry production goal is only a
few million per year, as in many state fish hatcheries, not enough
spawns will be incubated at any one time to justify the high egg
Downloaded by [Department Of Fisheries] at 23:24 19 March 2013
loading capacity of the See-Saw in its present configuration.
Given that the operational premise of the See-Saw is valid, a
smaller-scale incubator that is designed to move spawns through
the water in a similar manner could be useful for low-volume
hatcheries.
ACKNOWLEDGMENTS
Mention of trade names or commercial products in this ar-
ticle is solely for the purpose of providing specific informa-
tion and does not imply recommendation or endorsement by
the U.S. Department of Agriculture (USDA). Funding for this
project was provided by the USDA Agricultural Research Ser-
vice Current Research Information System (Project Number
6402-13320-004-00D) and the Southern Regional Aquaculture
Center (Reimbursable Cooperative Agreement SGA 6402-9-
345) and was made possible through Nonfunded Cooperative
Agreement 54-6402-349N with Needmore Fisheries, LLC. We
thank the management and employees of Needmore Fisheries,
S. Jones, and V. de Regt for their assistance with this project;
we also thank C. Tucker, N. Chatakondi, and J. Steeby for their
critical reviews of this manuscript.
REFERENCES
Avery, J. L., and J. A. Steeby. 2004. Hatchery management. Pages 145–165
in C. S. Tucker and J. A. Hargreaves, editors. Developments in aquaculture
and fisheries science, volume 34: biology and culture of Channel Catfish.
Elsevier, Amsterdam.
243
Carmichael, G. J., T. D. Bates, and T. R. Tiersch. 1994. Forced-air
incubation of catfish eggs. Journal of Applied Aquaculture 3:279–
284.
Clapp, A. 1929. Some experiments in rearing Channel Catfish. Transactions of
the American Fisheries Society 59:114–117.
Clemens, H. P., and K. E. Sneed. 1957. The spawning behavior of the Channel
Catfish Ictalurus punctatus. U.S. Fish and Wildlife Service Special Scientific
Report Fisheries 219.
Dhiyebi, H. A., M. J. O’Donnell, and P. A. Wright. 2013. Water chemistry
in the microenvironment of Rainbow Trout Oncorhynchus mykiss embryos
is affected by development, the egg capsule and crowding. Journal of Fish
Biology 82:444–457.
Mischke, C. C., T. E. Greenway, M. J. Griffin, and D. J. Wise. 2012. Effects of
fry age-at-stocking on growth and survival of Channel Catfish. Journal of the
World Aquaculture Society 43:135–139.
Ott, B. D., and E. L. Torrans. 2012. Effect of increased egg stocking den-
sity in existing and experimental catfish incubators. Proceedings of the An-
nual Conference Southeastern Association of Fish and Wildlife Agencies
64(2010):131–135.
Ringle, J. P., J. G. Nickum, and A. Moore. 1992. Chemical sepa-
ration of Channel Catfish egg masses. Progressive Fish-Culturist 54:
73–80.
SAS (Statistical Analysis Systems). 2008. SAS enterprise guide, v. 4.2
(4.22.0.9238). SAS Institute, Cary, North Carolina.
Shira, A. F. 1917. Notes on the rearing, growth, and food of the Channel Cat-
fish, Ictalurus punctatus. Transactions of the American Fisheries Society 46:
77–88.
Small, B. C., and T. D. Bates. 2001. Effect of low-temperature incuba-
tion of Channel Catfish Ictalurus punctatus eggs on development, sur-
vival, and growth. Journal of the World Aquaculture Society 32:189–
194.
Steeby, J. A., J. Nobile, and W. Wright. 2004. Safer high-density polyethylene
plastic paddles for hatching Channel Catfish eggs. North American Journal
of Aquaculture 66:334–335.
Torrans, L., B. Ott, R. S. Jones, R. Jones Jr., J. Baxter, B. McCollum, A. Wargo
III, and J. Donley. 2009. A vertical-lift incubator (the “Seesaw”) designed
for Channel Catfish egg masses. North American Journal of Aquaculture
71:354–359.
Torrans, L., and J. Steeby. 2006. Oxygen management at Channel Catfish
hatcheries. Global Aquaculture Advocate 9(June):56–58.
Torrans, L., and J. Steeby. 2008. Effects of dissolved oxygen concentration on
oxygen consumption and development of Channel Catfish eggs and fry: im-
plications for hatchery management. North American Journal of Aquaculture
70:286–295.
Tucker, C. S. 1991. Water quantity and quality requirements for Channel Catfish
hatcheries. Southern Regional Aquaculture Center, SRAC Publication 461,
Stoneville, Mississippi.
Tucker, C. S., and E. H. Robinson. 1990. Channel Catfish farming handbook.
Van Nostrand Reinhold, New York.
USDA (U.S. Department of Agriculture). 2010. Catfish 2010 part I: reference
of catfish health and production practices in the United States, 2009. USDA,
Animal and Plant Health Inspection Service, Veterinary Services, Report
580.1210, Fort Collins, Colorado.
This article was downloaded by: [Department Of Fisheries]
On: 19 March 2013, At: 23:25
Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,
37-41 Mortimer Street, London W1T 3JH, UK

North American Journal of Aquaculture

Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/unaj20

Decreased Hatchery Rearing Density Improves

Poststocking Harvest and Return to Spawning of
Landlocked Fall Chinook Salmon
a a a a
Michael E. Barnes , Matthew M. Wipf , Nola R. Domenici , Wendy M. Kummer & Robert
b
P. Hanten
a
South Dakota Department of Game, Fish and Parks, McNenny State Fish Hatchery, 19619
Trout Loop, Spearfish, South Dakota, 57783, USA
b
South Dakota Department of Game, Fish and Parks, 20641 South Dakota Highway 1806, Fort
Pierre, South Dakota, 57532, USA
Version of record first published: 19 Mar 2013.

To cite this article: Michael E. Barnes , Matthew M. Wipf , Nola R. Domenici , Wendy M. Kummer & Robert P. Hanten (2013):
Decreased Hatchery Rearing Density Improves Poststocking Harvest and Return to Spawning of Landlocked Fall Chinook
Salmon, North American Journal of Aquaculture, 75:2, 244-250

To link to this article: http://dx.doi.org/10.1080/15222055.2013.768573

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions

This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to
anyone is expressly forbidden.

The publisher does not give any warranty express or implied or make any representation that the contents
will be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses should
be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims,
proceedings, demand, or costs or damages whatsoever or howsoever caused arising directly or indirectly in
connection with or arising out of the use of this material.
 North American Journal of Aquaculture 75:244–250, 2013

C American Fisheries Society 2013

ISSN: 1522-2055 print / 1548-8454 online

DOI: 10.1080/15222055.2013.768573

COMMUNICATION

Decreased Hatchery Rearing Density Improves Poststocking

Harvest and Return to Spawning of Landlocked Fall
Chinook Salmon

Michael E. Barnes,* Matthew M. Wipf, Nola R. Domenici,

and Wendy M. Kummer
South Dakota Department of Game, Fish and Parks, McNenny State Fish Hatchery, 19619 Trout Loop,
Spearfish, South Dakota 57783, USA

Robert P. Hanten
Downloaded by [Department Of Fisheries] at 23:25 19 March 2013

South Dakota Department of Game, Fish and Parks, 20641 South Dakota Highway 1806, Fort Pierre,
South Dakota 57532, USA

procedures to improve the ability of hatchery-reared trout to sur-

Abstract
Juvenile landlocked fall Chinook Salmon Oncorhynchus
tshawytscha were coded-wire-tagged and reared in 1.8-m-diameter
circular tanks at a low or high density for up to 52 d prior to stock-
ing in Lake Oahe, South Dakota, during late May of 1999, 2003,
and 2004. Final hatchery rearing densities ranged from 7.29 to
11.72 kg/m3 for the low-density tanks and from 15.02 to 25.22 kg/m3
for the high-density tanks, with the higher densities being at least
double the lower densities in each year. Flows were adjusted to
maintain similar loadings (kg·L−1·min−1) between the treatments
each year. The fish used in this study came from spawns collected
during the October prior to stocking; TL at the end of hatchery
rearing ranged from 103 mm in 1999 to 124 mm in 2004. In each
year-class, the percentage of fish that were harvested by anglers
or that returned to spawn was significantly greater for the lower-
density treatment than for the higher-density treatment. Angler
harvest primarily consisted of age-3 fish, with a small number of
age-4 fish harvested as well. Feed conversion ratios were signifi-
cantly improved in the lower-density tanks relative to the higher-
density tanks in 1999 and 2003, and fish from the lower-density
tanks were also significantly longer just prior to stocking in 1999
and heavier prior to stocking in 2003. To maximize poststocking
survival for landlocked fall Chinook Salmon, lower rearing densi-
ties are recommended.
Poststocking survival of hatchery-reared salmonids released
into natural waters has long been known to be influenced by
hatchery practices. As early as the 1940s, Schuck (1948) identi-
fied the need for critical evaluation and modification of hatchery
vive after release into natural waters. Burrows and Chenoweth
(1970) also emphasized the need to make poststocking survival
the primary concern of hatchery personnel.
Rearing density (kg of fish per unit of rearing space) is one
hatchery practice that has been shown to affect postrelease sur-
vival (Tipping et al. 2004). Increased rearing density may in-
crease abnormal social interactions, stress, disease vulnerability,
and fin erosion and may decrease gill filament quality, physi-
ological development, and poststocking survival (Jones et al.
1996; Ewing et al. 1998; Hosfeld et al. 2009). Increased rear-
ing densities also can cause a reduction in water quality due to
a decrease in dissolved oxygen, an accumulation of metabolic
products and carbon dioxide, and a resulting decline in the pH
level of the water (Person-Le Ruyet et al. 2007).
Survival of Chinook Salmon Oncorhynchus tshawytscha af-
ter stocking appears to be significantly influenced by hatchery
rearing density. Denton (1988), Martin and Wertheimer (1989),
Banks (1994), and Ewing and Ewing (1995) all observed a
significant and uncompensated reduction in Chinook Salmon
survival to adulthood with increases in hatchery rearing density.
Ewing and Ewing (1995) found that percent survival to adult-
hood was negatively related to Chinook Salmon rearing density
in 14 of 15 brood years studied. Martin and Wertheimer (1989)
observed a negative relationship between rearing density and
poststocking survival for two different stocked sizes of Chinook
Salmon.

*Corresponding author: mike.barnes@state.sd.us

Received October 18, 2012; accepted January 14, 2013

244
 COMMUNICATION
The fall Chinook Salmon population in Lake Oahe, South
Dakota, became established in 1979 with fish likely originat-
ing from Lake Michigan broodstock and from the Abernathy
Fish Technology Center hatchery and Spring Creek National
Fish Hatchery in Washington (Lee et al. 1996; Lott et al.
1997). Since 1989, only progeny from Lake Oahe and another
genetically similar landlocked population upstream (Lake
Sakakawea, North Dakota) have been restocked into Lake Oahe.
Natural reproduction in the lake has never been documented
(Marrone and Stout 1997), and the fishery is maintained en-
tirely by hatchery stocking (Barnes et al. 2000). Angler use
of this fishery varies considerably from year to year; 16,384
angler-hours were recorded in 2010 (Longhenry et al. 2011).
Chinook Salmon in Lake Oahe appear to be phenotypically
different from those in the species’ native range (Barnes et al.
2000), and no research has been published concerning the
influence of hatchery rearing densities on the poststocking
Downloaded by [Department Of Fisheries] at 23:25 19 March 2013
survival of Lake Oahe Chinook Salmon. Rearing densities used
for landlocked Chinook Salmon in South Dakota hatcheries
have historically been over twice the densities recommended by
Martin and Wertheimer (1989), Banks (1994), and Ewing and
Ewing (1995) for optimum poststocking survival of Chinook
Salmon on the West Coast and over three times greater than
the optimum density indicated from Denton’s (1988) study
in Alaska. The objective of this study was to evaluate the
effect of reduced rearing densities on poststocking survival of
landlocked fall Chinook Salmon.
METHODS
General rearing procedures.—The rearing portion of this
experiment was conducted at McNenny State Fish Hatchery,
Spearfish, South Dakota, using the progeny of fall Chinook
Salmon that were collected from Lake Oahe. After the fish were
coded-wire-tagged, they were reared at two different densities
in circular tanks (1.8-m diameter; 0.8-m depth) for approxi-
mately 2 months prior to stocking into Lake Oahe. Well water
at a constant temperature of 11◦ C (total hardness as CaCO3 =
360 mg/L; alkalinity as CaCO3 = 210 mg/L; pH 7.6; total dis-
solved solids = 390 mg/L) was used throughout rearing. Flows
in each rearing unit were adjusted to maintain dissolved oxygen
at levels greater than 7.0 mg/L; at the end of hatchery rearing,
flows were approximately 57 L/min in the low-density tanks
and 114 L/min in the high-density tanks. Tag returns were ob-
tained via angler harvest and from adult Chinook Salmon that
ascended a fish ladder prior to spawning at Whitlock’s Spawning
Station (near Gettysburg, South Dakota) or that were obtained
via electrofishing just prior to spawning.
At the start and end of hatchery rearing, 20 fish/tank were
individually weighed to the nearest gram and measured for
TL to the nearest millimeter. To avoid pseudoreplication, the
weights and lengths from the 20 fish/tank were averaged, and
the mean values were used during subsequent data analysis.
The total weight of fish in each tank was recorded to the nearest
245
0.1 kg at the start of rearing and just prior to stocking. Feed
was uniformly dispensed from 0700 to 1900 hours in each
tank by using EWOS 505 automatic feeders (Norco-plast AB,
Sweden). The amount of feed administered per tank and the
number of fish mortalities in each tank were recorded for the
duration of the study. Coded wire tag return data were pooled
from the replicate tanks of each density treatment and were
analyzed using chi-square analysis. Hatchery rearing data were
analyzed with Student’s t-tests using tanks as the experimental
unit. Significance was predetermined at P < 0.05. This study
was replicated during three year-classes of stocking, with the
number of fish tagged and stocked varying from year to year.
All coded-wire-tagged fish of each group were stocked at the
same time each year near Whitlock’s Spawning Station.
1999 methods.—After coded-wire-tagging on April 9, 1999,
one group of 15,000 fish was divided among 12 circular tanks
(∼1,250 fish/tank) so that maximum rearing density was less
than 10 kg/m3 just prior to stocking. Another group of 15,000
fish was put into only five circular tanks (∼3,000 fish/tank),
thus achieving densities of greater than 20 kg/m3 just prior to
stocking. Hatchery rearing duration in 1999 was 52 d.
2003 methods.—After the fish were coded-wire-tagged
(April 8, 2003), a group of 4,000 fish was split into four circular
tanks (∼1,000 fish/tank), attaining a maximum rearing density
less than 8 kg/m3 just prior to stocking. Another group of 3,000
fish was put into only two circular tanks (∼1,500 fish/tank), re-
sulting in densities greater than 18 kg/m3 just prior to stocking.
Individual fish TLs and weights were recorded for 60 fish from
a common pool after coded-wire-tagging and prior to placement
in the circular tanks at the two densities. The hatchery rearing
period in 2003 was 51 d.
2004 methods.—After coded-wire-tagging on March 31,
2004, one group of 15,000 fish was apportioned among 12
circular tanks (∼1,250 fish/tank) so that the maximum rearing
density was less than 10 kg/m3 just prior to stocking. Another
group of 15,000 fish was divided among five circular tanks
(∼3,000 fish/tank), thereby resulting in densities of greater
than 20 kg/m3 just prior to stocking. Hatchery rearing duration
in 2004 was 39 d.
RESULTS
For each of the stocking years, the total number of coded
wire tags returned from the Chinook Salmon reared at lower
densities was significantly greater than the number returned
from fish reared at higher densities (Table 1). For the 1999
stocking year, coded wire tag returns from Chinook Salmon
of the lower-density rearing group were nearly six times the
returns from fish reared at higher densities. For this cohort, tags
were only collected during spawning operations; no tags were
recovered from angler-harvested fish during the period in which
the fish were present in Lake Oahe. For the 2003 stocking year,
tags were recovered from 1.35% of Chinook Salmon belonging
to the lower-density rearing group and from 0.60% of fish in the
 246 BARNES ET AL.

TABLE 1. Returns of coded-wire-tagged landlocked fall Chinook Salmon number of tagged adult returns per hatchery rearing tank was
that were reared at a low or high density prior to stocking in Lake Oahe, South greater from the low-density tanks than from the high-density
Dakota. For a given row, returns with different letters are significantly different
(P < 0.05).
tanks for the 1999 stocking year, but the number of tag returns
per tank was greater from the high-density tanks for the 2003
Rearing density and 2004 stocking years. Tags obtained from angler harvest were
predominantly from age-3 fish, with just a few tags returned
Year Variable Low High from age-4 fish. Tags from spawning Chinook Salmon were
1999 Initial density (kg/m3) 3.05 7.89 also obtained predominantly from age-3 fish, with a very minor
Final density (kg/m3) 7.29 15.02 contribution from age-2 and age-4 individuals.
Final loading (kg·L−1·min−1) 0.26 0.27 Final rearing tank loadings in 1999 were nearly equal be-
Fish coded-wire-tagged (N) 15,000 15,000 tween the treatments: 0.26 kg·L−1·min−1 in the low-density
Tagged fish stocked (N) 14,901 14,875 treatment and 0.27 kg·L−1·min−1 in the high-density treat-
Tag return from anglers (N) 0 0 ment. Final loadings were also very similar in 2004 at
Tag return from anglers (%) 0 0 0.42 kg·L−1·min−1 in the low-density treatment compared with
Tag return during spawning (N) 28 z 5y 0.45 kg·L−1·min−1 in the high-density tanks. The greatest differ-
Tag return during spawning (%) 0.19 0.03 ence in final loadings occurred during 2003, as the low-density
tanks had a final loading of 0.26 kg·L−1·min−1 compared with
Downloaded by [Department Of Fisheries] at 23:25 19 March 2013

Total tag returns (N) 28 z 5y

Total tag returns (%) 0.19 0.03 0.33 kg·L−1·min−1 for the high-density tanks.
Tag returns (N) per rearing tank 2.3 1.0 Feed conversion ratios were significantly different between
low- and high-density treatment tanks during 1999 and 2003 but
2003 Initial density (kg/m3) 4.73 13.30
not during 2004 (Table 2). There were no significant differences
Final density (kg/m3) 7.29 18.27
in mortality between the two density treatments in any of the
Final loading (kg·L−1·min−1) 0.26 0.33
3 years of the experiment. Individual TLs measured just prior to
Fish coded-wire-tagged (N) 4,000 6,000
stocking in 1999 were significantly greater for fish reared at the
Tagged fish stocked (N) 3,991 5,990
lower density than for fish reared at the higher density (Table 3).
Tag return from anglers (N) 15 z 3y
There were no other significant differences in fish TL, weight,
Tag return from anglers (%) 0.38 0.05
or condition factor at the end of hatchery rearing for any of the
Tag return during spawning (N) 39 33
study years.
Tag return during spawning (%) 0.98 0.55
Total tag returns (N) 54 z 36 y
Total tag returns (%) 1.35 0.60
DISCUSSION
Tag returns (N) per rearing tank 13.5 18
The greater adult returns of Lake Oahe fall Chinook Salmon
2004 Initial density (kg/m3) 5.57 13.00 when reared at lower densities in this study are supported by
Final density (kg/m3) 11.72 25.22 numerous other investigations of hatchery rearing densities and
Final loading (kg·L−1·min−1) 0.42 0.45 the poststocking survival of salmonid species in their native
Fish coded-wire-tagged (N) 15,000 15,000 ranges. Martin and Wertheimer (1989) compared final high-
Tagged fish stocked (N) 14,973 14,944 density (20.3 kg/m3) and low-density (6.6 kg/m3) treatments
Tag return from anglers (N) 41 32 similar to those we examined, and they found a significant in-
Tag return from anglers (%) 0.27 0.21 crease in returns of fall Chinook Salmon. Banks (1994) observed
Tag return during spawning (N) 108 z 55 y a poor adult return rate of only 0.13% from hatchery spring
Tag return during spawning (%) 0.72 0.37 Chinook Salmon reared for 276 d at a relatively high density
Total tag returns (N) 149 z 87 y (44 kg/m3) compared with a much greater return of 0.44% from
Total tag returns (%) 1.00 0.58 fish reared at a lower density (14.6 kg/m3). Conversely, Banks
Tag returns (N) per rearing tank 12.4 17.4 and LaMotte (2002) found only slight differences in returns of
Tule-strain fall Chinook Salmon reared in a hatchery for 86 d
at high (27.9 kg/m3) and low (7.0 kg/m3) densities, perhaps in-
higher-density rearing group. The number of Chinook Salmon dicating that density effects are influenced by the duration of
harvested by anglers in 2003 was significantly greater from the hatchery rearing. At only 39–52 d, the hatchery rearing peri-
lower-density group than from the higher-density group. With ods used in the present study were considerably shorter than
an overall tag return of 236 (0.79%), the largest number of tag the 86–365 d used in other poststocking evaluations of Chinook
recoveries was obtained from the fish stocked in 2004. For the Salmon density (Schreck et al. 1985; Bagley et al. 1994; Banks
2004 stocking year, the recovery of tagged fish from the lower- 1994). It is unknown whether the rearing of landlocked Chinook
density group was 1.00%, which was significantly greater than Salmon for longer durations at lower densities would produce
the return of 0.58% from fish reared at the higher density. The even greater poststocking returns.
 COMMUNICATION 247

TABLE 2. Mean (±SE) total tank rearing variables for landlocked fall Chinook Salmon that were reared at a low or high density prior to stocking in Lake Oahe.
Feed conversion ratio was calculated as (weight of food fed)/(fish weight gain). For a given row, means with different letters are significantly different (P < 0.05).

Rearing density
Year Variable Low High
1999 Number of tanks 12 5
Rearing period (d) 52 52
Fish per tank 1,250 3,000
Initial weight (kg) 6.2 16.0
Final weight (kg) 14.8 ± 0.2 30.5 ± 1.3
Weight gain (kg) 8.6 ± 0.2 14.5 ± 1.3
Food fed (kg) 10.3 25.5
Feed conversion ratio 1.21 ± 0.03 z 1.81 ± 0.02 y
Mortality (%) 0.83 ± 0.20 0.66 ± 0.08
2003 Number of tanks 4 2
Rearing period (d) 51 51
Downloaded by [Department Of Fisheries] at 23:25 19 March 2013

Fish per tank 1,000 3,000

Initial weight (kg) 9.6 27.0
Final weight (kg) 14.8 ± 0.2 37.1 ± 0.06
Weight gain (kg) 5.2 ± 0.2 10.1 ± 0.8
Food fed (kg) 4.9 14.1
Feed conversion ratio 0.90 ± 0.04 z 1.41 ± 0.12 y
Mortality (%) 0.23 ± 0.10 0.17 ± 0.10
2004 Number of tanks 12 5
Rearing period (d) 39 39
Fish per tank 1,250 3,000
Initial weight (kg) 11.3 26.4
Final weight (kg) 23.8 ± 0.3 51.2 ± 1.0
Weight gain (kg) 12.6 ± 0.3 24.7 ± 0.9
Food fed (kg) 10.8 25.4
Feed conversion ratio 0.90 ± 0.02 1.00 ± 0.03
Mortality (%) 0.18 ± 0.03 0.37 ± 0.10

Ewing and Ewing (1995) stated that the rearing of fall Chi-
nook Salmon at low density provides the most efficient bal-
ance between smolt production and adult yield. Those authors
also noted that adverse environmental conditions at the time of
stocking (e.g., available forage and water temperatures) may in-
tensify the adverse effects of high rearing densities. This likely
occurred during the 1999 stocking year, when food availability
was severely limited in Lake Oahe (Lott et al. 2000), which may
explain the low returns from both low- and high-density treat-
ments for the 1999 stocking year relative to the 2003 and 2004
stocking years. This may also explain why the higher tag returns
per hatchery tank for the low-density tanks in comparison with
the higher-density tanks were only observed for the 1999 stock-
ing year. The greater tag returns per tank from the high-density
tanks in the 2003 and 2004 stocking years indicate that in abso-
lute terms, the total number of Chinook Salmon surviving after
stocking could be enhanced by rearing at higher densities. How-
ever, this would likely only occur during years of favorable envi-
ronmental conditions and only if the numbers of eggs and subse-
quent fry are not limited—unlike the typical situation for Lake
Oahe Chinook Salmon (Barnes et al. 2003; Wipf and Barnes
2012). In addition, to maximize poststocking returns at high
densities, water flow to the rearing tanks cannot be limited and
increased expenditures on fish food would likely be necessary.
The results of this study were probably not affected by hatch-
ery loadings. Loadings were nearly identical between density
treatments in 1999, were very similar between treatments in
2004, and only varied by 0.07 kg·L−1·min−1 in 2003. Banks
(1994) noted that low loadings as a result of extremely high
flows at various rearing densities appeared to increase returns of
spring Chinook Salmon, although the results were inconsistent
and difficult to interpret. In contrast, Banks (1992) observed no
effect of raceway loadings on the poststocking survival of Coho
Salmon O. kisutch smolts.
The improvement in feed conversion for the lower-density
rearing units relative to the higher-density units is not sur-
prising. High hatchery rearing densities have been shown
to increase feed conversion ratios for Chinook Salmon and
 248 BARNES ET AL.

TABLE 3. Mean (±SE) individual fish TL, weight, and condition factor (K = 105 × [weight]/[TL3]) at the beginning and end of the rearing period (Table 2)
for fall Chinook Salmon that were reared at a low or high density prior to stocking in Lake Oahe. For a given row, means with different letters are significantly
different (P < 0.05).

Rearing density
Year Variable Low High
1999 Initial N (tanks) 12 5
Initial weight (g) 4.8 ± 0.1 z 5.4 ± 0.3 y
Initial TL (mm) 86.3 ± 0.4 z 89.6 ± 1.3 y
Initial K 0.72 ± 0.02 0.73 ± 0.03
Final N (tanks) 12 5
Final weight (g) 11.5 ± 0.4 10.5 ± 0.2
Final TL (mm) 107.9 ± 1.0 z 102.9 ± 0.3 y
Final K 0.88 ± 0.01 0.92 ± 0.02
2003 Initial N (individual fish) 60 60
Initial weight (g) 10.4 ± 0.3 10.4 ± 0.3
Downloaded by [Department Of Fisheries] at 23:25 19 March 2013

Initial TL (mm) 105.5 ± 1.1 105.5 ± 1.1

Initial K 0.88 ± 0.02 0.88 ± 0.02
Final N (tanks) 4 2
Final weight (g) 15.8 ± 1.0 13.3 ± 0.6
Final TL (mm) 120.1 ± 2.6 114.6 ± 1.1
Final K 0.87 ± 0.01 0.85 ± 0.02
2004 Initial N (tanks) 12 5
Initial weight (g) 7.5 ± 0.3 8.1 ± 0.4
Initial TL (mm) 101.4 ± 0.4 100.6 ± 0.6
Initial K 0.71 ± 0.03 0.78 ± 0.05
Final N (tanks) 12 5
Final weight (g) 20.3 ± 0.7 19.3 ± 0.6
Final TL (mm) 126.4 ± 1.1 124.1 ± 0.9
Final K 0.99 ± 0.01 0.99 ± 0.02

other salmonids (Piper et al. 1982; Mazur and Iwama 1993;
Mazur et al. 1993; Procarione et al. 1999). Holm et al. (1990)
and Ewing et al. (1998) concluded that growth rates during
hatchery production were inversely correlated with rearing
density. However, elevated loadings, rather than just increased
densities, may be the cause of decreased growth and feed
conversion in more-domesticated salmonids, such as Rainbow
Trout O. mykiss (Procarione et al. 1999; Ellis et al. 2002; North
et al. 2006; Person-Le Ruyet et al. 2007). Because flows in the
current study were adjusted in each tank to maintain similar
loadings between the treatments, it is likely that the differences
in feed conversion ratio between densities were due primarily
to density effects. The lack of a significant difference between
density treatments in 2004 may have been due to the relatively
short duration of the experimental rearing period (only 39 d).
Other short-duration (30–35 d) studies have found no density
effects on feed conversion (Wagner et al. 1996; Iguchi et al.
2003). However, studies with trial durations exceeding 270 d
have shown significant density effects on feed conversion
(Banks 1992, 1994; Bagley et al. 1994).
As was observed during the 1999 study year, individual
fish growth rates are often negatively related to rearing density
(Holm et al. 1990; Kebus et al. 1992; Procarione et al. 1999).
These density-dependent differences in growth may be due to
increased competition, increased aggression, or decreased vis-
ibility (Fenderson and Carpenter 1971; Fagerlund et al. 1981;
Iguchi et al. 2003; Hosfeld et al. 2009; Wipf and Barnes 2011).
Small sample sizes may have precluded the detection of signif-
icant differences in fish TL or weight prior to stocking in 2003,
and the relatively short rearing duration in 2004 may have been
insufficient for size differences to become apparent.
Fish that are reared in hatcheries are typically released into
natural waters in large numbers to maximize the probability of
survival (Aprahamian et al. 2003). However, Brockmark and
Johnsson (2010) and Brockmark et al. (2010) concluded that
Brown Trout Salmo trutta reared at natural densities were bet-
ter equipped for survival after release. The results from the
present study clearly indicate that lower rearing densities im-
prove poststocking survival in landlocked fall Chinook Salmon,
thereby allowing the rearing of fewer fish to achieve similar
 COMMUNICATION
harvest and return-to-spawn numbers. Thus, for situations in
which eggs or fish are limited, as previously occurred in Lake
Oahe (Hanten 2007), lower rearing densities would be desired.
Even if large numbers of eggs and fish are available, the rear-
ing of landlocked Chinook Salmon at higher densities would
appear to produce only minor improvements in poststocking
yield, depending on environmental conditions during the year of
stocking.
ACKNOWLEDGMENTS
We thank Amanda Davis, Rebekah Kelley, Eric Krebs, Will
Sayler, Rick Cordes, and Kyle Potter for their assistance with
this study.
REFERENCES
Aprahamian, M. W., K. Martin Smith, P. McGinnity, S. McKelvey, and J.
Downloaded by [Department Of Fisheries] at 23:25 19 March 2013
Taylor. 2003. Restocking of salmonids—opportunities and limitations. Fish-
eries Research 62:211–227.
Bagley, M. J., B. Bentley, and G. A. E. Gall. 1994. A genetic evaluation of
the influence of stocking density on the early growth of Rainbow Trout
(Oncorhynchus mykiss). Aquaculture 121:313–326.
Banks, J. L.1992. Effects of density and loading on Coho Salmon during hatch-
ery rearing and after release. Progressive Fish-Culturist 54:137–147.
Banks, J. L. 1994. Raceway density and water flow as factors affecting spring
Chinook Salmon (Oncorhynchus tshawytscha) during rearing and after re-
lease. Aquaculture 119:201–217.
Banks, J. L., and E. M. LaMotte. 2002. Effects of four density levels on Tule fall
Chinook Salmon during hatchery rearing and after release. North American
Journal of Aquaculture 64:24–33.
Barnes, M. E., R. J. Cordes, W. A. Sayler, and R. P. Hanten. 2003. Soft-
egg disease in landlocked fall Chinook Salmon eggs: possible causes and
therapeutic treatments. North American Journal of Aquaculture 65:126–133.
Barnes, M. E., R. P. Hanten, R. J. Cordes, W. A. Sayler, and J. Carreiro. 2000.
Reproductive performance of inland fall Chinook Salmon. North American
Journal of Aquaculture 62:203–211.
Brockmark, S., B. Adriaenssens, and J. I. Johnsson. 2010. Less is more: density
influences the development of behavioural life skills in trout. Proceedings of
the Royal Society of London B 277:3035–3043.
Brockmark, S., and J. I. Johnsson. 2010. Reduced hatchery rearing density in-
creases social dominance, postrelease growth, and survival in Brown Trout
(Salmo trutta). Canadian Journal of Fisheries and Aquatic Sciences 67:
288–295.
Burrows, R. E., and H. H. Chenoweth. 1970. The rectangular circulating rearing
pond. Progressive Fish-Culturist 32:67–80.
Denton, C. 1988. Marine survival of Chinook Salmon, Oncorhynchus
tshawytscha, reared at three densities. Alaska Department of Fish and Game,
Division of Fisheries Rehabilitation, Enhancement, and Development, Report
88, Juneau.
Ellis, T., B. North, A. P. Scott, N. R. Bromage, M. Porter, and D. Gadd. 2002.
The relationships between stocking density and welfare in farmed Rainbow
Trout. Journal of Fish Biology 61:493–531.
Ewing, R. D., and S. K. Ewing. 1995. Review of the effects of rearing density on
survival to adulthood for Pacific salmon. Progressive Fish-Culturist 57:1–25.
Ewing, R. D., J. E. Sheahan, M. A. Lewis, and A. N. Palmisano. 1998. Effects
of rearing density and raceway conformation on growth, food conversion,
and survival of juvenile spring Chinook Salmon. Progressive Fish-Culturist
60:167–178.
Fagerlund, U. H. M., J. R. McBride, and E. T. Stone. 1981. Stress-related effects
of hatchery rearing density on Coho Salmon. Transactions of the American
Fisheries Society 110:644–649.
249
Fenderson, O. C., and M. R. Carpenter. 1971. Effects of crowding on the
behaviour of juvenile hatchery and wild landlocked Atlantic Salmon (Salmo
salar L.). Animal Behaviour 19:439–447.
Hanten, R. P. 2007. 2005 and 2006 Whitlocks Bay spawning station report.
South Dakota Department of Game, Fish and Parks, Annual Report 07-05,
Pierre.
Holm, J. C., T. Refstie, and S. Bø. 1990. The effect of fish density and
feeding regimes on individual growth rate and mortality in Rainbow Trout
(Oncorhynchus mykiss). Aquaculture 89:225–232.
Hosfeld, C. D., J. Hammer, S. O. Handeland, S. Fivelstad, and S. O.
Stefansson. 2009. Effects of fish density on growth and smoltification in
intensive production of Atlantic Salmon (Salmo salar L.). Aquaculture 294:
236–241.
Iguchi, K., K. Ogawa, M. Nagae, and F. Ito. 2003. The influence of rearing
density on stress response and disease susceptibility of Ayu (Plecoglossus
altivelis). Aquaculture 220:515–523.
Jones, R. N., and W. H. Miller. 1996. An evaluation of rearing density in relation
to post-release smolt survival and adult returns of spring Chinook Salmon at
Dworshak National Fish Hatchery in Idaho. U.S. Fish and Wildlife Service,
Idaho Fishery Resource Office, Ahsahka.
Kebus, M. J., M. T. Collins, M. S. Brownfield, C. H. Amundson, T. B. Kayes,
and J. A. Malison. 1992. Effects of rearing density on the stress response and
growth of Rainbow Trout. Journal of Aquatic Animal Health 4:1–6.
Lee, J. D., G. J. Power, and J. C. Hendrickson. 1996. Lake Sakakawea’s Chi-
nook Salmon fishery, 1976–1995. North Dakota Game and Fish Department,
Fisheries Investigation Report 20, Bismarck.
Longhenry, C., K. Edwards, R. Hanten, and K. Potter. 2011. Annual fish popu-
lation and angler use, harvest, and preference surveys on Lake Oahe, South
Dakota, 2010. South Dakota Department of Game, Fish and Parks, Wildlife
Division, Annual Report 11-06, Pierre.
Lott, J., R. Hanten, B. Johnson, and W. Nelson-Stastny. 2000. Annual fish
population and angler use, harvest and preference surveys on Lake Oahe,
South Dakota, 1999. South Dakota Department of Game, Fish and Parks,
Wildlife Division, Annual Report 00-11, Pierre.
Lott, J., G. Marrone, and D. Stout. 1997. Influence of size-and-date at stocking,
imprinting attempts and growth on initial survival, homing ability, maturation
patterns and angler harvest of Chinook Salmon in Lake Oahe, South Dakota.
South Dakota Department of Game, Fish and Parks, Wildlife Division, Special
Report 97-20, Pierre.
Martin, R. M., and A. Wertheimer. 1989. Adult production of Chinook Salmon
reared at different densities and released as two smolt sizes. Progressive
Fish-Culturist 51:194–200.
Mazur, C. F., and G. K. Iwama. 1993. Handling and crowding stress reduces
number of plaque-forming cells in Atlantic Salmon. Journal of Aquatic Ani-
mal Health 5:98–101.
Mazur, C. F., D. Tillapaugh, J. R. Brett, and G. K. Iwama. 1993. The effect
of feeding level and rearing density on growth, feed conversion and sur-
vival in Chinook Salmon (Oncorhynchus tshawytscha) reared in salt water.
Aquaculture 117:129–140.
North, B. P., J. F. Turnbull, T. Ellis, M. J. Porter, H. Migaud, J. Bron, and N. R.
Bromage. 2006. The impact of stocking density on the welfare of Rainbow
Trout (Oncorhynchus mykiss). Aquaculture 255:466–479.
Person-Le Ruyet, J., N. Le Bayon, and S. Gros. 2007. How to assess fin damage
in Rainbow Trout, Oncorhynchus mykiss? Aquatic Living Resources 20:
191–195.
Piper, R. G., I. B. McElwain, L. E. Orme, J. P. McCraren, L. G. Fowler, and J. R.
Leonard. 1982. Fish hatchery management. U.S. Fish and Wildlife Service,
Washington, D.C.
Procarione, L. S., T. P. Barry, and J. A. Malison. 1999. Effects of high rearing
densities and loading rates on the growth and stress responses of juvenile
Rainbow Trout. North American Journal of Aquaculture 61:91–96.
Schreck, C. B., R. Patiño, C. K. Pring, J. R. Winton, and J. E. Holway. 1985.
Effects of rearing density on indices of smoltification and performance of
Coho Salmon, Oncorhynchus kisutch. Aquaculture 45:345–358.
 250 BARNES ET AL.
Schuck, H. A. 1948. Survival of hatchery trout in streams and possible methods
of improving the quality of hatchery trout. Progressive Fish-Culturist 10:
3–14.
Tipping, J. M., J. B. Byrne, and T. D. Hillson. 2004. The effect of juvenile
rearing density on adult returns of summer steelhead. North American Journal
of Aquaculture 66:319–324.
Wagner, E. J., S. S. Intelmann, and M. D. Routledge. 1996. The effects of
fry rearing density on hatchery performance, fin condition, and agonistic
Downloaded by [Department Of Fisheries] at 23:25 19 March 2013
behavior of Rainbow Trout Oncorhynchus mykiss fry. Journal of the World
Aquaculture Society 27:264–274.
Wipf, M. M., and M. E. Barnes. 2011. Competitor density and size effects on
aggression and feeding in Cutthroat Trout: implications for aquaculture. Open
Fish Science Journal 4:62–66.
Wipf, M. M., and M. E. Barnes. 2012. Parental male effects on landlocked fall
Chinook Salmon progeny survival. North American Journal of Aquaculture
74:443–448.
This article was downloaded by: [Department Of Fisheries]
On: 08 April 2013, At: 19:54
Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,
37-41 Mortimer Street, London W1T 3JH, UK

North American Journal of Aquaculture

Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/unaj20

Introduction to a Special Section: Lipids in Aquaculture

Nutrition and Physiology II
a b
Jesse T. Trushenski & Rebecca T. Lochmann
a
Fisheries and Illinois Aquaculture Center and Departments of Zoology and Animal Science,
Food and Nutrition, Southern Illinois University–Carbondale, 1125 Lincoln Drive, Carbondale,
Illinois, 62901, USA
b
Aquaculture/Fisheries Center of Excellence, University of Arkansas at Pine Bluff, 1200
North University Drive, Mail Slot 4912, Pine Bluff, Arkansas, 71610, USA
Version of record first published: 25 Mar 2013.

To cite this article: Jesse T. Trushenski & Rebecca T. Lochmann (2013): Introduction to a Special Section: Lipids in
Aquaculture Nutrition and Physiology II, North American Journal of Aquaculture, 75:2, 251-251

To link to this article: http://dx.doi.org/10.1080/15222055.2013.781905

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions

This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to
anyone is expressly forbidden.

The publisher does not give any warranty express or implied or make any representation that the contents
will be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses should
be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims,
proceedings, demand, or costs or damages whatsoever or howsoever caused arising directly or indirectly in
connection with or arising out of the use of this material.
 North American Journal of Aquaculture 75:251, 2013

C American Fisheries Society 2013

ISSN: 1522-2055 print / 1548-8454 online

DOI: 10.1080/15222055.2013.781905

SPECIAL SECTION: LIPIDS IN AQUACULTURE NUTRITION AND PHYSIOLOGY II

Introduction to a Special Section: Lipids in Aquaculture

Nutrition and Physiology II
Jesse T. Trushenski
Fisheries and Illinois Aquaculture Center and Departments of Zoology and Animal Science,
Food and Nutrition, Southern Illinois University–Carbondale, 1125 Lincoln Drive, Carbondale,
Illinois 62901, USA

Rebecca T. Lochmann
Aquaculture/Fisheries Center of Excellence, University of Arkansas at Pine Bluff,
1200 North University Drive, Mail Slot 4912, Pine Bluff, Arkansas 71610, USA
Downloaded by [Department Of Fisheries] at 19:54 08 April 2013

Beginning in 2009, the Fish Culture Section of the Ameri-
can Fisheries Society and the U.S. Aquaculture Society began
hosting joint symposia to highlight our evolving understanding
of lipid function and metabolism in fish and shellfish, with an
eye toward optimal nutrition and the judicious use of fish oil
and other lipids in aquafeeds. The papers from these early sym-
posia were published in 2011 in the North American Journal of
Aquaculture as a special section entitled “Lipids in Aquaculture
Nutrition and Physiology.” Although significant advances have
been made, the need for deeper and more comprehensive insight
into the roles of lipids and fatty acids is ongoing: fish oil and
its constitutive nutrients are currently recognized as the most
limiting resources in feeding cultured fish, and this is likely to
be the case into the future. Unless this bottleneck is adequately
addressed, our ability to produce fish—for food and for fish-
eries enhancement—will be constrained. Recognizing the need
for more and better information, our organizations continue to
collaborate in hosting annual symposia to address the issue of
lipids in aquafeeds and aquatic animals. To this end, we proudly
present “Lipids in Aquaculture Nutrition and Physiology II” to
fish culturists, aquaculture nutritionists, and the broader reader-
ship of the North American Journal of Aquaculture.

251
This article was downloaded by: [Department Of Fisheries]
On: 08 April 2013, At: 19:55
Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,
37-41 Mortimer Street, London W1T 3JH, UK

North American Journal of Aquaculture

Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/unaj20

The Effects of Diets Containing Standard Soybean

Oil, Soybean Oil Enhanced with Conjugated Linoleic
Acids, Menhaden Fish Oil, or an Algal Docosahexaenoic
Acid Supplement on Channel Catfish Performance,
Body Composition, Sensory Evaluation, and Storage
Characteristics
a b c a a
Jimmy Faukner , Steven D. Rawles , Andrew Proctor , Todd D. Sink , Ruguang Chen ,
a a
Harold Philips & Rebecca T. Lochmann
a
Department of Aquaculture and Fisheries, University of Arkansas at Pine Bluff, 1200 North
University Drive, Pine Bluff, Arkansas, 71601, USA
b
Harry K. Dupree Stuttgart National Aquaculture Research Center, Post Office Box 1050,
Stuttgart, Arkansas, 72160, USA
c
Department of Food Science, University of Arkansas, 2650 North Young Avenue,
Fayetteville, Arkansas, 72704, USA
Version of record first published: 25 Mar 2013.

To cite this article: Jimmy Faukner , Steven D. Rawles , Andrew Proctor , Todd D. Sink , Ruguang Chen , Harold Philips &
Rebecca T. Lochmann (2013): The Effects of Diets Containing Standard Soybean Oil, Soybean Oil Enhanced with Conjugated
Linoleic Acids, Menhaden Fish Oil, or an Algal Docosahexaenoic Acid Supplement on Channel Catfish Performance, Body
Composition, Sensory Evaluation, and Storage Characteristics, North American Journal of Aquaculture, 75:2, 252-265

To link to this article: http://dx.doi.org/10.1080/15222055.2012.713896

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions

This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to
anyone is expressly forbidden.

The publisher does not give any warranty express or implied or make any representation that the contents
will be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses should
be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims,
proceedings, demand, or costs or damages whatsoever or howsoever caused arising directly or indirectly in
connection with or arising out of the use of this material.
 North American Journal of Aquaculture 75:252–265, 2013

C American Fisheries Society 2013

ISSN: 1522-2055 print / 1548-8454 online

DOI: 10.1080/15222055.2012.713896

SPECIAL SECTION: LIPIDS IN AQUACULTURE NUTRITION AND PHYSIOLOGY II

The Effects of Diets Containing Standard Soybean Oil,

Soybean Oil Enhanced with Conjugated Linoleic Acids,
Menhaden Fish Oil, or an Algal Docosahexaenoic Acid
Supplement on Channel Catfish Performance, Body
Composition, Sensory Evaluation, and Storage
Characteristics
Jimmy Faukner
Downloaded by [Department Of Fisheries] at 19:55 08 April 2013

Department of Aquaculture and Fisheries, University of Arkansas at Pine Bluff,

1200 North University Drive, Pine Bluff, Arkansas 71601, USA

Steven D. Rawles
Harry K. Dupree Stuttgart National Aquaculture Research Center, Post Office Box 1050, Stuttgart,
Arkansas 72160, USA

Andrew Proctor
Department of Food Science, University of Arkansas, 2650 North Young Avenue, Fayetteville,
Arkansas 72704, USA

Todd D. Sink, Ruguang Chen, Harold Philips, and Rebecca T. Lochmann*

Department of Aquaculture and Fisheries, University of Arkansas at Pine Bluff,
1200 North University Drive, Pine Bluff, Arkansas 71601, USA

Abstract
Fish consumption is a common method of obtaining beneficial n-3 highly unsaturated fatty acids (LC-PUFAs),
but increased use of vegetable oils in fish diets to reduce dependence on fish oil dilutes the amounts of LC-PUFAs.
Conjugated linoleic acids (CLAs) are also considered beneficial for human health. Therefore, we investigated four
different lipid sources in Channel Catfish Ictalurus punctatus diets for their ability to enhance fatty acid profiles of
fillets to benefit human health while maintaining or improving fish performance. In a 175-d grow-out trial, Channel
Catfish (71.4 ± 0.1 g [mean ± SE]) were fed a commercial 32% protein diet supplemented with 2% lipid from
soybean oil (SO), soybean oil enhanced with conjugated linoleic acids, menhaden fish oil (FO), or an algal supplement
of Schizochytrium sp. high in 22:6(n-3) (docosahexaenoic acid, DHA). Diet effects were assessed by measuring fish
growth performance, muscle proximate and fatty acid composition, sensory characteristics of fillets, consumer taste
preferences, and oxidative stability of fillets during cold storage. There were no differences in fish growth performance
or proximate composition. Only fish fed the CLA diet contained CLAs in the muscle. Fish fed the FO and algal DHA
diets had higher concentrations of 22:6(n-3) in the muscle compared with fish fed the SO and CLA diets. Sensory
evaluation and consumer preference testing were more favorable for fillets from fish fed the SO and CLA diets than
from fish fed the FO and algal DHA diets. There were no differences in storage characteristics of fish refrigerated at
4◦ C for 2 weeks or frozen at −18◦ C for 4 weeks. Fillets from fish fed the FO diet yielded the highest concentration of
fatty acids for human health benefits, followed by the fillets from fish fed the algal DHA diets. The CLA diet produced
increased fillet concentrations of CLAs.

*Corresponding author: rlochmann@uaex.edu

Received March 29, 2012; accepted July 10, 2012
Published online March 25, 2013

252
 EFFECT OF LIPID SUPPLEMENTS ON CATFISH PERFORMANCE
The U.S. catfish industry generated US$500 million in rev-
enue during 2000 (Hanson and Sites 2006) and $423 million in
2011 (NASS 2012). Despite recent setbacks with global compe-
tition from other catfish species, rising feed costs, and increas-
ing fuel costs, Channel Catfish Ictalurus punctatus remains the
most important aquaculture finfish species in the United States.
The suitability of feed ingredients for commercial production
of both fish and feed are not only based on nutrient content,
but also economics and availability (Lochmann and Phillips
1995). The influence of dietary ingredients on product qual-
ity is also important, and the lipid composition of the fish is
relatively easy to manipulate through the diet. Known human
health benefits associated with n-3 highly unsaturated fatty acids
(LC-PUFAs), such as 20:5(n-3)1 and 22:6(n-3), include the re-
duction of cardiovascular disease, arthritis, atherosclerosis, dia-
betes, and cancer (Horrocks and Yeo 1999; Arterburn et al. 2006;
Simopoulos 2008). With increasing awareness of the health ben-
Downloaded by [Department Of Fisheries] at 19:55 08 April 2013
efits of different dietary lipids (Muskiet et al. 2004; Wang et al.
2006) the ability to market catfish enriched with n-3 or other
healthy fatty acids as a functional food could be a key factor
in restoring the competitiveness of the industry. The American
Dietetic Association defines functional foods as “whole foods
and fortified, enriched, or enhanced foods that have a potentially
beneficial effect on health when consumed as part of a varied
diet on a regular basis, at effective levels” (Hasler and Brown
2009).
Fish are the most common source of n-3 LC-PUFAs in most
human diets (Tocher 2003), but cultured Channel Catfish are
low in these fatty acids due to minimal use of marine fish meals
and oils in their diets (USDA–ARS 2012). Channel Catfish can
elongate and desaturate 18:3(n-3) to form LC-PUFAs so that 1–
2% dietary 18:3(n-3) or 0.5–0.75% n-3 LC-PUFAs will satisfy
the n-3 essential fatty acid requirement for growth (Satoh et al.
1989a, 1989b). Menhaden fish oil enhances catfish growth in
some studies (Stickney and Andrews 1972; Santha and Gatlin
1991) but not in others (Manning et al. 2007; Yildirim-Aksoy
et al. 2007). However, the concentration of n-3 LC-PUFAs in-
creases consistently in tissues of catfish fed dietary menhaden
fish oil. Unfortunately, the cost of menhaden fish oil is projected
to rise and there are no unexploited sources for this commodity
(Naylor et al. 2009). Nonfish sources of 22:6(n-3) (docosahex-
aenoic acid, DHA) such as algae (e.g., Schizochytrium sp.) have
also been tested in catfish diets. Li et al. (2009) found that catfish
fed diets with 1.0% or 1.5% Schizochytrium algae gained more
weight than fish fed diets with 0% or 0.5% algae. The n-3 LC-
PUFA concentration also increased in the fillet as algal inclusion
increased in the diet. However, algal oils are currently more ex-
pensive than fish oils, so alternative lipids are still needed to
1In fatty acid designations of this nature, the number to the left of the colon
is the number of carbon atoms in the compound, the number immediately to
the right of the colon is the number of double bonds, and the number after the
hyphen indicates the position of the first double bond from the methyl end.
253
support fish performance and enhance the quality of the fillets
while maintaining production profitability.
There are few alternative animal and plant lipids that contain
substantial amounts of n-3 LC-PUFAs (Gunstone 2011), but
there are other fatty acids with documented health benefits. For
instance, conjugated linoleic acids (CLAs) are linked to human
health benefits such as weight loss and reduced incidence of
cardiovascular problems and some cancers (Roche et al. 2001;
Benjamin and Spener 2009). The CLAs are n-6 fatty acids with
18 carbons and trans-double bonds that have been incorporated
into fillets of many cultured fish species to improve the fatty
acid profile without using fish oils (Berge et al. 2004; Bandarra
et al. 2006; Valente et al. 2007). Studies on catfish show that
dietary CLAs have few effects on fish weight gain and general
performance (Twibell and Wilson 2003; Manning et al. 2006).
However, Manning et al. (2006) documented a concentration-
dependent accumulation of total CLA isomers in the tissue,
which could be used to market catfish as a functional food.
Standard soybean oil (SO), CLA-enhanced soybean oil, algal
DHA, and menhaden fish oil (FO) have not been evaluated as
dietary lipid sources in Channel Catfish concurrently. Therefore,
we investigated the use of SO, CLA-enhanced soybean oil, algal
DHA, and FO in Channel Catfish diets to enhance fatty acid
profiles of fillets for human health benefits while maintaining
growth, survival, feed conversion, and sensory characteristics
and oxidative stability of the fillets.
METHODS
Diet manufacture.—Diets were prepared at the Harry K.
Dupree Stuttgart National Aquaculture Research Center, U.S.
Department of Agriculture, Agricultural Research Service,
Stuttgart, Arkansas. The basal diet was a commercial premium
extruded catfish feed with 32% protein and 5% lipid (ARKAT,
Dumas, Arkansas). Supplemental lipid sources were SO (con-
trol), SO enhanced with CLAs (CLA), an algal source of DHA
(Schizochytrium sp.) combined with soybean oil, and refined
FO. Fatty acid composition of the lipid supplements is shown
in Table 1. Standard soybean oil (Wesson) was used as a con-
trol because it contains neither conjugated linoleic acids nor n-3
LC-PUFAs and has a fatty acid profile similar to that of commer-
cial catfish diets. Soybean oil enhanced with CLAs (novel oil
product, University of Arkansas, Fayetteville; Jain and Proctor
2008) contained 12% total CLA isomers. The DHA algal sup-
plement (Aquagrow Gold, Advanced BioNutrition, Columbia,
Maryland; now Martek Biosciences) contained 19% 22:6(n-3)
from a nonfish source, and is potentially more sustainable than
marine fish oils. Thirteen grams of DHA algal supplement were
added to 7 g of soybean oil, and 20 g of the mixture was added
to each kilogram of diet to achieve supplementation of 2% total
lipid to the diet. Refined menhaden fish oil (Virginia Prime Gold,
Omega Protein, Houston, Texas) contained 28% n-3 LC-PUFAs
and 9% 22:6(n-3).
 254 FAUKNER ET AL.

TABLE 1. Selected mean ± SE fatty acid composition (percentage of total fatty acids by weight) of lipid supplements added to a 32% protein commercial
catfish diet to produce four experimental diets for Channel Catfish. The lipids (and diet designations) were soybean oil (SO), CLA-enriched soybean oil (CLA),
algal DHA (DHA), and menhaden fish oil (FO).a

Experimental diet
Fatty acid SO CLA DHA FO
Saturated fatty acidsb 16.2 ± 0.01 15.8 ± 0.0 19.9 37.2 ± 0.02
Monounsaturated fatty acidsc 21.9 ± 0.03 24.1 ± 0.3 1.9 29.4 ± 0.01
18:2(n-6) 53.9 ± 0.1 41.8 ± 0.2 0.0 1.3 ± 0.002
CLA: 9cis, 11trans 0.0 ± 0.0 1.0 ± 0.01 0.0 0.0 ± 0.0
CLA: 10trans, 12cis 0.0 ± 0.0 1.0 ± 0.04 0.0 0.0 ± 0.0
CLA trans isomersd 0.0 ± 0.0 8.4 ± 0.02 0.0 0.0 ± 0.0
CLA isomerse 0.0 ± 0.0 11.7 ± 0.1 0.0 0.0 ± 0.0
18:3(n-3) 7.3 ± 0.01 4.7 ± 0.1 0.0 1.3 ± 0.001
20:4(n-6) 0.0 ± 0.0 0.3 ± 0.004 0.2 1.9 ± 0.006
20:5(n-3) 0.0 ± 0.0 0.0 ± 0.0 0.5 15.8 ± 0.02
22:6(n-3) 0.0 ± 0.0 0.0 ± 0.0 19.4 8.9 ± 0.01
Downloaded by [Department Of Fisheries] at 19:55 08 April 2013

n-3f 7.3 ± 0.03 4.8 ± 0.1 26.7 29.4 ± 0.02

n-6g 53.9 ± 0.1 42.7 ± 0.2 0.5 3.9 ± 0.01
n-3 LC-PUFAsh 0.0 ± 0.0 0.0 ± 0.0 26.7 28.1 ± 0.01
n-6 LC-PUFAsi 0.0 ± 0.0 0.3 ± 0.004 0.2 2.2 ± 0.004
n-3 : n-6 ratio 0.1 ± 0.001 0.1 ± 0.001 53.5 7.6 ± 0.01
a
Fatty acid analysis was conducted in duplicate for the SO, CLA, and FO oils. Fatty acid analysis for DHA (Aquagrow Gold) is from the manufacturer (Martek Biosciences, Columbia,
Maryland) and from Doughman et al. (2007). All lipid supplements were 100% lipid except for the Aquagrow Gold supplement, which contained 55.57% crude fat.
b
Includes 14:0, 16:0, 18:0, and 20:0.
c
Includes 14:1, 16:1, 18:1, 20:1, 22:1, and 24:1.
d
Includes 9trans, 11trans; 10trans, 12trans; and 11trans, 13trans.
e
Includes 9cis,11trans; 10trans,12cis; 9cis,11cis/10cis, 12trans; 9trans, 11trans; 10trans,12trans; and 11trans, 13trans.
f
Includes 18:3(n-3), 20:5(n-3), 22:5(n-3), and 22:6(n-3).
g
Includes 18:2(n-6) (excluding CLA isomers), 18:3(n-6), 20:3(n-6), 20:4(n-6), and 22:4(n-6).
h
Includes 20:5n-3, 22:5n-3 and 22:6n-3.
i
Includes 20:4(n-6) and 22:4(n-6).

The basal diet was ground with a hammer mill (model LM6,
Kelly Duplex Mill and Manufacturing, Springfield, Ohio) to ob-
tain meal of the appropriate particle size (<1 mm). The basal
diet and 20 g of supplemental lipid source per kilogram of
diet were mixed for 45 min before the addition of 450 mL/kg
water, which was followed by 15 min of additional mixing in
a paddle mixer (model 6369, Marion Mixers, Marion, Iowa).
The diets were compression pelleted (6 mm) in a pellet mill
(model LAB2V41, Ambrette Machinery, Brooklyn, New York)
and dried in a conveyor convection oven (model JS 250, Mid-
dleby Marshall, Morton Grove, Illinois) for 25 min at 250◦ C
prior to drying in the sun to the desired moisture content (∼7%).
Due to discontinued production of the basal premium
extruded catfish feed with 32% protein and 5% lipid by the
manufacturer, an additional set of diets had to be manufactured
with a combination of premium and standard 32% protein
diets (ARKAT). The combination diet contained 710 g of
premium and 290 g of standard formula per kilogram of diet.
Premium and standard diets differed slightly in labeled lipid
contents and contained 5% and 3%, respectively. The con-
centration of supplemental lipid added to the combination
diet remained unchanged at 2%. The initial set of diets was
fed to the catfish for 22 weeks and the diets made from the
combination formula were fed for only the last 3 weeks. To
compensate for changes in the lipid concentration of the basal
diet, sample data of the diets were weighted dependent upon
the feeding duration of each set of diets. The weighted means
of the two sets of diets for proximate analysis and fatty acid
composition are shown in Tables 2 and 3.
Proximate and fatty acid analyses of diets.—Dry matter and
ash were analyzed according to standard methods (AOAC 1995).
Crude fiber was determined according to the ANKOM filter bag
technique (AOCS 2005). Crude protein was analyzed with the
Macro-Kjeldahl method of total nitrogen analysis. Total lipid
was extracted with chloroform–methanol and quantified (Folch
et al. 1957). A 10-mL aliquot from each lipid sample was evap-
orated under nitrogen prior to transesterfication with 14% boron
trifluoride. The resulting fatty acid methyl esters (FAMEs) were
analyzed by gas chromatography (Varian, model CP-3800 fit-
ted with a CP-8200 autosampler, Walnut Creek, California).
The FAMEs were quantified by comparison of retention time
and peak area to those of reference standards (GLC-473b, Nu-
Check Prep, Elysian, Minnesota). Reference standards for CLAs
were used to identify and quantify cis-9, trans-11; trans-9,
 EFFECT OF LIPID SUPPLEMENTS ON CATFISH PERFORMANCE 255

TABLE 2. Mean ± SE proximate composition (percent dry weight basis) outdoor 1,600-L tanks. Each experimental diet was assigned
of experimental diets formulated from a 32% protein commercial catfish diet randomly to four tanks. Tanks were filled with well water and
supplemented with 2% (by weight) of either soybean oil (SO), CLA-enriched
soybean oil (CLA), algal DHA (DHA), or menhaden fish oil (FO) that were fed
operated as a static system, except that full water exchange of
to Channel Catfish for 25 weeks. Values are weighted means of two batches of the tanks occurred twice a week. Each tank was supplied with
diets; the first set of diets was fed to the fish for 22 weeks and the second set for an air stone to provide supplemental aeration.
3 weeks. a,b Feeding and monitoring.—Fish were fed a fixed rate of 1.9%
of their body weight per day to maintain acceptable water
Diet Lipid Protein Dry matter Ash
quality. A subsample of 25 fish from each tank was weighed
Basal 5.1 ± 0.003 32.7 ± 0.2 90.6 ± 0.2 6.8 ± 0.03 every 2 weeks. Water temperature was measured daily and
SO 7.3 ± 0.1 32.3 ± 0.1 89.9 ± 0.1 6.7 ± 0.03 dissolved oxygen concentration was measured 5 d per week
CLA 7.4 ± 0.02 32.3 ± 0.2 90.4 ± 0.1 6.7 ± 0.02 with a YSI 55 dissolved oxygen meter (YSI, Yellow Springs,
DHA 7.1 ± 0.1 32.2 ± 0.2 89.9 ± 0.2 6.8 ± 0.01 Ohio). Mean daily water temperature was 28.1 ± 0.4◦ C and
FO 7.4 ± 0.03 32.4 ± 0.4 91.0 ± 0.1 6.8 ± 0.03 the mean dissolved oxygen concentration was 6.7 ± 0.3 mg/L.
a
Total ammonia, nitrite, pH, water hardness, and salinity were
Proximate composition data are means of triplicate analyses.
b
Crude fiber of the basal diet was 3.6 ± 0.2% (mean ± SE).
measured once a week. The mean ± SE values for the measured
water quality characteristics were as follows: pH, 7.9 ± 0.0;
cis-11/cis-10, trans-12; trans-10, cis-12 (Nu-Check Prep, total ammonia, 0.5 ± 0.3 mg/L; nitrite, 0.4 ± 0.3 mg/L;
Downloaded by [Department Of Fisheries] at 19:55 08 April 2013

Elysian, Minnesota); and trans-9, trans-11 (Matreya LLC,
Pleasant Gap, Pennsylvania) isomers.
Culture system and experimental design.—The feeding trial
lasted 25 weeks. Target final weight of individual fish was 500 g
(market size). Fifty Channel Catfish (individual initial weight,
71.4 ± 0.1 g [mean ± SE]) were stocked into each of 16
hardness, 43 ± 8.3 mg/L; and un-ionized ammonia, 0.002 ±
0.001 mg/L.
Sensory analysis and consumer preference testing.—At week
22, 20 fish per tank were harvested to reduce fish biomass as the
fish increased to harvestable size. Harvested fish were filleted,
and individual fillets from each fish were placed in a sealed

TABLE 3. Selected mean ± SE fatty acid composition (percentage of total fatty acids by weight) of diets formulated from a 32% protein commercial catfish
diet supplemented with 2% (by weight) of either soybean oil (SO), CLA-enriched soybean oil (CLA), algal DHA (DHA), or menhaden fish oil (FO) that were fed
to Channel Catfish for 25 weeks. Values are the weighted means of two batches of diets; the first set of diets was fed to the fish for 22 weeks and the second set for
3 weeks.

Diets
Fatty acid Basal SO CLA DHA FO
Saturated fatty acidsa 25.9 ± 0.2 22.9 ± 0.2 22.9 ± 0.4 25.4 ± 0.1 27.2 ± 0.02
Monounsaturated fatty acidsb 37.1 ± 0.8 32.4 ± 0.1 32.1 ± 0.8 33.2 ± 0.03 33.7 ± 0.03
18:2(n-6) 34.2 ± 0.9 39.4 ± 0.4 37.8 ± 0.6 33.8 ± 0.3 28.6 ± 0.01
CLA: 9cis, 11trans 0.0 ± 0.0 0.0 ± 0.0 0.3 ± 0.002 0.0 ± 0.0 0.0 ± 0.0
CLA: 10trans,12cis 0.0 ± 0.0 0.0 ± 0.0 0.3 ± 0.001 0.0 ± 0.0 0.0 ± 0.0
CLA trans isomersc 0.0 ± 0.0 0.0 ± 0.0 1.9 ± 0.02 0.0 ± 0.0 0.0 ± 0.0
CLA isomersd 0.0 ± 0.0 0.0 ± 0.0 2.8 ± 0.02 0.0 ± 0.0 0.0 ± 0.0
18:3(n-3) 2.6 ± 0.1 3.9 ± 0.1 3.3 ± 0.1 3.0 ± 0.04 2.5 ± 0.01
20:4(n-6) 0.1 ± 0.001 0.4 ± 0.02 0.4 ± 0.03 0.5 ± 0.02 0.7 ± 0.002
20:5(n-3) 0.0 ± 0.0 0.0 ± 0.0 0.0 ± 0.0 0.0 ± 0.0 3.4 ± 0.1
22:6(n-3) 0.0 ± 0.0 0.0 ± 0.0 0.0 ± 0.0 3.2 ± 0.04 1.9 ± 0.02
n-3e 2.6 ± 0.1 3.9 ± 0.1 3.3 ± 0.1 6.2 ± 0.1 8.5 ± 0.1
n-6f 34.3 ± 0.9 39.9 ± 0.4 38.0 ± 0.1 34.3 ± 0.3 29.4 ± 0.01
n-3 LC-PUFAsg 0.0 ± 0.0 0.0 ± 0.0 0.0 ± 0.0 3.2 ± 0.04 6.1 ± 0.1
n-6 LC-PUFAsh 0.1 ± 0.01 0.4 ± 0.02 0.4 ± 0.03 0.5 ± 0.02 0.7 ± 0.002
n-3 : n-6 ratio 0.1 ± 0.001 0.1 ± 0.001 0.1 ± 0.002 0.2 ± 0.002 0.3 ± 0.004
a
Fatty acid analysis of the diets was conducted in duplicate.
b
Includes 14:1, 16:1, 18:1, 20:1, 22:1, and 24:1.
c
Includes 9trans, 11trans; 10trans, 12trans; and 11trans, 13trans.
d
Includes 9cis,11trans; 10trans,12cis; 9cis,11cis/10cis, 12trans; 9trans, 11trans; 10trans,12trans; and 11trans, 13trans.
e
Includes 18:3(n-3), 20:5(n-3), 22:5(n-3), and 22:6(n-3).
f
Includes 18:2(n-6) (excluding CLA isomers), 18:3(n-6), 20:4(n-6), and 22:4(n-6).
g
Includes 20:5(n-3), 22:5(n-3), and 22:6(n-3).
h
Includes 20:4(n-6) and 22:4(n-6).
 256 FAUKNER ET AL.
plastic bag and frozen at −20◦ C. Fillets were later sent to
the Sensory Service Center at the University of Arkansas at
Fayetteville for sensory analysis. Prior to analysis, fillets were
thawed overnight under refrigeration. Three fillets were placed
in a glass cooking pan. The pan was sealed with plastic film
and fillets were cooked in a 1,100-W microwave oven for a
total of about 90 s, adjusted as needed for size; fillets were
cooked for 30 s, flipped and cooked another 30 s, and then
flipped again and cooked another 15–45 s depending on the size
of the fillet. Fillets were served to 10 panelists using a three-
digit code. Fish were served to panelists in random order and
samples were evaluated sequentially. Panelists were given a 7-
min palate cleansing period between each sample using bottled
spring drinking water and unsalted Saltine crackers. The trained
panelists evaluated samples using a modified sensory spectrum
method (Meilgaard et al. 1999). The scoring systems used for
both the sensory analysis and consumer preference tests as well
Downloaded by [Department Of Fisheries] at 19:55 08 April 2013
as sample questions are included in Appendix 1. Universal scale
references were available to panelists for the entire evaluation.
Panelists set terms used to describe samples during a 3-h orien-
tation session. Before evaluation, a preliminary sample of fish
was tasted by each panelist. As a group, panelists discussed at-
tribute scores provided by each person. This was the only time
during evaluation where discussion was allowed. This method
gave panelists an opportunity to adjust their scores if they were
not in agreement with other panelists. The preliminary sam-
ple also helped reduce bias since panelists often score their first
sample higher than other samples. In addition to the preliminary
sample, 15 samples were evaluated over a 3-h period with two
15-min breaks given. Each panelist tested a fish from each tank.
Seventy-five people who like fresh catfish and consume it at
least once every 6 months were selected for consumer preference
testing. Selected consumers were the primary shoppers of their
household. Age range was from 18 to over 65 and the proportion
of consumers was one-third male to two-thirds female. Fillets
were thawed and seasoned with salt, pepper, lemon, and butter
prior to cooking. Fillets were cooked in a convection oven at
177◦ C for 8–12 min. Sampling design was a balanced sequential
monadic Williams square and each consumer tasted a portion of
one fillet from each treatment. Participants evaluated the fillet
portions by answering 28 questions provided in a questionnaire
prepared by the Sensory Service Center.
Fish production performance.—Remaining fish in the study
were fed and maintained for three additional weeks after the ini-
tial harvest for sensory evaluation. However, during this time,
water temperatures fell below 18◦ C and fish had reached max-
imum growth for the season. Fish production performance cri-
teria consisted of weight gain, feed conversion ratio (FCR), and
survival. The following formulae were used to assess fish per-
formance:
Weight gain = (final weight − initial weight)/number of fish,
FCR = feed intake/weight gain, and
Survival = (number of fish at end of trial
/number of initial fish) × 100.
Proximate and fatty acid analysis of diets and tissues.—
Before the start of the feeding trial, three individual fish were
analyzed for proximate analysis and fatty acid composition us-
ing the same methods described for the diets. At the end of
the trial, three fish from each tank were euthanized with a so-
lution of tricaine methanesulfonate (>180 mg/L). Livers were
removed and weighed to determine hepatosomatic index, HSI =
[body weight (g)/liver weight (g)] × 100. Muscle tissue was re-
moved by filleting and weighed to determine muscle-to-body
weight ratio (muscle weight [g]/body weight [g]). Fish muscle
was then frozen at −20◦ C for subsequent proximate and fatty
acid analyses.
Oxidative stability of fillets.—After harvesting fish for prox-
imate analysis, three additional fish were harvested from each
tank to analyze oxidative stability of the muscle. Fish were
filleted and a 25–30-g median section from each fillet was indi-
vidually vacuum sealed. One fillet from each fish was subjected
to refrigeration (4◦ C) for 2 weeks, while the other was sub-
jected to frozen storage at −18◦ C for 1 month. After storage but
before analysis, samples were individually homogenized. Thio-
barbituric acid-reactive substances (TBARS) were quantified
(TBARS assay kit, Cayman Chemical Company, Ann Arbor,
Michigan) in catfish fillets as a measure of lipid peroxidation.
Values were converted to milligrams malondialdehyde (MDA)
per kilogram of fish muscle.
Statistical analysis.—Water quality measurements were sub-
jected to repeated-measures, mixed-model ANOVA using the
SAS version 9.0.1 program PROC MIXED (SAS Institute, Cary,
North Carolina) to determine whether there were differences in
water quality over time and among treatments (P ≤0.05). Weight
gain, FCR, survival, HSI, muscle ratio, proximate and fatty acid
composition data, and sensory and consumer preference testing
data were analyzed with mixed-model ANOVA using the SAS
version 9.0.1 software program PROC MIXED (SAS Institute)
followed by Tukey–Kramer pairwise comparisons to separate
treatment means. All data were natural log transformed prior to
analyses to account for potential skewedness or heteroscedas-
ticity. A probability level of P ≤ 0.05 was considered significant
in all statistical comparisons.
Step-wise discriminant analysis (SDA) was conducted on
the fatty acid profiles of the muscle with respect to profiles
in diets using the SAS program STEPDISC (SAS Institute).
The purpose of the SDA was to optimize the number of fatty
acids in the profile required to discriminate among classes. The
fatty acids and their abundances selected by SDA were then
subjected to canonical discriminant analysis (CDA) using the
SAS program CANDISC (SAS Institute). The CDA was used
to determine class differences based on the fatty acid profile.
Compared to univariate analysis, CDA accounts for interrela-
tionships among the fatty acids and class-dependent variables
 EFFECT OF LIPID SUPPLEMENTS ON CATFISH PERFORMANCE 257

(diets, catfish muscle, or liver tissue) and indicates the relative
contribution of fatty acids to class discrimination (Johnson and
Wichern 2002). During analysis, n canonical discriminant func-
tions (CDFs) were derived with n being the number of classes
(n = 9 in this study) that were linear combinations of the fatty
acid abundances from each class. For each class, the numerical
results of each CDF (e.g., CAN1, CAN2, . . . CANn) defined
the location, or centroid, of that class in n space based on all
the fatty acids in the profile. Canonical variates, or CDFs, are
uncorrelated although the underlying fatty acids may be highly
correlated and explain 100% of the between-class variance. The
CDFs also maximize the distances between class centroids so
that their separation in the canonical space indicates the sim-
ilarity of the fatty acid profiles among the classes. Only the
first CDFs for which the eigenvalue was greater than 1 and for
which more than 90% of the cumulative variance was explained
were used to simplify interpretation of the results. The pooled,
Downloaded by [Department Of Fisheries] at 19:55 08 April 2013
Proximate and Fatty Acid Analyses
Total lipid, protein, dry matter, and ash of fish muscle were
similar among treatments, as were MR and HSI (Table 5). Fatty
acid composition of muscle from fish fed the dietary treatments
is shown in Table 6. Saturated fatty acids were similar in the
muscle of fish fed CLA, FO, and DHA diets and lower in fish
fed the SO diet. Monounsaturated fatty acids were similar in
the muscle of fish fed SO, FO, and DHA diets and lower in
fish fed the CLA diet. Linoleic acid (18:2[n-6]) concentration
in fish muscle was different among treatments, and diets ranked
highest to lowest for 18:2(n-6) as follows: SO, CLA, DHA,
and FO. Only the CLA treatment had detectable concentrations
of CLA isomers. Linolenic acid (18:3[n-3]) concentration was
highest in the muscle of fish fed the SO diet, intermediate in
CLA and DHA diets, and lowest in the FO diet. Arachidonic
acid (20:4[n-6]) concentration was similar among treatments.
Eicosapentaenoic acid (20:5[n-3]) concentration was higher in

within-treatment, standardized canonical coefficients for each the muscle of fish fed the FO diet than fish fed SO, CLA,
CDF indicated the correlation between fatty acids and CDFs. or DHA diets. Docosahexaenoic acid (22:6[n-3]) concentration
The canonical coefficients for each class of fatty acid included was higher in the muscle of fish fed FO or DHA diets than in
in the model were then ordered and used to determine trends in those fed SO or CLA diets. Total n-3 fatty acids ranked highest
composition among the nine diet and muscle classes (Hair et al. to lowest for the diets as follows: FO, DHA, SO, and CLA. Diet
2005). All fatty acid abundances were log transformed prior to rankings for total n-6 fatty acid concentrations were as follows
discriminant analyses and P ≤ 0.05 was considered significant from highest to lowest: SO, CLA, DHA, and FO. The total n-3
in statistical comparisons. LC-PUFA concentration was highest in the muscle of fish fed
the FO diet, followed by the DHA diet, and lowest in SO and
CLA diets. The total n-6 LC-PUFA concentration was similar
RESULTS in muscle of fish fed the SO, CLA, and DHA diets, but lower in
Fish Growth Performance muscle of fish fed the FO diet than those fed the SO diet. The
There were no differences in weight gain, FCR, or survival ratio of n-3 to n-6 fatty acids was highest in the muscle of fish
among treatments (Table 4). There was 100% mortality in one
tank in the SO treatment in week 19, so values for that treatment
are means of only three replicates. The cause of the mortality TABLE 5. Mean proximate composition (% wet) of muscle, liver total lipid
(% wet), hepatosomatic index (HSI), and muscle ratio (MR) of Channel Catfish
is unknown but may be due to a temporary reduction in water fed a 32% protein commercial catfish diet supplemented with 2% (by weight)
quality that was not captured by routine monitoring. of either soybean oil (SO), CLA-enriched soybean oil (CLA), algal (DHA), or
menhaden fish oil (FO) for 25 weeks. Values are means of four replicate tanks
per treatment (three fish per tank) except for the SO treatment, which had three
TABLE 4. Mean weight gain, feed conversion ratio (FCR), and survival of
replicate tanks due to 100% mortality in one tank in week 19. Means within
Channel Catfish fed a 32% protein commercial catfish diet supplemented with
columns are not significantly different (P ≤ 0.05).
2% concentration of either soybean oil (SO), CLA-enriched soybean oil (CLA),
algal (DHA), or menhaden fish oil (FO) for 25 weeks. Values are means of
Diet treatment
four replicate tanks per treatment except for the SO treatment, which had three Diet Pooled
replicate tanks due to 100% mortality in one tank in week 19. Means within constituent SO CLA DHA FO SE P
columns are not significantly different (P ≤ 0.05).
Total lipid (%) 9.1 6.7 7.2 7.0 0.6 0.09
Diet and statistics Weight gain (g)a FCRb Survival (%) Protein (%) 17.0 17.1 16.7 16.5 0.2 0.08
SO 358.4 1.3 96 Dry matter 27.5 25.8 26.0 25.0 0.6 0.07
CLA 347.5 1.3 99 (%)
DHA 375.2 1.2 95 Ash (%) 1.1 1.14 1.18 1.08 0.003 0.06
FO 323.5 1.4 89 Liver total 10.3 10.2 9.5 9.0 0.4 0.06
Pooled SE 23.7 0.1 2.2 lipid (%)
P 0.84 0.51 0.07 HSIa 1.6 1.7 1.7 1.6 0.1 0.85
MRb 32.6 33.4 32.7 31.3 0.9 0.41
a
Weight gain = (final individual mean weight − mean individual initial weight)/number.
Initial individual weight was 71.4 ± 0.1 g (mean ± SE). a
HSI = [fish liver weight (g)/body weight (g)] × 100.
b
FCR = feed weight/fish weight gain. b
MR = fish muscle (g)/body weight (g).
 258 FAUKNER ET AL.

TABLE 6. Mean fatty acid composition (percentage of total fatty acids by weight) of Channel Catfish muscle from fish fed a 32% protein commercial catfish
diet supplemented with 2% (by weight) of either soybean oil (SO), CLA-enriched soybean oil (CLA), algal (DHA), or menhaden fish oil (FO) for 25 weeks.
Muscle data are means of four replicate tanks per treatment (three fish per tank) analyzed in duplicate except for the SO treatment, which had three replicate tanks
due to 100% mortality in one tank in week 19. Means within rows with different letters are significantly different (P ≤ 0.05).

Muscle
Fatty acid Initial fisha SO CLA DHA FO Pooled SE P
Saturated fatty acidsb 25.8 23.9 y 25.8 x 25.0 x 25.7 x 0.2 0.001
Monounsaturated fatty acidsc 51.5 49.8 x 48.0 y 50.5 x 50.8 x 0.4 0.006
18:2(n-6) 15.3 19.4 w 17.8 x 16.7 y 14.8 z 0.1 <0.001
CLA: 9cis,11trans 0.0 0.0 y 0.1 x 0.0 y 0.0 y 0.01 <0.001
CLA: 10trans,12cis 0.0 0.0 y 0.1 x 0.0 y 0.0 y 0.01 <0.001
CLA trans isomersd 0.0 0.0 y 1.1 x 0.0 y 0.0 y 0.02 <0.001
CLA isomerse 0.0 0.0 y 1.6 x 0.0 y 0.0 y 0.06 <0.001
18:3(n-3) 1.2 1.7 x 1.4 y 1.4 y 1.3 z 0.3 <0.001
20:4(n-6) 1.6 0.9 0.8 0.8 0.8 0.05 0.234
20:5(n-3) 0.2 0.1 y 0.1 y 0.1 y 1.0 x 0.03 <0.001
Downloaded by [Department Of Fisheries] at 19:55 08 April 2013

22:6(n-3) 1.1 0.5 y 0.6 y 2.4 x 2.3 x 0.1 <0.001

n-3f 2.8 2.5 y 2.3 z 4.1 x 5.3 w 0.1 <0.001
n-6g 17.6 20.8 w 19.1 x 17.9 y 15.8 z 15.8 <0.001
n-3 LC-PUFAsh 1.7 0.7 z 0.9 z 2.6 x 4.0 w 0.1 <0.001
n-6 LC-PUFAsi 1.8 1.0 y 0.9 yz 0.9 yz 0.8 z 0.05 <0.001
n-3 : n-6 ratio 0.15 0.11 z 0.11 z 0.22 y 0.32 x 0.006 <0.001
a
Initial fish data are means of two pooled muscle samples consisting of five fish each analyzed in duplicate.
b
Includes 14:0, 16:0, 18:0, and 20:0.
c
Includes 14:1, 16:1, 18:1, 20:1, 22:1, and 24:1.
d
Includes 9trans, 11trans; 10trans, 12trans; and 11trans, 13trans.
e
Includes 9cis,11trans; 10trans,12cis; 9cis,11cis/10cis, 12trans; 9trans, 11trans; 10trans,12trans; and 11trans, 13trans.
f
Includes 18:3(n-3), 20:5(n-3), 22:5(n-3), and 22:6(n-3).
g
Includes 18:2(n-6) (excluding CLA isomers), 18:3(n-6), 20:4(n-6), and 22:4(n-6).
h
Includes 20:5(n-3), 22:5(n-3) and 22:6(n-3).
i
Includes 20:4(n-6) and 22:4(n-6).

fed the FO diet, followed by fish fed the DHA diet, and lowest 60
in fish fed the SO and CLA diets. 50 DCLA
40 FCLA
Step-wise and Canonical Discriminant Analysis of Muscle
Results from the comparison of diet and muscle fatty acid 30
profiles indicated that they were significantly different (Wilks’ 20
lambda < 0.001) among the nine classes. The SDA results in-
CAN1

10
dicated all eight fatty acid groups significantly affected class
discrimination and were included in the final profiles for CDA. 0
The first three canonical variates (CAN1, CAN2, and CAN3) -10 FMFO
DMFO
accounted for 95% of the variance among the eight fatty acid Initial FSBO
-20
groups measured. When class means were ordered within each FDHA
DDHA
variate, CAN1 class means with respect to CAN2 maximally -30 DSBO
separated the CLA diet and muscle from the other diet and -40
catfish muscle profiles (Figure 1). The CAN2 class means max- -20 -10 0 10 20 30 40 50
imally separated the diet from the muscle profiles, and in par- CAN2
ticular the CLA, SO, and DHA diets were maximally separated
FIGURE 1. Scatter plot of centroids (class means) for the first and second
from the other diets and muscle profiles (Figure 2). The CAN3 canonical variates (CAN1, CAN2) separating the fatty acid profiles of the diets
class means maximally separated the FO diet from all the other and Channel Catfish muscle in trial 3 (25 weeks). The diet centroids are labeled
diets and muscle profiles (Figure 3). DSBO, DCLA, DDHA, and DMFO. The muscle centroids are labeled FSBO,
Pooled, within-class, standardized canonical coefficients FCLA, FDHA, and FMFO. Abbreviations: D = diet, F = fish, CLA = soy oil
(Table 7) revealed predominantly positive loading on CAN1 for enriched with CLA, DHA = algal DHA oil, FO = menhaden fish oil.
 EFFECT OF LIPID SUPPLEMENTS ON CATFISH PERFORMANCE 259

50 TABLE 7. Pooled within-class standardized canonical coefficients (loadings)

on the first three canonical variates estimated from the fatty acid profiles of diets
40 DCLA and Channel Catfish muscle.
DDHA
DSBO
Canonical variate
30
Fatty acid CAN1 CAN2 CAN3
20
CAN2

Saturated fatty acidsa 0.336 −0.733 −0.790

10 Monounsaturated fatty acidsb 0.109 −1.052 −0.737
DMFO
FSBO 18:2(n-6) −0.234 0.296 0.007
0
FDHA FCLA CLA isomersc 1.327 0.012 0.008
FMFO
18:3(n-3) −0.128 0.371 −0.372
-10
Initial 20:4(n-6) 0.591 −0.126 −0.637
20:5(n-3) 0.374 −0.684 1.121
-20
-10 0 10 20 30 40
22:6(n-3) 0.094 0.688 −0.437
CAN3 a
Includes 14:0, 16:0, 18:0, and 20:0.
b
Includes 14:1, 16:1, 18:1, 20:1, 22:1, and 24:1.
FIGURE 2. Scatter plot of centroids (class means) for the second and third
Downloaded by [Department Of Fisheries] at 19:55 08 April 2013

canonical variates (CAN2, CAN3) separating the fatty acid profiles of the diets
and Channel Catfish muscle in trial 3 (25 weeks). The diet centroids are labeled
DSBO, DCLA, DDHA, and DMFO. The muscle centroids are labeled FSBO,
FCLA, FDHA, and FMFO. Abbreviations are as described in Figure 1.
CLA isomers and negative loading for 18:2(n-6) and 18:3(n-3).
These loadings are reflected in the positive separation of CLA
diets and muscle from the other fatty acid profiles in Figure 1
and are due to the higher concentrations of CLA isomers mea-
sured in the CLA diet and muscle classes (Tables 3 and 6). The
negative loading was a result of lower concentrations of 18:2(n-
6) and 18:3(n-3) found in all muscle profiles compared with
diet profiles (Tables 3 and 6). The variate CAN2 showed strong
negative loading for monounsaturated fatty acids and positive
loading for 22:6(n-3). Muscle profiles were higher in monoun-
saturates and tended to be higher in 22:6(n-3) compared with
the diets. However, the initial fish and muscle of fish fed the FO
60
50 DCLA
40 FCLA
30
20
CAN1
c
Includes 9cis,11trans; 10trans,12cis; 9cis,11cis/10cis, 12trans; 9trans, 11trans;
10trans,12trans; and 11trans, 13trans.
or DHA diets (FMFO and FDHA, respectively) were signifi-
cantly different in those fatty acids from fish fed the SO or CLA
diets. In particular, the SO and CLA diets and fish muscle were
lower in 22:6(n-3) than the DHA and FO diet and fish muscle
profiles. Moreover, the concentration of 22:6(n-3) in muscle of
fish fed the DHA or FO diets was similar to that found in the
diets. The variate CAN3 was negatively loaded for saturated
fatty acids but positively loaded for 20:5(n-3). This coincides
with the large separation of the FO diet profile from the others
and to some extent the separation of the FMFO muscle profile
from the other muscle profiles (Figure 2). Higher concentrations
of these fatty acid classes were found in the FO diet and muscle
(Tables 3 and 6).
Sensory Analysis and Consumer Preference Testing
Analysis of sensory panelist scores revealed no differences
among treatments for the following flavor categories: salty,
sweet, sour, bitter, fish meat, fish oil and fat, musty–dusty–
old books, earthy–dirty–muddy, sulfury, metallic, blood serum,
chickeny, plastic, medicinal, or other flavors (Table 8). The FO
treatment did receive a higher score for green-grassy flavor com-

10 pared with SO, CLA, and DHA treatments.

0 Consumer preference scores from the first rating system re-
-10
Initial FMFO sulted in no differences among treatments for overall impres-
DMFO
FDHA sion, aroma, texture, or appearance (Table 9). However, scores
-20 FSBO
for flavor were lower for fish in the FO and DHA treatments
DDHA
-30 DSBO than in the SO and CLA treatments. Consumer preference scores
-40 from the second rating system resulted in no differences among
-10 0 10 20 30 40 treatments (data not shown).
CAN3
Oxidative Stability of Fillets
FIGURE 3. Scatter plot of centroids (class means) for the first and third Repeated-measures analysis of malondialdehyde (MDA)
canonical variates (CAN1, CAN3) separating the fatty acid profiles of the diets
and Channel Catfish muscle in trial 3 (25 weeks). The diet centroids are labeled
concentration in fillet tissue revealed no differences between
DSBO, DCLA, DDHA, and DMFO. The muscle centroids are labeled FSBO, refrigerated and frozen samples or among dietary treatments
FCLA, FDHA, and FMFO. Abbreviations are as described in Figure 1. (Table 10).
 260 FAUKNER ET AL.

TABLE 8. Means of sensory analysis scores from fillets of Channel Catfish TABLE 10. Mean malondialdehyde (MDA) values (mg MDA/kg muscle tis-
fed a 32% protein commercial catfish diet supplemented with 2% (by weight) of sue) from fillets of Channel Catfish fed a 32% protein commercial catfish diet
either soybean oil (SO), CLA-enriched soybean oil (CLA), algal DHA (DHA), supplemented with 2% (by weight) of either soybean oil (SO), CLA-enriched
or menhaden fish oil (FO) for 25 weeks. Values are means of scores given by soybean oil (CLA), algal DHA (DHA), or menhaden fish oil (FO) for 25 weeks.
10 panelists. Each panelist tasted one fillet sample from each tank. The scoring Initial samples consisted of two individual fillets per tank. Refrigerated and
system rated flavor and aroma intensities based on Universal Scale lexicons. frozen samples consisted of three individual fillets per tank. Fillet samples were
Means within rows with different letters are significantly different (P ≤ 0.05). tested for MDA concentration with a thiobarbituric acid-reactive substances kit
(see Methods). Means within columns are not significantly different (P ≤ 0.05).
Diet treatment
Flavors and Pooled MDA value (mg/kg)
aromas SO CLA DHA FO SE P
Initial Refrigerated Frozen
Salt 2.8 2.9 2.9 2.8 0.2 0.74 Diet and statistics sample sample sample
Sweet 0.3 0.3 0.3 0.3 0.2 0.97
Sour 1.4 1.5 1.5 1.5 0.3 0.74 SO 0.092 0.174 0.185
Bitter 0.6 0.7 0.7 0.7 0.1 0.70 CLA 0.081 0.158 0.162
Fish meat 4.1 4.2 4.4 4.3 0.3 0.12 DHA 0.093 0.187 0.188
Fish oil or fat 4.0 3.9 4.1 3.9 0.2 0.09 FO 0.073 0.21 0.21
Musty–dusty– 1.2 1.8 1.6 1.1 0.7 0.25 Pooled SE 0.02 0.04 0.04
Downloaded by [Department Of Fisheries] at 19:55 08 April 2013

old books P 0.53 0.61 0.43

Earthy–dirty– 3.3 3.4 3.6 3.6 0.6 0.62
muddy
Sulfury 0.2 0.3 0.3 0.2 0.4 0.87 lipid sources, which limit direct comparison of other study re-
Metallic 3.4 3.4 3.5 3.3 0.4 0.75 sults with those of this study. However, it is clear that dietary
Blood serum 0.1 0.2 0.5 0.1 0.4 0.17 lipid sources have variable effects on weight gain and feed ef-
Chickeny 0.2 0.4 0.1 0.1 0.3 0.08 ficiency of Channel Catfish. Manning et al. (2006) compared
Green–grassy 0.3 y 0.5 y 0.5 y 1.4 x 0.7 0.01 the effects of diets supplemented with different combinations
Plastic 0.1 0.1 0.1 0.0 0.2 0.49 of CLAs, corn oil, and menhaden fish oil in Channel Cat-
Medicinal 0.1 0.1 0.1 0.0 0.2 0.79 fish (initial weight, 57.4 g); weight gain was lower in fish fed
Other 0.0 0.0 0.1 0.0 0.2 0.45 a diet with CLAs and corn oil compared with fish fed diets
containing CLAs and corn oil in combination with menhaden
fish oil.
DISCUSSION
In general, the addition of CLAs to fish diets has limited
Fish Growth Performance effects on fish performance. Twibell and Wilson (2003) found
Many studies using CLAs in fish diets used a different con- that weight gain among Channel Catfish fed graded concentra-
centration of isomers, graded concentrations, or mixtures of tions of CLAs (up to 1.0%) and lipid (5.0–10.0%) were similar,
and studies with Atlantic Salmon Salmo salar (Berge et al.
TABLE 9. Consumer preference testing scores from fillets of Channel Catfish 2004), Rainbow Trout Oncorhynchus mykiss (Figueiredo-Silva
fed a 32% protein commercial catfish diet supplemented with 2% (by weight) of et al. 2005), and Nile Tilapia Oreochromis niloticus (Yasmin
either soybean oil (SO), CLA-enriched soybean oil (CLA), algal DHA (DHA) et al. 2004) produced similar results. In contrast, Common Carp
or menhaden fish oil (FO) for 25 weeks. Values are means of scores given by
75 individual participants. Each participant tasted one fillet sample from each
Cyprinis carpio fed diets with 1% CLAs had increased weight
treatment. The following response scale was used: 1 = dislike extremely, 2 = gain compared with carp fed diets without CLAs (Choi et al.
dislike very much, 3 = dislike moderately, 4 = dislike slightly, 5 = neither 1999), and Twibell et al. (2000) found that CLA concentra-
like nor dislike, 6 = like slightly, 7 = like moderately, 8 = like very much, 9 tions of up to 1% improved feed efficiency in hybrid Striped
= like extremely. Means within columns with different letters are significantly Bass (White Bass Morone chrysops × Striped Bass M. sax-
different (P ≤ 0.05).
atilis). Gilthead Seabream Sparus auratus fed diets with 2%
Mean test score or 4% CLAs (a 50:50 mixture of cis-9, trans-11 and trans-
10, cis-12 isomers) had decreased weight gain, but fish fed
Diet and Overall the diet with 2% CLAs had better feed efficiency (Diez et al.
statistics impression Flavor Aroma Texture Appearance 2007).
SO 6.3 6.3 x 6.1 6.3 6.3 Studies comparing algal DHA against other lipid sources in
CLA 6.1 6.1 x 6.2 6.3 6.4 the diets of Atlantic salmon (Miller et al. 2007) and juvenile
DHA 5.9 5.5 y 6.0 5.9 6.2 Cobia Rachycentron canadum (Salze et al. 2010) revealed no
FO 5.8 5.5 y 6.0 6.0 6.0 differences in fish performance. However, Li et al. (2009) tested
Pooled SE 0.22 0.23 0.17 0.21 0.19 graded levels of algal DHA (0–2.0%) in diets of juvenile Chan-
P 0.41 0.02 0.99 0.32 0.66 nel Catfish (initial weight, 20.4 g); fish fed diets with 0% or
0.5% algal DHA had lower weight gain than fish fed diets with
 EFFECT OF LIPID SUPPLEMENTS ON CATFISH PERFORMANCE
1.0% or 1.5% algal DHA. Feed efficiency ratio increased in fish
fed diets with 0.5, 1.0, or 1.5% algal DHA compared with fish
fed diets with 0% or 2.0% algal DHA (Li et al. 2009).
Proximate and Fatty Acid Composition
As expected, only Channel Catfish fed the CLA-enhanced
diet contained CLA isomers (1.6%) in the muscle. The majority
of the CLA isomers consisted of three trans isomers: trans-9,
trans-11; trans-10, trans-12; and trans-11, trans-13. Although
not as well documented as the human health benefits of cis-9,
trans-11 and trans-10, cis-12 isomers, there is recent evidence
that trans, trans CLA isomers may also provide human health
benefits (Gilbert et al. 2011). Manning et al. (2006) fed higher
concentrations of dietary CLAs to juvenile Channel Catfish and
obtained higher concentrations of CLAs (6.4%) in the muscle.
They used a CLA source that was higher in the cis-9, trans-11
Downloaded by [Department Of Fisheries] at 19:55 08 April 2013
and trans-10, cis-12 isomers, so that the diet with the high-
est CLA concentration contained 23.9% total CLA isomers.
Ramos et al. (2008) fed Rainbow Trout a diet with CLAs in
a time-course deposition study and found no difference in the
concentration of CLAs in muscle after 8 or 12 weeks. This sug-
gests that CLA deposition plateaus after a certain point and no
further increase in CLA deposition is likely. There may also be
differences in the deposition rate of different CLA isomers, as
demonstrated by Du et al. (1999), in the eggs of laying hens.
Recommendations for consumption of specific fatty acids for
human health benefits are based on a typical serving size of
57–85 g (2 to 3 oz) as defined by the Dietary Guidelines for
Americans (USDA USDHHS 2010). Quantitative recommen-
dations for daily intake of CLAs for human health benefits are
extremely variable and range from 61.3 mg (Brownbill et al.
2005) to 3,000 mg (Ip et al. 1994). Muscle lipid from fish fed
the CLA diet contained 178.7 mg of CLA isomers per 85 g of
fillet. Manning et al. (2006) reported that fish fed the diet with
the highest amount of CLAs (1.0%) provided over 283.0 mg
of CLAs per 85 g of fillet. Therefore, using a more concen-
trated CLA supplement increases fillet concentrations of CLAs
in less time, but the concentrated supplements are typically more
costly.
In Li et al. (2009) the highest concentration of dietary algal
DHA used was the same as in our trials (2%). However, the
total lipid concentration in the DHA diet in this study was 7.1%
compared with 4.7% in Li et al. (2009). Because 22:6(n-3) was
a higher percentage of total dietary lipids in the latter study,
there was a higher concentration of 22:6(n-3) in fish muscle.
In our trial, the FO diet contained substantially less 22:6(n-3)
than the DHA diet. However, the concentration of 22:6(n-3) was
similar in muscle of fish fed diets with FO or algal DHA. Satoh
et al. (1989b) demonstrated that Channel Catfish can produce
22:6(n-3) from 20:5(n-3). Therefore, it is likely that the higher
concentration of 20:5(n-3) in the FO diet (3.3%) compared with
the DHA diet (0.2%) provided more substrate that could be
converted into 22:6(n-3). It is also possible that 20:5(n-3) is
261
preferentially catabolized for energy and 22:6(n-3) is selectively
retained (Madsen et al. 1998; Yildirim-Aksoy et al. 2007).
Many of the dietary recommendations for the human health
benefits associated with n-3 LC-PUFAs include both 22:6(n-3)
and 20:5(n-3) (Smith and Sahyoun 2005; Gebauer et al. 2006;
Psota et al. 2006), and recommended intakes also vary widely.
The Dietary Guidelines for Americans (USDA USDHHS 2010)
recommends 250 mg. However, the American Heart Association
recommends a daily intake of 1 g of n-3 LC-PUFAs per day
(Harris 2004). Muscle from fish fed the SO and CLA diets
contained 81.3 mg and 56.3 mg of n-3 LC-PUFAs, respectively.
The amount of n-3 LC-PUFAs in the muscle of fish fed the algal
DHA and FO diets was 246.9 mg and 378.6 mg, respectively.
Therefore, muscle from fish fed the algal DHA and FO diets
would provide substantially more n-3 LC-PUFAs for human
health benefits than muscle from fish fed the SO or CLA diets.
In Manning et al. (2007), Channel Catfish raised in ponds and
fed a diet supplemented with 2% menhaden oil for 16 weeks
accumulated 421 mg of n-3 LC-PUFAs per 85 g of fillet. Fillets
from catfish fed a diet with 4% menhaden oil were only slightly
higher in n-3 LC-PUFAs. These findings indicate that the de-
position of n-3 LC-PUFAs in fish tissue does not necessarily
follow a dose–response pattern with respect to dietary lipid. In
Li et al. (2009), the calculated amount of n-3 LC-PUFAs in
muscle of fish fed a diet with 2% algal DHA was 253.3 mg per
85 g. These studies indicate that fish fed a diet with lower total
lipid content and higher concentrations of target fatty acids are
more effective for enhancing healthy fatty acids in fish muscle
compared with diet formulations used in this study. However,
total lipid content in the muscle of fish fed the algal DHA diet in
our study was higher than in the muscle of fish fed the 2% algal
DHA diet in Li et al. (2009). Therefore, the concentration of n-3
LC-PUFAs per unit of fillet was actually similar between stud-
ies. This shows that both the relative percentage of total fatty
acids in the muscle and the concentration of total lipid affect the
absolute amount of target fatty acids in fish muscle.
Canonical Discriminant Analysis of Fatty Acid Profiles of
Muscle
There were distinct differences in muscle fatty acid profiles
that can be attributed to the fatty acid profiles of the diets. The
large degree of separation on CAN1 for fatty acid profiles of the
CLA diet and muscle of fish fed the CLA diet compared with
other diets and fish muscle profiles is a result of no other diets
containing CLA isomers. The negative loading of 18:2(n-6) and
18:3(n-3) indicates that these fatty acids are not retained in fish
muscle in proportion to amounts in the diet. The negative load-
ing of monounsaturates on CAN2 indicates that these fatty acids
were selectively retained in fish muscle in a higher amount than
found in the diet. Major sources of potential energy in the mo-
nounsaturates group are 18:1(n-9), 20:1(n-9), and 22:1(n-11)
(Tocher 2003). The predominant monounsaturated fatty acid in
the muscle of fish was oleic acid (18:1[n-9]). Oleic acid predom-
inates in all tissues because it is intermediate in chain length
 262 FAUKNER ET AL.
and the position of its double bond provides flexibility in its
function. Fish can also synthesize 18:1(n-9) from shorter-chain
fatty acids, which could explain the higher muscle concentration
compared with that in the diet. The positive loading on CAN2
for 22:6(n-3) is due to the presence of this fatty acid in the DHA
and FO diets and the absence in the SO and CLA diets. The
negative loading on CAN3 for saturates indicates this class of
fatty acids was one of the distinguishing differences separating
fish fed the FO diet, which was significantly more positive rel-
ative to fish fed all other diets. This follows from the fact that
saturated fatty acids in fish muscle increased in fish fed the SO
and CLA diets, remained near constant in the muscle of fish fed
the DHA diet, and decreased in the muscle of fish fed the FO
diet. Possibly, the amount of saturates in the DHA diet was close
to the amount required by the fish for metabolic functions. The
positive loading of 20:5(n-3) on CAN3 is a result of the FO diet
being the only diet that contained measurable amounts of this
Downloaded by [Department Of Fisheries] at 19:55 08 April 2013
fatty acid along with the higher concentration of 20:5(n-3) in
the muscle of fish fed the FO diet compared with the muscle of
fish fed the other diets. Overall, the canonical plots and relative
distances between diets and resulting fish centroids suggest that
the muscle fatty acid profiles of fish fed the CLA diet were the
most profoundly changed with respect to their initial pretrial
profile.
Sensory Analysis and Consumer Preference Testing
Sensory evaluation of Channel Catfish fillets indicated that
fish fed the FO diet had more green-grassy flavor than fish fed
the other diets. The volatile aldehydes produced by oxidation of
n-3 LC-PUFAs are associated with flavors such as “green” and
“grassy” (Durnford and Shahidi 1998), which were stronger in
fillets of fish fed the FO diet than in other fillets. Virginia Prime
Gold, a refined and bleached menhaden fish oil product, was
used in the FO diet. Suja et al. (2012) also obtained reduced
sensory ratings for catfish fed diets with the same fish oil rel-
ative to soybean oil. In contrast, Manning et al. (2006) did not
find any differences in the sensory properties of fish fed different
combinations of refined menhaden fish oil, CLAs, and corn oil,
but they used a more refined menhaden fish oil. Rainbow Trout
fed graded concentrations of CLAs up to 1% of the diet had sim-
ilar sensory characteristics (Ramos et al. 2008). Manning et al.
(2007) conducted a sensory analysis of Channel Catfish fed diets
supplemented with 4% corn, canola, or deodorized, winterized,
or refined menhaden oil. Catfish fed diets with winterized or de-
odorized menhaden oil had higher fishy-flavor scores than fish
fed diets with corn, canola, or refined menhaden oil. Li et al.
(2009) did not find any differences in the sensory properties of
fish fed diets with graded concentrations of algal DHA.
In consumer preference testing, fillets of fish fed FO or al-
gal DHA received lower ratings for flavor than did fillets from
fish fed the SO or CLA diets. Fillets from the FO and DHA
treatments had higher concentrations of total n-3 LC-PUFAs
than fillets from CLA and SO treatments. The number of dou-
ble bonds in a fatty acid determines its susceptibility to attack
from free radicals (Trushenski and Lochmann 2009). Unstable
hydroperoxides resulting from oxidation of fatty acids are con-
verted to free radicals, which increase autoxidation rates. Sec-
ondary products of autoxidation are often aldehydes, ketones,
alcohols, carboxylic acids, and alkanes. It is typically these sec-
ondary products that contribute to oxidized flavor in fish (Gray
1978). Increased oxidation rates can affect sensory character-
istics (Bureau et al. 2008), and oxidative breakdown products
of both 20:5(n-3) and 22:6(n-3) is linked with the formation of
fish flavors (Karahadian and Lindsay 1989). Fowler et al. (1994)
reported that hybrid Striped Bass fed diets with elevated con-
centrations of 20:5(n-3) and 22:6(n-3) had higher fishy flavors
that were probably caused by these fatty acids being oxidized to
compounds with fishy flavors. Rainbow Trout fillets from fish
fed a diet with 30% menhaden oil and 300 mg/kg vitamin E had
higher TBARS values and fishy flavor than did fillets from fish
fed a diet with 30% menhaden oil and 1,500 mg/kg vitamin E.
However, fillets from fish fed a diet with 15% menhaden oil and
either 300 or 1,500 mg/kg vitamin E showed no differences in
sensory scores or TBARS values (Chaiyapechara et al. 2003).
This indicates that increased oxidation of fillets from fish fed
high-lipid diets may produce unfavorable sensory scores, espe-
cially when highly unsaturated lipids are used.
Oxidative Stability of Fillets
The TBARS values from fillets of Channel Catfish fed all
diets were well below the value (1 mg/kg MDA) given for devel-
opment of rancidity (Trushenski and Lochmann 2009). Rancid
flavors are typically detected between 0.5 and 2 mg of MDA per
kilogram of meat (Gray and Pearson 1987). Lower consumer
preference scores from FO and DHA diet treatments may be
attributed to factors other than oxidation of n-3 LC-PUFAs, as
measured by the TBARS assay. Many reactions may occur in
food products that can generate flavor and aroma compounds
that cannot be measured by the TBARS assay (Gordon 2004).
In summary, there was an overall increase in target fatty
acids (n-3 LC-PUFAs, CLAs) in Channel Catfish but not
to the degree that has been achieved in other studies using
different lipid supplements for a shorter period. Supplementing
a commercial catfish diet containing 5% intrinsic lipid with 2%
soybean oil enhanced with CLA isomers was not sufficient to
enhance fish fillets with enough CLAs (3 g) to provide health
benefits, according to Ha et al. (1989) or Ip et al. (1994). How-
ever, the CLA concentration in fillets from this study is consider-
ably higher than that in other products that are considered good
sources of CLAs. Pasture-fed beef (1.5% total lipid) produced
meat with 0.9% CLAs (Nuernberg et al. 2005), and cows fed
fresh pasture produced milk (4.0% total lipid) with 1.2% CLAs
(Kay et al. 2004). The soybean oil with CLA isomers is also high
in CLA trans isomers, which may have a specific health benefit
(reduction of cholesterol) that is not associated with products
with cis-9, trans-11 and trans-10, cis-12 CLA isomers (Gilbert
et al. 2011). The trans CLA isomers may have slower deposition
rates into fish tissues than other CLA isomers. Additional
 EFFECT OF LIPID SUPPLEMENTS ON CATFISH PERFORMANCE
studies are needed to determine the optimal amounts of dietary
trans CLA and finishing period needed to achieve sufficient
muscle concentrations in catfish to provide human health
benefits.
Sensory evaluation and consumer preference scores were
generally more favorable for fillets from Channel Catfish fed
the SO or CLA diets than those fed the FO or algal DHA diets.
However, similar TBARS values among treatments indicated
that sensory differences may not have been due to increased ox-
idation, or that non-TBARS oxidation products were involved.
Finally, lipid supplements that increase the cost of catfish
feed will not be accepted by the industry unless profitability is
maintained or increased. Currently, the cost of the algal DHA
supplement is highest, followed by menhaden oil and standard
soybean oil. The cost of the CLA-enriched soy oil is projected
to be between that of fish oil and soybean oil (A. Proctor, Uni-
versity of Arkansas at Fayetteville, personal communication).
Downloaded by [Department Of Fisheries] at 19:55 08 April 2013
Production costs of new products tend to be higher than those
of established products, and ingredient prices are volatile. Ulti-
mately, the use of different lipids will depend on the willingness
of consumers to pay higher prices for functional foods produced
with specialty diets.
ACKNOWLEDGMENTS
Menhaden fish oil was donated by Omega Protein, Inc.,
Aquagrow Gold was donated by Advanced Bionutrition Corp.
Brooke Benhest assisted with fatty acid analysis and identifica-
tion. J. F. Meullenet and Tonya S. P. Tokar provided valuable
insight on the sensory analysis and consumer preference testing
results. We thank the Arkansas Soybean Promotion Board for
funding. The manuscript was reviewed by Anita Kelly, David
Heikes, and Alf Haukenes.
REFERENCES
AOAC (Association of Official Analytical Chemists). 1995. Official methods of
analysis, 16th edition. AOAC, Arlington, Virginia.
AOCS (American Oil Chemists Society). 2005. Crude fiber analysis in feeds by
filter bag technique. AOCS approved procedure Ba 6a-05. AOCS, Urbana,
Illinois.
Arterburn, L. M., E. B. Hall, and H. Oken. 2006. Distribution, interconversion,
and dose response of n-3 fatty acids in humans. American Journal of Clinical
Nutrition 83(supplement):1467S–1476S.
Bandarra, M., M. L. Nunes, A. M. Andrade, J. A. M. Prates, S. Pereira, M.
Monteiro, P. Rema, and L. M. P. Valente. 2006. Effect of dietary conjugated
linoleic acid on muscle, liver and visceral lipid deposition in Rainbow Trout
juveniles Oncorhynchus mykiss. Aquaculture 254:496–505.
Benjamin, S., and F. Spener. 2009. Conjugated linoleic acids as functional food:
an insight into their health benefits. Nutrition and Metabolismm [online serial]
6:36.
Berge, G. M., B. Ruyter, and T. Asgard. 2004. Conjugated linoleic acid in
diets for juvenile Atlantic Salmon Salmo salar; effects on fish performance,
proximate composition, fatty acid and mineral content. Aquaculture 237:365–
380.
Brownbill, R. A., M. Petrosian, and J. Z. Ilich. 2005. Association between
dietary conjugated linoleic acid and bone mineral density in postmenopausal
women. Journal of the American College of Nutrition 24:177–181.
263
Bureau, D. P., K. Hua, and A. M. Harris. 2008. The effect of dietary lipid and
long chain n-3 PUFA levels on growth, energy utilization, carcass quality,
and immune function of Rainbow Trout Onchorhynchus mykiss. Journal of
the World Aquaculture Society 39:1–21.
Chaiyapechara, S. K., K. M. Liu, F. T. Barrows, R. W. Hardy, and F. M.
Dong. 2003. Proximate composition, lipid oxidation, and sensory charac-
teristics of fillets from Rainbow Trout Oncorhynchus mykiss fed diets con-
taining 10% to 30% lipid. Journal of the World Aquaculture Society 34:266–
277.
Choi, B. D., S. J. Kang, Y. L. Ha, and R. G. Ackman. 1999. Accumulation
of conjugated linoleic acid (CLA) in tissues of fish fed diets containing
various levels of CLA. Pages 61–71 in Y. L. Xiong, C. T. Ho, and F.
Shahidi, editors. Quality attributes of muscle foods. Kluwer and Plenum,
New York.
Diez, A., D. Menoyo, S. Perez-Benavente, J. A. Calduch-Giner, S. V. R. deCelis,
and A. Obach. 2007. Conjugated linoleic acid affects lipid composition,
metabolism, and gene expression in Gilthead Sea Bream Sparus aurata.
Journal of Nutrition 137:1363–1369.
Doughman, S. D., S. Krupanidhi, and C. B. Sanjeevi. 2007. Omega-3 fatty acids
for nutrition and medicine: considering microalgae oil as a vegetarian source
of EPA and DHA. Current Diabetes Reviews 3:198–203.
Du, M., D. U. Ahn, and J. L. Sell. 1999. Effect of dietary conjugated linoleic
acid on the composition of egg yolk lipids. Poultry Science 78:1639–1645.
Durnford, E., and F. Shahidi. 1998. Flavor of fish meat. Pages 130–158 in F.
Shahidi, editor. Flavor of meat, meat products and seafood. Blackie Academic
and Professional, London.
Figueiredo-Silva, A. C., P. Rema, N. M. Bandarra, M. L. Nunes, and L. M. P.
Valente. 2005. Effects of dietary conjugated linoleic acid on growth, nutri-
ent utilization, body composition, and hepatic lipogenesis in Rainbow Trout
juveniles Oncorhynchus mykiss. Aquaculture 248:163–172.
Folch, J., M. Lees, and G. H. Sloane-Stanley. 1957. A simple method for
the isolation and purification of total lipids from animal tissues. Journal of
Biological Chemistry 226:497–509.
Fowler, K. P., C. Karahadian, N. J. Greenberg, and R. M. Harrel. 1994. Compo-
sition and quality of aquacultured hybrid Striped Bass as affected by dietary
fatty acids. Journal of Food Science 59:70–75.
Gebauer, S. K., T. L. Psota, W. S. Harris, and P. M. Kris-Etherton. 2006.
N-3 fatty acid dietary recommendations and food sources to achieve essen-
tiality and cardiovascular benefits. American Journal of Clinical Nutrition
83(supplement):1526S–1535S.
Gilbert, W., V. Gadang, A. Proctor, V. Jain, and L. Devareddy. 2011. Trans-trans
conjugated linoleic acid enriched soybean oil reduces fatty liver and lowers
serum cholesterol in obese zucker rats. Lipids 46:961–968.
Gordon, M. H. 2004. Factors affecting lipid oxidation. Pages 128–140 in R.
Steel, editor. Understanding and measuring the shelf-life of food. CRC Press,
Boca Raton, Florida.
Gray, J. I. 1978. Measurement of lipid oxidation: a review. Journal of the
American Oil Chemists Society 55: 539–546.
Gray, J. I., and A. M. Pearson. 1987. Rancidity and warmed-over flavor. Pages
221–269 in A. M. Pearson and T. R. Dutson, editors. Restructured meat and
poultry products, advances in meat research. Van Nostrand Reinhold, New
York.
Gunstone, F. D. 2011. The world’s oils and fats. Pages 61–98 in G. M. Turchini,
W.-K. Ng, and D. R. Tocher, editors. Fish oil replacement and alternative
lipid sources in aquaculture feeds. CRC Press, Boca Raton, Florida.
Ha, Y., N. Grimm, and M. Pariza. 1989. Newly recognized anticarcinogenic
fatty acids: identification and quantification in natural and processed cheeses.
Journal of Agricultural and Food Chemistry 37:75–81.
Hair, J. F., R. E. Anderson, R. L. Tatham, and W. C. Black. 2005. Multivariate
data analysis, 6th edition. Prentice Hall, Englewood Cliffs, New Jersey.
Hanson, T., and D. Sites. 2006. 2005 catfish database. Mississippi State Univer-
sity, Information Report 2006–01, Starkville.
Harris, W. S. 2004. Fish oil supplementation: evidence for health benefits.
Cleveland Clinic Journal of Medicine 71:208–221.
 264 FAUKNER ET AL.
Hasler, C. M., and A. C. Brown. 2009. Position of the American Dietetic
Association: functional foods. Journal of the American Dietetic Association
109:735–746.
Horrocks, L. A., and Y. K. Yeo. 1999. Health benefits of docosahexaenoic acid
(DHA). Pharmacological Research 40:211–225.
Ip, C., M. Singh, H. J. Thompson, and J. A. Scimeca. 1994. Conjugated linoleic
acid suppresses mammary carcinogenesis and proliferative activity of the
mammary gland in the rat. Cancer Research 54:1212–1215.
Jain, V. P., and A. Proctor. 2008. Pilot scale production of conjugated linoleic
acid-rich soy oil by photo-irradiation. Journal of Food Science 73:183–192.
Johnson, R. A., and D. W. Wichern. 2002. Applied multivariate statistical data
analysis, 5th edition. Prentice Hall, Englewood Cliffs, New Jersey.
Karahadian, C., and R. C. Lindsay. 1989. Evaluation of compounds contributing
characterizing fishy flavors in fish oils. Journal of the American Oil Chemists
Society 66:953–960.
Kay, J. K., T. R. Mackle, M. J. Auldist, N. A. Thomson, and D. E. Bauman.
2004. Endogenous synthesis of cis-9, trans-11 conjugated linoleic acid in
dairy cows fed fresh pasture. Journal of Dairy Science 87:369–378.
Li, M. H., E. H. Robinson, C. S. Tucker, B. B. Manning, and L. Khoo. 2009.
Effects of dried algae Schizochytrium sp., a rich source of docosahexaenoic
Downloaded by [Department Of Fisheries] at 19:55 08 April 2013
acid, on growth, fatty acid composition, and sensory quality of Channel
Catfish. Aquaculture 292:232–236.
Lochmann, R. T., and H. Phillips. 1995. Comparison of rice bran oil, poultry
fat, and cod liver oil as supplemental lipids in feeds for Channel Catfish and
Golden Shiner. Journal of Applied Aquaculture 5:47–56.
Madsen, L., L. Frøyland, E. Dyrøy, K. Helland, and R. K. Berge. 1998. Do-
cosahexaenoic and eicosapentaenoic acid are differently metabolized in rat
liver during mitochondria- and peroxisome proliferation. Journal of Lipid
Research 39:583–594.
Manning, B. B., M. H. Li, and E. H. Robinson. 2007. Feeding of Channel Catfish
Ictalurus punctatus diets with refined marine fish oil elevates omega-3 highly
unsaturated fatty acids. Journal of the World Aquaculture Society 38:49–58.
Manning, B. B., M. H. Li, E. H. Robinson, and B. C. Peterson. 2006. Enrich-
ment of Channel Catfish Ictalurus punctatus fillets with conjugated linoleic
acid and omega-3 fatty acids by dietary manipulation. Aquaculture 261:337–
342.
Meilgaard, M., G. V. Civille, and B. T. Carr. 1999. Sensory evaluation tech-
niques, 3rd edition. CRC Press, Boca Raton, Florida.
Miller, M. R., P. D. Nichols, and C. G. Carter. 2007. Replacement of fish oil
with thraustochytrium Schizochytrium sp. L. oil in Atlantic Salmon parr Salmo
salar diets. Comparative Biochemistry and Physiology A 148:382–392.
Muskiet, F. A. J., M. R. Fokkema, A. Schaafsma, E. R. Boersma, and M. A.
Crawford. 2004. Is docosahexaenoic acid (DHA) essential? Lessons from
DHA status reglation, our ancient diet, epidemiology and randomized con-
trolled trials. Journal of Nutrition 134:183–186.
NASS (National Agricultural Statistics Service). 2012. Catfish production, Jan-
uary 2012. U.S. Department of Agriculture, Washington, D.C.
Naylor, R. L., R. W. Hardy, D. P. Bureau, A. Chiu, M. Elliott, A. P. Farrell,
I. Forster, D. M. Gatlin, R. J. Goldburg, K. Hua, and P. D. Nichols. 2009.
Feeding aquaculture in an era of finite resources. Proceedings of the National
Academy of Sciences of the USA 106:15103–15110.
Nuernberg, G., K. Ender, D. Dannenberger, K. Nuernberg, K. Ender, J. Voigt,
N. D. Scollan, J. D. Wood, G. R. Nute, and R. I. Richardson. 2005. Effect of a
grass-based and a concentrate feeding system on meat quality characteristics
and fatty acid composition of longissimus muscle in different cattle breeds.
Livestock Production Science 94:137–147.
Psota, T. L., S. K. Gebauer, and P. Kris-Etherton. 2006. Dietary omega-3
fatty acid intake and cardiovascular disease. American Journal of Cardiology
98(supplement):3S–18S.
Ramos, A., N. M. Bandarra, P. Rema, P. Vas-Pires, M. L. Nunes, A. M. An-
drade, A. R. Cordeiro, and L. M. P. Valente. 2008. Time course deposition of
conjugated linoleic acid in market size Rainbow Trout Oncorhynchus mykiss
muscle. Aquaculture 274:366–374.
Roche, H. M., E. N. Noone, A. Nugent, and M. J. Gibney. 2001. Conju-
gated linoleic acid: a novel therapeutic nutrient? Nutrition Research Reviews
14:173–187.
Salze, G., E. McLean, P. R. Battle, M. H. Schwarz, and S. R. Craig. 2010.
Use of soy protein concentrate and novel ingredients in the total elimination
of fish meal and fish oil in diets for juvenile cobia Rachycentron canadum.
Aquaculture 298:294–299.
Santha, C. R., and D. M. Gatlin. 1991. Growth response and fatty acid composi-
tion of Channel Catfish fry fed practical diets supplemented with menhaden
fish oil. Progressive Fish-Culturist 53:135–140.
Satoh, S., W. E. Poe, and R. P. Wilson. 1989a. Studies on the essential fatty acid
requirements of Channel Catfish Ictalurus punctatus. Aquaculture 79:121–
128.
Satoh, S., W. E. Poe, and R. P. Wilson. 1989b. Effect of dietary n-3 fatty acids on
weight gain and liver polar lipid fatty acid composition of fingerling Channel
Catfish. Journal of Nutrition 119:23–28.
Simopoulos, A. 2008. The importance of the omega-6/omega-3 fatty acid ratio
in cardiovascular disease and other chronic diseases. Experimental Biology
and Medicine 233:674–688.
Smith, K. M., and N. R. Sahyoun. 2005. Fish consumption: recommendations
versus advisories, can they be reconciled? Nutrition Reviews 63:39–46.
Stickney, R. R., and J. W. Andrews. 1972. Effects of dietary lipids on growth,
food conversion, lipid and fatty acid composition of Channel Catfish. Journal
of Nutrition 102:249–258.
Suja, B., R. Lochmann, T. Sink, H. Phillips, and R. Chen. 2012. Effect of diets
supplemented with soybean, flaxseed, or menhaden fish oil on the growth,
feed utilization, immune status, and sensory properties of channel catfish in
a recirculating system at 22◦ C. Journal of Applied Aquaculture 24:16–31.
Tocher, D. R. 2003. Metabolism and functions of lipids and fatty acids in teleost
fish. Reviews in Fisheries Science 11:107–184.
Trushenski, J. T., and R. T. Lochmann. 2009. Potential, implications and so-
lutions regarding the use of rendered animal fats in aquafeeds. American
Journal of Animal and Veterinary Sciences 4:108–128.
Twibell, R. G., B. A. Watkins, L. Rogers, and P. B. Brown. 2000. Effects
of dietary conjugated linoleic acids on hepatic and muscle lipids in hybrid
Striped Bass. Lipids 35:155–161.
Twibell, R. G., and R. P. Wilson. 2003. Effects of dietary conjugated linoleic
acids and total dietary lipid concentrations on growth responses of juvenile
Channel Catfish Ictalurus punctatus. Aquaculture 221:621–628.
USDA–ARS (U.S. Department of Agriculture, Agricultural Research Service).
2012. USDA national nutrient database for standard reference, release 25.
Available: http://www.ars.usda.gov/ba/bhnrc/ndl.
USDA USDHHS (U.S. Department of Agriculture/U.S. Department of Health
and Human Services). 2010. Dietary guidelines for Americans 2010. U.S.
Government Printing Office, Washington, D.C.
Valente, L. M. P., N. M. Bandarra, A. C. Figueiredo-Silva, P. Rema, P. Vaz-Pires,
S. Martins, J. A. M. Prates, and M. L. Nunes. 2007. Conjugated linoleic acid in
diets for large-size Rainbow Trout Oncorhynchus mykiss: effects on growth,
chemical composition and sensory attributes. British Journal of Nutrition
97:289–297.
Wang, C., W. S. Harris, M. Chung, A. H. Lichtenstein, E. M. Balk, B. Kupelnick,
H. S. Jordan and J. Lau. 2006. N-3 fatty acids from fish or fish-oil supple-
ments but not α-linolenic acid, benefit cardiovascular disease outcomes in
primary- and secondary-prevention studies: a systematic review. American
Journal of Clinical Nutrition 84:5–17.
Yasmin A., T. Takeuchi, M. Hayashi, T. Hirota, W. Ishizuka, and S. Ishida.
2004. Effect of conjugated linoleic and docosahexaenoic acids on growth of
juvenile tilapia Oreochromis niloticus. Fisheries Science 70:473–481.
Yildirim-Aksoy, M., R. Shelby, C. Lim, and P. H. Klesius. 2007. Growth perfor-
mance and proximate and fatty acid compositions of Channel Catfish Ictalurus
punctatus fed for different durations with a commercial diet supplemented
with various levels of menhaden fish oil. Journal of the World Aquaculture
Society 38:461–473.
 EFFECT OF LIPID SUPPLEMENTS ON CATFISH PERFORMANCE 265

APPENDIX 1. Sample questions and scoring systems used for consumer preference testing of fillets of Channel
Catfish fed diets with different supplemental lipids for 25 weeks.

Part 1. A 9-point scoring system was used to evaluate general traits of the samples.
Sample question: Please observe and taste this sample. All things considered, which statement best describes your OVERALL
IMPRESSION of this product?

Dislike Dislike Dislike Dislike Neither like Like Like Like Like
extremely very much moderately slightly nor dislike slightly moderately very much extremely

The 9-point scoring system was used for additional questions about the appearance, aroma, flavor, and texture of the product.
Downloaded by [Department Of Fisheries] at 19:55 08 April 2013

Part 2. The JAR (just about right) scale was used to score samples on particular product traits.
Sample question: Considering only the COLOR of the product, which statement best describes your impression of this product?

Much too A little Just about A little Much too

weak too weak right too strong strong

The JAR system was used for additional questions on the aroma, overall flavor intensity, overall fish flavor, saltiness, moistness,
firmness, and flakiness of the product. In addition, panelists were asked whether they noticed an off-flavor (not expected in fish)
in the sample (yes or no) and were given one open-response question in which they were asked to comment on the sample and
report what they liked or disliked about it.
This article was downloaded by: [Department Of Fisheries]
On: 08 April 2013, At: 19:56
Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,
37-41 Mortimer Street, London W1T 3JH, UK

North American Journal of Aquaculture

Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/unaj20

Alternative Feeding Strategies to Maximize Fish Oil and

Fish Meal Sparing in Largemouth Bass Culture while
Maintaining Production Performance and Product Value
a a a
Andrew R. Coursey , Jesse T. Trushenski & Christopher C. Kohler
a
Fisheries and Illinois Aquaculture Center and Department of Zoology, Southern Illinois
University–Carbondale, 1125 Lincoln Drive, Carbondale, Illinois, 62901-6511, USA
Version of record first published: 02 Apr 2013.

To cite this article: Andrew R. Coursey , Jesse T. Trushenski & Christopher C. Kohler (2013): Alternative Feeding Strategies
to Maximize Fish Oil and Fish Meal Sparing in Largemouth Bass Culture while Maintaining Production Performance and Product
Value, North American Journal of Aquaculture, 75:2, 266-276

To link to this article: http://dx.doi.org/10.1080/15222055.2012.713895

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions

This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to
anyone is expressly forbidden.

The publisher does not give any warranty express or implied or make any representation that the contents
will be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses should
be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims,
proceedings, demand, or costs or damages whatsoever or howsoever caused arising directly or indirectly in
connection with or arising out of the use of this material.
 North American Journal of Aquaculture 75:266–276, 2013

C American Fisheries Society 2013

ISSN: 1522-2055 print / 1548-8454 online

DOI: 10.1080/15222055.2012.713895

SPECIAL SECTION: LIPIDS IN AQUACULTURE NUTRITION AND PHYSIOLOGY II

Alternative Feeding Strategies to Maximize Fish Oil and Fish

Meal Sparing in Largemouth Bass Culture while Maintaining
Production Performance and Product Value

Andrew R. Coursey, Jesse T. Trushenski,* and Christopher C. Kohler

Fisheries and Illinois Aquaculture Center and Department of Zoology,
Southern Illinois University–Carbondale, 1125 Lincoln Drive, Carbondale, Illinois 62901-6511, USA
Downloaded by [Department Of Fisheries] at 19:56 08 April 2013

Abstract
Feeding trials were conducted to determine (1) the optimal inclusion of poultry byproduct meal (PBM) and poultry
fat (PF) as alternatives to fish meal (FM) and fish oil (FO) in diets for Largemouth Bass Micropterus salmoides and
(2) whether the poultry-based formulations could be used in conjunction with FO-based finishing feeds in a strategy
to minimize the use of marine-derived ingredients while maintaining performance and fillet composition. In trial 1,
fish (mean ± SE = 16.0 ± 0.5 g) were reared for 9 weeks on diets (∼40% protein, ∼10% lipid) containing graded
levels of FM (0, 7.5, or 15%) and FO (0, 3, or 6%), with PBM and PF replacing FM and FO, respectively. The results
indicated that complete replacement of marine feedstuffs with PBM and PF had no significant effects on production
performance; all diets yielded equivalent weight gain (mean ± SE = 277 ± 45%), feed conversion ratio (FCR;
1.4 ± 0.4), and feed consumption (39 ± 6 g/fish), although fillet fatty acid profile reflected the diet. Based on these
results, the 7.5% FM–0% FO and 0% FM–0% FO diets were further evaluated. In trial 2, fish (mean ± SE = 39 ±
3 g) were reared for 16 weeks on the 7.5% FM–0% FO or 0% FM–0% FO feeds, followed by an additional 12
weeks on the assigned diet or a 15% FM–6% FO finishing diet. The performance and fillet composition of these fish
were compared with those of their counterparts that only received the finishing feed or a commercially available
Largemouth Bass feed. All approaches yielded equivalent weight gain (mean ± SE = 496 ± 65%), FCR (2.0 ± 0.2),
and feed consumption (407 ± 65 g/fish). Fillet fatty acid profiles reflected dietary intake, but finishing significantly
increased the n-3 fatty acid and long-chain polyunsaturated fatty acid levels. The use of poultry-byproduct-based
Largemouth Bass feeds may offer cost savings, and undesired effects on fillet nutritional value may be attenuated by
the use of finishing feeds.

Worldwide population growth will lead to increased depen-
dence on farmed fish as a protein source. In 2005, 142 mil-
lion metric tons of food fish were produced worldwide, with
aquaculture contributing nearly half of the total volume (FAO
2010). Production of farmed fish for consumption is becoming
increasingly popular but also increases the demand for the raw
materials used in making aquafeeds. Aquafeed manufacturing
has long relied on the use of fish meal (FM) and fish oil (FO) as
sources of protein, energy, and essential nutrients (NRC 2011).
Marine feedstuffs are also commonly used in aquafeeds for
their palatability and attractant properties. However, increasing
demand, rising prices for competing commodities (i.e., grains
and oilseeds), and variability in reduction fishery landings have
given rise to dramatic increases in FM and FO prices, which
have grown by 400% over the last 20 years, including a twofold
increase since 2004 (FAO 2008). As a result of increasing feed
prices and their large contribution to overall production costs
in the aquaculture industry, the Food and Agriculture Organi-
zation of the United Nations (FAO 2008) recently described
marine feedstuffs as the number-one market constraint limit-
ing the growth of aquaculture. Simultaneous replacement of
FM and FO may reduce harvest pressure on reduction fisheries
and would limit dependence on an expensive commodity in a
fluctuating market.

*Corresponding author: saluski@siu.edu

Received January 21, 2012; accepted July 13, 2012
Published online April 2, 2013

266
 ALTERNATIVE FEEDSTUFFS IN LARGEMOUTH BASS NUTRITION
Competition for marine ingredients (i.e., between the aqua-
culture industry and other livestock industries) and recent in-
creases in FM and FO prices have led to an increased interest in
land-based feedstuffs (FAO 2011a, 2011b). Nutritionists have
investigated ways to reduce the level of marine feedstuffs in
aquafeeds by using plant feedstuffs as alternative protein and
lipid sources (Trushenski et al. 2006; Gatlin et al. 2007). Partial
replacement of marine feedstuffs is possible for many species
(Gallagher 1994; Bell et al. 2001; Kaushik et al. 2004; Lewis
and Kohler 2008), but complete replacement of marine feed-
stuffs is typically less successful. Omnivorous species readily
accept feeds with little FM or FO (Robinson and Li 1994; El-
Saidy and Gaber 2002), but FM and FO replacement has proven
much more difficult for carnivorous species (Davis et al. 1995;
Lunger et al. 2006; Turchini et al. 2009). Carnivorous species are
unable to utilize carbohydrates as an energy source as success-
fully as omnivorous and herbivorous fish species (Gatlin et al.
Downloaded by [Department Of Fisheries] at 19:56 08 April 2013
2007); carnivorous species also require a higher level of pro-
tein, with some species requiring n-3 fatty acids and long-chain
(LC) polyunsaturated fatty acids (PUFAs) for development and
optimal growth (NRC 2011).
Replacement of FM and FO with plant feedstuffs poses a
problem for carnivorous fish species, such as Largemouth Bass
Micropterus salmoides, because of the need for diets with a
high protein content and low dietary carbohydrate content. Al-
though widely recognized as a sport fish, Largemouth Bass are
also growing in popularity as food fish. Live markets for Large-
mouth Bass are found in cities with large ethnic populations,
and the fish are commonly sold live for approximately US$10
per kilogram (∼$5 per pound; Tidwell et al. 2005; P. Hitchens,
Southern Illinois University Carbondale, personal communica-
tion). Feed costs are currently the highest production expense
in aquaculture (Adelizi et al. 1998; Webster et al. 1999; Naylor
et al. 2000); in Largemouth Bass culture, feed costs represent
an estimated 50% of production costs (Tidwell et. al. 2000).
Sparing of marine-derived feedstuffs in Largemouth Bass diets
is obviously advantageous in this context, but Largemouth Bass
are exclusively carnivorous and require diets with high protein
levels and low carbohydrate content (Tidwell et al. 1996). To-
tal replacement of FM with plant feedstuffs is not considered
a viable option for practical Largemouth Bass diets because
these fish are unable to tolerate the high carbohydrate content
(>20%) associated with most plant feedstuffs (Mitchell et al.
2002). However, many promising animal-derived feedstuffs are
available, including meat-and-bone meal, feather meal, fisheries
bycatch and seafood processing offal derivatives, blood meal,
poultry byproduct meal (PBM), and various rendered animal
fats (Meeker 2009). Because these ingredients generally do not
present the same challenges as plant-derived feedstuffs (e.g.,
reduced digestibility, palatability, and antinutritional factors),
they may be especially appropriate for replacing FM and FO
in Largemouth Bass feeds. Replacement of FM with poultry-
based feedstuffs has shown promise already (Tidwell et al. 2005;
Subhadra et al. 2006), but joint replacement of FM and FO has
267
not been comprehensively addressed. Furthermore, while it is
known that the fatty acid composition of Largemouth Bass tis-
sues reflects dietary intake (Subhadra et al. 2006; Laporte and
Trushenski 2011), strategies to correct for diet-related changes
in the fatty acid composition and associated nutritional value of
Largemouth Bass fillets have not been assessed. Accordingly,
we conducted trials to determine (1) the optimal inclusion of
PBM and poultry fat (PF) as alternatives to FM and FO in diets
for Largemouth Bass and (2) whether the poultry-based formu-
lations could be used in conjunction with FO-based finishing
feeds in a feeding strategy to minimize the use of marine-derived
ingredients while maintaining production performance and fillet
fatty acid composition of Largemouth Bass.
METHODS
Trial 1.—Isoproteic, isolipidic diets (40% protein, 10% fat;
Table 1) were formulated using Mixit 3 software (Agricultural
Software, Kingsville, Texas) based on a diet formulated by Tid-
well et al. (2005). Diets contained 0, 7.5, or 15% FM and 0, 3, or
6% FO in a 3 × 3 factorial design, with PBM replacing FM and
PF replacing FO as the marine-derived components decreased
in the formulations. The nine dietary treatments (expressed as%
FM–% FO) included the following: 0FM–0FO, 7.5FM–0FO,
15FM–0FO, 0FM–3FO, 7.5FM–3FO, 15FM–3FO, 0FM–6FO,
7.5FM–6FO, and 15FM–6FO.
Although attempts were made to ensure similarity among the
formulations, some variation in other feedstuffs was necessary
to ensure isoproteic, isolipidic composition due to differences
in the proximate composition of the test ingredients. Diets were
prepared by following standard practices used at the Fisheries
and Illinois Aquaculture Center (Carbondale, Illinois). Briefly,
feedstuffs were homogenized by using a cutter–mixer (Model
CM450; Hobart Corp., Troy, Ohio) and pelleted with a food
grinder (commercial-grade electric grinder; Cabela’s, Inc., Syd-
ney, Nebraska). Pellets were air dried at ambient temperature
and were stored frozen (−20◦ C) in airtight plastic containers
prior to and during the trial.
Triplicate samples of each diet were subjected to standard
analytical methods to determine proximate and fatty acid com-
position (Table 2). Diet samples were analyzed using stan-
dard (AOAC 1995) methods for moisture (Official Method
950.46), crude protein (Official Method 984.13), and ash (Of-
ficial Method 920.153). Crude lipid content was determined
gravimetrically after extraction by using the method of Folch
et al. (1957). Total lipid was subjected to acid-catalyzed trans-
methylation, which was performed overnight at 50◦ C as de-
scribed previously by Christie (1982). The resultant fatty acid
methyl esters were separated by using gas chromatography as
described in detail by Laporte and Trushenski (2011).
Largemouth Bass were spawned in earthen ponds and the
resulting fingerlings were feed-trained in an indoor aquaculture
system according to the methods described by Heidinger (2000).
After feed-training, approximately 600 fish were stocked into a
 268 COURSEY ET AL.

TABLE 1. Formulation (%) of experimental Largemouth Bass diets containing graded levels of poultry byproduct meal and poultry fat as replacements for fish
meal (FM) and fish oil (FO).

Dietary treatment (% FM/% FO)

Feedstuff 0FM–0FO 0FM–3FO 0FM–6FO 7.5FM–0FO 7.5FM–3FO 7.5FM–6FO 15FM–0FO 15FM–3FO 15FM–6FO
Menhaden FMa 0.0 0.0 0.0 7.5 7.5 7.5 15.0 15.0 15.0
Poultry byproduct mealb 32.0 32.0 32.0 23.4 23.4 23.4 14.7 14.7 14.7
Menhaden FOc 0.0 2.6 5.3 0.0 2.8 5.5 0.0 2.9 5.8
Poultry fatd 5.3 2.6 0.0 5.5 2.8 0.0 5.8 2.9 0.0
Soybean meal 34.3 34.3 34.3 34.0 34.0 34.0 34.0 34.0 34.0
Wheat middlings 17.2 17.2 17.2 18.4 18.4 18.4 19.4 19.4 19.4
Methionine 0.2 0.2 0.2 0.1 0.1 0.1 0.0 0.0 0.0
Othere 11.1 11.1 11.1 11.1 11.1 11.1 11.1 11.1 11.1
a
Special Select Menhaden FM (Omega Protein, Houston, Texas).
b
Pet food grade and antioxidant stabilized (Tyson Foods, Inc., Springdale, Arkansas).
c
Virginia Prime Gold (Omega Protein).
d
Antioxidant stabilized (Tyson Foods).
Downloaded by [Department Of Fisheries] at 19:56 08 April 2013

e
Comprised the following: 4.5% cellulose, 2.5% dicalcium phosphate, 2% carboxymethyl cellulose, 0.9% sodium phosphate, 0.6% choline chloride, 0.2% Stay-C (35%), 0.2% mineral
mix (7,000-mg/kg copper, 70,000-mg/kg iron, 100,000-mg/kg manganese, 200,000-mg/kg zinc, and 0.24% iodine in a cellulose base), and 0.2% vitamin mix (99.8-mg/kg selenium,
2,200-mg/kg folic acid, 88,000-mg/kg niacin, 35,200-mg/kg pantothenic acid, 11,000-mg/kg vitamin B6, 13,200-mg/kg riboflavin, 11,000-mg/kg thiamin, 11,000-mg/kg vitamin B12,
66,000-mg/kg vitamin E, 4,400-mg/kg vitamin K, 4,400,000-IU/kg vitamin A, 2,200,000-IU/kg vitamin D, and 100,000-mg/kg vitamin C in a cellulose base).

recirculating aquaculture system comprised of thirty-six 150-L
tanks and associated mechanical and biological filtration units.
Fish were acclimated to laboratory conditions (photoperiod =
16 h light : 8 h dark; water temperature = 23.0 ± 0.2◦ C) and
were fed a composite mixture of equal parts of the experimental
diets for 2 weeks to ensure feed acceptance prior to initiation of
the trial. After the acclimation period, fish were group-weighed
(initial individual weight [mean ± SE] = 16.0 ± 0.5 g) and
were stocked at 15 fish/tank. Tanks were randomly assigned to
one of the nine dietary treatments (4 tanks/treatment; N = 4).
All fish were fed to apparent satiation twice daily (at 0800 and
1600 hours) for 9 weeks. Water quality parameters were main-
tained within optimal growth conditions for Largemouth Bass
culture throughout the study period. Temperature and dissolved
oxygen were monitored daily by using a YSI Model 550 oxy-
gen meter (Yellow Springs Instruments, Yellow Springs, Ohio).
Ammonia, nitrite, nitrate, and alkalinity were monitored weekly
by use of commercially available water quality testing supplies
(Hach Co., Loveland, Colorado).
After completion of the study period, all fish were counted
and group-weighed by tank to calculate survival, average indi-
vidual feed consumption, feed conversion ratio (FCR = [feed
intake, dry weight]/[wet weight gain]), and average weight
gain (weight gain [%] = 100 × [final average individual weight
− initial average individual weight]/[initial average individual
weight]) for each replicate tank. A subsample of 5 fish/tank was
euthanized via single cranial pithing, and the fish were dissected
to remove livers, intraperitoneal fat (IPF) masses, and samples
of muscle tissue for further analysis. Livers and IPF masses
from these individuals were weighed to the nearest 0.1 g to
calculate the hepatosomatic index (HSI [%] = 100 × [liver wet
weight]/[whole-body wet weight]) and the liposomatic index
(LSI [%] = 100 × [IPF wet weight]/[whole-body wet weight]).
Muscle samples were stored at −80◦ C prior to lyophilization
(Freezone 6; Labconco Corp., Kansas City, Missouri) and fatty
acid analysis in the same manner as described for feed samples.
All data were analyzed by mixed-model, two-way ANOVA to
determine the significance of FM replacement and FO replace-
ment as main effects and to test for a significant interaction
between these factors (Statistical Analysis System version 9.2;
SAS Institute, Cary, North Carolina). When significant main
effects or interaction effects were observed, Tukey’s honestly
significant difference (HSD) tests were used for post hoc pair-
wise comparison of means. Although each tank represented
multiple fish, individual tanks were used as experimental units
(4 tanks/treatment; N = 4), and effects or differences were con-
sidered significant at P-values less than 0.05.
Trial 2.—Three of the diets evaluated in trial 1 were also
evaluated in trial 2 (15FM–6FO, 7.5FM–0FO, and 0FM–0FO;
formulations are shown in Table 1). The 15FM–6FO diet served
as the control and finishing diet, and the 7.5FM–0FO and 0FM–
0FO diets were used as experimental grow-out diets. For pur-
poses of comparison with current Largemouth Bass culture prac-
tices in the southern Illinois region, a commercially available
extruded feed (Aquamax; 41% protein and 12% lipid; Purina
Mills, LLC, St. Louis, Missouri) that is commonly used by lo-
cal Largemouth Bass growers was also evaluated. All feeds were
analyzed as previously described to confirm proximate compo-
sition and fatty acid composition (Table 3). Feeds were used
exclusively or in combination to generate six feeding strategies:
(1) the 0FM–0FO diet administered exclusively for 28 weeks
(0FM–0FO); (2) the 0FM–0FO diet administered for 16 weeks,
followed by finishing with the 15FM–6FO diet for 12 weeks
(0FM–0FO + Finishing); (3) the 7.5FM–0FO diet adminis-
tered exclusively for 28 weeks (7.5FM–0FO); (4) the 7.5FM–
0FO diet administered for 16 weeks, followed by finishing with
 ALTERNATIVE FEEDSTUFFS IN LARGEMOUTH BASS NUTRITION 269

TABLE 2. Analyzed proximate composition and fatty acid composition of experimental diets given to Largemouth Bass in trial 1 (least-squares means ± SEs;
SEs < 0.1 are reported as 0.0; FM = fish meal; FO = fish oil; FAMEs = fatty acid methyl esters). Only major fatty acids (≥1% in at least one treatment group)
are reported.

Dietary treatment (% FM/% FO)

Component or
fatty acid 0FM–0FO 0FM–3FO 0FM–6FO 7.5FM–0FO 7.5FM–3FO 7.5FM–6FO 15FM–0FO 15FM–3FO 15FM/6FO

Proximate composition (% dry matter basis, except moisture)

Moisture 16.7 ± 0.0 16.8 ± 0.0 14.3 ± 0.1 12.8 ± 0.0 18.9 ± 0.2 16.2 ± 0.1 18.0 ± 0.1 17.0 ± 0.1 14.3 ± 0.0
Lipid 9.6 ± 0.0 9.9 ± 0.0 9.6 ± 0.0 9.7 ± 0.0 10.2 ± 0.0 9.4 ± 0.0 9.6 ± 0.0 9.9 ± 0.0 10.0 ± 0.0
Protein 43.2 ± 0.5 43.3 ± 0.5 43.1 ± 0.7 43.5 ± 0.4 43.3 ± 0.6 44.2 ± 0.5 43.1 ± 0.5 43.3 ± 0.4 43.6 ± 0.4
Ash 11.4 ± 0.1 11.3 ± 0.2 11.6 ± 0.2 11.8 ± 0.2 12.2 ± 0.2 9.9 ± 0.1 12.0 ± 0.2 11.9 ± 0.2 10.8 ± 0.1
Fatty acid composition (% FAMEs)
14:0 0.8 ± 0.2 3.0 ± 0.0 5.7 ± 0.0 1.3 ± 0.0 3.8 ± 0.0 6.6 ± 0.1 1.9 ± 0.0 4.6 ± 0.0 5.5 ± 0.1
16:0 24.5 ± 0.1 23.6 ± 0.1 22.5 ± 0.2 24.6 ± 0.0 23.1 ± 0.1 21.9 ± 0.1 24.5 ± 0.0 23.0 ± 0.0 22.5 ± 0.1
18:0 6.6 ± 0.0 6.1 ± 0.0 5.5 ± 0.1 6.6 ± 0.0 5.8 ± 0.0 5.1 ± 0.1 6.3 ± 0.1 5.4 ± 0.0 5.3 ± 0.0
SFAsa 32.1 ± 0.1 32.9 ± 0.1 33.9 ± 0.3 32.6 ± 0.0 32.9 ± 0.1 33.8 ± 0.2 32.8 ± 0.1 33.2 ± 0.0 33.6 ± 0.2
16:1(n-7) 5.8 ± 0.1 7.4 ± 0.0 9.7 ± 0.1 6.0 ± 0.0 7.9 ± 0.0 10.1 ± 0.1 6.4 ± 0.0 8.3 ± 0.0 9.6 ± 0.1
18:1(n-7) 2.0 ± 0.0 2.4 ± 0.0 2.8 ± 0.0 2.1 ± 0.0 2.4 ± 0.0 2.8 ± 0.0 2.1 ± 0.0 2.5 ± 0.0 2.7 ± 0.0
Downloaded by [Department Of Fisheries] at 19:56 08 April 2013

18:1(n-9) 36.1 ± 0.5 29.1 ± 0.1 21.4 ± 0.2 34.4 ± 0.1 26.6 ± 0.1 18.0 ± 0.1 32.4 ± 0.1 23.7 ± 0.0 21.6 ± 0.0
MUFAsb 44.2 ± 0.4 39.3 ± 0.2 34.5 ± 0.3 42.9 ± 0.1 37.4 ± 0.1 31.7 ± 0.2 41.3 ± 0.1 35.1 ± 0.0 34.5 ± 0.1
18:2(n-6) 21.0 ± 0.3 17.3 ± 0.1 12.3 ± 0.0 20.3 ± 0.1 15.8 ± 0.1 11.1 ± 0.1 19.2 ± 0.0 15.8 ± 0.0 13.1 ± 0.0
20:4(n-6) 0.5 ± 0.0 0.7 ± 0.0 0.9 ± 0.0 0.6 ± 0.0 0.8 ± 0.0 1.0 ± 0.0 0.6 ± 0.0 0.8 ± 0.0 0.9 ± 0.0
n-6c 21.8 ± 0.2 18.3 ± 0.1 13.5 ± 0.0 21.2 ± 0.1 17.0 ± 0.1 12.5 ± 0.2 20.1 ± 0.1 16.9 ± 0.1 14.3 ± 0.1
18:3(n-3) 1.2 ± 0.0 1.3 ± 0.0 1.3 ± 0.0 1.2 ± 0.0 1.3 ± 0.0 1.4 ± 0.0 1.2 ± 0.0 1.4 ± 0.0 1.3 ± 0.0
18:4(n-3) 0.1 ± 0.0 0.7 ± 0.0 1.4 ± 0.0 0.2 ± 0.0 0.9 ± 0.0 1.6 ± 0.0 0.3 ± 0.0 1.0 ± 0.0 1.3 ± 0.0
20:5(n-3) 0.3 ± 0.2 3.5 ± 0.1 7.4 ± 0.3 0.9 ± 0.2 4.9 ± 0.1 9.0 ± 0.2 1.7 ± 0.0 5.6 ± 0.0 7.0 ± 0.1
22:5(n-3) 0.0 ± 0.0 0.5 ± 0.0 0.9 ± 0.0 0.2 ± 0.0 0.6 ± 0.0 1.1 ± 0.0 0.3 ± 0.0 0.8 ± 0.0 1.0 ± 0.0
22:6(n-3) 0.2 ± 0.1 2.2 ± 0.1 4.4 ± 0.2 0.7 ± 0.0 3.2 ± 0.0 5.6 ± 0.1 1.4 ± 0.0 3.7 ± 0.0 4.1 ± 0.1
n-3d 1.4 ± 0.5 6.5 ± 0.2 13.3 ± 0.6 1.7 ± 0.0 9.1 ± 0.1 16.4 ± 0.4 3.5 ± 0.1 10.6 ± 0.1 12.8 ± 0.3
n-3 : n-6 ratio 0.1 ± 0.0 0.4 ± 0.0 1.0 ± 0.0 0.1 ± 0.0 0.5 ± 0.0 1.3 ± 0.0 0.2 ± 0.0 0.6 ± 0.0 0.9 ± 0.0
PUFAse 23.5 ± 0.3 26.7 ± 0.3 29.7 ± 0.6 24.4 ± 0.3 28.4 ± 0.3 32.2 ± 0.4 25.2 ± 0.1 30.2 ± 0.0 30.0 ± 0.3
LC-PUFAsf 1.2 ± 0.5 7.4 ± 0.2 14.4 ± 0.6 2.5 ± 0.0 10.1 ± 0.1 17.6 ± 0.4 4.4 ± 0.1 11.6 ± 0.0 13.9 ± 0.3
C18 PUFAsg 22.2 ± 0.2 19.2 ± 0.1 15.1 ± 0.1 21.7 ± 0.2 18.1 ± 0.2 14.5 ± 0.1 20.8 ± 0.0 18.5 ± 0.0 16.0 ± 0.0
a
Sum of all fatty acids without double bonds; includes 20:0 in addition to the individually reported SFAs.
b
Sum of all fatty acids with one double bond; includes 20:1(n-9) and 24:1(n-9) in addition to the individually reported MUFAs.
c
Sum of all n-6 fatty acids; includes 20:2(n-6) and 20:3(n-6) in addition to the individually reported n-6 fatty acids.
d
Sum of all n-3 fatty acids; includes 20:3(n-3) and 20:4(n-3) in addition to the individually reported n-3 fatty acids.
e
Sum of all fatty acids with two or more double bonds; includes 18:3(n-4), 20:2(n-6), 20:3(n-6), 20:3(n-3), and 20:4(n-3) in addition to the individually reported PUFAs.
f
Sum of all fatty acids with three or more double bonds and a chain length of 20 or more carbon atoms; includes 20:3(n-6), 20:3(n-3), and 20:4(n-4) in addition to the individually
reported LC-PUFAs.
g
Sum of all PUFAs with a chain length of 18 carbon atoms; includes 18:3(n-4) in addition to the individually reported C18 PUFAs.

the 15FM–6FO diet for 12 weeks (7.5FM–0FO + Finishing);
(5) the 15FM–6FO diet administered exclusively for 28 weeks
(15FM–6FO Control); and (6) the commercial diet administered
exclusively for 28 weeks (Commercial).
Each feeding strategy was randomly assigned to five tanks
(N = 5) in a recirculating aquaculture system consisting of
thirty 240-L tanks with associated mechanical and biological
filtration units. The system was stocked with a reference pop-
ulation of fish (∼600 fish; ∼20 fish/tank) that were sourced
from a commercial vendor (Logan Hollow Fish Farm, Gorham,
Illinois). Fish were acclimated to laboratory conditions (pho-
toperiod = 24 h light : 0 h dark) and were fed a combination
of all experimental diets (equal amounts of each experimental
diet were combined to form a composite ration that was fed at
∼3% body weight/d; the commercial feed was not used during
acclimation) for 2 weeks prior to initiation of the feeding trial.
After the acclimation period, all fish were collected from the
system, subsamples of fish were group-weighed to determine
average individual weight (mean ± SE = 39 ± 3.2 g), and
the system was restocked at a rate of 15 fish/tank to start the
formal feeding trial. Fish were cultured for 28 weeks as pre-
viously described for trial 1 except that uneaten commercial
feed pellets were removed and counted 30 min after feeding,
and the recorded feed consumption was adjusted to account for
uneaten feed. Although this approach was possible for the float-
ing, uniform commercial pellets, the sinking, irregularly shaped
experimental feeds could not be effectively enumerated; thus,
consumption of these diets was based on the amount offered to
each tank. Although water temperatures were maintained at a
relatively consistent level during trial 1, the temperatures in trial
 270 COURSEY ET AL.

TABLE 3. Analyzed proximate and fatty acid composition of experimental and commercial diets given to Largemouth Bass in trial 2 (least-squares means ±
SEs; SEs < 0.1 are reported as 0.0; FM = fish meal; FO = fish oil; FAMEs = fatty acid methyl esters). Only major fatty acids (≥1% in at least one treatment
group) are reported. All fatty acid abbreviations and groupings are as defined in Table 2.

Diet (% FM–% FO)

Component or fatty acid 0FM–0FO 7.5FM–0FO 15FM–6FO Commercial
Proximate composition (% dry matter basis, except moisture)
Moisture 10.2 ± 0.0 9.9 ± 0.0 12.5 ± 0.0 7.0 ± 0.0
Crude lipid 10.3 ± 0.6 11.0 ± 0.6 10.0 ± 0.5 10.2 ± 0.4
Protein 41.5 ± 0.2 41.8 ± 0.1 41.3 ± 0.2 42.4 ± 0.1
Ash 10.2 ± 0.2 9.9 ± 0.1 10.8 ± 0.0 11.4 ± 0.1
Fatty acid composition (% FAMEs)
14:0 0.6 ± 0.0 1.1 ± 0.0 6.7 ± 0.1 4.3 ± 0.0
16:0 23.6 ± 0.1 23.9 ± 0.1 21.9 ± 0.2 19.5 ± 0.2
18:0 6.9 ± 0.0 6.8 ± 0.0 4.9 ± 0.0 4.9 ± 0.0
SFAs 31.3 ± 0.0 31.9 ± 0.1 33.6 ± 0.4 29.1 ± 0.3
Downloaded by [Department Of Fisheries] at 19:56 08 April 2013

16:1(n-7) 4.9 ± 0.0 5.3 ± 0.0 9.5 ± 0.1 5.9 ± 0.0

18:1(n-7) 1.9 ± 0.0 2.0 ± 0.0 2.9 ± 0.0 2.4 ± 0.0
18:1(n-9) 35.5 ± 0.1 34.0 ± 0.1 15.4 ± 0.1 18.7 ± 0.2
MUFAs 42.6 ± 0.1 41.6 ± 0.1 28.7 ± 0.1 27.5 ± 0.3
18:2(n-6) 23.4 ± 0.1 22.2 ± 0.1 9.9 ± 0.1 24.5 ± 0.1
20:4(n-6) 0.5 ± 0.0 0.6 ± 0.0 1.1 ± 0.0 0.8 ± 0.0
n-6 24.2 ± 0.1 23.0 ± 0.2 11.5 ± 0.1 25.6 ± 0.1
18:3(n-3) 1.5 ± 0.0 1.4 ± 0.0 1.8 ± 0.0 3.1 ± 0.0
18:4(n-3) 0.0 ± 0.0 0.2 ± 0.0 2.0 ± 0.0 1.3 ± 0.0
20:5(n-3) 0.0 ± 0.0 0.7 ± 0.0 8.9 ± 0.4 5.3 ± 0.1
22:5(n-3) 0.0 ± 0.0 0.1 ± 0.1 1.7 ± 0.1 0.9 ± 0.0
22:6(n-3) 0.0 ± 0.0 0.6 ± 0.0 7.9 ± 0.3 5.2 ± 0.2
n-3 0.1 ± 0.1 1.4 ± 0.1 20.6 ± 0.9 12.1 ± 0.4
n-3:n-6 ratio 0.0 ± 0.0 0.1 ± 0.0 1.8 ± 0.1 0.5 ± 0.0
PUFAs 25.7 ± 0.1 26.0 ± 0.3 36.3 ± 1.1 42.2 ± 0.4
LC-PUFAs 0.8 ± 0.1 2.2 ± 0.1 22.0 ± 1.0 13.1 ± 0.4
C18 PUFAs 24.9 ± 0.1 23.8 ± 0.1 14.2 ± 0.1 29.0 ± 0.1

2 were allowed to vary with seasonal changes in temperature 30

so that the conditions would approximate the more extensive
Largemouth Bass culture practices (e.g., pond culture) common 25
in the Midwestern United States (Figure 1). Throughout the 28-
Temperature (°C)

week study period, fish were group-weighed by tank each month 20

and 1 fish/tank was sampled (as previously described for trial 1)
to assess changes in fillet fatty acid profile over time (note that 15
fillet samples collected at weeks 8 and 12 were lost and were not
included in the analysis). At the completion of the feeding trial, 10 Trial 1
all remaining fish were sampled to determine overall produc- Trial 2
tion performance according to the metrics previously described, 5
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28
and the proximate composition and final fillet fatty acid profiles
Duraon of Trial (weeks)
were also determined.
Production performance data were analyzed by mixed- FIGURE 1. Mean weekly water temperature (◦ C) for trial 1 (9 weeks) and
model, one-way ANOVA, and post hoc Tukey’s HSD tests were trial 2 (28 weeks).
 ALTERNATIVE FEEDSTUFFS IN LARGEMOUTH BASS NUTRITION 271

TABLE 4. Trial 1 production performance and fillet fatty acid composition (least-squares means ± SEs; SEs < 0.1 are reported as 0.0) of Largemouth Bass that
were fed graded levels of poultry byproduct meal and poultry fat in replacement of fish meal (FM) and fish oil (FO). Within rows, means with letters in common
are not significantly different (P > 0.05). Only major fatty acids (≥1% fatty acid methyl esters [FAMEs] in at least one treatment group) are reported. All fatty
acid abbreviations and groupings are as defined in Table 2.

Performance Dietary treatment (% FM–% FO)

variable
or fatty acid 0FM–0FO 0FM–3FO 0FM–6FO 7.5FM–0FO 7.5FM–3FO 7.5FM–6FO 15FM–0FO 15FM–3FO 15FM/6FO

Production performance
Initial weight (g) 15.9 ± 1.1 16.7 ± 0.5 15.8 ± 0.5 14.7 ± 0.6 16.7 ± 1.2 16.1 ± 1.2 16.2 ± 0.6 16.0 ± 0.2 17.2 ± 0.7
Final weight (g) 48.6 ± 2.7 43.6 ± 1.4 45.4 ± 0.3 43.6 ± 1.9 47.1 ± 1.3 39.9 ± 1.3 44.0 ± 2.0 42.8 ± 1.9 45.4 ± 2.1
Weight gain (%) 308 ± 11 262 ± 12 288 ± 7 297 ± 9 285 ± 9 250 ± 9 271 ± 7 268 ± 5 267 ± 11
FCRa 1.51 ± 0.1 1.33 ± 0.1 1.49 ± 0.1 1.35 ± 0.1 1.49 ± 0.1 1.50 ± 0.1 1.3 ± 0.1 1.32 ± 0.1 1.43 ± 0.1
Consumption (g/fish) 36.6 ± 1.2 30.2 ± 0.8 31.3 ± 1.7 34.0 ± 1.4 32.4 ± 1.8 32.6 ± 0.4 30.3 ± 1.0 30.0 ± 0.9 30.6 ± 0.8
HSIb 1.90 ± 0.1 1.83 ± 0.1 1.67 ± 0.1 1.98 ± 0.1 1.89 ± 0.1 1.78 ± 0.1 2.12 ± 0.1 2.20 ± 0.2 2.02 ± 0.1
LSIc 3.34 ± 0.2 2.75 ± 0.3 2.97 ± 0.2 3.26 ± 0.2 2.97 ± 0.2 3.12 ± 0.2 3.22 ± 0.2 3.16 ± 0.2 3.11 ± 0.3
Survival (%) 93 ± 1.7 95 ± 3.2 85 ± 6.7 97 ± 1.9 90 ± 5.8 95 ± 3.2 88 ± 3.3 97 ± 3.2 98 ± 1.7

Fillet fatty acid composition (% FAMEs)

14:0 1.4 ± 0.1 x 2.8 ± 0.1 y 3.6 ± 0.3 z 2.0 ± 0.1 x 2.9 ± 0.2 y 4.2 ± 0.2 z 2.2 ± 0.2 x 3.3 ± 0.3 y 3.9 ± 0.3 z
16:0 21.2 ± 0.2 21.0 ± 0.2 20.9 ± 0.2 21.3 ± 0.2 20.6 ± 0.2 20.6 ± 0.3 21 ± 0.1 21.1 ± 0.3 20.7 ± 0.2
18:0 5.0 ± 0.3 4.8 ± 0.2 5.3 ± 0.3 5.1 ± 0.3 4.8 ± 0.6 4.8 ± 0.3 4.9 ± 0.3 4.9 ± 0.5 4.9 ± 0.3
Downloaded by [Department Of Fisheries] at 19:56 08 April 2013

SFAs 27.7 ± 0.2 28.8 ± 0.2 29.8 ± 0.3 28.5 ± 0.3 28.4 ± 0.3 29.6 ± 0.4 28.2 ± 0.2 29.4 ± 0.4 29.5 ± 0.2
16:1(n-7) 6.0 ± 0.4 7.5 ± 0.2 8.0 ± 0.5 6.6 ± 0.3 7.6 ± 0.3 9 ± 0.4 6.7 ± 0.4 8.1 ± 0.6 8.4 ± 0.5
18:1(n-7) 2.6 ± 0.0 wv 2.8 ± 0.0 yxwv 3.1 ± 0.0 zy 2.7 ± 0.04 w 2.8 ± 0.3 xw 3.1 ± 0.3 zy 2.8 ± 0.0 x 3.0 ± 0.1 zy 3.1 ± 0.0 z
18:1(n-9) 35.9 ± 4.8 x 26.4 ± 0.5 yx 20.1 ± 0.7 zy 29.4 ± 0.6 x 24.7 ± 0.0 y 19.7 ± 0.6 z 28.0 ± 0.9 x 24.0 ± 1.1 zy 19.9 ± 0.7 z
MUFAs 45.8 ± 5.0 y 37.8 ± 0.7 z 32.2 ± 1.2 z 40.0 ± 0.9 z 36.1 ± 0.8 z 32.7 ± 0.3 z 38.7 ± 1.3 z 36.1 ± 1.7 z 32.4 ± 1.2 z
18:2(n-6) 20.1 ± 0.1 w 16.7 ± 0.3 yx 12.2 ± 0.4 z 18.6 ± 0.2 xw 15.5 ± 0.0 y 11.7 ± 0.0 z 17.9 ± 0.2 x 14.3 ± 0.5 zy 12.9 ± 0.4 z
20:4(n-6) 1.4 ± 0.2 1.2 ± 0.1 1.6 ± 0.0 1.2 ± 0.1 1.3 ± 0.1 1.4 ± 0.1 1.3 ± 0.2 1.2 ± 0.2 1.5 ± 0.2
n-6 22.9 ± 0.4 w 18.9 ± 0.2 x 14.7 ± 0.2 z 20.9 ± 0.3 x 17.7 ± 0.2 yx 13.8 ± 0.2 z 20.3 ± 0.2 x 16.4 ± 0.4 zy 15.2 ± 0.2 z
18:3(n-3) 0.8 ± 0.0 y 1.1 ± 0.0 z 1.1 ± 0.1 z 0.9 ± 0.0 y 1.1 ± 0.0 z 1.2 ± 0.1 z 1.0 ± 0.0 y 1.1 ± 0.1 z 1.3 ± 0.1 z
18:4(n-3) 0.2 ± 0.0 xw 0.4 ± 0.0 yx 0.7 ± 0.1 zy 0.3 ± 0.0 x 0.5 ± 0.0 y 0.9 ± 0.0 z 0.3 ± 0.0 x 0.6 ± 0.0 y 0.9 ± 0.1 z
20:5(n-3) 0.8 ± 0.1 x 1.8 ± 0.08 y 3.6 ± 0.1 z 1.2 ± 0.1 x 2.3 ± 0.1 y 4.0 ± 0.1 z 1.3 ± 0.1 x 2.4 ± 0.2 y 3.4 ± 0.1 z
22:5(n-3) 0.7 ± 0.1 xw 1.6 ± 0.1 x 3.0 ± 0.1 z 1.0 ± 0.1 x 2.1 ± 0.0 y 3.4 ± 0.1 zy 1.3 ± 0.1 x 2.3 ± 0.1 y 2.9 ± 0.1 z
22:6(n-3) 5.1 ± 0.7 y 8.8 ± 0.6 z 13.6 ± 1.2 z 6.8 ± 0.6 z 10.7 ± 0.8 z 12.8 ± 0.9 z 8.3 ± 1.0 z 10.4 ± 1.5 z 13.0 ± 1.3 z
n-3 6.8 ± 0.8 x 12.7 ± 0.7 y 21.0 ± 1.3 z 9.2 ± 0.7 x 15.8 ± 0.8 y 21.1 ± 1.1 z 11.3 ± 1.1 yx 15.9 ± 1.8 zy 20.2 ± 1.4 z
PUFAs 30.7 ± 1.1 y 33.4 ± 0.6 z 38.0 ± 1.1 z 31.5 ± 0.8 z 35.5 ± 0.6 z 37.7 ± 0.9 z 33.1 ± 1.1 z 34.5 ± 1.6 z 38.1 ± 1.1 z
LC-PUFAs 9.0 ± 1.0 x 14.3 ± 0.8 yx 22.9 ± 1.5 z 11.0 ± 0.8 x 17.4 ± 1 zyx 22.8 ± 1.2 zy 13.0 ± 1.3 x 17.5 ± 2.0 zy 22.0 ± 1.5 z
C18 PUFAs 21.2 ± 0.1 x 18.6 ± 0.3 y 14.6 ± 0.5 z 20.0 ± 0.2 yx 17.6 ± 0.4 zy 14.5 ± 0.4 zy 19.5 ± 0.3 yx 16.5 ± 0.6 zy 15.6 ± 0.5 z

a
Feed conversion ratio = (feed intake, dry weight)/(wet weight gain).
b
Hepatosomatic index = [(liver wet weight)/(individual fish wet weight)] × 100.
c
Liposomatic index = [(intraperitoneal fat wet weight)/(individual fish wet weight)] × 100.

used for pairwise comparisons (Statistical Analysis System ver-
sion 9.2). To determine the effects of feeding strategy and time
on fillet fatty acid composition, a repeated-measures one-way
ANOVA was used—in conjunction with post hoc Tukey’s HSD
tests where appropriate. In all cases, tanks were considered the
experimental units (5 tanks/treatment; N = 5) and effects or
differences were considered significant at P-values less than
0.05.
RESULTS
Trial 1
Replacement of FM and FO with PBM and PF had no sig-
nificant effect on the production performance of juvenile Large-
mouth Bass in trial 1 (Table 4). Percent survival was not affected
by inclusion of PBM or PF, and all mortalities were fish that
jumped from tanks or were cannibalized. No significant differ-
ences were observed for FCR, weight gain, or feed consumption.
The FCR was within an acceptable range for Largemouth Bass
culture and was similar to values reported from other studies
of juvenile Largemouth Bass. No significant differences were
observed for HSI or LSI.
The fatty acid content of fillet tissue was significantly al-
tered by both FM and FO replacement as main effects, but
replacement of dietary FO with PF had the strongest influence
on fillet composition (Table 4). Fillet tissue reflected the fatty
acid profiles of the diets (Tables 2, 4). Fish that were fed diets
containing little to no FO showed a significant reduction in fillet
levels of FO-associated fatty acids (i.e., n-3 LC-PUFAs, includ-
ing 20:5[n-3] and 22:6[n-3]; in fatty acid notation, the number
of carbon atoms is given to the left of the colon, the number of
double bonds is given to the right of the colon, and the posi-
tion of the first double bond from the methyl end is given after
the hyphen) and an increase in PF-associated fatty acids (i.e.,
monounsaturated fatty acids [MUFAs], including 18:1[n-9]) in
comparison with the 15FM–6FO (control) diet. However, statis-
tically equivalent levels of LC-PUFAs were maintained among
fish that were fed the 15FM–3FO, 7.5FM–6FO, and 7.5FM–3FO
 272 COURSEY ET AL.

TABLE 5. Trial 2 production performance and fillet fatty acid composition (least-squares means ± SEs; SEs < 0.1 are reported as 0.0) of Largemouth Bass that
received a commercial feed, a marine-feedstuff-based feed (15% fish meal [15FM]–6% fish oil [6FO]), or poultry-byproduct-based feeds alone or in combination
with a finishing feed (the 15FM–6FO feed). Within rows, means with letters in common are not significantly different (P > 0.05). All production parameters were
calculated as described in Table 4. Only major fatty acids (≥1% fatty acid methyl esters [FAMEs] in at least one treatment group) are reported. All fatty acid
abbreviations and groupings are as defined in Table 2.

Dietary treatment (% FM–% FO)

Performance
variable 15FM–6FO 0FM–0FO + 7.5FM–0FM +
or fatty acid 0FM–0FO 7.5FM–0FO Control Finishing Finishing Commercial

Production performance
Initial weight (g) 38.2 ± 1.2 38.2 ± 1.0 39.6 ± 0.3 39.7 ± 0.6 40.2 ± 0.2 39.2 ± 0.8
Final weight (g) 239.5 ± 10.8 257.0 ± 12.4 248.2 ± 3.6 234.6 ± 15.4 249.7 ± 13.6 213.1 ± 17.6
Weight gain (%) 503 ± 38 481 ± 27 540 ± 24 506 ± 22 512 ± 20 446 ± 32
Consumption (g/fish) 442.1 ± 13.8 zy 466.0 ± 14.7 y 412.5 ± 12.9 z 409.8 ± 4.0 zy 430.6 ± 17.0 zy 277.0 ± 15.3 z
FCR 2.21 ± 0.1 y 2.19 ± 0.1 y 2.02 ± 0.1 zy 2.14 ± 0.2 y 2.07 ± 0.1 y 1.63 ± 0.1 z
HSI 1.50 ± 0.1 y 1.70 ± 0.1 y 1.50 ± 0.1 y 1.44 ± 0.1 y 1.53 ± 0.1 y 3.19 ± 0.2 z
LSI 2.14 ± 0.1 2.01 ± 0.2 1.41 ± 0.1 1.92 ± 0.2 1.93 ± 0.1 2.25 ± 0.1
Survival (%) 89 ± 3 85 ± 2 91 ± 4 88 ± 3 91 ± 3 83 ± 3
Downloaded by [Department Of Fisheries] at 19:56 08 April 2013

Fatty acid composition (% FAMEs)

14:0 0.95 ± 0.1 x 1.4 ± 0.3 x 2.8 ± 0.3 z 1.5 ± 0.1 yx 1.7 ± 0.1 yx 2.2 ± 0.2 zy
16:0 20.2 ± 0.1 y 20.1 ± 0.2 y 20.9 ± 0.2 z 20.6 ± 0.2 z 20.4 ± 0.2 z 19.5 ± 0.2 y
18:0 4.4 ± 0.1 4.6 ± 0.2 5.1 ± 0.2 4.7 ± 0.2 4.6 ± 0.3 4.2 ± 0.2
SFAs 25.6 ± 0.2 x 26.0 ± 0.2 yx 29.0 ± 0.5 z 26.9 ± 0.3 y 26.7 ± 0.4 y 26.0 ± 1.8 x
16:1(n-7) 5.6 ± 0.3 5.7 ± 0.5 6.9 ± 0.5 5.6 ± 0.4 6.1 ± 0.4 6.4 ± 0.5
18:1(n-7) 2.6 ± 0.1 y 2.6 ± 0.1 y 3.2 ± 0.1 z 2.6 ± 0.0 y 2.7 ± 0.0 y 3.0 ± 0.0 z
18:1(n-9) 33.8 ± 1.1 z 29.4 ± 0.9 z 18.3 ± 1.5 x 26.3 ± 1.7 zy 27.3 ± 1.2 zy 21.8 ± 0.9 yx
MUFAs 43.2 ± 1.3 z 38.7 ± 1.5 zy 29.4 ± 1.7 x 35.6 ± 2.1 y 37.2 ± 1.6 y 32.3 ± 2.5 x
18:2(n-6) 22.7 ± 0.5 z 21.4 ± 1.0 z 9.7 ± 0.7 x 17.7 ± 0.4 y 16.9 ± 0.3 y 18.3 ± 0.4 y
20:4(n-6) 1.9 ± 0.2 1.5 ± 0.2 1.7 ± 0.1 2.0 ± 0.2 1.5 ± 0.2 1.5 ± 0.1
n-6 26.3 ± 0.8 z 24.3 ± 1.2 z 12.1 ± 0.6 x 21.0 ± 0.5 y 19.5 ± 0.3 y 20.8 ± 1.4 y
18:3(n-3) 0.6 ± 0.0 x 0.9 ± 0.0 y 1.1 ± 0.1 y 0.8 ± 0.0 yx 0.9 ± 0.0 y 1.5 ± 0.1 z
18:4(n-3) 0.0 ± 0.0 x 0.1 ± 0.0 x 0.8 ± 0.1 z 0.3 ± 0.0 y 0.4 ± 0.0 y 0.4 ± 0.0 y
20:5(n-3) 0.4 ± 0.0 x 1.1 ± 0.2 x 5.2 ± 0.4 z 2.6 ± 0.3 y 2.7 ± 0.2 y 2.7 ± 0.2 y
22:5(n-3) 0.4 ± 0.1 x 0.8 ± 0.2 x 3.4 ± 0.2 z 1.7 ± 0.2 y 1.5 ± 0.1 y 2.1 ± 0.1 y
22:6(n-3) 2.9 ± 0.4 x 7.1 ± 0.8 yx 16.5 ± 1.4 z 9.8 ± 1.3 y 9.6 ± 1.1 y 12.8 ± 1.0 zy
n-3 3.7 ± 0.4 w 9.3 ± 1.1 x 26.1 ± 2.0 z 14.6 ± 1.8 y 14.3 ± 1.4 y 18.3 ± 1.7 y
n-3 : n-6 ratio 0.1 ± 0.0 w 0.4 ± 0.1 xw 2.3 ± 0.2 z 0.7 ± 0.1 yx 0.7 ± 0.1 y 0.9 ± 0.1 y
PUFAs 30.6 ± 1.2 x 34.7 ± 1.4 yx 40.6 ± 1.5 z 37.0 ± 2.0 zy 35.4 ± 1.4 zy 41.2 ± 2.9 zy
LC-PUFAs 6.8 ± 0.8 w 11.6 ± 1.2 x 28.1 ± 2.1 z 17.3 ± 2.0 y 16.3 ± 1.6 y 20.1 ± 1.8 y
C18 PUFAs 23.2 ± 0.5 z 22.4 ± 0.8 z 12.1 ± 0.7 x 19.1 ± 0.4 y 18.6 ± 0.4 y 20.5 ± 1.4 y

feeds. An increase in fillet n-6 fatty acids (particularly 18:2[n-
6]) was also observed among fish that received the PF-based
feeds. Fillet levels of 18:0, 20:0, and 20:4(n-6) were not af-
fected by PF inclusion. Replacement of FM with PBM had a
similar—although less pronounced—effect on fatty acid profile,
with greater levels of PBM inclusion generally associated with
the loss of n-3 fatty acids and LC-PUFAs and the accumulation
of n-6 fatty acids and MUFAs.
Trial 2
Production performance did not vary among fish that were
given the experimental feeds exclusively or in combination;
however, significant differences were associated with adminis-
tering the Commercial feed (Table 5). Fish that were fed the
Commercial diet exhibited significantly lower FCR values than
fish that were fed according to any other feeding strategy ex-
cept the 15FM–6FO Control strategy; no significant differences
in FCR were observed among the other feeding strategies. No
significant differences were observed for percent weight gain.
There was a significant difference for dry matter consumption,
as fish that were given the Commercial feed apparently con-
sumed significantly lower amounts of dry matter than fish that
received the experimental diets. A significant difference in HSI
was also observed; fish that were fed the Commercial diet had
a significantly higher HSI than those that were given the ex-
perimental diets. The LSI did not significantly differ among
treatments. Survival was not affected by feeding strategy. Most
of the mortalities were fish that jumped from tanks, whereas
other mortalities were idiopathic but not apparently related to
dietary treatments.
Repeated-measures analysis revealed significant effects of
feeding strategy and time on fillet fatty acid profile for all
 ALTERNATIVE FEEDSTUFFS IN LARGEMOUTH BASS NUTRITION
30
25
20
LC-PUFA (% FAME)
15
10
0
Downloaded by [Department Of Fisheries] at 19:56 08 April 2013
0 2 4 6 8 10 12
Trial Duraon (weeks)
FIGURE 2. Mean long-chain polyunsaturated fatty acid (LC-PUFA) levels (% fatty acid methyl esters [FAMEs]) in fillets of Largemouth Bass that were fed
a commercial feed, a marine-feedstuff-based feed (15% fish meal [15FM]–6% fish oil [6FO]), or poultry-byproduct-based feeds alone or in combination with a
finishing feed (the 15FM–6FO feed). The points represent means of single individuals sampled from five replicate tanks; lines indicate moving averages to illustrate
general trends.
major fatty acids and fatty acid groupings except 20:4(n-6),
which was significantly affected by time but not feeding strategy,
and 22:5(n-3), which was significantly affected by feeding strat-
egy but not time. In particular, LC-PUFAs were affected by the
duration of rearing on grow-out and finishing feeds (Figure 2)
and apparently were affected by water temperature (Figure 1;
other temporal fatty acid data are not shown). At harvest, fillet
fatty acid composition reflected the fatty acid profile of the con-
sumed feeds (Table 5). Diets containing no FO yielded higher
fillet concentrations of 18:1(n-9) and 18:2(n-6) and lower con-
centrations of n-3 fatty acids and LC-PUFAs. Significant differ-
ences were observed in all fatty acids except 16:1(n-7), 18:0, and
20:4(n-6). Although consumption rates were reduced through-
out trial 2 due to lower water temperatures (Figure 1), finish-
ing with the 15FM–6FO diet still significantly increased fillet
LC-PUFA and n-3 fatty acid content for fish that were fed the
0FM–0FO or 7.5FM–0FO diet (Table 5).
DISCUSSION
We found that the FM and FO in diets for Largemouth Bass
could be completely replaced without affecting production per-
formance. Growth and FCR values observed in both trials were
excellent and were similar to ranges reported in other Large-
mouth Bass feeding studies (Tidwell et al. 2005; Subhadra et al.
2006). Each of these studies, including our trial 1, involved
small juveniles (starting weight ∼ 3–16 g), which readily took
to the experimental feeds and exhibited roughly 250–1,400%
weight gain over 9–12 weeks, depending largely on the initial
size of the fish. Although short-term feeding trials (e.g., trial
1 and those from the referenced studies) often yield equiva-
273
0 FM/0 FO
0 FM/0 FO + Finishing
7.5 FM/0 FO
7.5 FM/0 FM + Finishing
15 FM/6 FO
Commercial
0 FM/0 FM Moving Average
0 FM/0 FM + Finishing Moving Average
7.5 FM/0 FM Moving Average
7.5 FM/0 FM + Finishing Moving Average
15 FM/6 FO Moving Average
Commercial Moving Average
14 16 18 20 22 24 26 28
lent growth performance results that can be misleading, the fact
that the same results were also observed in a long-term growth
trial provides strong evidence that Largemouth Bass are toler-
ant of poultry-byproduct-based diets containing little to no FM
or FO. These results are also supported by a longer-term study
described by Cochran et al. (2009), who assessed reduced-FM
Largemouth Bass feeds containing a blend of PBM, soybean
meal, and blood meal in a long-term (180-d) growth trial. Those
authors reported no major differences in growth performance
among fish that were fed a commercial diet or experimental di-
ets containing 8–45% FM. The opportunistic feeding nature of
Largemouth Bass may play a role in the wide variety of protein
and fat sources that are accepted and the high alternative in-
gredient inclusion rates that are tolerated. Although lower FCR
values were achieved by using the Commercial feed in trial 2,
this is likely due to the physical attributes of the pellets rather
than to differences in composition of the various feeds. Because
the commercial pellets floated and were uniform, it was pos-
sible to remove uneaten feed and to calculate more accurate
consumption values. In contrast, the experimental pellets were
irregularly shaped and sank rapidly; thus, uneaten feed could not
be accounted for in the experimental treatments, likely leading
to artificially elevated estimates of consumption without the cor-
responding growth and in turn leading to elevated FCR values
for the experimental diets. Poultry byproduct meal is a highly
digestible feedstuff, and its apparent digestibility coefficients for
crude energy, crude protein, and dry matter are similar to those
of FM (Portz and Cyrino 2004; Masagounder et al. 2009). Al-
though the apparent digestibility coefficient of protein in PBM
is less than that of soybean meal, the need for diets containing
 274 COURSEY ET AL.
less than 20% carbohydrate to prevent liver necrosis precludes
the use of soybean meal as a complete replacement for FM in
Largemouth Bass feeds (Portz and Cyrino 2004; Masagounder
et al. 2009). No overt signs of liver necrosis were observed as
a result of feeding any of the experimental feeds in this study,
and HSI values associated with experimental feeds were well
within the ranges typically reported for Largemouth Bass cul-
ture (Tidwell et al. 2000). Although signs of liver necrosis were
not observed, significantly increased HSI values were associated
with the Commercial feed. This observation suggests that car-
bohydrate intake exceeded appropriate levels for Largemouth
Bass in the Commercial treatment (Mitchell et al. 2002; Amoah
et al. 2008). The specific formulation of the Commercial feed
is proprietary and unknown, but the guaranteed analysis of the
feed reported nearly 20% carbohydrate in the formulation. In
addition, the extrusion process that was used to manufacture the
Commercial feed (but not the other experimental diets) may have
Downloaded by [Department Of Fisheries] at 19:56 08 April 2013
made the carbohydrate fraction more readily digested, further
exacerbating any issues associated with carbohydrate excess
from this feed.
Although it was possible to replace FO with PF without af-
fecting production performance, the fillet fatty acid profile was
significantly altered. Numerous studies have shown that tissue
fatty acid composition is influenced by the fatty acid profile of
the diet for many fishes (Bell et al. 2001, 2004; Brandsen et al.
2003; Regost et al. 2003; Wonnacott et al. 2004; Lane et al.
2006; Lewis and Kohler 2008; Trushenski et al. 2011a, 2011b;
reviewed by Turchini et al. 2009), including Largemouth Bass
(Laporte and Trushenski 2011; Trushenski and Kohler 2011).
As long as essential fatty acid requirements are met, including
any requirements for LC-PUFAs, fish growth is rarely affected
by FO sparing or replacement (Turchini et al. 2009). Thus, it
is perhaps unsurprising that FO replacement with PF has been
accomplished without suppressing growth in a wide variety of
fish species (Trushenski and Lochmann 2009; Trushenski and
Kohler 2011; Trushenski et al. 2011a), including Largemouth
Bass (Subhadra et al. 2006; Tidwell et al. 2005), which re-
portedly have low or negligible requirements for FO-associated
LC-PUFAs (Tidwell et al. 2007). However, in each of these
cases, dietary inclusion of PF at the expense of FO has led to
an increase in fillet MUFAs and n-6 fatty acids and a reduc-
tion in fillet LC-PUFAs and n-3 fatty acids. Replacement of FM
with PBM did not have a significant effect on fatty acid com-
position, but minor numeric differences were observed, most
likely due to the amount of residual lipids found in both protein
meals as has been previously noted when administering PBM
(Bright et al. 2005; Subhadra et al. 2006; Trushenski and Kohler
2011). Although fish that are fed high levels of PF and PBM
would still provide a healthy, low-fat protein source, their fillets
would not provide maximal levels of beneficial n-3 fatty acids
and LC-PUFAs. However, applying a finishing feed 12 weeks
prior to harvest had a significant restorative effect on fillet lev-
els of LC-PUFAs and n-3 fatty acids. The n-3 fatty acid and
LC-PUFA levels observed among fish that were given the fin-
ishing diet were comparable to those associated with feeding
the Commercial diet. Finishing diets containing high levels of
FO have been effective in restoring 22:6(n-3) and total n-3 LC-
PUFAs in European Sea Bass Morone (Dicentrarchus) labrax
(Mourente et al. 2005), Atlantic Salmon Salmo salar (Bell et al.
2004), Red Seabream Chrysophrys (Pagrus) auratus (Glencross
et al. 2003), and White Bass Morone chrysops × Striped Bass
Morone saxatilis hybrids (Lane et al. 2006). These previous
studies have shown that the amount of marine feedstuffs given
to fish during production periods can be reduced without af-
fecting growth and that a finishing feed containing high levels
of FO can be used to maintain healthy n-3 fatty acid and LC-
PUFA concentrations at harvest. Utilization of finishing diets
may reduce the amounts of FM and FO used (and the accompa-
nying production costs) while maintaining a healthy product for
consumers.
To provide an abundant and quality food source, fish produc-
ers will need to reduce costs as well as provide a healthy and
affordable product for the marketplace. Use of PBM and PF as
feed ingredients in production diets could limit the overall costs
to producers while maintaining normal production growth. Sig-
nificant savings to producers are possible due to the low cost
of PBM ($300 per metric ton) in comparison with FM ($1,600
per metric ton) and the low cost of PF ($550 per metric ton) in
comparison with FO ($1,100 per metric ton; all prices are ap-
proximate and current as of June 2011; FAO 2012; C. Malone,
Tyson Foods, Inc., Springdale, Arkansas, personal communica-
tion). For example, Cochran et al. (2009) reported a significant
savings in cost per unit weight gain among Largemouth Bass that
were given reduced-FM feeds containing PBM, soybean meal,
or blood meal ($0.72–0.73 per kilogram) relative to fish that re-
ceived an FM-based experimental diet ($0.83 per kilogram) or a
commercial diet ($1.04 per kilogram; all formulations contained
FO). Largemouth Bass require at least two growing seasons to
reach a marketable size for the live market, and any reduction in
cost to producers could expand production and marketability. By
using PBM- and PF-based diets in combination with finishing
diets that are high in n-3 fatty acids and LC-PUFAs, producers
would be able to provide an affordable product with the benefits
of desirable fatty acids, such as 20:5(n-3) and 22:6(n-3). As the
costs of FM and FO continue to increase, PBM and PF appear
to be quality alternatives for producing carnivorous fish that re-
quire a low carbohydrate percentage in the diet. When coupled
with finishing diets that are high in n-3 fatty acids and LC-
PUFAs, these diets can provide low-cost, healthy alternatives to
diets that use only FM and FO as protein and lipid sources in
the culture of freshwater and marine fishes.
Although our study demonstrates that complete replace-
ment of FM and FO with poultry byproducts may not affect
growth performance, it is important to recognize the limitations
of our data sets and the risk associated with extrapolation to
commercial culture of Largemouth Bass or other species. The
Largemouth Bass production cycle requires two growing sea-
sons (2 years) for the fish to reach marketable size. Arguably,
 ALTERNATIVE FEEDSTUFFS IN LARGEMOUTH BASS NUTRITION
minor fatty acid (specifically LC-PUFA) deficiencies not ob-
served in this work could manifest over the length of a longer
feeding period. Thus, feeds that are low in marine ingredients
or that are marine ingredient free may not be appropriate for all
phases of Largemouth Bass production. Furthermore, in our tri-
als the Largemouth Bass were reared under optimal conditions,
whereas in a commercial production setting the fish are likely to
be periodically exposed to suboptimal conditions. Largemouth
Bass are typically cultured in earthen ponds, where they may
be exposed to variable and extreme temperatures, low dissolved
oxygen levels, pathogens, parasites, and other challenges associ-
ated with intensive fish culture. Additionally, Largemouth Bass
are commonly sold to live markets after being transported hun-
dreds of miles. Given the importance of certain nutrients (e.g.,
amino acids and fatty acids) in maintaining fish health and vigor
(Montero et al. 2003; Li et al. 2009), intensively reared Large-
mouth Bass may not perform as well on alternative-feedstuff-
Downloaded by [Department Of Fisheries] at 19:56 08 April 2013
based diets. Differences in dietary fatty acid intake have been
linked to differences in disease resistance for some cultured
fishes (Fracalossi and Lovell 1994; Li et al. 1994; Thompson
et al. 1996). Thus, seasonal variations in water temperature and
other environmental challenges could possibly lead to a ne-
cessity for specialized diets for Largemouth Bass (e.g., use of
high-LC-PUFA feeds prior to winter months to ensure overwin-
ter survival; or use of nutraceutical-supplemented diets prior to
handling or stocking). Future research should determine whether
replacement of FM and FO leads to lessened survivability during
overwintering or transport. Further research should (1) examine
whether complete replacement of FM and FO is possible in di-
ets for Largemouth Bass that are raised in earthen ponds; and
(2) incorporate stress challenges to determine whether PF and
PBM inclusion rates affect the stress tolerance and robustness
of Largemouth Bass.
REFERENCES
Adelizi, P. D., R. R. Rosati, K. Warner, Y. V. Wu, T. R. Muench, M. R. White,
and P. B. Brown. 1998. Evaluation of fish meal free diets for Rainbow Trout,
Oncorhynchus mykiss. Aquaculture Nutrition 4:225–262.
Amoah, A., S. D. Coyle, C. D. Webster, R. M. Durborow, L. A. Bright, and
J. H. Tidwell. 2008. Effects of graded levels of carbohydrate on growth and
survival of Largemouth Bass, (Micropterus salmoides). Journal of the World
Aquaculture Society 39:397–405.
AOAC (Association of Official Analytical Chemists). 1995. Official methods of
analysis. AOAC, Arlington, Virginia.
Bell, J. G., R. J. Henderson, D. R. Tocher, and J. R. Sargent. 2004. Replacement
of dietary fish oil with increasing levels of linseed oil: modification of flesh
fatty acid compositions in Atlantic Salmon (Salmo salar) using a fish oil
finishing diet. Lipids 39:223–232.
Bell, J. G., J. McEvoy, D. R. Tocher, F. McGee, P. J. Campbell, and J. R.
Sargent. 2001. Replacement of fish oil with rapeseed oil in diets of Atlantic
Salmon (Salmo salar) affects tissue lipid composition and hepatocyte fatty
acid metabolism. Journal of Nutrition 131:1535–1543.
Brandsen, M. P., C. G. Carter, and P. D. Nichols. 2003. Replacement of fish
oil with sunflower oil in feeds for Atlantic Salmon (Salmo salar L.): effect
on growth performance, tissue fatty acid composition and disease resistance.
Comparative Biochemistry and Physiology 135:611–625.
Bright, L. A., S. D. Coyle, and J. H. Tidwell. 2005. Effect of dietary lipid level
and protein energy ration on growth and body composition of Largemouth
275
Bass (Micropterus salmoides). Journal of the World Aquaculture Society
36:129–134.
Christie, W. W. 1982. Lipid analysis, 2nd edition. Pergamon, Oxford, UK.
Cochran, N. J., S. D. Coyle, and J. H. Tidwell. 2009. Evaluation of re-
duced fish meal diets for second year growout of the Largemouth Bass,
Micropterus salmoides. Journal of the World Aquaculture Society 40:735–
743.
Davis, A. D., D. Jirsa, and C. R. Arnold. 1995. Evaluation of soybean proteins
as replacements for menhaden fish meal in practical diets for the Red Drum
(Sciaenops ocellatus). Journal of the World Aquaculture Society 26:48–58.
El-Saidy, D. M. S. D., and M. M. A. Gaber. 2002. Complete replacement of fish
meal by soybean meal with dietary L-lysine supplementation for Nile Tilapia
Oreochromis niloticus (L.) fingerlings. Journal of the World Aquaculture
Society 33:297–306.
FAO (Food and Agriculture Organization of the United Nations). 2008. State of
world fisheries and aquaculture report—2008. FAO, Rome.
FAO (Food and Agriculture Organization of the United Nations). 2010. State of
world fisheries and aquaculture report—2010. FAO, Rome.
FAO (Food and Agriculture Organization of the United Nations). 2011a.
Fishmeal market report—August 2011. Available: http://www.globefish.
org/fishmeal.html. (June 2012).
FAO (Food and Agriculture Organization of the United Nations). 2011b. Fish
oil market report—August 2011. Available: http://www.globefish.org/fish-
oil-august-2011.html. (June 2012).
FAO (Food and Agriculture Organization of the United Nations). 2012. Com-
modity price index. Available: http://www.fao.org/economic/est/prices. (June
2011).
Folch, J., M. Lees, and G. H. S. Stanley. 1957. A simple method for the isolation
and purification of total lipids from animal tissues. Journal of Biological
Chemistry 226:497–459.
Fracalossi, D. M., and R. T. Lovell. 1994. Dietary lipid sources influence
responses of Channel Catfish (Ictalurus punctatus) to challenge with the
pathogen Edwardsiella ictaluri. Aquaculture 119:287–298.
Gallagher, M. L. 1994. The use of soybean meal as a replacement for fish meal in
diets for hybrid Striped Bass (Morone saxatilis × M. chrysops). Aquaculture
126:119–127.
Gatlin, D. M. III, S. D. Rawles, and T. G. Gaylord. 2007 Digestibility of gross
nutrients by Sunshine Bass in animal by products and commercially blended
products used as fish meal replacements. North American Journal of Aqua-
culture 68:74–80.
Glencross, B. D., W. E. Hawkins, and J. G. Curnow. 2003. Restoration of the
fatty acid composition of Red Seabream (Pagrus auratus) using a fish oil
finishing diet after grow-out on plant oil based diets. Aquaculture Nutrition
9:409–418.
Heidinger, R. C. 2000. A white paper on the status and needs of a Largemouth
Bass culture in the north central region. North Central Regional Aquaculture
Center, East Lansing, Michigan.
Kaushik, S. J., D. Coves, G. Dutto, and D. Blanc. 2004. Almost total re-
placement of fish meal by plant protein sources in the diet of a marine
teleost, the European Seabass, (Dicentrarchus labrax). Aquaculture 230:391–
404.
Lane, R. L., J. T. Trushenski, and C. C. Kohler. 2006. Modification of fillet
composition and evidence of differential fatty acid turnover in Sunshine Bass
(Morone chrysops × M. saxatilis) following change in dietary lipid source.
Lipids 41:1029–1038.
Laporte, J., and J. T. Trushenski. 2011. Growth performance and tissue fatty
acid composition of Largemouth Bass fed diets containing fish oil or blends
of fish oil and soy-derived lipids. North American Journal of Aquaculture
73:435–444.
Lewis, H. A., and C. C. Kohler. 2008. Corn gluten meal partially replaces
dietary fish meal without compromising growth or the fatty acid composition
of Sunshine Bass. North American Journal of Aquaculture 70:50–60.
Li, M. H., D. J. Wise, M. R. Johnson, and E. H. Robinson. 1994. Dietary
menhaden oil reduced resistance of Channel Catfish (Ictalurus punctatus) to
Edwardsiella ictaluri. Aquaculture 128:335–344.
 276 COURSEY ET AL.
Li, P., K. Mai, J. T. Trushenski, and G. Wu. 2009. New developments in
fish amino acid nutrition: towards function and environmentally oriented
aquafeeds. Amino Acids 37:43–53.
Lunger, A. N., S. R. Craig, and E. McLean. 2006. Replacement of fish meal in
Cobia (Rachycentron canadum) diets using an organically certified protein.
Aquaculture 257:393–399.
Masagounder, K., J. D. Firman, R. S. Hayward, S. Sun, and P. B. Brown.
2009. Apparent digestibilities of common feedstuffs for Bluegill (Lepomis
macrochirus) and Largemouth Bass (Micropterus salmoides) using individual
test ingredients. Aquaculture Nutrition 15:29–37.
Meeker, D. L. 2009. North American rendering-processing high quality protein
and fats for feed. Revista Brasileira de Zootecnia 38:432–440.
Mitchell, A. J., A. E. Goodwin, R. T. Lochman, and D. M. Tieman. 2002.
Massive hepatic necrosis and nodular regeneration in Largemouth Bass fed
diets high in available carbohydrate. Journal of the World Aquaculture Society
33:466–477.
Montero, D., T. Kalinowski, A. Obach, L. Robaina, L. Tort, M. J. Caballero, and
M. S. Izquirdo. 2003. Vegetable lipid sources for Gilthead Seabream (Sparus
aurata): effects on fish health. Aquaculture 225:353–370.
Mourente, G., J. E. Good, and J. G. Bell. 2005. Partial substitution of fish
Downloaded by [Department Of Fisheries] at 19:56 08 April 2013
oil with rapeseed, linseed and olive oils in diets for European Sea Bass
(Dicentrarchus labrax L.): effects on flesh fatty acid composition, plasma
prostaglandins E2 and F2α , immune function and effectiveness of a fish oil
finishing diet. Aquaculture Nutrition 11:25–40.
Naylor, R. L., R. J. Goldburg, J. H. Primavera, N. Kautsky, M. C. M. Beveridge,
J. Clay, C. Folke, J. Lubchenko, H. Looney, and M. Troell. 2000. Effect of
aquaculture on world fish supplies. Nature 405:1017–1024.
NRC (National Research Council). 2011. Nutrient requirements of fish and
shrimp. National Academies Press, Washington, D.C.
Portz, L., and J. E. P. Cyrino. 2004. Digestibility of nutrients and amino acids of
different protein sources in practical diets by Largemouth Bass (Micropterus
salmoides) (Lacepede, 1802). Aquaculture Research 35:312–320.
Regost, C., J. Arzel, J. Robin, G. Rosenlund, and S. J. Kaushik. 2003. Total
replacement of fish oil by soybean or linseed oil with a return to fish oil in
Turbot (Psetta maxima) 1. Growth performance, flesh fatty acid profile, and
lipid metabolism. Aquaculture 217:465–482.
Robinson, E. H., and M. H. Li. 1994. Use of plant proteins in catfish feeds:
replacement of soybean meal with cottonseed meal and replacement of fish
meal with soybean meal and cottonseed meal. Journal of the World Aquacul-
ture Society 25:271–276.
Subhadra, B., R. Lochmann, S. Rawles, and R. Chen. 2006. Effect of fish meal
replacement with poultry by-product meal on the growth, tissue composition
and hematological parameters of Largemouth Bass (Micropterus salmoides)
fed diets containing different lipids. Aquaculture 260:221–231.
Thompson, K. D., M. F. Tatner, and R. J. Henderson. 1996. Effects of dietary
(n-3) and (n-6) polyunsaturated fatty acid ratio on the immune response of
Atlantic Salmon, Salmo salar L. Aquaculture Nutrition 2:21–31.
Tidwell, J. H., S. Coyle, and L. A. Bright. 2007. Effects of different types of
dietary lipids on growth and fatty acid composition of Largemouth Bass.
North American Journal of Aquaculture 69:257–264.
Tidwell, J. H., S. D. Coyle, L. A. Bright, and D. Yasharian. 2005. Evaluation
of plant and animal source proteins in practical diets for Largemouth Bass
(Micropterus salmoides). Journal of the World Aquaculture Society 36:454–
463.
Tidwell, J. H., S. D. Coyle, and T. A. Woods. 2000. Species profile: Largemouth
Bass. South Regional Aquaculture Center, Publication 722, Stoneville, Mis-
sissippi.
Tidwell, J. H., C. D. Webster, and S.D Coyle. 1996. Effects of dietary pro-
tein level on second year growth and water quality for Largemouth Bass
(Micropterus salmoides) raised in ponds. Aquaculture 145:213–223.
Trushenski, J. T., C. S. Kasper, and C. C. Kohler. 2006. Challenges and oppor-
tunities in finfish nutrition. North American Journal of Aquaculture 68:122–
140.
Trushenski, J. T., and C. C. Kohler. 2011. Joint replacement of marine feedstuffs
with stabilized poultry protein meal and fat in practical diets for Sunshine
Bass (Morone chrysops × M. saxatilis). Journal of Applied Aquaculture
23:329–350.
Trushenski, J. T., and R. T. Lochmann. 2009. Potential, implications and so-
lutions regarding the use of rendered animal fats in aquafeeds. American
Journal of Animal and Veterinary Sciences 4:108–128.
Trushenski, J. T., J. Rosenquist, and B. R. Gause. 2011a. Growth performance,
tissue fatty acid composition, and consumer appeal of Rainbow Trout reared
on feeds containing terrestrially derived rendered fats. North American Jour-
nal of Aquaculture 73:468–478.
Trushenski, J. T., M. Schwarz, H. Lewis, J. Laporte, B. Delbos, R. Takeuchi,
and L. A. Sampaio. 2011b. Effect of replacing dietary fish oil with soybean
oil on production performance and fillet lipid and fatty acid composition
of juvenile Cobia (Rachycentron canadum). Aquaculture Nutrition 17:437–
447.
Turchini, G. M., B. E. Torstensen, and W. Ng. 2009. Fish oil replacement in
finfish nutrition. Reviews in Aquaculture 1:10–57.
Webster, C. D., L. G. Tiu, A. M. Morgan, and A. Gannam. 1999. Effect of partial
and total replacement of fish meal on growth and total body composition of
Sunshine Bass (Morone saxatilis × Morone chrysops) fed practical diets.
Journal of the World Aquaculture Society 30:443–453.
Wonnacott, E. J., R. L. Lane, and C. C. Kohler. 2004. Influence of dietary
replacement of menhaden oil with canola oil on fatty acid composition of
Sunshine Bass. North American Journal of Aquaculture 66:243–250.
This article was downloaded by: [Department Of Fisheries]
On: 08 April 2013, At: 19:56
Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,
37-41 Mortimer Street, London W1T 3JH, UK

North American Journal of Aquaculture

Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/unaj20

Evaluation of the Effects of Menhaden Oil and Soybean

Oil Levels in Purified Diets on Growth and Gonad
Production in Adult Sea Urchin Lytechinus variegatus
a d a a a
Victoria K. Gibbs , Mickie L. Powell , Hugh S. Hammer , Warren T. Jones , Stephen A.
a b c
Watts , Addison L. Lawrence & John M. Lawrence
a
Department of Biology, University of Alabama at Birmingham, 1300 University Boulevard,
Birmingham, Alabama, 35294, USA
b
Texas AgriLife Mariculture Research Laboratory, Texas A&M University System, 1300 Port
Street, Port Aransas, Texas, 78373, USA
c
Department of Integrative Biology, University of South Florida, 4202 East Fowler Avenue,
Tampa, Florida, 33620, USA
d
Department of Biology, Villanova University, 800 Lancaster Avenue, Villanova,
Pennsylvania, 19085, USA
Version of record first published: 02 Apr 2013.

To cite this article: Victoria K. Gibbs , Mickie L. Powell , Hugh S. Hammer , Warren T. Jones , Stephen A. Watts , Addison
L. Lawrence & John M. Lawrence (2013): Evaluation of the Effects of Menhaden Oil and Soybean Oil Levels in Purified Diets
on Growth and Gonad Production in Adult Sea Urchin Lytechinus variegatus , North American Journal of Aquaculture, 75:2,
277-284

To link to this article: http://dx.doi.org/10.1080/15222055.2012.741559

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions

This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to
anyone is expressly forbidden.

The publisher does not give any warranty express or implied or make any representation that the contents
will be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses should
be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims,
proceedings, demand, or costs or damages whatsoever or howsoever caused arising directly or indirectly in
connection with or arising out of the use of this material.
 North American Journal of Aquaculture 75:277–284, 2013

C American Fisheries Society 2013

ISSN: 1522-2055 print / 1548-8454 online

DOI: 10.1080/15222055.2012.741559

SPECIAL SECTION: LIPIDS IN AQUACULTURE NUTRITION AND PHYSIOLOGY II

Evaluation of the Effects of Menhaden Oil and Soybean Oil

Levels in Purified Diets on Growth and Gonad Production
in Adult Sea Urchin Lytechinus variegatus

Victoria K. Gibbs,*1 Mickie L. Powell, Hugh S. Hammer, Warren T. Jones,

and Stephen A. Watts
Department of Biology, University of Alabama at Birmingham, 1300 University Boulevard, Birmingham,
Alabama 35294, USA

Addison L. Lawrence
Downloaded by [Department Of Fisheries] at 19:56 08 April 2013

Texas AgriLife Mariculture Research Laboratory, Texas A&M University System, 1300 Port Street,
Port Aransas, Texas 78373, USA

John M. Lawrence
Department of Integrative Biology, University of South Florida, 4202 East Fowler Avenue, Tampa,
Florida 33620, USA

Abstract
Development of a standardized reference diet will facilitate the determination of nutritional requirements for sea
urchins. A purified diet, containing only chemically defined ingredients, provides consistency in diet formulations
for reproducibility across different laboratories. In the present study, the growth performance of small Lytechinus
variegatus (11.9 ± 1.3 g [mean ± SD] wet weight, 28.4 ± 1.1 mm test diameter; 16 per treatment) fed ad libitum
daily (16 weeks) one of seven purified diets differing in the level (0, 2.4, or 4.8% as fed) and source of neutral lipid
(refined menhaden oil and soy oil) was compared with the growth performance of individuals fed a semipurified diet
previously shown to support high growth rates and gonad production. Survival was ≥88% for all dietary treatments,
and the growth rates of individuals fed the 1.2% menhaden oil and 1.2% soybean oil purified diet (215% wet weight
gain) were approximately 65% of those for individuals fed the semipurified diet (326% wet weight gain). The growth
rates and wet organ weights for individuals fed purified diets were not significantly affected by lipid source. However,
increased total dietary lipid resulted in higher dry gut weights. Ovary weight was higher than testis weight for
all treatments. Although no significant differences were detected among the purified diet treatments, qualitative
performance (survival, weight gain, and gonad production) for individuals fed the purified diet containing 1.2%
menhaden oil and 1.2% soybean oil was best under the conditions of this study. The purified diet used in this study is
adequate to evaluate specific nutrients on sea urchin weight gain and organ production.

Sea urchin fisheries worldwide are on the decline due to
overexploitation of wild stocks (Keesing and Hall 1998; Lesser
and Walker 1998; Andrews et al. 2002; Robinson 2004). Global
capture production reported by the Food and Agriculture Orga-
nization of the United Nations (FAO) for sea urchins and other
echinoderms had decreased 24% by 2010 from its peak in 1995.
Alternatively, reported world aquaculture production of echino-
derms (primarily Japanese sea cucumbers Stichopus japonicus)
continues to rise and has at times exceeded capture produc-
tion by more than 35% (FAO 2012). Commercial culture of sea

*Corresponding author: victoria.gibbs@villanova.edu

1
Current address: Department of Biology, Villanova University, 800 Lancaster Avenue, Villanova, Pennsylvania 19085, USA.
Received April 4, 2012; accepted October 14, 2012
Published online April 2, 2013
277
 278 GIBBS ET AL.
urchins will require nutritionally complete diets formulated to
provide the daily nutritional requirements for supporting opti-
mal growth at all life stages. The nutritional requirements for
growth and/or gonad production in a sea urchin species are diffi-
cult to assess from published literature due to inconsistencies in
diet formulations across laboratories, the failure to use purified
ingredients for formulations, and the potential for differential
dietary requirements among species. Watts et al. (2010) offered
specific criteria to consider when designing and implementing
nutritional studies, with the objective of establishing standard-
ization of experimental design among studies, including the use
of a reference diet.
Semipurified diets, i.e., mixtures of purified and practical in-
gredients, are commonly used for nutritional studies and can be
used to quantify requirements for some nutrients (Watts et al.
2010). Practical ingredients, such as animal and plant meals,
contain a variety of nutrients (proteins, carbohydrates, lipids,
Downloaded by [Department Of Fisheries] at 19:56 08 April 2013
etc.) and may vary in composition among different batches or
lots. New lots of practical ingredients must be analyzed for
nutrient composition, and diet formulations must be adjusted
accordingly to maintain desired nutrient levels. D’Abramo and
Castell (1997) recommended that studies designed to determine
qualitative and quantitative nutrient requirements for a species
be conducted using purified diets. The American Institute of Nu-
trition, Experimental Animal Nutrition Committee (EANCAIN
1987), defined purified diets as being composed of commer-
cially refined proteins, carbohydrates, and lipids with added
mineral and vitamin mixtures. The higher purity of ingredients
characteristic of purified diets allows greater control of the nu-
trient composition of the diet to achieve reproducibility over
time (D’Abramo and Castell 1997). The use of purified diets
offers the ability to vary single nutrients while keeping all other
nutrients constant. A purified diet that promotes adequate sur-
vival and growth has the potential to be used as a standardized
reference diet for nutritional research.
Studies using purified diets in sea urchins are limited.
Akiyama et al. (2001) reported the use of a casein-based pu-
rified diet to determine the optimum dietary protein level for
the growth of small Pseudocentrotus depressus. The purified
diets used in that study supported growth and precocious gonad
production, and growth performance was greater for urchins fed
the purified diets than for those fed a brown alga diet (Akiyama
et al. 2001).
Dietary lipid requirements can be difficult to determine be-
cause lipids comprise many diverse compounds that have a va-
riety of functions (NRC 2011). The National Research Council
(2011) recommended that dietary lipids be supplied across a
range, with the lower limit defined as the amount required to
supply the requirements for essential fatty acids (EFAs) and the
upper limit as the amount that results in unwanted deposition
of lipid in the body. The level of dietary lipid needed to supply
EFAs will depend on the fatty acid profile of the lipid source
(NRC 2011). Essential fatty acids are composed of the linoleic
n-6 and linolenic n-3 families of polyunsaturated fatty acids
(PUFAs). The linolenic family typically has the highest EFA
value for marine animals (D’Abramo 1997). Marine fish oils
are important sources of the long-chain polyunsaturated fatty
acids (LC-PUFAs) such as eicosapentaenoic acid (20:5[n-3]1)
and docosahexaenoic acid (22:6[n-3]) (De Silva et al. 2011).
In an attempt to develop sustainable diet formulations, ef-
forts have been made to partially or completely replace fish oils
with plant oils. The most common plant oil sources, soybean
oil and corn oil, are high in linoleic acid (18:2[n-6]) and have
very low levels of linolenic acid (18:3[n-3]) (Brown and Hart
2011). The LC-PUFAs characteristic of marine fish oils typi-
cally have higher nutritive values for marine crustaceans than
18:2(n-6) (D’Abramo 1997). Sea urchins are capable of syn-
thesizing 20:4(n-6) (arachidonic acid) and 20:5(n-3) from C18
PUFAs (Bell et al. 2001; Castell et al. 2004; Liu et al. 2007a,
2007b; González-Durán et al. 2008). Dietary lipid studies with
Lytechinus variegatus fed semipurified diets suggest that low
levels of marine derived oils should be included to support opti-
mal growth for juveniles and small adults (Hammer et al. 2010;
Gibbs 2011).
In this study, the growth performance and gonad production
of small adult L. variegatus fed a purified diet were compared
with those of small adults fed a semipurified diet that is reported
to support weight gain and gonad production (Gibbs et al. 2010;
Hammer et al. 2010, 2012). Additionally, the performance of
purified diets that differ in levels of menhaden oil (FO) and
soy oil (SO), exclusively or in combination, were compared to
determine the level that best supports weight gain and gonad
production in small adult L. variegatus.
METHODS
Collection, initial measurements, and culture system.—Adult
L. variegatus (∼28 mm test diameter) were collected in Octo-
ber 2004 from Saint Joseph Bay, Florida (30◦ N, 85.5◦ W), and
transported to the Texas AgriLife Mariculture Research Labo-
ratory in Port Aransas, Texas. The sea urchins were held for
approximately 1 month in 750-L tanks at approximately 32 ±
2 g/kg salinity and 22 ± 1◦ C. Natural seawater was filtered
using stratified sand filtration and a Diamond water filter (Di-
amond Water Conditioning, Horton, Wisconsin) before being
piped to the systems under flow-through conditions (exchange
rate, approximately 150% daily). During this holding period,
the sea urchins were fed a maintenance ration (approximately
once every 3 d) of a formulated diet (Hammer et al. 2012).
Sea urchins (11.9 ± 1.3 g wet weight [mean ± SD], 28.4 ±
1.1 mm test diameter, 16 per treatment) were randomly selected
from the collection population for entry into the growth trial.
Initial wet weights were determined by blotting individuals for
1In fatty acid designations of this nature, the number to the left of the colon
is the number of carbon atoms in the compound, the number immediately to
the right of the colon is the number of double bonds, and the number after the
hyphen indicates the position of the first double bond from the methyl end.
 MENHADEN VERSUS SOYBEAN OIL IN SEA URCHIN DIETS 279

10 s on paper towels to remove excess water and weighing them TABLE 1. Calculated nutrient levels on an as-fed basis for the semipurified
to the nearest milligram on a Mettler balance (Mettler Toledo and base purified diets used in this study. The nutrient profiles for ingredients
provided by vendors were used to calculate the nutrient levels of diets. Aster-
Scales, Dublin, Ohio). Individual test diameters were measured isks denote levels empirically determined by Eurofins Scientific, Inc., for the
to the nearest millimeter at two perpendicular points across the semipurified diet only.
ambitus using callipers.
The culture system used for the growth trial was described Semipurified Base purified
by Jones et al. (2010). Briefly, the sea urchins were placed in- Nutrient diet diet
dividually into cylindrical enclosures constructed from plastic
Crude protein (%)* 28.38 28.04
4-mm open mesh (∼30 cm high, 12 cm in diameter). The mesh
Carbohydrate (%) 32.62 37.35
enclosures were fitted into 11.5-cm (interior diameter) PVC cou-
Crude fiber (%)* 2.40 2.00
plings, and small plastic spacers (∼0.5 cm thick) were placed
Total ash (%)* 22.25 18.64
under the bottom of each coupling to allow water circulation
Crude fat (%)* 7.54 3.29
beneath the enclosures. The floor of the mesh enclosure was ap-
Cholesterol (%) 0.22 0.19
proximately 5.5 cm from the bottom of the tank. Four cylindrical
Carotenoid (%) 0.42 0.92
enclosures were placed into each fiberglass tank with a 0.07 m2
Calcium (%)* 2.71 2.55
bottom surface area and with 20 L water volume (4 tanks per
Phosphorus (%)* 2.01 1.94
treatment × 4 individuals per tank). Seawater was supplied
Downloaded by [Department Of Fisheries] at 19:56 08 April 2013

Sodium (%) 1.29 1.35

to each enclosure at a ca. rate of 25 L/h. Tanks were con-
Potassium (%) 1.63 1.62
nected within a temperature-controlled, semirecirculating aqua-
Magnesium (%) 0.39 0.40
culture system with mechanical and biological filtration, foam
Iron (ppm) 319 326
fractionation, and UV sterilization. Seawater was exchanged
Zinc (ppm) 91 91
in the semirecirculating system at an approximate rate of 10%
Manganese (ppm) 71 69
volume/d.
Copper (ppm) 47 47
Water parameters were maintained daily at 32 ± 2 g/kg salin-
Selenium (ppm) 0.228 0.227
ity, 22 ± 1◦ C, and dissolved oxygen at 7 ± 2 mg/L. Photoperiod
Arginine (%) 2.15 2.03
was maintained at 12 h light : 12 h dark. Ammonia, nitrite, ni-
Histidine (%) 0.62 0.70
trate, and pH levels were checked weekly by chemical titration.
Isoleucine (%) 1.12 1.24
Diet preparation and feeding.—A semipurified diet contain-
Leucine (%) 2.00 2.37
ing practical and purified ingredients (approximately 28% ma-
Lysine (%) 1.69 1.98
rine source ingredients, 34.6% plant source ingredients, 6.5%
Methionine (%) 0.49 0.65
crude fat, 1.1% carotenoids, 0.7% vitamin premix, 18.9% min-
Cystine (%) 0.27 0.20
eral premix, 10.2% binder antifungal antioxidant; Table 1) and
Phenylalanine (%) 1.29 1.33
seven purified diets (Table 1) that varied in the level and source
Tyrosine (%) 0.96 1.20
of neutral lipid (Table 2) were prepared as follows: In puri-
Threonine (%) 1.05 0.96
fied diets, menhaden oil and soybean oil were reciprocally ex-
Tryptophan (%) 0.26 0.29
changed by weight with wheat starch and all other nutrients
Valine (%) 1.19 1.41
were held constant. Ingredients were blended with a twin shell
Vitamin A (IU) 4,800 4,800
dry blender (Patterson-Kelley Co., East Stroudsburg, Pennsyl-
Vitamin D (IU) 3,000 3,000
vania) for 10 min and then mixed in a Hobart mixer (Model
Vitamin E (ppm) 240 240
A-200; Hobart Corporation, Troy, Ohio) for 40 min. De-ionized
Vitamin C (ppm) 349 349
water (500 mL/kg) with binders was then added to the dry in-
Thiamine (ppm) 36 36
gredients and mixed an additional 10 min to achieve a mash
Riboflavin (ppm) 48 48
consistency appropriate for extrusion. Diets were extruded at
Pyridoxine (ppm) 96.0 96.3
room temperature (26–28◦ C) using a meat chopper attachment
Niacin (ppm) 99.0 99.3
(Model A-200; Hobart Corporation) fitted with a 3.2-mm die.
Pantothenic acid (ppm) 36.0 36.5
Moist, rod-like pellets were dried in a forced air oven at 35◦ C to
Biotin (ppm) 1 0.971
a moisture content of 8–10%, placed in plastic bags, and stored
at 4◦ C. Small aliquots of feed pellets (∼50 g) were stored at Inositol (%) 0.10 1.04
room temperature (23◦ C) for daily feeding activities. The prox- Choline (%) 0.10 0.8
imate composition of the diets was empirically determined by Folic acid (ppm) 24.0 24.0
Eurofins Scientific, Inc. (Des Moines, Iowa; Table 1). Individual Vitamin B12 (ppm) 0.2 0.2
sea urchins were fed their respective treatment diets ad libitum Gross energy (cal/g) 3,823 3,658
daily for 16 weeks. Twenty-four hours after the animals were
 280 GIBBS ET AL.

TABLE 2. Ingredients and levels (as fed) used to formulate purified diets. The levels of menhaden oil (FO) and/or soy oil (SO) were exchanged with wheat
starch to maintain the levels of other nutrients.

FO (%) : SO (%)
Ingredient (%) 0:0 1.2 : 1.2 2.4 : 0 0 : 2.4 2.4 : 2.4 4.8 : 0 0 : 4.8
Wheat starch 36.8 34.4 34.4 34.4 32.0 32.0 32.0
Casein (vitamin free) 15.0 15.0 15.0 15.0 15.0 15.0 15.0
Soy protein isolate 10.0 10.0 10.0 10.0 10.0 10.0 10.0
Dicalcium phosphate 7.2 7.2 7.2 7.2 7.2 7.2 7.2
Gelatin 4.0 4.0 4.0 4.0 4.0 4.0 4.0
Acid-washed diatomaceous earth 3.5 3.5 3.5 3.5 3.5 3.5 3.5
Magnesium sulfate 3.5 3.5 3.5 3.5 3.5 3.5 3.5
Soy lecithin (97% acetone insolubles) 3.3 3.3 3.3 3.3 3.3 3.3 3.3
Menhaden oil 0.0 1.2 2.4 0.0 2.4 4.8 0.0
Soy oil 0.0 1.2 0.0 2.4 2.4 0.0 4.8
Alginate 3.0 3.0 3.0 3.0 3.0 3.0 3.0
Downloaded by [Department Of Fisheries] at 19:56 08 April 2013

Potassium chloride 3.0 3.0 3.0 3.0 3.0 3.0 3.0

Cellulose 2.5 2.5 2.5 2.5 2.5 2.5 2.5
Calcium carbonate 2.2 2.2 2.2 2.2 2.2 2.2 2.2
Sodium chloride 2.2 2.2 2.2 2.2 2.2 2.2 2.2
Beta carotene 1.0 1.0 1.0 1.0 1.0 1.0 1.0
Sodium hexametaphosphate 1.0 1.0 1.0 1.0 1.0 1.0 1.0
Vitamin premixa 0.6 0.6 0.6 0.6 0.6 0.6 0.6
Arginine 0.3 0.3 0.3 0.3 0.3 0.3 0.3
Cholesterol 0.3 0.3 0.3 0.3 0.3 0.3 0.3
Ethoxiquine 0.2 0.2 0.2 0.2 0.2 0.2 0.2
Iron sulfate 0.11 0.11 0.11 0.11 0.11 0.11 0.11
Methionine 0.10 0.10 0.10 0.10 0.10 0.10 0.10
Stabilized vitamin C 0.10 0.10 0.10 0.10 0.10 0.10 0.10
Zinc sulfate 0.03 0.03 0.03 0.03 0.03 0.03 0.03
Copper sulfate 0.02 0.02 0.02 0.02 0.02 0.02 0.02
Manganese sulfate 0.01 0.01 0.01 0.01 0.01 0.01 0.01
Gross energy (cal/g) 3,658 3,837 3,934
a
Vitamin premix contained the following levels per kilogram of feed: retinol, 800,000 IU; ascorbic acid, 350 mg; cholecalciferol, 500,000 IU; tocopherol, 40,000 mg; thiamine,
6,000 mg; riboflavin, 8,000 mg; pyridoxine, 16,000 mg; niacin, 16,000 mg; pantothenic acid, 6,000 mg; biotin, 160 mg; folic acid, 4,000 mg; cyanocobalamine, 30mg.

proffered feed, uneaten feed was removed from the enclosure by
siphon. Feces were removed from tanks by siphon each week.
Growth and organ production.—The wet weight and test
diameter of each individual was determined as previously de-
scribed at initial stocking and at 4, 8, 12, and 16 weeks. The
percentage wet weight gain over the 16-week study period was
calculated as
{[final wet weight (g) − initial wet weight (g)]/
initial wet weight (g)} × 100.
removed from the coelomic cavity and rinsed in clean culture
water to remove feed pellets. The gonads were removed from the
coelomic cavity, and a wet mount of gonad tissue was checked
for each individual to determine sex. The gut and gonads were
blotted on paper towels to remove excess water, each trans-
ferred to a preweighed aluminum weigh pan and weighed to
determine organ wet weight to the nearest milligram. Organs
were then dried to a constant dry weight in a forced air oven
at 50◦ C (72 h). Organ indices were calculated for the gut and
gonad based on dry weight as

At the end of 16 weeks, individuals were dissected. A circu-
lar incision around the peristomial membrane allowed the oral
surface (including Aristotle’s lantern) of the sea urchin to be
separated from the body. The esophagus was cut at Aristotle’s
lantern, and the lantern was separated from the oral surface of
the test. The gut (esophagus, stomach, and intestine) was then
[dry organ weight/total dry weight] × 100.
Statistics.—The semipurified dietary treatment was com-
pared with the seven purified dietary treatments using one-way
analysis of variance (ANOVA) to evaluate organismal growth
and organ weights exclusive of the gonad. A two-way ANOVA
 MENHADEN VERSUS SOYBEAN OIL IN SEA URCHIN DIETS 281

was used to compare ovaries and testes from the semipurified

diet treatment with those from the purified diet treatment. A
P-value < 0.05 was considered statistically significant. Tukey’s
adjustment for multiple comparisons was used to compare spe-
cific differences in treatment means. The assumptions of nor-
mality and equality of variance were satisfied for all data.
For the purified dietary treatments, general linear models
(GLMs) were created to determine how much of the variation
in the dependent (outcome) variables could be explained by the
predictor variables using the “glm” function in R version 2.12.1
(R Foundation for Statistical Computing). Outcomes associated
with organismal growth and organ weights were each included
as dependent variables in the GLMs with one or more of the
following predictor variables: FO level, SO level, total lipid
level, and sex. In the GLMs, organ weights were size-adjusted
using the weight of the test. Nonsignificant predictor variables
were removed from the model, and parameter estimates for
Downloaded by [Department Of Fisheries] at 19:56 08 April 2013

predictors with a P value < 0.05 were recorded. Normality and

homogeneity of variance of the residuals were confirmed for all
data using the “glm.diag.plots” function in the library “boot.”

RESULTS
Water Quality and Survival
Water quality parameters were maintained as follows: tem-
perature, 22 ± 2◦ C; salinity, 32 ± 2 g/kg; ammonia, 0.1 ±
0.05 mg/L; nitrite, 0.1 ± 0.05 mg/L; nitrate, 5 ± 2 mg/L; and
pH, 8 ± 0.3. Survival was 100% for the semipurified diet and
all purified diet treatments containing FO. Survival for the 0%
FO–2.4% SO diet treatment was 94%, and survival for the 0%
FO–4.8% SO treatment was 88%.
Organismal Growth
Wet weights and diameters were not different between the
sexes and were combined for analysis. Wet weight gain among
all treatments was >150% after 16 weeks (Figure 1A). Wet
weight gain for individuals fed the semipurified diet was higher
than that for individuals fed a purified diet (ANOVA; model
df = 7, F = 14.01, P < 0.0001; Figure 1A). Diameter gain
(%) was also higher for individuals fed the semipurified diet
(ANOVA; model df = 7, F = 6.512, P < 0.001; Figure 1B). No
significant differences in weight gain or diameter were found
among the purified dietary treatments.
Organ Weights
Gut weights were not different between sexes and were com-
bined for analysis. Wet gut weights (Table 3) for individuals
fed the semipurified diet were higher than those for individu-
als fed purified diets except the diets containing 0% FO–4.8%
SO or 1.2% FO–1.2% SO (ANOVA; model df = 7, F = 4.21,
P = 0.0004). Dry gut weights (Table 3) were also higher for
individuals fed the semipurified diet than those for fed puri-
fied diets except the diets containing 0% FO–4.8% SO, 2.4%
FO–2.4% SO, and 2.4% FO–0% SO (ANOVA; model df =
FIGURE 1. (A) Wet weight gain (%) and (B) diameter gain (%) for adult
Lytechinus variegatus fed either purified diets containing different levels of
menhaden and soybean oils (% as fed) or a semipurified diet for 16 weeks. Values
are means ± SDs (n = 14–16 per treatment); asterisks indicate significantly
different treatments (ANOVA; P < 0.0001).
7, F = 3.16, P = 0.0051). Total dietary lipid level was a sig-
nificant predictor of dry gut weight when adjusted for total
size using the dry test weight (GLM; P = 0.0409). Parameter
estimates indicate that a 1% increase in total dietary lipid in-
creased dry gut weight by 0.004 g. The dry gut index (Table 4)
was not different among dietary treatments (ANOVA; model
df = 7, F = 0.58, P = 0.7741) and was not significantly af-
fected by dietary lipid level or source.
Gonad weights differed between females and males, so go-
nads were separated by sex (ovary and testis) for analysis. Wet
ovary and testis weights (Table 3) were higher for individuals
fed the semipurified diet than for individuals fed a purified diet
(ANOVA; model df = 7, F = 5.69, P < 0.0001). Among the
purified diets, the parameter estimates indicate that, when ad-
justed for wet test weight, the wet ovary weight of females was
1.02 g higher than the wet testis weight of males (GLM; P =
0.0001). Dry ovary and testis weights (Table 3) were higher for
individuals fed the semipurified diet (ANOVA; model df = 7,
 282 GIBBS ET AL.

TABLE 3. Wet and dry organ weights of adult Lytechinus variegatus fed either purified diets containing different levels of menhaden (FO) and soy oils (SO)
(% as fed) or a semipurified diet for 16 weeks. Values are means ± SDs (n = 14–16 per treatment). Within columns, values with the same lowercase letters are
not significantly different.

FO SO Gut Ovary Testis

Wet weight (g)
0 0 0.63 ± 0.12 y 4.78 ± 1.4 y 2.60 ± 1.4 y
1.2 1.2 0.73 ± 0.13 zy 5.17 ± 1.7 y 3.65 ± 1.3 y
2.4 0 0.71 ± 0.16 y 5.16 ± 1.8 y 4.15 ± 2.1 y
0 2.4 0.67 ± 0.15 y 4.89 ± 0.8 y 3.52 ± 1.1 y
2.4 2.4 0.71 ± 0.15 y 4.44 ± 1.2 y 4.25 ± 1.1 y
4.8 0 0.64 ± 0.16 y 4.08 ± 1.7 y 3.64 ± 1.1 y
0 4.8 0.75 ± 0.23 zy 4.45 ± 2.0 y 3.18 ± 1.6 y
Semipurified diet 0.90 ± 0.22 z 7.81 ± 1.6 z 5.24 ± 1.5 z
Dry weight (g)
0 0 0.14 ± 0.03 y 1.36 ± 0.50 y 0.84 ± 0.46 y
1.2 1.2 0.15 ± 0.03 y 1.63 ± 0.61 y 1.22 ± 0.49 y
Downloaded by [Department Of Fisheries] at 19:56 08 April 2013

2.4 0 0.16 ± 0.04 zy 1.48 ± 0.52 y 1.42 ± 0.82 y

0 2.4 0.15 ± 0.04 y 1.53 ± 0.30 y 1.20 ± 0.35 y
2.4 2.4 0.16 ± 0.04 zy 1.26 ± 0.40 y 1.45 ± 0.41 y
4.8 0 0.15 ± 0.04 y 1.27 ± 0.60 y 1.16 ± 0.38 y
0 4.8 0.17 ± 0.06 zy 1.30 ± 0.61 y 0.95 ± 0.44 y
Semipurified diet 0.20 ± 0.05 z 2.43 ± 0.54 z 1.76 ± 0.48 z

F = 6.12, P < 0.0001). Among the purified diets, the param-
eter estimates indicate that, when adjusted for dry test weight,
the dry ovary weights of females were 0.21 g higher than the
dry testis weights of males (GLM; P = 0.0143). The results
from the two-way ANOVA indicated a significant effect of sex
(model df = 1, F = 10.13, P = 0.0019) and diet treatment
(model df = 7, F = 2.36, P = 0.0281) on gonad index. The dry
ovary indices were higher than the dry testis indices (Table 4).
No significant interaction between sex and diet was observed
(two-way ANOVA; model df = 7, F = 0.937, P = 0.4813).
Individuals fed the semipurified diet had higher gonad indices
than individuals fed purified diets containing 0% FO–0% soy oil
or 0% FO–4.8% soy oil (Table 4). Dietary lipid level or source
in the purified diets did not significantly affect wet or dry ovary
or testis weights or dry indices.
DISCUSSION
Water quality parameters were maintained within an optimal
range and within safe levels for nitrogen (no negative effect on
growth) according to Basuyaux and Mathieu (1999). Further, the
high growth and survival of the sea urchins fed the semipurified
diet was similar to that observed in other published reports
(Gibbs et al. 2010; Hammer et al. 2010, 2012) and confirm that
water quality was not limiting in this study.
Among their criteria for standardizing sea urchin nutrition
studies, Watts et al. (2010) stated that an adequate feed for the

TABLE 4. Organ indices for adult Lytechinus variegatus fed either a semipurified diet or a purified diet differing in menhaden (FO) and soy oil (SO) levels
(% as fed) for 16 weeks. Values are means ± SDs (n = 14–16 per treatment). Within columns, values with same lowercase letters are not significantly different.

FO SO Dry gut index (%) Dry ovary index (%) Dry testis index (%)
0 0 1.69 ± 0.36 z 15.5 ± 5.2 y 10.8 ± 5.1 y
1.2 1.2 1.71 ± 0.32 z 17.3 ± 5.2 zy 13.7 ± 4.9 zy
2.4 0 1.76 ± 0.32 z 16.1 ± 3.3 zy 16.1 ± 6.4 zy
0 2.4 1.67 ± 0.36 z 16.4 ± 1.8 zy 13.7 ± 3.5 zy
2.4 2.4 1.82 ± 0.24 z 14.1 ± 3.3 zy 14.4 ± 3.9 zy
4.8 0 1.84 ± 0.34 z 14.7 ± 4.9 zy 14.2 ± 3.8 zy
0 4.8 1.91 ± 0.35 z 14.5 ± 5.4 y 10.8 ± 3.6 y
Semipurified diet 1.77 ± 0.48 z 20.5 ± 2.9 z 15.4 ± 2.7 z
 MENHADEN VERSUS SOYBEAN OIL IN SEA URCHIN DIETS
determination of dietary nutrient requirements required survival
>90% and growth rates that are no less than 80% of the growth
rates for animals living in the wild under comparable abiotic
conditions. In the present study, the semipurified diet and all
purified diets containing FO led to survival of 100%. The mean
growth rates for individuals fed the 1.2% FO–1.2% SO purified
diet was 65% of the growth rate observed for individuals fed the
semipurified diet. Beddingfield and McClintock (1998) reported
a 5–18% increase in test diameter for small L. variegatus held
in the laboratory for 28 weeks and fed agar pellets comprised
of natural seagrasses from the wild. Test diameter increases for
individuals in the present study exceeded these values across all
treatments. The survival and growth rates reported in the current
study suggest that the purified diet is adequate for determining
dietary nutrient requirements and can be used to vary a single
nutrient while maintaining the levels of all other nutrients.
Under the conditions of this study, the weight gain of indi-
Downloaded by [Department Of Fisheries] at 19:56 08 April 2013
viduals fed the purified diets was not significantly influenced
by dietary lipid level or source. None of the purified diets were
lipid-free because 3.3% phospholipid (soybean lecithin) was
included in all purified diets and 3% soybean lecithin was in-
cluded in the semipurified diet. When evaluating purified diets
in Common Carp Cyprinus carpio larvae, Radünz-Neto et al.
(1994) found that the inclusion of phospholipids in the diet was
more important for early survival and growth than the n-3 fatty
acids provided by cod liver oil. Conklin et al. (1980) reported
that soybean lecithin inclusion in purified diets was critical for
the survival of juvenile American lobster Homarus americanus
and hybrid American lobster × European lobster H. gammarus.
When fed a diet containing 5% corn + linseed + menhaden
oils (1:1:1) with 3% soybean lecithin, small adult green sea
urchins Strongylocentrotus droebachiensis exhibited improved
growth and lipid deposition in gonadal tissues (González-Durán
et al. 2008). Juvenile Lytechinus variegatus fed semipurified di-
ets containing dietary levels of soybean lecithin as low as 1%
(as fed) stored phospholipids from the diet as neutral lipid in the
gut and gonad (Gibbs et al. 2009). These lipid stores in the gut
and gonad may be used as an energy source during periods of
starvation (Giese 1966; Lawrence et al. 1966). The dietary phos-
pholipid level in purified diets should be reduced or eliminated
from some treatments to better observe any potential effect of
neutral lipid sources and levels.
In general, gut size was isometric with body size, indicating
larger urchins (those fed semipurified diets) had larger guts.
For individuals fed the purified diets, higher total lipids in the
diet resulted in higher gut weights. Similar effects of increased
dietary lipid level on gut size have been reported for juvenile
and small L. variegatus fed semipurified diets (Gibbs et al. 2009;
Gibbs 2011). The sea urchin gut can provide short-term nutrient
storage, and lipids are the predominant proximate nutrient stored
in the gut and used during periods of starvation (Lawrence et al.
1966; Klinger et al. 1988; Bishop and Watts 1992).
Ovary weight was consistently higher than testis weight
among all but one of the purified diets. The sex-specific dif-
283
ferences in gonad weight in the present study were similar to
those reported by Gibbs (2011) for small L. variegatus. In the
present study, the dietary lipid level in purified diets did not
significantly influence ovary or testis size for small adult L.
variegatus. However, in small L. variegatus developing virginal
gonad tissue, ovary and testis weights were reduced by a higher
level of SO in semipurified diets (Gibbs 2011). Reduced dry
ovary and testis indices for individuals fed purified diets con-
taining low dietary lipids or high levels from SO suggest that
gonad production is reduced when dietary lipids are limiting or
in excess.
While dietary lipid level and source in purified diets did
not significantly affect weight gain, the purified diet containing
1.2% FO and 1.2% SO supported better qualitative growth (i.e.,
no apparent deficiencies in survival, growth, or gonad weight)
than the other purified diets in this study. A balance of n-3 and n-
6 fatty acids from dietary oils may be needed to support growth
in sea urchins. Growth rates and feed efficiencies were improved
for prawns fed purified diets containing 6% lipid with 3:1 or 1:1
mixtures of pollack liver oil and soy oil (Deshimaru et al. 1979).
A mixture of C18 PUFAs and LC-PUFAs may provide a balance
of lipids for energy and the EFAs needed for metabolism.
In conclusion, the purified diet developed for this study sup-
ports growth and survival in L. variegatus. However, growth
performance was greater for individuals fed the semipurified
diet. The nutrient profiles of the semipurified and purified di-
ets are very similar; however, the degree of digestibility and
bioavailability of specific nutrients from practical and purified
ingredients may be very different. The digestibility of specific
diet ingredients remains to be determined in sea urchins, and
the results from such studies would lead to stepwise improve-
ments in the purified formulation. Despite these caveats, the use
of purified diets (or other standardized formulations) should be
considered for sea urchin nutrition research.
ACKNOWLEDGMENTS
We would like to thank Jeff Barry, Anthony Siccardi, Patty
Waits Beasley, Frank Castille, Woody Lawson, and Scott Walker
at the Texas A&M Shrimp Mariculture Research Lab for their
technical support. Thanks to Adele Cunningham, Samiksha
Raut, Rebecca Jones, Anna Morris Taylor, and Randy Watts
of the University of Alabama at Birmingham for their as-
sistance and support. Thanks also to Robert Makowsky for
statistical consultation. This research was funded in part by
a National Science Foundation GAANN fellowship and by
Texas Sea Grant. This publication was supported by the Na-
tional Sea Grant College Program of the U.S. Department of
Commerce’s National Oceanic and Atmospheric Administra-
tion under grants NA16RG2258 and NA06OAR4170078, the
Mississippi–Alabama Sea Grant Consortium projects R/SP-9
and R/SP-15, respectively, and the University of Alabama at
Birmingham. The views expressed herein do not necessarily
reflect those of any of those organizations.
 284 GIBBS ET AL.
REFERENCES
Akiyama, T., T. Unuma, and T. Yamamoto. 2001. Optimum protein level in a
purified diet for young red sea urchin Pseudocentrotus depressus. Fisheries
Science 67(2):361–363.
Andrews, N. L., Y. Agatsuma, E. Ballesteros, A. G. Bazhin, E. P. Creaser, D.
K. A. Barnes, L. W. Botsford, A. Bradbury, A. Campbell, J. D. Dixon, S.
Einarsson, P. K. Gerring, K. Hebert, M. Hunter, S. B. Hur, C. R. Johnson,
M. A. Juinio-Menez, P. Kalviss, R. J. Miller, C. A. Moreno, J. S. Palleiro, D.
Rivas, S. M. L. Robinson, S. C. Schroeter, R. S. Steneck, R. L. Vadas, D. A.
Woodby, and Z. Xiaoqi. 2002. Status and management of world sea urchin
fisheries. Oceanography and Marine Biology Annual Review 40:343–425.
Basuyaux, O., and M. Mathieu. 1999. Inorganic nitrogen and its effect on growth
of the abalone Haliotis tuberculata Linnaeus and the sea urchin Paracentrotus
lividus Lamarck. Aquaculture 174(1–2):95–107.
Beddingfield, S. D., and J. B. McClintock. 1998. Differential survivorship, re-
production, growth, and nutrient allocation in the regular echinoid Lytechinus
variegatus (Lamarck) fed natural diets. Journal of Experimental Marine Bi-
ology and Ecology 226:195–215.
Bell, M., J. Dick, and M. Kelly. 2001. Biosynthesis of eicosapentaenoic acid in
the sea urchin Psammechinus miliaris. Lipids 36(1):79–82.
Downloaded by [Department Of Fisheries] at 19:56 08 April 2013
Bishop, C. D., and S. A. Watts. 1992. Biochemical and morphometric study of
growth in the stomach and intestine of the echinoid Lytechinus variegatus
(Echinodermata). Marine Biology 114:459–467.
Brown, P. B., and S. D. Hart. 2011. Soybean oil and other n-6 polyunsaturated
fatty acid–rich vegetable oils. Pages 133–160 in G. M. Turchini, W.-K. Ng,
and D. R. Tocher, editors. Fish oil replacement and alternative lipid sources
in aquaculture feeds. CRC Press, Boca Raton, Florida.
Castell, J. D., E. J. Kennedy, S. M. C. Robinson, G. J. Parsons, T. J. Blair,
and E. Gonzalez-Duran. 2004. Effect of dietary lipids on fatty acid com-
position and metabolism in juvenile green sea urchins (Strongylocentrotus
droebachiensis). Aquaculture 242:417–435.
Conklin, D. E., L. R. D’Abramo, C. E. Bordner, and N. A. Baum. 1980. A
successful purified diet for the culture of juvenile lobsters: the effect of
lecithin. Aquaculture 21(3):243–249.
D’Abramo, L. R. 1997. Triacylglycerols and fatty acids. Pages 71–84 in D. E.
Conklin, L. R. D’Abramo, and D. M. Akiyama, editors. Crustacean nutrition:
advances in world aquaculture, volume 6. World Aquaculture Society, Baton
Rouge, Louisiana.
D’Abramo, L. R., and J. D. Castell. 1997. Research methodology. Pages 3–25 in
D. E. Conklin, L. R. D’Abramo, and D. M. Akiyama, editors. Crustacean nu-
trition: advances in world aquaculture, volume 6. World Aquaculture Society,
Baton Rouge, Louisiana.
De Silva, S. S., D. S. Francis, and A. G. J. Tacon. 2011. Fish oils in aquaculture in
retrospect. Pages 1–20 in G. M. Turchini, W.-K. Ng, and D. R. Tocher, editors.
Fish oil replacement and alternative lipid sources in aquaculture feeds. CRC
Press, Boca Raton, Florida.
Deshimaru, O., K. Kuroki, and Y. Yone. 1979. Studies on a purified diet for
prawn, XV. The composition and level of dietary lipid appropriate for growth
of prawn. Bulletin of the Japanese Society of Scientific Fisheries 45:591–594.
EANCAIN (Experimental Animal Nutrition Committee of the American In-
stitute of Nutrition). 1987. Guidelines for describing diets for experimental
animals. Journal of Nutrition 117(1):16–17.
FAO (Food and Agriculture Organization of the United Nations). 2012. The
state of world aquaculture 2012. FAO, Rome.
Gibbs, V. K. 2011. An evaluation of dietary lipids on growth performance in
the sea urchin, Lytechinus variegatus (Echinodermata: Echinoidea). Doctoral
dissertation. University of Alabama at Birmingham, Birmingham.
Gibbs, V. K., M. L. Powell, H. S. Hammer, W. T. Jones, S. A. Watts, A. L.
Lawrence, and J. M. Lawrence. 2010. Effects of dietary phospholipids and
cholesterol on growth and organ production in the sea urchin, Lytechinus
variegatus. Pages 369–375 in L. G. Harris, S. A. Boettger, C. W. Walker, and
M. P. Lesser, editors. Echinoderms: Durham. CRC/Balkema, Leiden, The
Netherlands.
Gibbs, V. K., S. A. Watts, A. L. Lawrence, and J. M. Lawrence. 2009. Dietary
phospholipids affect growth and production of juvenile sea urchin Lytechinus
variegatus. Aquaculture 292(1–2):95–103.
Giese, A. C. 1966. Lipids in the economy of marine invertebrates. Physiology
Review 46(2):244–298.
González-Durán, E., J. D. Castell, S. M. C. Robinson, and T. J. Blair. 2008.
Effects of dietary lipids on the fatty acid composition and lipid metabolism of
the green sea urchin Strongylocentrotus droebachiensis. Aquaculture 276(1–
4):120–129.
Hammer, H. S., M. L. Powell, W. T. Jones, V. K. Gibbs, A. L. Lawrence, J. M.
Lawrence, and S. A. Watts. 2012. Effect of feed protein and carbohydrate lev-
els on feed intake, growth, and gonad production of the sea urchin, Lytechinus
variegatus. Journal of the World Aquaculture Society 43(2):145–158.
Hammer, H. S., M. L. Powell, V. K. Gibbs, W. T. Jones, S. A. Watts, A. L.
Lawrence, J. M. Lawrence, and L. R. D’Abramo. 2010. Effect of dietary
menhaden oil and soy oil on consumption, somatic growth and gonad pro-
duction of the sea urchin, Lytechinus variegatus. Pages 377–383 in L. G.
Harris, S. A. Boettger, C. W. Walker, and M. P. Lesser, editors. Echinoderms:
Durham. CRC Press/Balkema, Leiden, The Netherlands.
Jones, W. T., M. L. Powell, V. K. Gibbs, H. S. Hammer, J. Fox, A. L. Lawrence,
and S. A. Watts. 2010. The effect of dietary selenium on weight gain and
gonad production of the sea urchin Lytechinus variegatus. Journal of the
World Aquaculture Society 41(5):675–686.
Keesing, J. K., and K. C. Hall. 1998. Review of harvests and status of world sea
urchin fisheries points to opportunities for aquaculture. Journal of Shellfish
Research. 17:1597–1604.
Klinger, T. S., S. A. Watts, and D. Forcucci. 1988. Effect of short-term feeding
and starvation on storage and synthetic capacities of gut tissues of Lytechinus
variegatus (Lamarck) (Echinodermata: Echinoidea). Journal of Experimental
Marine Biology and Ecology 117:187–195.
Lawrence, J. M., A. L. Lawrence, and A. C. Giese. 1966. Role of the gut as a
nutrient-storage organ in the purple sea urchin (Strongylocentrotus purpura-
tus). Physiological Zoology 39(4):281–290.
Lesser, M. P., and C. W. Walker. 1998. Introduction to the special sec-
tion on sea urchin aquaculture. Journal of Shellfish Research. 17:1505–
1506.
Liu, H., M. S. Kelly, E. J. Cook, K. Black, H. Orr, J. X. Zhu, and S. L. Dong.
2007a. The effect of diet type on growth and fatty acid composition of sea
urchin larvae, I. Paracentrotus lividus (Lamarck, 1816) (Echinodermata).
Aquaculture 264(1–4):247–262.
Liu, H., M. S. Kelly, E. J. Cook, K. Black, H. Orr, J. X. Zhu, and S. L. Dong.
2007b. The effect of diet type on growth and fatty acid composition of the
sea urchin larvae, II. Psammechinus miliaris (Gmelin). Aquaculture 264(1–
4):263–278.
NRC (National Research Council). 2011. Lipids. Pages 102–134 in Nutrient
requirements of fish and shrimp. National Academy Press, Washington, D.C.
Radünz-Neto, J., G. Corraze, C. Nicole, and P. Bergot. 1994. Lipid supplemen-
tation of casein-based purified diets for carp (Cyprinus carpio L.) larvae.
Aquaculture 128(1–2):153–161.
Robinson, S. M. 2004. The evolving role of aquaculture in the global pro-
duction of sea urchins. Pages 343–357 in J. M. Lawrence and O. Guzman,
editors. Sea urchins: fisheries and ecology. DEStech Publications, Lancaster,
Pennsylvania.
Watts, S. A., J. M. Lawrence, and A. L. Lawrence. 2010. Approaches to the study
of sea urchin nutrition. Pages 331–346 in L. G. Harris, S. A. Boettger, C. W.
Walker, and M. P. Lesser, editors. 12th International Echinoderm Conference.
CRC Press/Balkema, Durham, New Hampshire.
This article was downloaded by: [Department Of Fisheries]
On: 08 April 2013, At: 19:56
Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,
37-41 Mortimer Street, London W1T 3JH, UK

North American Journal of Aquaculture

Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/unaj20

Effects of Increasing Docosahexaenoic Acid (DHA) and

Arachidonic Acid (ARA) in Brood Diets of Monodactylus
sebae on Fecundity, Egg and Larval Quality, and Egg
Fatty Acid Composition
a b a b f b a b c
C. L. Ohs , M. A. DiMaggio , S. W. Grabe , J. S. Broach , C. A. Watson , N. E.
d e
Breen & F. T. Barrows
a
Program in Fisheries and Aquatic Sciences, School of Forest Resources and Conservation,
Institute of Food and Agricultural Sciences, University of Florida, 7922 Northwest 71st
Street, Gainesville, Florida, 32653, USA
b
Indian River Research and Education Center, Institute of Food and Agricultural Sciences,
University of Florida, 2199 South Rock Road, Fort Pierce, Florida, 34945, USA
c
Tropical Aquaculture Laboratory, University of Florida, 1408 24th Street Southeast, Ruskin,
Florida, 33570, USA
d
Department of Chemistry, Roger Williams University, One Old Ferry Road, Bristol, Rhode
Island, 02809, USA
e
U.S. Department of Agriculture, Fish Technology Center, Agricultural Research Service,
4050 Bridger Canyon Road, Bozeman, Montana, 59715, USA
f
Department of Biological Sciences, University of New Hampshire, 175 Spaulding Hall,
Durham, New Hampshire, 03824, USA
Version of record first published: 02 Apr 2013.

To cite this article: C. L. Ohs , M. A. DiMaggio , S. W. Grabe , J. S. Broach , C. A. Watson , N. E. Breen & F. T. Barrows (2013):
Effects of Increasing Docosahexaenoic Acid (DHA) and Arachidonic Acid (ARA) in Brood Diets of Monodactylus sebae on
Fecundity, Egg and Larval Quality, and Egg Fatty Acid Composition, North American Journal of Aquaculture, 75:2, 285-294

To link to this article: http://dx.doi.org/10.1080/15222055.2012.746248

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions

This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to
anyone is expressly forbidden.

The publisher does not give any warranty express or implied or make any representation that the contents
will be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses should
be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims,
proceedings, demand, or costs or damages whatsoever or howsoever caused arising directly or indirectly in
connection with or arising out of the use of this material.
 North American Journal of Aquaculture 75:285–294, 2013

C American Fisheries Society 2013

ISSN: 1522-2055 print / 1548-8454 online

DOI: 10.1080/15222055.2012.746248

SPECIAL SECTION: LIPIDS IN AQUACULTURE NUTRITION AND PHYSIOLOGY II

Effects of Increasing Docosahexaenoic Acid (DHA) and

Arachidonic Acid (ARA) in Brood Diets of Monodactylus
sebae on Fecundity, Egg and Larval Quality, and Egg Fatty
Acid Composition

C. L. Ohs* and M. A. DiMaggio1

Program in Fisheries and Aquatic Sciences, School of Forest Resources and Conservation,
Institute of Food and Agricultural Sciences, University of Florida, 7922 Northwest 71st Street,
Gainesville, Florida 32653, USA; and Indian River Research and Education Center,
Downloaded by [Department Of Fisheries] at 19:56 08 April 2013

Institute of Food and Agricultural Sciences, University of Florida, 2199 South Rock Road, Fort Pierce,
Florida 34945, USA

S. W. Grabe
Indian River Research and Education Center, Institute of Food and Agricultural Sciences,
University of Florida, 2199 South Rock Road, Fort Pierce, Florida 34945, USA

J. S. Broach
Program in Fisheries and Aquatic Sciences, School of Forest Resources and Conservation,
Institute of Food and Agricultural Sciences, University of Florida, 7922 Northwest 71st Street,
Gainesville, Florida 32653, USA; and Indian River Research and Education Center,
Institute of Food and Agricultural Sciences, University of Florida, 2199 South Rock Road, Fort Pierce,
Florida 34945, USA

C. A. Watson
Tropical Aquaculture Laboratory, University of Florida, 1408 24th Street Southeast, Ruskin,
Florida 33570, USA

N. E. Breen
Department of Chemistry, Roger Williams University, One Old Ferry Road, Bristol,
Rhode Island 02809, USA

F. T. Barrows
U.S. Department of Agriculture, Fish Technology Center, Agricultural Research Service,
4050 Bridger Canyon Road, Bozeman, Montana 59715, USA

Abstract
The Guinean Fingerfish Monodactylus sebae is a popular euryhaline ornamental fish species with limited aquacul-
ture production. One of the bottlenecks to commercial production is lack of knowledge of the nutritional requirements
for broodstock. Therefore, three broodfish diets were formulated and fed to Guinean Fingerfish broodstock to de-
termine their quantitative and qualitative effects on egg production and egg and larval morphology. The dietary

*Corresponding author: cohs@ufl.edu

1
Present address: Department of Biological Sciences, University of New Hampshire, 175 Spaulding Hall, Durham, New Hampshire 03824,
USA.
Received April 23, 2012; accepted October 27, 2012
Published online April 2, 2013
285
 286 OHS ET AL.

treatments consisted of a control, a diet with increased docosahexaenoic acid (DHA), and a diet with increased DHA
and arachidonic acid (DHA + ARA). Broodfish fed the DHA + ARA diet spawned more frequently than broodfish
fed the DHA diet and those fed the control diet. The greatest egg production was also observed from broodfish fed
the DHA + ARA diet. The mean hatching success of floating eggs was not significantly different among diets. The
mean egg and oil globule diameters for both floating and sinking eggs were significantly smaller for broodfish fed the
DHA + ARA diet rather than the other diets. At both 24 and 48 h, survival was significantly greater for fish fed the
control diet than for those fed the DHA and DHA + ARA diets. At both 24 and 48 h posthatch, notochord length
was significantly shorter for larvae from broodfish fed the DHA + ARA diet than for larvae from fish fed the control
diet. The fatty acid profiles of the eggs closely resembled the fatty acid profiles of the diets with respect to DHA and
ARA levels. Guinean Fingerfish females appear to have the ability to regulate the levels of DHA and ARA assimilated
into developing eggs, although there is not clear evidence that they can elongate and desaturate C18 fatty acids into
DHA and ARA.

Ornamental fish production, like all other segments of aqua- tein, 15% lipid), and although occasional spawns were observed,
culture, relies heavily on the successful spawning, hatching, suboptimal broodfish nutrition resulting in poor egg production
and survival of larvae. Little is known about the nutrient re- and egg and larval quality was suspected.
Downloaded by [Department Of Fisheries] at 19:56 08 April 2013

quirements of most ornamental fish species. There are no com-
mercially available pelleted diets that are specifically designed
to feed broodstock ornamental species. Currently, the artificial
diets available to commercial ornamental fish producers are pri-
marily based on nutritional trials conducted on one or more of
the major food fish species and do not always match ornamental
producer needs for particle size, buoyancy, moisture content,
palatability, and shelf life.
The Guinean Fingerfish Monodactylus sebae is a euryhaline,
catadromous fish species that is usually captured wild from
coastal rivers of the western coast of Africa and marketed as
a freshwater ornamental species. In recent years, their popular-
ity and relatively high market value in the ornamental industry
has prompted interest in the development of commercial culture
methods. Successful spawning was first reported by Akatsu et al.
(1977), who described the spawning behavior, egg developmen-
tal stages, larval developmental stages, and salinity tolerance of
the larvae. Larvae were successfully cultured by feeding rotifers
Brachionus plicatilis for 19 d, then Artemia nauplii for 15 d,
with weaning onto tubifex worms by 33 d posthatch (dph)
(Akatsu et al. 1977). Ogasawara et al. (1978) described juvenile
development and reported that juveniles were euryhaline and
could tolerate transfer to freshwater only after 30 dph.
Guinean Fingerfish are broadcast spawners and produce
buoyant eggs in seawater that can be collected from surface-
skimming egg collectors (Akatsu et al. 1977). Spawning of
Guinean Fingerfish by Akatsu et al. (1977) occurred at 5–
10-d intervals for up to 70 d, indicating these fish are multi-
batch, asynchronous spawners. Spawning occurred at 25◦ C and
a salinity of 25% of full-strength seawater (Cl, 4.59–4.64 g/L).
More recently, the University of Florida Tropical Aquaculture
Laboratory spawned Guinean Fingerfish at 25◦ C and salinities
greater than 25 g/L, although high hatching percent and lar-
val survival was not consistently observed in repeated spawning
(Craig Watson, University of Florida, personal communication).
Broodstock were fed a commercially available (Zeigler Bros.
Inc., Gardners, Pennsylvania) “salmon starter” diet (50% pro-
Ornamental fish producers in Florida have identified brood-
stock nutrition to be a major bottleneck in commercialization.
Little is known about the nutritional requirements of ornamental
fish species. It is assumed that it is important to include essential
fatty acids in their diets, as it is with other species (Izquierdo
et al. 2001). Thus, current industry standards rely heavily on
feeding a wide variety of foods, often resulting in increased
production costs. Frozen and live feeds including blood worms
(chironomid larvae), shrimp, squid, tubifex worms, and black
worms Lumbriculus variegatus are commonly fed to ornamental
fishes. These are expensive, can vary in quality, and, if cultured,
require additional space and resources. The other option used in
the freshwater ornamental fish aquaculture industry is to condi-
tion broodstock in outdoor ponds, where they can feed on the
natural productivity to supplement their dietary provision. This
increases the space and cost required for production. There is
a strong need for a commercially available diet formulated to
contain adequate levels of essential fatty acids. Development of
such a diet will reduce required nutritional supplementation and
thus served as the impetus for this research.
The provision of lipids to developing oocytes and their accu-
mulation in the yolk are essential to the successful reproduction
and development of embryos (Brooks et al. 1997). The impor-
tance of long-chain polyunsaturated fatty acids (LC-PUFAs, i.e.,
fatty acids with 20 or more carbon atoms and 3 or more dou-
ble bonds), including the n-3 LC-PUFAs docosahexaenoic acid
(DHA; 22:6[n-3]2) and eicosapentaenoic acid (EPA; 20:5[n-3])
and the n-6 LC-PUFA arachidonic acid (ARA; 20:4[n-6]) in
broodstock nutrition have been studied extensively with multiple
species of fish. In general, low n-3 LC-PUFA levels in brood-
stock diets decrease egg and larval quality (Watanabe 1982;
Harel et al. 1994; Li et al. 2005). Conversely, studies have shown
2In fatty acid designations of this nature, the number to the left of the colon
is the number of carbon atoms in the compound, the number immediately to
the right of the colon is the number of double bonds, and the number after the
hyphen indicates the position of the first double bond from the methyl end.
 EFFECTS OF DHA AND ARA ON BROODFISH DIETS
that an excess of n-3 LC-PUFAs can have detrimental effects
on egg quality (Furuita et al. 2002, 2003; Li et al. 2005). Both
ARA and EPA serve as precursors for the eicosanoid synthe-
sis associated with final oocyte maturation and have repeatedly
been shown to affect egg quality (Mustafa and Srivastava 1989;
Bell et al. 1997; Tocher 2003; Lane and Kohler 2006; Ling et al.
2006).
Many freshwater fish possess the enzymatic pathways to
elongate and desaturate shorter-chain fatty acids (El-Sayed et al.
2005; Kumaran et al. 2007; Sink and Lochmann 2008; Durland
et al. 2010; Quintero et al. 2011; Wing-Keong and Wang 2011),
such as linolenic (18:3[n-3]) and linoleic (18:2[n-6]) acid, into
n-3 and n-6 LC-PUFAs, respectively. Most marine fish are pre-
sumed to have a low capability or inability to bioconvert C18
fatty acids into LC-PUFAs. Therefore, these fatty acids must be
included in the diet of marine fish (Tocher 2003).
The LC-PUFA composition of eggs and early larval stages
Downloaded by [Department Of Fisheries] at 19:56 08 April 2013
can be highly influenced by LC-PUFA levels and sources in
broodstock diets. This is evident with many marine species
including Red Seabream (also known as Madai) Pagrus ma-
jor, Yellowtail (also known as Buri) Seriola quinqueradiata,
Japanese Flounder (also known as Olive Flounder) Paralichthys
olivaceus, Striped Jack (also known as White Trevally) Pseudo-
caranx dentex (Watanabe and Vassallo-Agius 2003), and Yel-
lowfin Seabream Acanthopagrus latus (Zakeri et al. 2011).
This has also been shown with freshwater species, including
Channel Catfish Ictalurus punctatus (Durland et al. 2010) and
Nile Tilapia Oreochromis niloticus (Wing-Keong and Wang
2011).
Zakeri et al. (2011) fed Yellowfin Seabream broodstock three
diets varying in their DHA, EPA, and ARA content by using fish
oil, sunflower oil, and a 1:1 fish oil : sunflower oil combination
as the fatty acid sources. The highest levels of DHA, EPA, and
ARA in eggs from these broodstock were from those of fish fed
the diet containing only fish oil, which had the highest levels of
these three fatty acids (the levels of the three fatty acids were
positively correlated with the levels in the different diets). The
DHA levels were actually almost 10% higher in the eggs than in
the diet, while the EPA and ARA levels were 2.6% and 12.3%
lower, respectively, in the eggs. Positive correlations between
egg and dietary LC-PUFA levels have also been reported in
other species (Zakeri et al. 2011). Nile Tilapia were fed four
diets varying in DHA, EPA, and ARA content by using fish oil,
crude palm oil, linseed oil, and a 1:1 fish oil : crude palm oil
combination as the fatty acid sources (Wing-Keong and Wang
2011). Broodstock fed the diet containing only fish oil (which
had the highest levels of DHA, EPA, and ARA) produced eggs
with the highest DHA levels (5% above the level in the diet).
However, the highest EPA and ARA levels in eggs were observed
among broodfish fed the diets containing linseed oil and crude
palm oil, respectively, and the levels in the eggs were elevated
compared with their respective diets. The authors attributed this
observation to the Nile Tilapia’s ability to maintain specific fatty
acid levels in the gonads and eggs and to generate LC-PUFAs
287
from the shorter-chain fatty acid precursors palmitic acid (16:0)
and linolenic acid.
Selective pressure to maintain specific LC-PUFA levels in
eggs and other tissues despite their levels and precursor levels
in diets has been reported in other species (Izquierdo et al. 2001;
Tocher 2003); however, LC-PUFA deficiencies and exhaustion
within eggs can occur during the spawning season. Significant
declines in the DHA levels of eggs from wild Common Snook
Centropomus undecimalis have been observed throughout their
5-month spawning season (Yanes-Roca et al. 2009). Brood-
fish diets must be able to consistently supply the required LC-
PUFAs, especially to multibatch, asynchronous spawners who
have limited time to elongate and desaturate shorter-chain fatty
acids and incorporate them into their eggs.
Little research has been conducted on the nutritional require-
ments of ornamental broodstock fishes, and no data are currently
available on the nutritional requirements of Guinean Fingerfish.
Therefore, a suite of quantitative and qualitative measurements
were collected to elucidate the dietary effects on various cul-
ture aspects essential to Guinean Fingerfish production. The re-
sults of this experiment should help producers formulate brood-
stock diets for more efficient captive reproduction of Guinean
Fingerfish.
METHODS
Culture protocol.—Brood Guinean Fingerfish were stocked
into three independent systems at a density of 30 fish per
tank. The sex ratio was assumed to be 1:1 because there is
no known sexual dimorphism in this species and intraovarian
catheterization could have resulted in ovarian regression or in-
jury in this fragile species. Each system consisted of a 2,000-
L blue polyethylene rectangular tank equipped with a trickle
filter, heaters, supplemental aeration, and an ultraviolet ster-
ilizer. Illumination was provided by diffuse room lights with
a cycle of 12 h light : 12 h dark, and the temperature was
maintained at >20◦ C with submersible heaters. The appropri-
ate salinity was attained by mixing sterilized natural seawater
(35 g/L) with sterilized well water (2 g/L). Dissolved oxygen
(DO), pH, temperature, and salinity were recorded daily (ap-
proximately 0900 hours). Temperature and DO were measured
using a YSI 550A meter, and salinity was determined using a
YSI 30 salinity–conductivity meter (YSI, Inc., Yellow Springs,
Ohio). The pH was measured using a Hach sensION1 portable
pH meter (Hach Co., Loveland, Colorado). Total ammonia ni-
trogen, nitrite, and nitrate were evaluated weekly using standard
methods and a Hach DR 4800 spectrophotometer.
Guinean Fingerfish broodfish were stocked at an age of 1
year and were cultured in these conditions for one additional
year. To induce spawning, the salinity of the broodfish tanks
was increased by 5 g/L every 3 days until a salinity of 25 g/L
was attained. When that salinity and a temperature of 24◦ C
was attained, natural volitional spawning initiated. Eggs were
collected in a floating air lift egg collector constructed with
 288 OHS ET AL.

TABLE 1. Formulations and proximate composition of experimental diets average temperature of 83◦ C. The pressure at the die head was
with varying levels of DHA and ARA. All values are presented as percentages approximately 3,103 kPa. The pellets were then dried in a pulse
of the total.
bed drier (Buhler AG) for ∼18 min at 90◦ C and cooled with
Ingredient Control DHA DHA + ARA forced air at ambient air temperatures to reach final moisture
levels of <10%. Diets were stored in plastic-lined paper bags
Formulation at room temperature until shipped to Florida, where they were
Menhaden meala 61.08 61.08 61.08 stored in lidded buckets inside a dark, air-conditioned room.
Wheat flourb 20.3 14.91 13.9 Fatty acid analysis of the diets and eggs.—The fatty acid
AlgaMac 3050c 0 8.7 8.7 composition of the experimental diets and floating eggs was
AlgaMac ARAd 0 0 4.22 determined following the method of Drillet et al. (2006)
Lecithin 4.9 4.9 4.9 by extraction of the lipids using a mixture of chloroform
Corn protein concentratee 4.62 4.62 4.62 and methanol (2:1) followed by a one-step acetyl chloride–
Fish oil, menhadenf 3.0 3.0 3.0 catalyzed transesterification of the lipids in methanol. The fatty
Spirulinag 3.0 3.0 3.0 acid methyl esters (FAMEs) were subsequently analyzed us-
Vitamin premixh 2.0 2.0 2.0 ing gas chromatography–mass spectrometry (GC–MS). A 2:1
Choline CL 0.6 0.6 0.6 chloroform–methanol mixture (10 mL) was transferred to a cul-
Stay-C 0.3 0.3 0.3 ture tube with 40–50 mg of lyophilized sample. The mixture
Downloaded by [Department Of Fisheries] at 19:56 08 April 2013

Trace mineral premixi 0.1 0.1 0.1 in the culture tube was sonicated at 0◦ C for 45 min, followed
Astaxanthin 0.1 0.1 0.1 by evaporation of the solvent at 65◦ C by applying a flow of
Proximate composition warm nitrogen. A solution (3 mL) of toluene, methanol, and
Crude protein 48.51 48.39 48.02 acetyl chloride (40:50:10) was added, followed by 20 µL of
Crude lipid 11.07 11.21 12.07 1.0-mg/mL internal standard (17:0) and the mixture was heated
Moisture 6.29 5.40 5.81 for 1 h at 65◦ C. Aqueous NaHCO3 (5% by weight, 1 mL) was
a
Omega Proteins, Inc., Menhaden Special Select; 628 g/kg protein. added to the culture tube, and the upper layer containing the fatty
b
Manildra Milling; 4 g/kg protein. acid methyl esters was removed. The water phase was extracted
c
Aquafauna Bio-Marine, Inc.; 176 g/kg protein, 562 g/kg lipid. twice more with heptane (1 mL), and the combined organic lay-
d
Aquafauna Bio-Marine, Inc.; 140 g/kg protein, 300 g/kg lipid.
e
Cargill, Empyreal 75; 761.0 g/kg protein. ers were collected and evaporated at 65◦ C by a stream of warm
f
Omega Proteins, Inc. nitrogen. The methyl esters were resuspended in chloroform
g
Earthrise Nutritionals; 665 g/kg protein.
h
Contributions (mg/kg of diet): zinc, 40; manganese, 13; iodine, 5; and copper, 9.
i
Contributions per kilogram of diet: vitamin A, 19,300 IU; vitamin D, 13,200 IU;
vitamin E, 264 IU; vitamin K3 , 2.2 mg: thiamin mononitrate, 18.2 mg; riboflavin, 19.2 mg;
pyridoxine hydrochloride, 27.4 mg; pantothenate DL-calcium, 93.0 mg; cyancobalamin, TABLE 2. Composition of the AlgaMac products added to diets, as reported
0.06 mg; nicotinic acid, 43.6 mg; biotin, 0.68 mg; folic acid, 5.0 mg; and inositol, 1,200 IU. by the manufacturer. The proximate and fatty acid compositions of the products
are presented as percentages of the total; NA = not available.

a surface-skimming pipe which drew water into a collection Component AlgaMac 3050 AlgaMac ARA
basket fitted with a net constructed of 500-µm nylon mesh.
Protein 17.6 14.0
Dietary treatments.—Three diets, a control and two experi-
Fat 56.2 30.0
mental diets, were fed to brood Guinean Fingerfish once daily
Carbohydrates 15.9 20.0
to apparent satiation. The three diets were formulated to have
Ash 8.2 20.0
different levels of DHA and ARA (Table 1). The analyzed nu-
Moisture 2.1 7.0
trient composition of each diet is also presented in Table 1. The
Fiber NA 1.0
control diet was based on the nutrient specifications and ingre-
Fatty acids
dient list developed by a commercial feed mill for ornamental
14:0 8.85 NA
producers, although it was never commercially produced. The
16:0 26.6 13.2
formulation of the DHA and DHA + ARA diets was altered
16:1 0.42 NA
by adding the commercially available algal additives Algamac
18:0 0.64 10.3
3050 Flake and Algamac-ARA (Table 2) to increase the DHA
18:1 0.11 10.8
and ARA contents, respectively.
18:2 NA 8.7
Diet preparation.—The control diet and both experimental
20:4 NA 37
diets were produced with commercial manufacturing methods
20:3(n-6) 0.22 NA
using a twin-screw cooking extruder (DNDL-44; Buhler AG,
20:5(n-3) 2.88 0
Uzwil, Switzerland) at the Fish Technology Center, Bozeman,
22:5(n-6) 17.04 0
Montana. Diet mash was exposed to an average of 114◦ C for 18 s
22:6(n-3) 43.27 2.4
in five barrel sections, and the last section was water cooled to an
 EFFECTS OF DHA AND ARA ON BROODFISH DIETS
(1.0 mL) and filtered, and an aliquot of the sample (1 µL) was
analyzed by GC–MS.
The GC–MS instrument was composed of an Agilent 6850
gas chromatograph equipped with an Agilent DB-23 column
(60 m long, 0.250-mm interior diameter with a 0.25-µm film
thickness) and an Agilent 5975B mass selective detector. The
inlet temperature was 250◦ C, and helium was used as the carrier
gas, with a 2.1-mL/min flow rate and splitless injection. The
oven temperature was initially set to 50◦ C for 1 min, then in-
creased to 160◦ C at a rate of 25◦ C/min, and then increased to
240◦ C at a rate of 2◦ C/min and held for 5 min. The concentration
of FAMEs in each sample was determined by using the quanti-
zation feature on the ChemStation (Agilent Technologies) and
comparing the peak areas of known FAME standards (FAME
mix C4 –C24 unsaturates; Supelco, Inc.) with those observed
from the samples. FAME peaks were identified by comparing
retention times and mass spectrographs with those of known
Downloaded by [Department Of Fisheries] at 19:56 08 April 2013
FAMEs and corroborating with the NIST MS Search 2.0 mass
spectral library.
Proximate analysis of the diets.—Feed samples were dried
and analyzed using AOAC (1995) methods for proximate com-
position, with the exception of protein and crude lipid. Dried
samples were finely ground by mortar and pestle and analyzed
for crude protein (total nitrogen × 6.25) using a LECO nitro-
gen determinator (TruSpec N; LECO Corporation, St. Joseph,
Michigan). Crude lipid was analyzed using a soxhlet extraction
apparatus (Soxtec System HT; Foss Tecator AB, Hoganas, Swe-
den) with methylene chloride as the extracting solvent and ash
by incineration at 550◦ C in a muffle furnace.
Spawning performance and egg and larval quality.—The
dietary effects on spawning performance were measured by
quantifying the number of spawns, egg quantity, fertilization
percent, hatch percent, egg morphology (egg diameter and oil
droplet diameter), larval survival at 24 and 48 h posthatch, larval
morphology (oil droplet diameter, yolk volume, and notochord
length) at hatch, and larval morphology (notochord length) at
24 and 48 h posthatch.
Following each spawn, both floating and sinking eggs were
removed from the egg collector and placed into an appropriate-
sized graduated cylinder and allowed to separate. Three 1-mL
samples of both the floating and sinking eggs were quantified
under a dissecting microscope, and this count was used to vol-
umetrically estimate the total quantity of floating and sinking
eggs. A spawn was defined as the collection of >1.0 mL of
floating eggs. Fertilization percent was calculated by observing
100 floating eggs and 100 sinking eggs under a dissecting mi-
croscope and recording the percentage undergoing successful
embryological development.
To determine the 24-h hatching percent, a sample of floating
eggs (n = 50) was stocked into each of three replicate 1-L screen
bottom containers (55 µm screen) and suspended in a shallow
aerated tank (453 L) to achieve a final water volume of 750 mL
in each hatching container. After 24 h, larvae and unhatched
eggs were quantified. To determine 24- and 48-h (first-feeding)
289
survival, three replicate 1-L containers were similarly stocked
with samples of floating eggs (n = 50) into the shallow aerated
453-L tank. At 24 and 48 h posthatch, larvae were transferred to
petri dishes for observation; larvae which moved were consid-
ered alive and were quantified. The shallow aerated 453-L tank
was heated to 25◦ C, the salinity was 24 g/L, and it was exposed
to diffuse room lighting on a 12 h light : 12 h dark cycle. Wa-
ter quality variables were measured using the same protocols
as with the broodfish tanks. Subsamples of the floating eggs
for each spawn not stocked in the containers were triple-rinsed
with de-ionized water and stored in a –80◦ C freezer for fatty
acid analysis.
Egg and oil globule diameters for both floating and sinking
eggs were obtained by placing a random egg sample on separate
sedgewick rafter cells with calibrated grids and photographed
using a dissecting microscope equipped with a digital camera.
Floating and sinking egg morphology (n = 100) was measured
from the catalogued photographs using SigmaScan Pro 5.0 soft-
ware (Systat Software, Inc., Point Richmond, California).
Larvae were photographed at hatch and 24 and 48 h (first
feeding) posthatch using the same method as used for the eggs.
At hatch the notochord length, yolk volume, and oil droplet
diameter were measured. Yolk volume was determined by mea-
suring the length and height and was calculated using the volume
of a prolate sphere (V = π/6LH2, where V is volume, L is length,
and H is height) according to Avila and Juario (1987). At 24 and
48 h posthatch, only notochord length was measured.
Statistical analysis.—All statistical analyses were performed
with SAS version 9.2 (SAS Institute, Inc., Cary, North Carolina).
The residuals were tested for normality prior to analysis, and
percent data were arcsine–square root transformed. All egg (fer-
tilization percent, hatch percent, morphometric, and fatty acid
level) and larval (survival and morphometric) variables mea-
sured were compared among diets using analysis of variance
(ANOVA) followed by Tukey’s honestly significant difference
means separation tests, where each spawn served as a replicate.
Regression models were fitted for both DHA and ARA levels of
floating eggs with diet, date, and a diet × date interaction term
as predictor variables. Statistical differences were considered
significant if P ≤ 0.05. All numerical data are reported as the
means ± standard deviations (SDs).
RESULTS
During the observed spawning period, dissolved oxygen
ranged from 5.4 to 7.1 mg/L; temperature ranged from 24.1◦ C
to 29.5◦ C; salinity ranged from 0 to 28.2 g/L during the 1-year
culture period and from 24 to 28.1 g/L during spawning; pH
ranged from 6.4 to 7.1 (sodium bicarbonate was added when
pH was found to be <6.8); total ammonia nitrogen ranged from
0 to 0.96 mg/L; nitrite ranged from 0.004 to 1.05 mg/L; an
nitrate ranged from 0 to 160 mg/L.
Broodfish fed the DHA + ARA diet spawned more fre-
quently (67 times, 46 with viable eggs) than broodfish fed the
 290 OHS ET AL.

TABLE 3. Fertilization rates for floating and sinking eggs of Guinean Fingerfish, hatch rates for floating eggs, and egg and oil diameters for floating and sinking
eggs. Values are means ± SDs; within columns, significant differences are indicated by different lowercase letters.

Floating egg Sinking egg Floating egg Floating egg Floating egg Sinking egg Sinking egg
fertilization fertilization hatch diameter oil diameter diameter oil diameter
Diet (%) (%) (%) (mm) (mm) (mm) (mm)
Control 96.5 ± 6.1 zy 83.9 ± 24.8 57.4 ± 34.8 0.644 ± 0.042 z 0.380 ± 0.029 z 0.639 ± 0.037 z 0.348 ± 0.025 z
DHA 93.6 ± 7.4 y 70.6 ± 36.5 55.2 ± 32.8 0.640 ± 0.038 y 0.360 ± 0.029 y 0.633 ± 0.049 y 0.347 ± 0.032 z
DHA + ARA 97.2 ± 3.7 z 88.4 ± 16.7 47.5 ± 35.4 0.637 ± 0.038 x 0.351 ± 0.026 x 0.629 ± 0.037 x 0.339 ± 0.026 y
Number of eggs 11,281 11,281 8,942 8,942
measured
P-value 0.0237 0.0512 0.0747 <0.0001 <0.0001 <0.0001 <0.0001

DHA diet (33 times, 30 with viable eggs) and those fed the from broodfish fed the DHA + ARA diet than for those fed the
control diet (49 times, 39 with viable eggs). The greatest egg control diet (Table 4).
production was observed among broodfish fed the DHA + The fatty acid profiles of the diets and floating eggs spawned
Downloaded by [Department Of Fisheries] at 19:56 08 April 2013

ARA diet, with 1,260,255 eggs being spawned, while fish fed
the DHA diet spawned 521,211 eggs and those fed the control
diet spawned 758,282 eggs. Standardizing egg production per
spawn with the number of females in each tank and assum-
ing equal contributions by all females, those fed the DHA +
ARA diet produced 2,813 eggs per spawn per female, those fed
the DHA diet produced 1,959, and those fed the control diet
produced 2,160.
The mean percent fertilization of floating eggs was signifi-
cantly different among experimental diets (Table 3). There was
no significant difference for the mean percent fertilization of
sinking eggs among experimental diets (Table 3). The mean
hatching success of floating eggs was not significantly different
among diets (Table 3). The mean egg and oil globule diameters
for both floating and sinking eggs were significantly smaller for
broodfish fed the DHA + ARA diet than for those fed the other
diets (Table 3).
At hatch, mean notochord length and oil droplet diameter
were significantly smaller for larvae from broodfish fed the
DHA + ARA diet, although the yolk volume was not signif-
icantly different among diets (Table 4). At both 24 and 48 h,
survival was significantly greater for the control diet than the
DHA and DHA + ARA diets (Table 4). At both 24 and 48 h
posthatch, notochord length was significantly smaller for larvae
from fish fed each diet are presented in Tables 5 and 6, respec-
tively. The fatty acid profiles of the eggs closely resembled the
fatty acid profiles of the diets with respect to DHA and ARA lev-
els. Many significant differences were observed in the fatty acid
levels of floating eggs for multiple fatty acids and their ratios,
including DHA and ARA. DHA levels were elevated ≥10.0%
in both the diet and eggs for the DHA and DHA + ARA
treatments compared with the control treatment. DHA levels of
floating eggs were significantly different (P < 0.0001) among
treatments, with the highest levels being observed in eggs from
broodfish fed the DHA + ARA diet (27.6%) and the lowest
in eggs from the broodfish fed the control diet (14.52%). ARA
levels were elevated ≥4.0% in both the diet and eggs for the
DHA + ARA treatment compared with the control and DHA
treatments. The ARA levels of floating eggs were significantly
different (P < 0.0001) among treatments, with the highest levels
being observed in eggs from broodfish fed the DHA + ARA
diet (6.07%) and the lowest in eggs from the broodfish fed the
control diet (1.44%). EPA levels were similar in all diets, al-
though the EPA levels of floating eggs from broodfish fed the
control diet (7.22%) were significantly elevated (P < 0.0001)
relative to those from broodfish fed the DHA and DHA + ARA
diets. The DHA : EPA ratios of floating eggs were significantly
different (P < 0.0001) among treatments, with the highest levels

TABLE 4. Oil diameter, yolk volume, and notochord length of newly hatched Guinean Fingerfish larvae and mean survival and notochord length of larvae 24
and 48 h posthatch over the course of the 88-d spawning period, by diet. Values are means ± SDs; within columns, significant differences are indicated by different
lowercase letters.

24-h 48-h
Oil diameter Yolk volume Notochord Number 24-h 48-h Notochord Number of Notochord Number of
at spawn at spawn length at of larvae Survival Survival length larvae length larvae
Diet (mm) (mm3) spawn (mm) measured (%) (%) (mm) measured (mm) measured

Control 0.020 ± 0.005 z 0.023 ± 0.007 2.230 ± 0.241 z 789 62.0 ± 41.3 z 45.3 ± 41.2 z 2.193 ± 0.207 z 823 2.184 ± 0.216 z 651
DHA 0.021 ± 0.005 z 0.022 ± 0.007 2.246 ± 0.202 z 610 42.8 ± 35.3 y 30.5 ± 35.4 y 2.210 ± 0.184 z 441 2.165 ± 0.202 zy 317
DHA + ARA 0.019 ± 0.005 y 0.024 ± 0.055 2.201 ± 0.236 y 862 46.4 ± 45.0 y 30.3 ± 35.2 y 2.119 ± 0.211 y 646 2.149 ± 0.189 y 494
P-value <0.0001 0.6396 0.0007 0.0033 0.0047 <0.0001 0.0155
 EFFECTS OF DHA AND ARA ON BROODFISH DIETS 291

TABLE 5. Fatty acid composition of experimental diets expressed as percent-

ages of total fatty acids.

Fatty acid Control DHA DHA + ARA

12:0 0.03 0.06 0.23
13:0 0.02 0.02 0.02
14:0 4.89 6.72 7.09
14:1 0.01 0.01 0.02
15:0 0.30 0.36 0.33
15:1 0.00 0.01 0.00
16:0 21.51 23.73 22.29
16:1 7.43 7.17 6.64
17:0 0.83 0.88 0.55
17:1 0.00 0.01 0.00
FIGURE 1. Percentage DHA levels in floating eggs of Guinean Fingerfish
18:0 4.35 3.64 4.05 fed different diets throughout the 88-d spawning period. The control, DHA,
18:1 0.01 0.01 0.00 and DHA + ARA diets are represented by a squares, triangles, and circles,
18:1(n-9)cis 11.88 8.60 8.05 respectively. The equations for the three treatments are as follows: control: y =
Downloaded by [Department Of Fisheries] at 19:56 08 April 2013

18:2(n-6)trans 0.00 0.00 0.00 0.0056x – 212.91 (R2 = 0.0148); DHA: y = 0.0081x – 303.23 (0.1284); and
DHA + ARA: y = 0.0082x – 303.66 (0.0372).
18:2(n-6)cis 22.22 10.83 4.42
18:3(n-6) 0.32 0.52 1.09
18:3(n-3) 2.37 1.22 0.44 DISCUSSION
20:0(n-3) 0.07 0.16 0.20 This study represents the first investigation into broodstock
20:1(n-3) 0.36 0.65 0.34 nutrition with Guinean Fingerfish and the first to evaluate the
20:2(n-3) 0.01 0.04 0.02 effects of DHA and ARA on the biochemical composition of
21:0(n-3) 0.00 0.00 0.01 eggs, egg quality, egg morphometrics, larval survival, and larval
20:3(n-6) 0.02 0.06 0.53 morphometrics. Guinean Fingerfish have good aquaculture po-
20:4(n-6) 0.57 0.57 6.77 tential and will volitionally spawn repeatedly if the temperature
20:3(n-3) 0.01 0.02 0.01 is >24◦ C, the salinity is >25 g/L, and the broodfish are fed diet
20:5(n-3) 14.10 13.69 12.10 formulations similar to those in this experiment.
22:0(n-3) 0.13 0.05 0.24 Brood Guinean Fingerfish fed increased DHA + ARA pro-
22:1(n-9) 0.00 0.02 0.01 duced the greatest number of eggs and spawns. Many studies
22:2(n-3) 0.01 0.00 0.00 have indicated positive results for various reproductive parame-
23:0(n-3) 0.01 0.01 0.01 ters by including LC-PUFAs in broodstock diets. Recent studies
24:0(n-3) 0.07 0.07 0.17 with female Channel Catfish fed diets with increased levels of
22:6(n-3) 9.22 21.68 24.68 DHA, EPA, and ARA resulted in higher fecundity and greater
24:1(n-9) 0.09 0.10 0.22

being observed in eggs from broodfish fed the DHA + ARA

diet (4.33%) and the lowest in eggs from the broodfish fed the
control diet (2.03%). The EPA : ARA ratios of floating eggs were
also significantly different (P < 0.0001) among treatments, with
the highest levels being observed in eggs from broodfish fed the
control diet (5.03%) and the lowest in eggs from the broodfish
fed the DHA + ARA diet (1.08%).
The relationship between DHA and ARA content in floating
eggs from broodfish fed each diet throughout the 88-d spawning
period is depicted in Figures 1 and 2. For both the DHA and FIGURE 2. Percentage ARA levels in floating eggs of Guinean Fingerfish fed
different diets throughout the 88-d spawning period. The control, DHA, and
ARA regression models, no significant interaction between date DHA + ARA treatments are represented by squares, triangles, and circles,
and diet was detected, nor was date found to be a significant respectively. The equations for the three treatments are as follows: control: y =
factor (P = 0.3675 and P = 0.3546 in the regression models for –0.0001x + 5.7191 (R2 = 0.0009); DHA: y = –0.0026x + 106.96 (0.1372);
ARA and DHA, respectively). and DHA + ARA: y = 0.0083x – 330.38 (0.1462).
 292 OHS ET AL.

TABLE 6. Fatty acid composition of floating eggs as percentages of total fatty acids from Guinean Fingerfish females fed the experimental diets. Values are
means ± SDs; within rows, significant differences are indicated by different lowercase letters.

Fatty acid Control DHA DHA + ARA

12:0 0.01 ± 0.00 y 0.01 ± 0.00 y 0.02 ± 0.01 z
13:0 0.00 ± 0.00 y 0.01 ± 0.00 z 0.01 ± 0.00 z
14:0 1.51 ± 0.17 y 1.73 ± 0.20 z 1.80 ± 0.20 z
14:1 0.03 ± 0.01 0.03 ± 0.01 0.03 ± 0.01
15:0 0.12 ± 0.01 y 0.12 ± 0.01 zy 0.13 ± 0.02 z
15:1 0 ± 0 0 ± 0 0 ± 0
16:0 9.58 ± 0.45 9.61 ± 0.47 9.56 ± 0.63
16:1 9.43 ± 0.51 z 9.43 ± 0.47 z 8.95 ± 0.43 y
17:0 7.56 ± 1.71 8.20 ± 1.45 7.34 ± 2.39
17:1 0.27 ± 0.19 0.31 ± 0.16 0.31 ± 0.17
18:0 3.36 ± 0.99 3.19 ± 0.42 3.22 ± 0.39
18:1 0.01 ± 0.01 0.01 ± 0.01 0.01 ± 0.01
18:1(n-9)cis 28.83 ± 1.63 z 27.01 ± 1.48 y 24.45 ± 1.71 x
Downloaded by [Department Of Fisheries] at 19:56 08 April 2013

18:2(n-6)trans 0 ± 0 0 ± 0 0 ± 0
18:2(n-6)cis 19.45 ± 0.7 z 11.79 ± 0.62 y 8.10 ± 0.87 x
18:3(n-6) 0.93 ± 0.19 y 0.73 ± 0.11 x 1.14 ± 0.15 z
18:3(n-3) 1.71 ± 0.13 z 1.08 ± 0.13 y 0.62 ± 0.16 x
20:0 0.05 ± 0.02 0.05 ± 0.01 0.04 ± 0.02
20:1 0.40 ± 0.07 z 0.38 ± 0.07 zy 0.34 ± 0.10 y
20:2 0.35 ± 0.08 z 0.23 ± 0.06 y 0.16 ± 0.06 x
21:0 0 ± 0 0 ± 0 0 ± 0
20:3(n-6) 0.64 ± 0.10 y 0.48 ± 0.11 x 0.82 ± 0.18 z
20:4(n-6) 1.44 ± 0.08 x 1.96 ± 0.20 y 6.07 ± 0.79 z
20:3(n-3) 0.05 ± 0.03 z 0.04 ± 0.02 y 0.02 ± 0.01 x
20:5(n-3) 7.22 ± 0.61 z 6.72 ± 0.54 y 6.50 ± 0.80 y
22:0 0.01 ± 0.00 0.01 ± 0.00 0.00 ± 0.00
22:1 0.01 ± 0.01 0.01 ± 0 0.01 ± 0.01
22:2 0 ± 0 0 ± 0 0 ± 0
23:0 0.01 ± 0.00 0 ± 0.00 0 ± 0.00
24:0 0.07 ± 0.04 0.06 ± 0.03 0.05 ± 0.03
22:6(n-3) 14.52 ± 1.11 x 25 ± 0.64 y 27.6 ± 0.97 z
24:1 0.01 ± 0.00 0.01 ± 0.00 0.01 ± 0.01
DHA : EPA 2.03 ± 0.28 x 3.74 ± 0.29 y 4.33 ± 0.82 z
EPA : ARA 5.03 ± 0.52 z 3.47 ± 0.55 y 1.08 ± 0.13 x

fry production (as well as eggs with higher levels of DHA and
EPA) than did diets with decreased levels of the three fatty acids
(Durland et al. 2010). Female Yellowfin Seabream have also
achieved higher relative fecundities when fed increased levels
of DHA, EPA, and ARA (Zakeri et al. 2011). The fecundity
of many other sparids has been improved by including optimal
levels of n-3 LC-PUFAs in the diet, while negative effects of
excess n-3 LC-PUFAs on fecundity have also been reported
(Izquierdo et al. 2001). Spawning frequency and egg output
were improved for Guinean Fingerfish by increasing both the
levels of DHA and ARA in the diet (rather than the DHA level
alone) relative to the control. This could be indicative of the
greater importance of ARA level or LC-PUFA ratio than DHA
level alone for fecundity and spawning frequency in Guinean
Fingerfish.
Increasing either the DHA level alone or the DHA and ARA
levels in brood Guinean Fingerfish diets resulted in smaller
eggs with smaller oil reserves, as well as smaller larvae at hatch
(with smaller oil reserves) and at 24 and 48 h posthatch. More-
head et al. (2001) have observed smaller eggs with smaller yolk
reserves for Striped Trumpeters Latris lineata fed diets with
increased levels of DHA and EPA compared with a standard
commercial salmon diet. Wing-Keong and Wang (2011) ob-
served no differences in egg diameter, volume, or weight or
larvae weight and length for Nile Tilapia broodstock fed diets
varying in DHA, EPA, and ARA. In contrast, increased levels
 EFFECTS OF DHA AND ARA ON BROODFISH DIETS
of DHA, EPA, and ARA in the diets of female Channel Cat-
fish have resulted in larger eggs (Durland et al. 2010; Quintero
et al. 2011). In the present study, the results suggest a negative
effect of high DHA as well as ARA levels on the aforemen-
tioned egg and larval morphometric characteristics. However,
it is likely that the reduced egg sizes and nutritional reserves
(in the form of small oil diameters for eggs and larvae) and
notochord lengths at varying stages are just a manifestation of
the increased number of eggs per spawn per female and the
allocation of resources to compensate for this number of eggs
(Brooks et al. 1997). Also, the differences observed in the egg
and larval morphometrics in the present study appear to be very
small (between 2 and 100 µm) and the results of the statistical
analysis were likely influenced by the large number of replicate
spawns and measurements made.
Fertilization rates in wild Common Snook were positively
correlated with DHA level within the egg throughout a 4-year
Downloaded by [Department Of Fisheries] at 19:56 08 April 2013
study (Yanes-Roca et al. 2009). In contrast, Fernandez-Palacios
et al. (1995) found positive correlations between fertilization
and EPA and ARA levels for Gilthead Seabream (also known
as Gilthead Bream) Sparus auratus. The effects of EPA and
ARA on fertilization are attributed to their role in the membrane
structure of sperm as well as their role in prostaglandin syn-
thesis and involvement in hormone and pheromone production
during steroidogenesis and spawning, respectively (Izquierdo
et al. 2001). Although no significant differences in fertilization
were observed in floating eggs in the present study, the high-
est fertilization rates were associated with increasing both the
levels of DHA and ARA in the diet. This could once again be
indicative of the greater importance of ARA than DHA (and pos-
sibly their ratio) in Ginean Fingerfish broodstock performance.
Our results clearly indicate that both floating and sinking eggs
were fertilized and viable and that diet had no obvious effect on
fertilization. Therefore, it is recommended that producers col-
lect both floating and sinking eggs in order to more efficiently
achieve production goals.
Morehead et al. (2001) observed a decrease in egg hatcha-
bility for Striped Trumpeters fed diets with increased levels of
DHA and EPA. Decreases in egg hatchability have also been
observed for Nile Tilapia when broodfish were fed diets with
high levels of DHA, EPA, and ARA (Wing-Keong and Wang
2011). In contrast, Zakeri et al. (2011) observed increases in egg
hatchability as well as larval survival to 3 d posthatch (≥16%
and 25%, respectively) for Yellowfin Seabream broodstock fed
diets with increased levels of DHA, EPA, and ARA. Yanes-Roca
et al. (2009) observed a positive correlation between the DHA
level in wild Common Snook eggs and hatching percentage and
larval survival. In the present study, hatchability as well as 24-
and 48-h larval survival decreased as the levels of DHA and
ARA increased in Guinean Fingerfish broodstock diets. These
decreases may be attributed to excessive DHA and/or ARA lev-
els in the diet as well as to a subsequent increase in antioxidant
nutrient requirements (Izquierdo et al. 2001). The decreases in
survival may also be a result of a correlation between egg size
293
and nutrient availability in the egg (Brooks et al. 1997). Further
research is needed to evaluate this.
The DHA and DHA + ARA diets fed to brood Guinean Fin-
gerfish did not increase the 48-h survival of larvae. Additionally,
the larvae from broodfish females fed the DHA and DHA +
ARA diets had the lowest oil volume, yolk volume, and noto-
chord length at hatch, although the differences are very small
and the results of the statistical analysis were influenced by the
large number of replicate spawns. Reduced nutritional reserves
and notochord length may be a manifestation of the higher num-
bers of eggs per spawn. The DHA and DHA + ARA diets fed
to broodfish also produced larvae with the shortest notochord
length at 24 and 48 h posthatch. Again, these differences are
very small and the results of the statistical analysis were likely
influenced by the large sample size. Therefore, the application of
these small differences may be limited in a commercial context.
The fatty acid profiles of floating Guinean Fingerfish eggs
were positively correlated with the fatty acid profiles of the
diets with respect to DHA, EPA, and ARA, although the ac-
tual levels in eggs were slightly elevated for DHA, similar for
ARA, and drastically reduced for EPA (∼6.0%; nearly half the
level included in the diets) compared with the diets. The posi-
tive correlations and similarities in level observed for DHA and
ARA in the diets and eggs, coupled with the lack of significant
changes for these two fatty acids over time, suggest that based
on the feeding regime used in this study Guinean Fingerfish fe-
males are able to maintain these fatty acids in eggs at levels near
those in the diet throughout a protracted spawning season. There
were no major differences in the levels of linolenic and linoleic
acid between diet and egg for any diet, suggesting that Guinean
Fingerfish females have limited ability to utilize these C18 pre-
cursors to synthesize LC-PUFAs that may then be incorporated
into eggs. The large decrease in EPA concentrations in the eggs
relative to the levels in the diet and the associated increase in
DHA in the eggs suggests that Guinean Fingerfish females have
the ability to regulate the levels of these LC-PUFAs and their
assimilation into developing eggs.
CONCLUSIONS
This study has demonstrated that LC-PUFA levels in Guinean
Fingerfish broodstock diets can affect both the qualitative and
quantitative aspects of eggs and larvae. It appears that overall
spawning frequency and egg output are improved by increasing
the levels of DHA and ARA in diets, whereas most of the egg
and larval morphometric and viability parameters are negatively
impacted. The positive correlations and similarities in levels
observed for LC-PUFAs in the diets and eggs suggest the ease
of manipulating these fatty acids for future dose–response diet
studies with Guinean Fingerfish broodstock in order to optimize
egg production and larval output. Future dose–response studies
of LC-PUFAs as well as shorter-chain fatty acids could provide
more insight into the ability of Guinean Fingerfish broodstock
to manipulate and maintain specific fatty acid levels in tissues
and eggs.
 294 OHS ET AL.
REFERENCES
Akatsu, S., Y. Ogasawara, and F. Yasuda. 1977. Spawning behavior and devel-
opment of eggs and larvae of the Striped Fingerfish, Monodactylus sebae.
Japanese Journal of Ichthyology 23(4):208–214.
AOAC (Association of Official Analytical Chemists). 1995. Official methods of
analysis of AOAC International, 16th edition. AOAC International, Arlington,
Virginia.
Avila, E. M., and J. V. Juario. 1987. Yolk and oil globule utilization and devel-
opmental morphology of the digestive tract epithelium in larval Rabbitfish,
Siganus guttatus (Bloch). Aquaculture 65:319–331.
Bell, J. B., B. M. Farndale, M. P. Bruce, J. M. Navas, and M. Carillo. 1997.
Effects of broodstock dietary lipid on fatty acid compositions of eggs from
Sea Bass (Dicentrarchus labrax). Aquaculture 149:107–119.
Brooks, S., C. R. Tyler, and J. P. Sumpter. 1997. Egg quality in fish: what makes
a good egg? Reviews in Fish Biology and Fisheries 7:387–416.
Drillet, G., M. H. Iversen, T. F. Sørensen, R. Hans, L. Torben, and B. W. Hansen.
2006. Effect of cold storage upon eggs of a calanoid copepod, Acartia tonsa
(Dana), and their offspring. Aquaculture 254:714–729.
Durland, E. R., H. E. Quintero, D. A. Davis, and R. A. Dunham. 2010. Influence
of forage fish and dietary lipid supplements on egg quality and fry production
Downloaded by [Department Of Fisheries] at 19:56 08 April 2013
in Channel Catfish (Ictalurus punctatus) × Blue Catfish (Ictalurus furcatus)
hybridization. Aquaculture Nutrition 16:153–162.
El-Sayed, A.-F. M., C. R. Manour, and A. A. Ezzat. 2005. Effects of dietary
lipid source on spawning performance of Nile Tilapia (Oreochromis niloti-
cus) broodstock reared at different water salinities. Aquaculture 248:187–
196.
Fernandez-Palacios, H., M. S. Izquierdo, L. Robaina, A. Valencia, M. Salhi, and
J. Vergara. 1995. Effect of n-3 HUFA level in broodstock diets on egg quality
of Gilthead Seabream (Sparus aurata L.). Aquaculture 132:325–337.
Furuita, H., H. Tanaka, T. Yamamoto, N. Suzuki, and T. Takeuchi. 2002. Effects
of high levels of n-3 HUFA in broodstock diet on egg quality and egg fatty
acid composition of Japanese Flounder Paralichthys olivaceus. Aquaculture
210:323–333.
Furuita, H., T. Yamamoto, T. Shima, N. Suzuki, and T. Takeuchi. 2003. Effect
of arachidonic acid levels in broodstock diet on larval and egg quality of
Japanese Flounder Paralichthys olivaceus. Aquaculture 220:725–735.
Harel, M., A. Tandler, G. W. Kissil, and S. Applebaum. 1994. The kinetics
of nutrient incorporation into body tissues of Gilthead Sea Bream S. aurata
females and subsequent effects on egg composition and egg quality. British
Journal of Nutrition 72:45–58.
Izquierdo, M. S., H. Fernandez-Palacios, and A. G. J. Tacon. 2001. Effect
of broodstock nutrition on reproductive performance of fish. Aquaculture
197:25–42.
Kumaran, S., R. Lochmann, N. Stone, A. Kachowski, and Y. W. Lee. 2007.
Effects of diets with or without menhaden fish meal and oil on egg size,
hatchability, and fry size for rosy red Fathead Minnows. North American
Journal of Aquaculture 69:419–428.
Lane, R. L., and C. C. Kohler. 2006. Effects of dietary lipid and fatty acids on
White Bass reproduction performance, egg hatchability, and overall quality
of progeny. North American Journal of Aquaculture 68:141–150.
Li, Y.-Y., W.-Z. Chen, Z.-W. Sun, J.-H. Chen, and K.-G. Wu. 2005. Effects of
n-3 HUFA content in broodstock diet on spawning performance and fatty
acid composition of eggs and larvae in Plectorhunchus cinctus. Aquaculture
245:263–272.
Ling, S., M.-K. Kuah, T. S. T. Muhammad, S. Kolkovski, and A. C. Shu-
Chien. 2006. Effect of dietary HUFA on reproductive performance, tissue
fatty acid profile, and desaturase and elongase mRNAs in female Swordtail
Xiphophorus helleri. Aquaculture 261:204–214.
Morehead, D. T., P. R. Hart, G. A. Dunstan, M. Brown, and N. W. Pankhurst.
2001. Differences in egg quality between wild Striped Trumpeter (Latris lin-
eata) and captive Striped Trumpeter that were fed different diets. Aquaculture
192:39–53.
Mustafa, T., and K. C. Srivastava. 1989. Prostaglandins (eicosanoids) and their
role in ectothermic organisms. Advances in Comparative and Environmental
Physiology 5:157–207.
Ogasawara, Y., S. Akatsu, and Y. Taki. 1978. Juvenile stages and effect of salinity
on the survival of larvae and juveniles in the Striped Fingerfish, Monodactylus
sebae. Japanese Journal of Ichthyology 24(4):246–250.
Quintero, H. E., E. Durland, D. A. Davis, and R. Dunham. 2011. Effect of lipid
supplementation on reproductive performance of female Channel Catfish, Ic-
talurus punctatus, induced and strip-spawned for hybridization. Aquaculture
Nutrition 17:117–129.
Sink, T. D., and R. T. Lochmann. 2008. Effects of dietary lipid source and con-
centration on Channel Catfish (Ictalurus punctatus) egg biochemical com-
position, egg and fry production, and egg and fry quality. Aquaculture 283:
68–76.
Tocher, D. R. 2003. Metabolism and functions of lipids and fatty acids in teleost
fish. Reviews in Fisheries Science 11:107–184.
Watanabe, T. 1982. Lipid nutrition in fish. Comparative Biochemistry and Phys-
iology 73A:3–15.
Watanabe, T., and R. Vassallo-Agius. 2003. Broodstock nutrition research on
marine finfish in Japan. Aquaculture 227:35–61.
Wing-Keong, N., and Y. Wang. 2011. Inclusion of crude palm oil in the brood-
stock diets of female Nile Tilapia, Oreochromis niloticus, resulted in enhanced
reproductive performance compared to broodfish fed diets with added fish oil
or linseed oil. Aquaculture 314:122–131.
Yanes-Roca, C., N. Rhody, M. Nystrom, and K. L. Main. 2009. Effects of
fatty acid composition and spawning season patterns on egg quality and
larval survival in Common Snook (Centropomus undecimalis). Aquaculture
287:335–340.
Zakeri, M., P. Kochanian, J. G. Marammazi, V. Yavari, A. Savari, and M. Haghi.
2011. Effects of dietary n-3 HUFA concentrations on spawning performance
and fatty acid composition of broodstock, eggs, and larvae in Yellowfin Sea
Bream, Acanthopagrus latus. Aquaculture 310:388–394.
This article was downloaded by: [Department Of Fisheries]
On: 08 April 2013, At: 19:57
Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,
37-41 Mortimer Street, London W1T 3JH, UK

North American Journal of Aquaculture

Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/unaj20

Traditional and Modified Soy Oils as Substitutes for Fish

Oil in Feeds for Hybrid Striped Bass
a a
Jesse T. Trushenski & Kenson L. Kanczuzewski
a
Fisheries and Illinois Aquaculture Center and Department of Zoology, Southern Illinois
University–Carbondale, 1125 Lincoln Drive, Room 173, Carbondale, Illinois, 62901-6511, USA
Version of record first published: 02 Apr 2013.

To cite this article: Jesse T. Trushenski & Kenson L. Kanczuzewski (2013): Traditional and Modified Soy Oils as Substitutes for
Fish Oil in Feeds for Hybrid Striped Bass, North American Journal of Aquaculture, 75:2, 295-304

To link to this article: http://dx.doi.org/10.1080/15222055.2012.732673

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions

This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to
anyone is expressly forbidden.

The publisher does not give any warranty express or implied or make any representation that the contents
will be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses should
be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims,
proceedings, demand, or costs or damages whatsoever or howsoever caused arising directly or indirectly in
connection with or arising out of the use of this material.
 North American Journal of Aquaculture 75:295–304, 2013

C American Fisheries Society 2013

ISSN: 1522-2055 print / 1548-8454 online

DOI: 10.1080/15222055.2012.732673

SPECIAL SECTION: LIPIDS IN AQUACULTURE NUTRITION AND PHYSIOLOGY II

Traditional and Modified Soy Oils as Substitutes for Fish Oil

in Feeds for Hybrid Striped Bass
Jesse T. Trushenski* and Kenson L. Kanczuzewski
Fisheries and Illinois Aquaculture Center and Department of Zoology,
Southern Illinois University–Carbondale, 1125 Lincoln Drive, Room 173, Carbondale,
Illinois 62901-6511, USA
Downloaded by [Department Of Fisheries] at 19:57 08 April 2013

Abstract
In order to assess the relative merits of different soy oils as alternatives to fish oil, we evaluated the production
performance and fillet fatty acid profiles of hybrid Striped Bass fed feeds containing fish oil (FO) or 50:50 blends of
fish oil and standard (STD SO), saturated fatty acid-enriched (SFA SO), low alpha-linolenic acid (LOW ALA SO), or
hydrogenated (HYD SO) soy oil for 12 weeks (4 tanks/diet, 8 fish/tank; initial weight = 67.8 ± 0.2 g, grand mean ±
SE). Weight gain (214 ± 5%) and specific growth rate (1.4 ± 0.0% body weight/d) were unaffected by diet. Although
some differences were observed for feed conversion ratio, none of the soy-fed groups were significantly different from
the FO-fed group (1.2 ± 0.0). Feeding the soy diets altered fillet fatty acid composition, but coefficient of distance
values indicated the SFA SO (Djh = 3.9) and HYD SO (Djh = 9.1) feeds yielded fillet fatty acid profiles that were
more similar to the FO control than the STD SO (Djh = 17.3) and LOW ALA SO (Djh = 17.6) feeds. All of the soy
oils assessed were acceptable as partial substitutes for fish oil in hybrid Striped Bass feeds in terms of production
performance; however, SFA SO appears more suitable for maintaining fillet fatty acid profile.

With aquaculture production increasing rapidly and fish oil
resources remaining static (Turchini et al. 2009; De Silva et al.
2011), it is evident that alternative lipid resources are needed
to ensure that the continued growth of aquaculture will remain
both profitable and sustainable. Various lipid alternatives such
as plant-based lipids have been proven effective in the grow
out of many species of fish in an aquaculture setting (Turchini
et al. 2011b). For example, hybrid Striped Bass (White Bass
Morone chrysops × Striped Bass M. saxatilis) grow equally
well when fed diets containing fish oil or blends of fish oil
and various alternative lipids such as canola, coconut, corn,
grapeseed, flaxseed, or poultry oils (Wonnacott et al. 2004;
Lane et al. 2006; Lewis and Kohler 2008; Trushenski et al.
2008a). Although these and other common alternative lipids
yield adequate growth in fish, they contain little or no long-
chain polyunsaturated fatty acids (LC-PUFAs; C20–22 fatty acids
with three or more double bonds) and ultimately affect the fatty
acid profile of the edible product (Glencross 2009; Turchini
et al. 2009; Trushenski and Lochmann 2009; Rosenlund et al.
2011). Sparing fish oil with alternative lipids results in an in-
crease in saturated fatty acids (SFAs; fatty acids with no dou-
ble bonds), monounsaturated fatty acids (MUFAs; fatty acids
with a single double bond), and medium-chain polyunsatu-
rated fatty acids (C18 PUFAs; C18 fatty acids with two or
more double bonds) in fish feeds and, in turn, the tissues
of fish fed these diets. Although some plant oils, such as
flaxseed oil, contain high levels of n-3 fatty acids (primarily
18:3[n-3]1), they do not contain n-3 LC-PUFAs such as eicos-
apentaenoic (20:5[n-3]) and docosahexaenoic acids (22:6[n-
3]), which have been shown to reduce cardiovascular diseases,
inflammatory bowel disease, cancer, and rheumatoid arthritis
in humans (Wall et al. 2010). Consequently, fish oil sparing

*Corresponding author: saluski@siu.edu

1
In fatty acid designations of this nature, the number to the left of the colon is the number of carbon atoms in the compound, the number
immediately to the right of the colon is the number of double bonds, and the number after the hyphen indicates the position of the first double
bond from the methyl end.
Received April 10, 2012; accepted September 12, 2012
Published online April 2, 2013
295
 296 TRUSHENSKI AND KANCZUZEWSKI
affects the nutritional value of farmed fish (Hunter and Roberts
2000).
Finishing feeds offer an alternative method for raising fish
with alternative-lipid sources while maintaining sufficient levels
of LC-PUFAs in the fillets at harvest (Robin et al. 2003; Jobling
2004; Glencross and Turchini 2011). Recent studies have sug-
gested that hybrid Striped Bass and Nile Tilapia Oreochromis
niloticus possess the ability to selectively utilize SFAs and
MUFAs as energy sources for growth while selectively integrat-
ing LC-PUFAs into fillet tissues, thereby increasing the amount
of LC-PUFAs conserved during grow out or restored to the fillet
during finishing (Trushenski et al. 2008a, 2008b, 2009). Con-
versely, these studies have shown that diet compositions high in
C18 PUFAs tend to negatively impact LC-PUFA conservation
and/or restoration. The fatty acid composition of the alternative
lipid, specifically its saturated versus unsaturated fatty acid con-
tent, appears to determine the magnitude of the lipid’s effect on
Downloaded by [Department Of Fisheries] at 19:57 08 April 2013
tissue fatty acid profiles and fillet LC-PUFA loss.
Soy oil is a low-cost and widely distributed lipid source,
with soybeans comprising of 90% total oilseed production in
the United States (USDA 2010). However, soy oil also is high
in C18 PUFAs, primarily 18:2(n-6) (>50% total fatty acid methyl
esters [FAMEs]), which, as noted above, has been shown to neg-
atively impact LC-PUFA conservation and/or restoration. Soy
oil also contains 18:3(n-3) (∼7% FAMEs), which is perhaps
less problematic than 18:2(n-6) but can still interfere with LC-
PUFA deposition. However, soy oil may be modified through
selective breeding or processing to decrease its C18 PUFA con-
tent and increase its MUFA and SFA content. For example,
soy oils can be partially or fully hydrogenated to increase
their MUFA and SFA content; such oils have already shown
promise as fish oil alternatives in feeds for Rainbow Trout On-
corhynchus mykiss (Trushenski et al. 2011a), Largemouth Bass
Micropterus salmoides (Laporte and Trushenski 2011), and Nile
Tilapia (Mulligan and Trushenski 2013).
Further research is thus needed in order to determine the
effectiveness of various types of soy oil and their effects on
different fishes. Hybrid Striped Bass are one of the most reared
fish in the United States, but rising feed costs have been the
industry’s primary concern for several years (M. Turano, up-
dates on U.S. hybrid Striped Bass industry presented at Aqua-
culture America, 2011 and 2012). Cost savings may be real-
ized through fish oil sparing, but it is important to avoid or
minimize any negative effects of alternative lipid–based feeds
on production performance and fillet nutritional value. Accord-
ingly, we evaluated the production performance and fillet fatty
acid profiles of juvenile hybrid Striped Bass fed practical feeds
containing fish oil or blends of fish oil and standard (rich in
C18 PUFAs, particularly 18:2[n-6] and 18:3[n-3]), low alpha-
linolenic acid (rich in C18 PUFAs, particularly 18:2[n-6]), par-
tially hydrogenated (rich in MUFAs, particularly 18:1[n-7] and
18:1[n-9]), or fully hydrogenated (rich in SFAs, particularly
16:0 and 18:0) soy oil. These oils were selected because they
represent a range of fatty acid profiles that might provide greater
or lesser advantage in terms of conserving fillet fatty acid pro-
file while maintaining the growth performance of hybrid Striped
Bass.
METHODS
Preparation and analyses of diets.—Diets were prepared
based on a previously validated hybrid Striped Bass feed for-
mulation containing 9.8% fish oil (FO; Virginia Gold, Omega
Protein, Inc., Houston, Texas), which served as the control
diet in the present study. Four experimental feeds were for-
mulated to include 4.9% fish oil and 4.9% soy oil as standard
soy (STD SO), low-alpha linolenic-acid (LOW ALA SO), par-
tially hydrogenated (HYD SO), or fully hydrogenated (SFA SO)
soy-derived lipids (Archer Daniels Midland Company, Decatur,
Illinois). Feed formulations and proximate composition are pro-
vided in Table 1, and lipid and dietary fatty acid profiles are sum-
marized in Tables 2 and 3, respectively. Feed ingredients were
blended using a cutter–mixer (Model CM450; Hobart Corpora-
tion, Troy, Ohio), pelleted using a food grinder (1.5-hp [1 hp =
746 W] electric grinder; Cabela’s, Sidney, Nebraska), and dried
using a commercial food dehydrator for approximately 6 h at
100◦ C (Harvest Saver R-5A; Commercial Dehydrator Systems,
Eugene, Oregon). Feeds were manufactured and stored frozen
(–20◦ C) for the duration of the study. Feeds were analyzed
in triplicate to confirm their proximate composition. Moisture
content was determined gravimetrically after freeze-drying for
∼24 h (Freezone 6; Labconco Corporation, Kansas City, Mis-
souri); freeze-dried samples were pulverized and used for the
determination of ash, protein, and lipid content. Ash was de-
termined following incineration in a muffle furnace for 4 h at
650◦ C. Lipid content was determined gravimetrically following
modified Folch extraction (Folch et al. 1957). Protein content
was determined using a protein analyzer (FP-528; Leco Cor-
poration, St. Joseph, Michigan). Fatty acid composition was
determined using standardized procedures used in our labora-
tory. Briefly, total lipid was subjected to acid-catalyzed trans-
methylation overnight at 50◦ C, as described by Christie (1982).
The resultant fatty acid methyl esters were separated using a
Shimadzu GC-170A gas chromatograph (Shimadzu Scientific
Instruments, Kyoto, Japan) equipped with a flame ionization
detector fitted with a permanently bonded polyethylene glycol,
fused silica capillary column (Omegawax 250, 30 m × 0.25 mm
interior diameter, 0.25 µm film). Helium was used as the car-
rier gas (30 cm/s; 205◦ C) with an injection volume of 1.0 µL
and A temperature of 250◦ C. A split injection technique (100:1)
was used with a temperature setting as follows: 50◦ C for 2 min,
increased to 220◦ C at 4◦ C/min and held for 15 min. Individ-
ual FAMEs were identified by reference to external standards
(Supelco 37 Component FAME Mix, PUFA-1, and PUFA-3;
Supelco, Bellefonte, Pennsylvania). All solvents were of high-
performance liquid chromatography grade and obtained from
Sigma Diagnostics Inc. (St. Louis, Missouri).
Experimental design and feeding trial.—A recirculation sys-
tem consisting of 170-L tanks equipped with mechanical (sand
filter) and biological filtration (trickle-down biofilters) units
was used for the feeding trial. Supplemental aeration was
 SOY OILS IN HYBRID STRIPED BASS FEEDS 297

TABLE 1. Dietary formulation and proximate composition of diets fed to hybrid Striped Bass. Abbreviations are as follows: FO = diet containing fish oil only;
STD SO = diet containing a 50:50 blend of fish oil and standard soy oil; LOW ALA SO = diet containing a 50:50 blend of fish oil and low alpha linolenic acid
soy oil; SFA SO = diet containing a 50:50 blend of fish oil and fully hydrogenated soy oil; and HYD SO = diet containing a 50:50 blend of fish oil and partially
hydrogenated soy oil.

Ingredient FO STD SO LOW ALA SO SFA SO HYD SO

Dietary formulation
Menhaden fish meala 20.0 20.0 20.0 20.0 20.0
Soybean meal 30.0 30.0 30.0 30.0 30.0
Menhaden fish oila 9.8 4.9 4.9 4.9 4.9
Soy oilb 0.0 4.9 4.9 4.9 4.9
Wheat bran 20.18 20.18 20.18 20.18 20.18
Corn gluten meal 14.0 14.0 14.0 14.0 14.0
Sodium phosphate 1.5 1.5 1.5 1.5 1.5
Dicalcium phosphate 1.5 1.5 1.5 1.5 1.5
Vitamin premixc 0.12 0.12 0.12 0.12 0.12
Mineral premixd 0.1 0.1 0.1 0.1 0.1
Downloaded by [Department Of Fisheries] at 19:57 08 April 2013

Stay-Ce 0.2 0.2 0.2 0.2 0.2

Choline chloride 0.6 0.6 0.6 0.6 0.6
Carboxymethyl cellulose 2.0 2.0 2.0 2.0 2.0
f
Proximate composition
Dry matter 95.5 ± 0.1 94.9 ± 0.1 95.1 ± 0.1 94.1 ± 0.1 94.8 ± 0.1
Protein 41.0 ± 0.6 41.5 ± 0.6 40.9 ± 0.6 40.9 ± 0.6 39.5 ± 0.6
Lipid 15.3 ± 0.7 16.2 ± 0.7 15.3 ± 0.7 16.0 ± 0.7 15.0 ± 0.7
Ash 10.3 ± 0.5 10.7 ± 0.5 10.6 ± 0.5 10.0 ± 0.5 10.1 ± 0.5
a
Omega Protein, Inc., Houston, Texas.
b
Archer Daniels Midland, Decatur, Illinois.
c
Formulated to contain 25.000% L-ascorbyl-2-polyphosphate, 14.000% RRR-alpha tocopheryl acetate, 13.160% vitamin K, 12.500% inositol, 12.500% nicotinic acid, 7.500%
riboflavin, 6.250% calcium pantothenate, 2.500% pyridoxine hydrochloride, 1.250% thiamine mononitrate, 1.000% vitamin A palmitate, 0.500% cyanocobalamin, 0.450% folic acid,
0.125% biotin, and 0.010% cholecalciferol in a cellulose base.
d
Formulated to contain 24.897% zinc oxide, 14.933% ferrous sulfate, 3.470% manganese oxide, 0.967% cupric carbonate, 0.262% potassium iodide, 0.060% sodium selenate, and
0.030% cobalt carbonate in a cellulose base.
e
Roche Vitamins, Inc., Parsippany, New Jersey.
f
Values are least-squares means ± SEs of triplicate samples and report percent composition on a dry matter basis (except dry-matter content).

provided to each individual tank using air stones connected
to a blower. Dietary treatments were randomly assigned to
quadruplicate tanks of fish (8 fish/tank; 67.8 ± 0.2 g), and
feeds were offered to apparent satiation once daily for 12 weeks.
Satiation-based feeding was approximated by slowly offering
feed until the fish no longer surfaced to actively consume pellets
when offered. Typically, this resulted in a few pellets remain-
ing uneaten; no attempts were made to recover uneaten pel-
lets because of concern about inadvertent collection of fecal
matter along with the feed. Temperature and dissolved oxygen
were monitored daily (always before feeding, and occasionally
again after feeding) throughout the study using a YSI oxy-
gen meter (Model 55; Yellow Springs, Ohio), whereas alkalin-
ity, ammonia, and nitrite-nitrogen were measured periodically
(weekly throughout the trial, except for two missed samplings)
using a spectrophotometer and appropriate reagents (HACH
DR/2800; Hach Co., Loveland, Colorado). Throughout the trial,
the temperature was cooler than is considered optimal for hy-
brid Striped Bass growth (20.1 ± 1.4◦ C [mean ± SD]; range =
16.2–23.1◦ C) but still within the range appropriate for cultur-
ing this fish. Recordings of other water quality data were lost;
however, when operated according to standard operating proce-
dures, the water quality in this system is well within the ranges
appropriate for hybrid Striped Bass culture (Kohler 2000). Fish
husbandry practices were conducted according to the standards
of the Southern Illinois University Institutional Animal Care and
Use Committee under Animal Care and Use Protocol 08-027.
Sample collection and analyses.—After the end of the 12-
week feeding trial, all fish were counted and group-weighed by
tank. Production performance metrics were calculated using the
following equations:
Weight Gain (%)
(average final weight-average initial weight)
= 100 ×
average initial weight
Feed Conversion Ratio (FCR)
average individual dry matter feed intake
=
average individual weight gain
Specific Growth Rate (SGR, % body weight/day)
= 100 ×
loge average final weight-loge average initial weight
days of feeding
 298 TRUSHENSKI AND KANCZUZEWSKI

TABLE 2. Fatty acid composition of lipid sources used in the experimental diets (% fatty acid methyl esters). Values are least-squares means ± SEs of triplicate
samples. See Table 1 for abbreviations.

Fatty acid(s) FO STD SO LOW ALA SO SFA SO HYD SO

14:0 8.7 ± 0.0 0.1 ± 0.0 0.1 ± 0.0 0.6 ± 0.0 0.1 ± 0.0
16:0 18.9 ± 0.0 11.1 ± 0.0 11.6 ± 0.1 21.2 ± 0.0 12.0 ± 0.1
18:0 3.4 ± 0.0 4.4 ± 0.1 4.2 ± 0.0 76.4 ± 3.7 8.6 ± 0.1
Total SFAsa 33.1 ± 0.0 16.1 ± 0.5 16.1 ± 0.1 98.5 ± 0.5 20.9 ± 0.1
16:1(n-7) 11.4 ± 0.0 0.1 ± 0.0 0.1 ± 0.0 0.0 ± 0.0 0.0 ± 0.0
18:1(n-7) 3.2 ± 0.0 1.5 ± 0.0 1.4 ± 0.0 0.0 ± 0.0 16.2 ± 0.1
18:1(n-9) 5.9 ± 0.0 21.8 ± 0.5 23.4 ± 0.1 0.8 ± 0.3 59.5 ± 0.1
Total MUFAsb 21.8 ± 0.0 23.5 ± 0.0 24.9 ± 0.2 0.8 ± 0.3 75.8 ± 0.1
18:2(n-6) 1.7 ± 0.0 52.9 ± 0.7 56.5 ± 0.3 0.2 ± 0.1 3.0 ± 0.0
20:4(n-6) 0.9 ± 0.0 0.0 ± 0.0 0.0 ± 0.0 0.0 ± 0.0 0.0 ± 0.0
n-6c 3.4 ± 0.0 52.9 ± 0.7 56.5 ± 0.3 0.2 ± 0.1 3.0 ± 0.0
18:3(n-3) 2.0 ± 0.0 7.4 ± 0.3 2.3 ± 0.0 0.1 ± 0.1 0.2 ± 0.0
18:4(n-3) 4.4 ± 0.0 0.1 ± 0.0 0.0 ± 0.0 0.0 ± 0.0 0.0 ± 0.0
Downloaded by [Department Of Fisheries] at 19:57 08 April 2013

20:4(n-3) 1.7 ± 0.1 0.0 ± 0.0 0.0 ± 0.0 0.0 ± 0.0 0.0 ± 0.0
20:5(n-3) 12.9 ± 0.0 0.0 ± 0.0 0.0 ± 0.0 0.0 ± 0.0 0.0 ± 0.0
22:5(n-3) 2.3 ± 0.0 0.0 ± 0.0 0.0 ± 0.0 0.0 ± 0.0 0.0 ± 0.0
22:6(n-3) 15.3 ± 0.1 0.0 ± 0.0 0.0 ± 0.0 0.0 ± 0.0 0.0 ± 0.0
n-3d 38.8 ± 0.1 7.5 ± 0.3 2.3 ± 0.0 0.1 ± 0.1 0.2 ± 0.0
Total PUFAse 45.1 ± 0.1 60.4 ± 1.1 58.9 ± 0.3 0.7 ± 0.1 3.3 ± 0.0
Total LC-PUFAsf 33.5 ± 0.1 0.0 ± 0.0 0.0 ± 0.0 0.0 ± 0.0 0.0 ± 0.0
Total C18 PUFAsg 8.9 ± 0.0 60.4 ± 1.1 58.8 ± 0.3 0.4 ± 0.1 3.2 ± 0.0
(n-3):(n-6) 11.3 ± 0.1 0.1 ± 0.0 0.0 ± 0.0 0.4 ± 0.2 0.0 ± 0.0
a
Sum of all fatty acids without double bonds, including 8:0, 10:0, 12:0, 15:0, 17:0, 20:0, 22:0, and 24:0.
b
Sum of all fatty acids with a single double bond, including 14:1, 15:1, 17:1, 20:1(n-9), 22:1(n-9), 22:1(n-11), and 24:1(n-9).
c
Sum of all n-6 fatty acids, including 18:3(n-6), 20:2(n-6), and 20:3(n-6).
d
Sum of all n-3 fatty acids, including 20:3(n-3).
e
Sum of all fatty acids with two or more double bonds, including 16:2(n-4), 16:3(n-4), 18:3(n-4), 18:3(n-6), 20:3(n-3), 20:2(n-6), and 20:3(n-6).
f
Sum of C20–22 fatty acids with three or more double bonds, including 20:3(n-3) and 20:3(n-6).

Coefficient of distance (Djh) values (Turchini et al. 2006)

Feed Intake (% body weight/day) = 100
average individual dry matter feed intake
×
(initial individual weight × final individual weight)0.5 /days of feeding
Three fish were randomly selected from each tank and euth-
anized by an overdose of tricaine methanesulfonate (MS-222;
∼200 mg/L bath immersion until cessation of opercular move-
ment) followed by single cranial pithing. Individual weights
were determined, and livers and intraperitoneal fat (IPF) masses
were dissected from each fish and weighed in order to calcu-
late the hepatosomatic index (HSI; 100 × [liver weight/body
weight]) and liposomatic index (LSI; 100 × [IPF weight/whole
body weight]). Fillet samples were collected from each fish,
stored in sterile sample bags (Whirlpak; Nasco, Fort Atkinson,
Wisconsin) and frozen (–80◦ C) for later fatty acid analysis. Prior
to analysis, fillets were cut into approximately 8-cm × 3-cm ×
1-cm (length × width × thickness) pieces and freeze-dried for
∼48 h using the equipment previously described. After fillet
samples were freeze-dried and pulverized, the crude lipid con-
tent and fatty acid composition were determined using the same
methods described previously for feed samples.
comparing overall fillet fatty acid profiles between the
experimental treatments and the FO control group were cal-
culated as follows:
 1/2

n
Djh = (Pij-Pih) 2
,
i=1
where Pij is the percent content of fatty acid i in the control
treatment (FO) and Pih is the percent content of fatty acid i in an
experimental treatment. Only major individual fatty acids (>1%
of total quantified FAMEs; no fatty acid groupings, e.g., SFAs,
MUFAs) were used in the calculation of Djh.
Statistical analyses.—One-way analysis of variance
(ANOVA) was used to determine whether significant differ-
ences existed among treatment means for production perfor-
mance, fillet composition, and fatty acid profile. Tanks were
used as the experimental units (N = 4). For metrics exhibit-
ing significant treatment effects, Tukey’s honestly significant
difference tests were used for pairwise comparison of treatment
means. Coefficient of distance data were determined using mean
fatty acid composition data, yielding a single value for each
 SOY OILS IN HYBRID STRIPED BASS FEEDS 299

TABLE 3. Fatty acid composition of experimental diets (% fatty acid methyl esters). Values are least-squares means ± SEs of triplicate samples.

Fatty acid(s) FO STD SO LOW ALA SO SFA SO HYD SO

14:0 7.7 ± 0.1 4.3 ± 0.1 4.4 ± 0.1 4.6 ± 0.1 5.4 ± 0.1
16:0 19.4 ± 0.1 16.5 ± 0.1 16.8 ± 0.1 17.4 ± 0.1 20.2 ± 0.1
18:0 3.8 ± 0.1 4.2 ± 0.1 4.1 ± 0.1 35.7 ± 0.1 7.0 ± 0.1
Total SFAsa 31.2 ± 0.2 25.3 ± 0.2 25.7 ± 0.2 58.8 ± 0.2 33.1 ± 0.2
16:1(n-7) 9.9 ± 0.1 5.5 ± 0.1 5.6 ± 0.1 5.4 ± 0.1 6.8 ± 0.1
18:1(n-7) 2.7 ± 0.3 2.2 ± 0.3 2.2 ± 0.3 1.5 ± 0.3 4.4 ± 0.3
18:1(n-9) 8.3 ± 0.1 14.4 ± 0.1 14.8 ± 0.1 5.7 ± 0.1 18.9 ± 0.1
Total MUFAsb 21.7 ± 0.3 22.6 ± 0.3 23.1 ± 0.3 13.1 ± 0.3 30.7 ± 0.3
18:2(n-6) 8.8 ± 0.1 28.9 ± 0.1 30.3 ± 0.1 7.8 ± 0.1 10.9 ± 0.1
20:4(n-6) 1.1 ± 0.0 0.6 ± 0.0 0.6 ± 0.0 0.6 ± 0.0 0.7 ± 0.0
n-6c 10.4 ± 0.1 29.6 ± 0.1 31.0 ± 0.1 8.4 ± 0.1 11.7 ± 0.1
18:3(n-3) 2.1 ± 0.0 4.4 ± 0.0 2.3 ± 0.0 1.4 ± 0.0 1.7 ± 0.0
18:4(n-3) 2.9 ± 0.0 1.5 ± 0.0 1.5 ± 0.0 1.5 ± 0.0 1.9 ± 0.0
20:4(n-3) 1.5 ± 0.0 0.8 ± 0.0 0.8 ± 0.0 0.8 ± 0.0 1.0 ± 0.0
Downloaded by [Department Of Fisheries] at 19:57 08 April 2013

20:5(n-3) 12.4 ± 0.1 6.5 ± 0.1 6.4 ± 0.1 6.5 ± 0.1 8.1 ± 0.1
22:5(n-3) 2.4 ± 0.0 1.2 ± 0.0 1.2 ± 0.0 1.3 ± 0.0 1.6 ± 0.0
22:6(n-3) 13.0 ± 0.2 6.8 ± 0.2 6.7 ± 0.2 6.9 ± 0.2 8.6 ± 0.2
n-3d 34.4 ± 0.4 21.2 ± 0.4 19.0 ± 0.4 18.4 ± 0.4 22.8 ± 0.4
Total PUFAse 47.1 ± 0.4 52.1 ± 0.4 21.3 ± 0.4 28.0 ± 0.4 36.1 ± 0.4
Total LC-PUFAsf 30.7 ± 0.3 15.9 ± 0.3 15.8 ± 0.3 16.1 ± 0.3 20.0 ± 0.3
Total C18 PUFAsg 13.9 ± 0.1 34.8 ± 0.1 34.1 ± 0.1 10.7 ± 0.1 14.6 ± 0.1
(n-3):(n-6) 3.3 ± 0.1 0.7 ± 0.1 0.6 ± 0.1 2.2 ± 0.1 1.9 ± 0.1
a
Sum of all fatty acids without double bonds, including 10:0, 12:0, and 20:0.
b
Sum of all fatty acids with a single double bond.
c
Sum of all n-6 fatty acids, including 20:3(n-6).
d
Sum of all n-3 fatty acids.
e
Sum of all fatty acids with two or more double bonds.
f
Sum of C20–22 fatty acids with three or more double bonds, including 20:3(n-6).
g
Sum of C18 fatty acids with two or more double bonds.

experimental dietary treatment group. Accordingly, these data
were not analyzed using formal statistical procedures, All sta-
tistical procedures were completed using the GLM framework
of the Statistical Analysis System 9.2 (SAS Institute, Cary,
North Carolina). Differences were considered significant at
P < 0.05.
RESULTS
The production performance of hybrid Striped Bass was
largely unaffected by dietary lipid source (Table 4). No sig-
nificant differences were observed for weight gain (214 ± 5%
[grand mean ± SE]) or SGR (1.4 ± 0.0% body weight/d),
though treatments did vary in terms of feed intake and FCR.
The HYD SO and SFA SO treatments differed in terms of feed
intake (1.6% versus 1.8% body weight/d, respectively) and in
FCR (1.1 versus 1.2). Despite these relatively small differences
among groups, no significant differences in production perfor-
mance were observed between experimental treatments and the
FO control group.
Fillet fatty acid composition was generally observed to reflect
dietary fatty acid composition and intake (Table 5). The HYD
SO feed contained high levels of MUFAs, and the fillets of fish
fed this diet contained significantly elevated levels of MUFAs
(31.2% FAMEs) in comparison with the other treatments (24.9–
27.0% FAMEs). Similarly, the STD SO and LOW ALA SO
feeds contained high levels of C18 PUFAs (34.8% and 34.1%
FAMEs, respectively), and the fillets of fish fed these diets were
significantly enriched in C18 PUFAs (26.0% and 25.5% FAMEs,
respectively). Although the SFA SO feed contained high levels
of SFAs (58.8% FAMEs), SFAs were not significantly elevated
in the fillets of fish fed this diet (28.8%) instead of the FO feed
(28.8%). The fillets of fish fed STD SO, LOW ALA SO, SFA
SO, and HYD SO exhibited decreased levels of n-3 fatty acids
(23.7–30.8% FAMEs) as well as higher levels of n-6 fatty acids
(11.8–25.6% FAMEs) than those of fish fed the FO feed ([n-
3] = 34.0% FAMEs, [n-6] = 9.8% FAMEs). Fillet levels of
LC-PUFAs were highest among those fed the FO feed (32.5%
FAMEs), lowest among those fed the STD SO and LOW ALA
SO feeds (21.2% and 22.4% FAMEs), and intermediate among
those fed the SFA SO and HYD SO feeds (29.4% and 27.8%
FAMEs). Coefficient of distance values indicated that the SFA
SO (Djh = 3.9) and HYD SO (Djh = 9.1) feeds generally yielded
fillet fatty acid profiles that were more similar to those of the
FO control than the STD SO (Djh = 17.3) and LOW ALA SO
(Djh = 17.6) feeds did.
 300 TRUSHENSKI AND KANCZUZEWSKI

TABLE 4. Production performance by dietary treatment. Values are the least-squares means of the values associated with quadruplicate tanks. A pooled SE and
P-value are provided for each metric; SEs < 0.1 are reported as 0.0. Within rows, means with common lowercase letters are not significantly different.

Performance metric FO STD SO LOW ALA SO SFA SO HYD SO Pooled SE P-value

Survival (%) 100 100 100 100 100
Initial individual weight (g) 67.8 68.5 67.8 67.8 67.3 0.5 0.358
Final individual weight (g) 210.9 213.0 214.1 218.6 207.5 11.2 0.893
Weight gain (%) 211.3 211.1 215.5 222.2 208.3 15.4 0.905
SGR (% body weight/d) 1.4 1.4 1.4 1.4 1.4 0.1 0.906
Total consumption (g/fisha) 169.3 168.0 170.1 186.5 159.3 10.5 0.186
Feed intake (% body weight/d) 1.7 yz 1.7 yz 1.7 yz 1.8 z 1.6 y 0.1 0.036
FCR 1.2 yz 1.2 yz 1.2 yz 1.2 z 1.1 y 0.0 0.045
LSI 1.4 1.6 1.4 1.3 1.2 0.2 0.107
HSI 1.2 1.4 1.4 1.3 1.3 0.1 0.423
a
Dry matter.
Downloaded by [Department Of Fisheries] at 19:57 08 April 2013

TABLE 5. Fatty acid composition of total lipid in fillets (% fatty acid methyl esters). Values are least-squares means of triplicate samples from quadruplicate
tanks. A pooled SE and P-value are provided for each fatty acid or fatty acid grouping. Within rows, means with common lowercase letters are not significantly
different.

Fatty acid(s) FO STD SO LOW ALA SO SFA SO HYD SO Pooled SE P-value

14:0 5.0 z 3.2 w 3.1 w 4.6 y 3.6 x 0.1 <0.001
16:0 20.3 z 17.9 y 18.3 y 20.3 z 19.6 z 0.3 <0.001
18:0 3.4 x 3.7 y 3.7 xy 3.8 y 4.4 z 0.1 <0.001
Total SFAsa 28.8 z 24.9 x 25.2 x 28.8 z 27.6 y 0.3 <0.001
16:1(n-7) 8.9 z 5.6 w 5.5 w 8.1 y 6.2 x 0.2 <0.001
18:1(n-7) 3.4 z 2.7 y 2.5 y 3.3 z 3.6 z 0.2 <0.001
18:1(n-9) 12.6 w 16.6 y 16.0 y 14.3 x 20.3 z 0.4 <0.001
20:1(n-9) 1.2 yz 1.0 wx 1.0 w 1.2 z 1.1 xy 0.0 <0.001
Total MUFAsb 26.1 xy 25.9 xy 24.9 x 27.0 y 31.2 z 0.7 <0.001
18:2(n-6) 7.5 x 22.1 z 23.0 z 9.9 y 9.5 y 0.4 <0.001
20:2(n-6) 0.8 x 1.3 z 1.3 z 1.0 y 0.7 x 0.0 <0.001
20:4(n-6) 1.4 z 1.0 y 1.1 y 1.4 z 1.4 z 0.1 <0.001
n-6c 9.8 x 24.6 z 25.6 z 12.8 y 11.8 y 0.4 <0.001
18:3(n-3) 1.5 y 2.9 z 1.6 y 1.6 y 1.2 x 0.1 <0.001
18:4(n-3) 1.7 z 1.0 w 0.9 w 1.5 y 1.2 x 0.0 <0.001
20:4(n-3) 1.2 z 0.7 x 0.7 x 1.0 y 0.8 x 0.1 <0.001
20:5(n-3) 10.1 z 6.3 x 6.5 x 9.6 z 8.2 y 0.2 <0.001
22:5(n-3) 2.7 z 1.8 x 1.8 x 2.6 z 2.2 y 0.0 <0.001
22:6(n-3) 16.9 z 11.3 y 12.2 y 14.6 z 15.0 z 0.8 <0.001
n-3d 34.0 z 24.0 x 23.7 x 30.8 y 28.7 y 0.8 <0.001
Total PUFAse 45.1 y 49.1 z 49.9 z 44.3 y 41.2 x 0.7 <0.001
Total LC-PUFAsf 32.5 z 21.2 x 22.4 x 29.4 y 27.8 y 0.9 <0.001
Total C18 PUFAsg 10.6 x 26.0 z 25.5 z 12.9 y 11.9 y 0.5 <0.001
(n-3):(n-6) 3.5 z 1.0 x 0.9 x 2.5 y 2.4 y 0.1 <0.001
a
Sum of all fatty acids without double bonds, including 10:0, 12:0, and 20:0.
b
Sum of all fatty acids with a single double bond.
c
Sum of all n-6 fatty acids, including 20:3(n-6).
d
Sum of all n-3 fatty acids.
e
Sum of all fatty acids with two or more double bonds.
f
Sum of C20–22 fatty acids with three or more double bonds, including 20:3(n-6).
g
Sum of C18 fatty acids with two or more double bonds.
 SOY OILS IN HYBRID STRIPED BASS FEEDS
DISCUSSION
Past studies have shown that many fish species can readily
accept alternative lipid sources in aquafeeds as long as essen-
tial fatty acid requirements are met (Torstensen et al. 2000;
Bell et al. 2001; Wonnacott et al. 2004; Turchini et al. 2009).
Hybrid Striped Bass do not possess the ability to produce suf-
ficient amounts of LC-PUFAs from C18 precursors and thus
require dietary provision of 20:4(n-6), 20:5(n-3), and 22:6(n-
3) (Nematipour and Gatlin 1993). In our study, each of the
feeds contained at least 4.9% fish oil and 20% fish meal, which
provided ample amounts of these LC-PUFAs. Thus, produc-
tion performance was unaffected by 50% replacement of fish
oil with standard, low alpha-linolenic acid, highly saturated,
or hydrogenated soy oil. Although the growth achieved during
the trial is somewhat less than reported in other trials with ju-
venile hybrid Striped Bass (e.g., Gallagher 1994; Trushenski
2009; Laporte and Trushenski 2012), our results are consistent
Downloaded by [Department Of Fisheries] at 19:57 08 April 2013
with those reported in other studies (Rawles and Gatlin 1998;
Gaylord and Rawles 2005), and undoubtedly the growth
achieved in any particular study is a function of initial starting
weight (a bit larger in the present work) and water temperature
(a bit cooler than optimal in the present work). Previous studies
utilizing alternative lipid sources such as canola, coconut, corn,
grapeseed, linseed, or poultry oils (partial or full replacement) in
hybrid Striped Bass feeds did not report substantial differences
in growth (Wonnacott et al. 2004; Lane et al. 2006; Lewis and
Kohler 2008; Trushenski et al. 2011b). Furthermore, past stud-
ies have shown that various soy oil derivatives, including the
ones used in this study, yield adequate production performance
in other taxa (Trushenski et al. 2011a; Laporte and Trushenski
2011; Mulligan and Trushenski 2013). Accordingly, our results
agree with previous reports addressing these particular soy oils
and various other fish oil alternatives.
In general, the fatty acid profile of fish tissues closely mim-
ics the fatty acid composition of aquafeeds (Robin et al. 2003;
Jobling 2004; Turchini et al. 2009). This phenomenon has
been observed in many taxa, including hybrid Striped Bass
(Nematipour and Gatlin 1993; Fair et al. 1993; Wonnacott et al.
2004; Lewis and Kohler 2008; Trushenski et al. 2008a, 2008b).
In general, our data are consistent with these observations. For
example, fish fed diets such as the STD SO and LOW ALA SO
that were high in C18 PUFAs and n-6 fatty acids exhibited high
levels of these fatty acids in their fillets. Conversely, fish fed
the FO control that was low in C18 PUFAs and n-6 fatty acids
exhibited low levels of these fatty acids. Although fatty acid
profiles generally reflect dietary intake, increasing consumption
of SFAs has been shown to have a less dramatic effect on fil-
let fatty acid profile (Henderson 1996; Frøyland et al. 2000;
Trushenski et al. 2008a, 2009; Trushenski 2009; Turchini et al.
2011a). Feeding diets based on SFA-rich lipids such as SFA-
enriched soybean and cottonseed lipids, coconut oil, and palm
oils did not result in proportionate SFA enrichment in the fillets
of Rainbow Trout (Trushenski et al. 2011a), Largemouth Bass
(Laporte and Trushenski 2011), Murray Cod Muccullochella
301
peelii (Turchini et al. 2011a), or Nile Tilapia (Mulligan and
Trushenski 2013). Furthermore, the provision of SFA-rich diets
has been linked to greater deposition or retention of LC-PUFAs
in fish (Trushenski et al. 2011a; Turchini et al. 2011a), includ-
ing hybrid Striped Bass (Trushenski et al. 2008a, 2008b). These
studies have shown that some fish taxa possess the ability to
effectively utilize SFAs and MUFAs for energy production via
beta oxidation while retaining LC-PUFAs in the fillets. Some
have referred to this effect as the “omega-3 sparing effect” of
MUFAs and, to a lesser extent, SFAs (Turchini et al. 2011a).
Conversely, feeding diets rich in C18 PUFAs has the opposite
effect: it has been suggested that C18 PUFAs, particularly 18:2(n-
6), effectively “out-compete” LC-PUFAs for tissue deposition,
and feeding C18 PUFA–rich feeds exacerbates the loss of LC-
PUFAs typically associated with fish oil sparing (Trushenski
et al. 2008b). Our study supports these concepts: fish fed the SFA
SO (SFA-rich) and HYD SO (MUFA-rich) feeds most closely
resembled those fed the LC-PUFA–rich FO feed, whereas those
fed the STD SO and LOW ALA SO (both C18 PUFA–rich)
feeds had strikingly different fillet fatty acid profiles. Although
the low alpha linolenic acid soybean oil contains less 18:3(n-3)
than standard soybean oil, it contains a higher percentage of
18:2(n-6) and thus roughly equivalent levels of total C18 PU-
FAs. For example, the SFA SO diet, which contained roughly
half (∼16% FAMEs) the LC-PUFA content of the FO control
(∼31% FAMEs), resulted in an only slightly reduced LC-PUFA
concentration in fillet fatty acids (29.4% versus 32.5% FAMEs).
Although the slightly elevated concentrations of LC-PUFAs in
the SFA SO and HYD SO diets may partially explain the higher
levels of LC-PUFAs observed in these groups (16.1% and 20.0%
FAMEs versus 15.9% and 15.8%), the dietary differences do not
fully explain the magnitude of the fillet differences observed
(29.4% and 27.8% versus 21.2% and 22.4%). Conversely, the
STD SO and LOW ALA SO treatments high in C18 PUFAs
produced significantly lower LC-PUFA levels within the fil-
lets. Although the STD SO and LOW ALA SO fillet LC-PUFA
levels exceed dietary concentrations and indicate some level
of selective retention of these fatty acids, the levels are lower
than those observed in the SFA SO and HYD SO fillets, most
likely because of the high concentrations of 18:2(n-6) found in
the STD SO and LOW ALA SO feeds. In mammalian studies,
18:2(n-6) has been shown to be a moderately or highly pre-
ferred substrate for uptake and incorporation into lipids (Lands
et al. 1982; Huang et al. 1987; Emken et al. 1994); similar re-
sults have been observed in studies employing teleost models,
which reported moderate to high rates of label 18:2(n-6) uptake
and esterification into lipid by isolated teleost cells (Pérez et al.
1999; Rodrı́guez et al. 2002). Elsewhere, we have reported the
consequences of selective metabolism of 18:2(n-6) in hybrid
Striped Bass, including tissue incorporation and retention levels
exceeding those that would be anticipated based on dietary in-
take (Lane et al. 2006; Trushenski et al. 2008b, 2011b). Others
have reported the same, and in a recent review of fish oil spar-
ing in aquafeeds Turchini et al. (2009) said of the overriding
 302 TRUSHENSKI AND KANCZUZEWSKI
influence of dietary 18:2(n-6) on fillet composition, “the con-
tent of 18:2(n-6) in a potential alternative lipid source must be
considered as one of the most informative (negative) parameters
to be considered because this fatty acid is responsible for most
of the detrimental modifications to the fatty acid composition
of farmed fish fillets.” Further, the effects of 18:2(n-6) on tissue
composition have led some to suggest that, by “outcompeting”
other essential fatty acids, a dietary excess of 18:2(n-6) may
induce deficiencies in fish (Blanchard et al. 2008; Trushenski
et al. 2013, this issue).
Conversely, both the HYD SO and SFA SO feeds were low in
18:2(n-6) and high in SFAs, but the SFAs were not proportion-
ately reflected in the fillets. Although the SFA SO diet contained
nearly a twofold increase in SFAs compared with the FO con-
trol, the fillets of fish fed these two diets had equivalent levels of
SFAs. Similar effects have been reported by others investigating
SFA-rich lipids (Turchini et al. 2009). Arguably, these effects
Downloaded by [Department Of Fisheries] at 19:57 08 April 2013
could be a consequence of impaired digestibility and absorption
or catabolism of SFAs. Limited digestibility of SFA-rich lipids
has been observed in marine coldwater species, which was at-
tributed to differences in melting points and the limited PUFA
content of the alternative lipids (Lie et al. 1987; Olsen et al.
1998; Ng et al. 2004; Fonseca-Madrigal et al. 2005). This sce-
nario seems unlikely given the lack of growth differences and the
moderate water temperatures used in the present work. Selec-
tive catabolism of dietary SFAs seems a more likely explanation,
particularly as selective catabolism of SFAs has been reported
in several fish species (Henderson and Sargent 1985; Kiessling
and Kiessling 1993; Henderson 1996; Frøyland et al. 2000). Al-
though the HYD SO treatment yielded results broadly similar to
those of the SFA SO treatment, the HYD SO treatment exhib-
ited significantly elevated MUFA fillet retention with reduced
LC-PUFA concentration. Furthermore, the SFA SO treatment
was the only treatment out of the four SO-based feeds that did
not yield significant differences in either 20:5(n-3) or 22:6(n-3)
compared with the FO control. Saturated fatty acids are appar-
ently inferior substrates for tissue deposition, as opposed to C18
PUFAs and LC-PUFAs. Although LC-PUFAs may be the pre-
ferred substrate, C18 PUFAs are incorporated within the tissues
when they are abundant. Accordingly, experimental diets high
in SFAs and low in C18 PUFAs closely mirrored the FO con-
trol in terms of growth performance and fatty acid profiles; diets
low in SFAs and high in C18 PUFAs yielded the opposite results.
Increased consumption of LC-PUFAs has been recom-
mended by both medical and nutritional communities to im-
prove the health and well-being of human populations (Kris-
Etherton et al. 2002). For example, consumption of 5.5 g of
20:5(n-3) and 22:6(n-3) per month has been found to reduce
the risk of cardiac arrest in humans by 50% (Siscovick et al.
2000). The commercial feeds of carnivorous fish species rely
heavily on reduction fisheries for feed manufacturing, which
have raised concerns regarding the economic and environmen-
tal sustainability of aquaculture production (Pauly et al. 1998).
Partial or complete replacement of fish oil with plant-derived
lipids can be an effective strategy to reduce costs and reliance
on limited marine resources, but doing so yields seafood prod-
ucts that are less valuable as a source of LC-PUFAs for human
consumers. Based on the present results and the recent work
of others, it is becoming increasingly apparent that the use of
alternative lipids high in SFAs and low in C18 PUFAs is an ef-
fective strategy for minimizing the loss of beneficial LC-PUFAs
among fish reared on reduced or fish-oil-free feeds. Depend-
ing on the recommending authority, adults are encouraged to
consume as much as 1–2 g of 20:5(n-3) + 22:6(n-3) per day,
though recommendations in the 250–500 mg per day range are
more common (Aranceta and Pérez-Rodrigo 2012). Assuming
that hybrid Striped Bass fillets contain 2% lipid, the portion
size is 85 g (3 oz), and marketable-sized fish have the same fatty
acid composition as we report herein, each portion of fish would
contain the following amounts of 20:5(n-3) + 22:6(n-3): FO =
0.43 g, STD SO = 0.28 g, LOW ALA SO = 0.30 g, SFA SO =
0.38 g, and HYD SO = 0.37 g (weight percent of FAME con-
verted to weight percent of lipid using a conversion factor of
0.93 g FAME/g fish lipid [Weihrauch et al. 1977]). Thus, to
consume 250 mg of 20:5(n-3) + 22:6(n-3) per day, consumers
would have to eat as few as ∼4 portions per week of the FO
fillets or more than 6 portions per week of the STD SO fillets.
Although this may be inconsistent with other recommendations
to “eat fish twice a week,” it is important to recognize that this
rule of thumb is based on consumption of fat-fleshed fish (i.e.,
trout and salmon), which contain significantly more lipid than
lean-fleshed fish like hybrid Striped Bass. Hybrid Striped Bass
remain an excellent source of lean protein, but their value as a
source of (n-3) LC-PUFAs is limited by their low fat content
and, potentially, the lipid sources they are fed. Although all of
the soy oils evaluated were adequate in terms of maintaining the
growth performance of hybrid Striped Bass, the SFA SO lipid
may hold some strategic advantage in terms of conserving the
fatty acid profile and the associated nutritional value of these
fish.
ACKNOWLEDGMENTS
Our sincere thanks to the Illinois Soybean Association for
supporting this research project under grant 09-ISA-35-409-3.
We also thank Archer Daniels Midland for providing the soy
oils evaluated as well as Omega Protein for the donation of the
fish oil and fish meal used in the experimental feeds. We further
thank the many students and staff of the Fisheries and Illinois
Aquaculture Center who assisted us with data collection, partic-
ularly Curtis Crouse, who also provided substantial assistance
in the preparation of fillet samples for analysis.
REFERENCES
Aranceta, J., and C. Pérez-Rodrigo. 2012. Recommended dietary reference
intakes, nutritional goals and dietary guidelines for fat and fatty acids: a
systematic review. British Journal of Nutrition 107(Supplement S2):8–22.
Bell, J. G., J. McEvoy, D. R. Tocher, F. McGhee, P. J. Campbell, and J. R.
Sargent. 2001. Replacement of fish oil with rapeseed oil in diets of Atlantic
 SOY OILS IN HYBRID STRIPED BASS FEEDS
Salmon (Salmo salar) affects tissue lipid compositions and hepatocyte fatty
acid metabolism. Journal of Nutrition 131:1535–1543.
Blanchard, G., J. G. Makombu, and P. Kestemont. 2008. Influence of different
dietary 18:3n-3/18:2n-6 ratio on growth performance, fatty acid composition
and hepatic ultrastructure in Eurasian Perch, Perca fluviatilis. Aquaculture
284:144–150.
Christie, W. W. 1982. Lipid analysis, 2nd revised edition. Pergamon,
Oxford, UK.
De Silva, S. S., D. S. Francis, and A. G. J. Tacon. 2011. Fish oils in aquaculture:
in retrospect. Pages 1–20 in G. M. Turchini, W.-K. Ng, and D. R. Tocher,
editors. Fish oil replacement and alternative lipid sources in aquaculture feeds.
CRC Press, Boca Raton, Florida.
Emken, E. A., R. O. Adlof, and R. M. Gulley. 1994. Dietary linoleic acid influ-
ences desaturation and acylation of deuterium-labeled linoleic and linolenic
acids in young adult males. Biochimica et Biophysica Acta, Lipids and Lipid
Metabolism 1213:277–288.
Fair, P. H., W. P. Williams, and T. I. J. Smith. 1993. Effect of dietary menhaden
oil on growth and muscle fatty acid composition of hybrid Striped Bass,
Morone chrysops × M. saxatilis. Aquaculture 116:171–189.
Folch, J., M. Lees, and G. H. Sloane Stanley. 1957. A simple method for
Downloaded by [Department Of Fisheries] at 19:57 08 April 2013
the isolation and purification of total lipides from animal tissues. Journal of
Biological Chemistry 226:497–509.
Fonseca-Madrigal, J., V. Karalazos, P. J. Campbell, J. G. Bell, and D. R. Tocher.
2005. Influence of dietary palm oil on growth, tissue fatty acid composi-
tions, and fatty acid metabolism in liver and intestine in Rainbow Trout
(Oncorhynchus mykiss). Aquaculture Nutrition 11:241–250.
Frøyland, L., Ø. Lie, and R. K. Berge. 2000. Mitochondrial and peroxisomal
β-oxidation capacities in various tissues from Atlantic Salmon Salmo salar.
Aquaculture Nutrition 6:85–89.
Gallagher, M. L. 1994. The use of soybean meal as a replacement for fish meal in
diets for hybrid Striped Bass (Morone saxatilis × M. chrysops). Aquaculture
126:119–127.
Gaylord, T. G., and S. D. Rawles. 2005. The modification of poultry by-product
meal for use in hybrid Striped Bass Morone chrysops × M. saxatilis diets.
Journal of the World Aquaculture Society 36:363–374.
Glencross, B. D. 2009. Exploring the nutritional demand for essential fatty acids
by aquaculture species. Reviews in Aquaculture 1:71–124.
Glencross, B. D., and G. M. Turchini. 2011. Fish oil replacement in starter,
grow-out, and finishing feeds for farmed aquatic animals. Pages 373–404 in
G. M. Turchini, W. K. Ng, and D. R. Tocher, editors. Fish oil replacement
and alternative lipid sources in aquaculture feeds. CRC Press, Boca Raton,
Florida.
Henderson, R. J. 1996. Fatty acid metabolism in freshwater fish with particular
reference to polyunsaturated fatty acids. Archives of Animal Nutrition 49:
5–22.
Henderson, R. J., and J. R. Sargent. 1985. Chain-length specificities of mitochon-
drial and peroxisomal ß-oxidation of fatty acids in livers of Rainbow Trout
(Salmo gairdneri). Comparative Biochemistry and Physiology 82B:79–85.
Huang, Y. S., R. L. Hancock, and D. F. Horrobin. 1987. Selective incorporation
of n-3 and n-6 fatty acids in essential fatty acid deficient rats in response to
short-term oil feeding. Biochemistry International 14:659–666.
Hunter, B. J., and D. C. K. Roberts. 2000. Potential impact of the fat composition
of farmed fish on human health. Nutrition Research 20:1047–1058.
Jobling, M. 2004. “Finishing” feeds for carnivorous fish and the fatty acid
dilution model. Aquaculture Research 35:706–709.
Kiessling, K. H., and A. Kiessling. 1993. Selective utilization of fatty acids in
Rainbow Trout (Oncorhynchus mykiss Walbaum) red muscle mitochondria.
Canadian Journal of Zoology 71:248–251.
Kohler, C. C. 2000. Striped Bass and hybrid Striped Bass culture. Pages 898–907
in R. R. Stickney, editor. Encyclopedia of aquaculture. Wiley, New York.
Kris-Etherton, P. M., W. S. Harris, and L. J. Appel. 2002. Fish consumption, fish
oil, omega-3 fatty acids, and cardiovascular disease. Circulation 106:2747–
2757.
303
Lands, W. E. M., M. Inoue, Y. Sugiura, and H. Okuyama. 1982. Selective
incorporation of polyunsaturated fatty acids into phosphatidylcholine by rat
liver microsomes. Journal of Biological Chemistry 257:14968–14972.
Lane, R. L., J. T. Trushenski, and C. C. Kohler. 2006. Modification of fillet
composition and evidence of differential fatty acid turnover in Sunshine Bass
Morone chrysops × M. saxatilis following change in dietary lipid source.
Lipids 41:1029–1038.
Laporte, J., and J. T. Trushenski. 2011. Growth performance and tissue fatty
acid composition of Largemouth Bass fed diets containing fish oil or blends
of fish oil and soy-derived lipids. North American Journal of Aquaculture
73:435–444.
Laporte, J., and J. T. Trushenski. 2012. Production performance, stress tolerance,
and intestinal integrity of Sunshine Bass fed increasing levels of soybean
meal. Journal of Animal Physiology and Animal Nutrition 96:513–526.
Lewis, H. A., and C. C. Kohler. 2008. Minimizing fish oil and fish meal with
plant-based alternatives in Sunshine Bass diets without negatively impact-
ing growth and muscle fatty acid profile. Journal of the World Aquaculture
Society 39:573–585.
Lie, Ø., E. Lied, and G. Lambertsen. 1987. Lipid digestion in cod (Gadus
morhua). Comparative Biochemistry and Physiology 88B:697–700.
Mulligan, B. L., and J. T. Trushenski. 2013. Use of standard or modi-
fied plant-derived lipids as alternatives to fish oil in feeds for juvenile
Nile Tilapia. Journal of Aquatic Food Product Technology 22:47–57. DOI:
10.1080/10498850.2011.623336.
Nematipour, G. R., and D. M. Gatlin III. 1993. Effects of different kinds of
dietary lipid on growth and fatty acid composition of juvenile Sunshine Bass,
Morone chrysops ♀× M. saxatilis ♂. Aquaculture 114:141–154.
Ng, W. K., T. Sigholt, and J. G. Bell. 2004. The influence of environmental
temperature on the apparent nutrient and fatty acid digestibility in Atlantic
Salmon (Salmo salar L.) fed finishing diets containing different blends of
fish oil, rapeseed oil and palm oil. Aquaculture Research 35:1228–1237.
Olsen, R. E., R. J. Henderson, and E. Ringø. 1998. The digestion and selective
absorption of dietary fatty acids in Arctic Charr, Salvelinus alpinus. Aqua-
culture Nutrition 4:13–21.
Pauly, D., V. Christensen, J. Dalsgaard, R. Froese, and F. Torres Jr. 1998. Fishing
down marine food webs. Science 279:860–863.
Pérez, J. A., C. Rodrı́guez, and R. J. Henderson. 1999. The uptake and esterifica-
tion of radiolabelled fatty acids by enterocytes isolated from Rainbow Trout
(Oncorhynchus mykiss). Fish Physiology and Biochemistry 20:125–134.
Rawles, S. D., and D. M. Gatlin III. 1998. Carbohydrate utilization in Striped
Bass (Morone saxatilis) and Sunshine Bass (M. chrysops ♀ × M. saxatilis
♂). Aquaculture 161:201–212.
Robin, J. H., C. Regost, J. Arzel, and S. J. Kaushik. 2003. Fatty acid profile
of fish following a change in dietary fatty acid source: model of fatty acid
composition with a dilution hypothesis. Aquaculture 225:283–293.
Rodrı́guez, C., J. A. Pérez, and R. J. Henderson. 2002. The esterification and
modification of n-3 and n-6 polyunsaturated fatty acids by hepatocytes and
liver microsomes of turbot (Scophthalmus maximus). Comparative Biochem-
istry and Physiology 132B:559–570.
Rosenlund, G., G. Corraze, M. Izquierdo, and B. E. Torstensen. 2011. The effects
of fish oil replacement on nutritional and organoleptic qualities of farmed fish.
Pages 487–522 in G. M. Turchini, W. K. Ng, and D. R. Tocher, editors. Fish
oil replacement and alternative lipid sources in aquaculture feeds. CRC Press,
Boca Raton, Florida.
Siscovick, D. S., T. E. Raghunathan, I. King, S. Weinmann, V. E. Bovbjerg, L.
Kushi, L. A. Cobb, M. K. Copass, B. M. Psaty, R. Lemaitre, B. Retzlaff, and
R. H. Knopp. 2000. Dietary intake of long-chain n-3 polyunsaturated fatty
acids and the risk of primary cardiac arrest. American Journal of Clinical
Nutrition 71:208S–212S.
Torstensen, B. E., Ø. Lie, and L. Frøyland. 2000. Lipid metabolism and tissue
composition in Atlantic Salmon (Salmo salar L.): effects of capelin oil, palm
oil, and oleic acid–enriched sunflower oil as dietary lipid sources. Lipids
35:653–664.
 304 TRUSHENSKI AND KANCZUZEWSKI
Trushenski, J. T. 2009. Saturated lipid sources in feeds for Sunshine Bass:
alterations in production performance and tissue fatty acid composition. North
American Journal of Aquaculture 71:363–373.
Trushenski, J. T., P. Blaufuss, B. Mulligan, and J. Laporte. 2011a. Growth
performance and tissue fatty acid composition of Rainbow Trout reared on
feeds containing fish oil or equal blends of fish oil and traditional or novel
alternative lipids. North American Journal of Aquaculture 73:194–203.
Trushenski, J. T., J. Boesenberg, and C. C. Kohler. 2009. Influence of grow-
out feed fatty acid composition on finishing success in Nile Tilapia. North
American Journal of Aquaculture 71:242–251.
Trushenski, J. T., B. Gause, and H. A. Lewis. 2011b. Selective fatty acid
metabolism, not the sequence of dietary fish oil intake, prevails in fillet fatty
acid profile change in Sunshine Bass. North American Journal of Aquaculture
73:204–211.
Trushenski, J. T., H. A. Lewis, and C. C. Kohler. 2008a. Fatty acid profile of
Sunshine Bass: I. profile change is affected by initial composition and differs
among tissues. Lipids 43:629–641.
Trushenski, J. T., H. A. Lewis, and C. C. Kohler. 2008b. Fatty acid profile of
Sunshine Bass: II. profile change differs among fillet lipid classes. Lipids
43:643–653.
Downloaded by [Department Of Fisheries] at 19:57 08 April 2013
Trushenski, J. T., and R. T. Lochmann. 2009. Potential, implications and so-
lutions regarding the use of rendered animal fats in aquafeeds. American
Journal of Animal and Veterinary Sciences 4:108–128.
Trushenski, J. T., B. Mulligan, D. Jirsa, and M. Drawbridge. 2013. Sparing fish
oil with soybean oil in feeds for White Seabass: effects of inclusion rate and
soybean oil composition. North American Journal of Aquaculture 75:xx–xx.
Turano, M. 2011. 2011 U.S. hybrid Striped Bass industry update. Presentation
given at Aquaculture America, New Orleans, Louisiana. World Aquaculture
Society, Baton Rouge, Louisiana.
Turano, M. 2012. 2012 U.S. hybrid Striped Bass industry update. Presenta-
tion given at Aquaculture America, Las Vegas, Nevada. World Aquaculture
Society, Baton Rouge, Louisiana.
Turchini, G. M., D. S. Francis, and S. S. De Silva. 2006. Modification of tissue
fatty acid composition in Murray Cod (Maccullochella peelii peelii, Mitchell)
resulting from a shift from vegetable oil diets to a fish oil diet. Aquaculture
Research 37:570–585.
Turchini, G. M., D. S. Francis, S. P. S. D. Senadheera, T. Thanuthong, and S. S.
De Silva. 2011a. Fish oil replacement with different vegetable oils in Murray
Cod: evidence of an “omega-3 sparing effect” by other dietary fatty acids.
Aquaculture 315:250–259.
Turchini, G. M., W. K. Ng, and D. R. Tocher, editors. 2011b. Fish oil replacement
and alternative lipid sources in aquaculture feeds. CRC Press, Boca Raton,
Florida.
Turchini, G. M., B. E. Torstensen, and W. K. Ng. 2009. Fish oil replacement in
finfish nutrition. Reviews in Aquaculture 1:10–57.
USDA (U.S. Department of Agriculture). 2010. Soybeans and oil crops:
background. USDA, Economic Research Service, Washington, D.C.
Available: www.ers.usda.gov/Briefing/SoybeansOilcrops/background.htm.
(March 2010).
Wall, R., R. P. Ross, G. F. Fitzgerald, and C. Stanton. 2010. Fatty acids from fish:
the anti-inflammatory potential of long-chain omega-3 fatty acids. Nutrition
Reviews 68:280–289.
Weihrauch, J. L., L. P. Posati, B. A. Anderson, and J. Exler. 1977. Lipid con-
version factors for calculating fatty acid contents of foods. Journal of the
American Oil Chemists’ Society 54:36–40.
Wonnacott, E. J., R. L. Lane, and C. C. Kohler. 2004. Influence of dietary
replacement of menhaden oil with canola oil on fatty acid composition of
Sunshine Bass. North American Journal of Aquaculture 66:243–250.
This article was downloaded by: [Department Of Fisheries]
On: 08 April 2013, At: 19:57
Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,
37-41 Mortimer Street, London W1T 3JH, UK

North American Journal of Aquaculture

Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/unaj20

Sparing Fish Oil with Soybean Oil in Feeds for White

Seabass: Effects of Inclusion Rate and Soybean Oil
Composition
a a b b
Jesse Trushenski , Bonnie Mulligan , David Jirsa & Mark Drawbridge
a
Fisheries and Illinois Aquaculture Center and Department of Zoology, Southern Illinois
University–Carbondale, 1125 Lincoln Drive, Carbondale, Illinois, 62901, USA
b
Hubbs-SeaWorld Research Institute, 2595 Ingraham Street, San Diego, California, 92109,
USA
Version of record first published: 02 Apr 2013.

To cite this article: Jesse Trushenski , Bonnie Mulligan , David Jirsa & Mark Drawbridge (2013): Sparing Fish Oil with
Soybean Oil in Feeds for White Seabass: Effects of Inclusion Rate and Soybean Oil Composition, North American Journal of
Aquaculture, 75:2, 305-315

To link to this article: http://dx.doi.org/10.1080/15222055.2012.720650

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions

This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to
anyone is expressly forbidden.

The publisher does not give any warranty express or implied or make any representation that the contents
will be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses should
be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims,
proceedings, demand, or costs or damages whatsoever or howsoever caused arising directly or indirectly in
connection with or arising out of the use of this material.
 North American Journal of Aquaculture 75:305–315, 2013

C American Fisheries Society 2013

ISSN: 1522-2055 print / 1548-8454 online

DOI: 10.1080/15222055.2012.720650

SPECIAL SECTION: LIPIDS IN AQUACULTURE NUTRITION AND PHYSIOLOGY II

Sparing Fish Oil with Soybean Oil in Feeds for White

Seabass: Effects of Inclusion Rate and Soybean Oil
Composition

Jesse Trushenski* and Bonnie Mulligan

Fisheries and Illinois Aquaculture Center and Department of Zoology,
Southern Illinois University–Carbondale, 1125 Lincoln Drive, Carbondale, Illinois 62901, USA

David Jirsa and Mark Drawbridge

Hubbs-SeaWorld Research Institute, 2595 Ingraham Street, San Diego, California 92109, USA
Downloaded by [Department Of Fisheries] at 19:57 08 April 2013

Abstract
Fish oil sparing has proven difficult for some fish species, especially marine carnivores like White Seabass Atrac-
toscion nobilis that require one or more long-chain polyunsaturated fatty acids (LC-PUFAs). Recent studies have
suggested that the use of saturated fatty acid (SFA)–rich lipids instead of C18 polyunsaturated fatty acid–rich (C18
PUFA) lipids may be advantageous in maintaining tissue levels of LC-PUFAs; SFA-rich lipids may also offer a strate-
gic advantage in terms of meeting the LC-PUFA requirements of marine carnivores while minimizing dietary fish
oil inclusion. Accordingly, we assessed the performance and tissue fatty acid composition of White Seabass (3.8 ±
0.2 g [mean ± SE]) fed diets containing fish oil or graded levels of C18 PUFA–rich standard soy oil or SFA-rich
hydrogenated soy oil (replacing 25, 50, 75, or 100% of dietary fish oil) for 8 weeks. Feed conversion ratio, weight gain,
and specific growth rate were not impaired by partial or complete replacement of dietary fish oil with hydrogenated
soy oil; however, fish oil sparing with standard soy oil was associated with declining performance. The tissue fatty
acid profiles of fish fed the hydrogenated soy oil–based diets were very similar to those of fish fed the fish oil–based
feed, but the standard soy oil–based feeds resulted in concomitant loss of n-3 fatty acids and LC-PUFAs. In all cases,
the magnitude of the dietary effect was greater among liver and fillet tissues than among brain and eye tissues. These
data suggest a limitation, potentially related to LC-PUFA deficiency, associated with replacing fish oil with standard
soybean oil, but not with hydrogenated soybean oil. Our data suggest that the LC-PUFA requirements of White
Seabass can be effectively reduced by feeding SFA-rich alternative lipids, allowing for a greater level of fish oil sparing
without growth impairment or tissue profile modification than is possible with C18 PUFA–rich lipids.

White Seabass Atractoscion nobilis are a highly valued com-
mercial and sport fish in southern California and are consid-
ered an excellent fish for aquaculture. White Seabass have been
cultured by Hubbs-SeaWorld Research Institute (HSWRI) un-
der contract to the California Department of Fish and Game
for stock enhancement in Southern California waters as part
of the Ocean Resources Enhancement and Hatchery Program
since 1986. In recent years, interest in commercializing the cul-
ture of the species has intensified, particularly in the context
of offshore aquaculture. Within the existing stock enhancement
project, White Seabass are typically cultured to an average size
of 20–25 cm prior to release. During this period they are reared
on commercially available diets formulated for other species
because little work has been devoted to the development of
practical diets specifically tailored to this species. There is lim-
ited published information on the nutritional requirements of
this species: López et al. (2006, 2009) evaluated the perfor-
mance of White Seabass fed graded levels of lipid in the context
of high protein (61–67%) diets, whereas Durazo et al. (2010)
focused on estimating the protein demand in diets containing

*Corresponding author: saluski@siu.edu

Received April 16, 2012; accepted July 30, 2012
Published online April 2, 2013
305
 306 TRUSHENSKI ET AL.
a moderate amount of lipid (14–15%). Although little is cur-
rently known about the nutrient requirements of this species,
researchers are already beginning to investigate alternatives to
fish meal and fish oil as ingredients in the feeds used in White
Seabass culture. For example, Jirsa et al. (2010) evaluated the
inclusion of soy meals as the protein source in practical diets for
White Seabass juveniles, and research assessing fish oil sparing
has been initiated by other collaborating teams (Jirsa et al., in
press).
Fish oil sparing has proven particularly difficult for species
that require one or more of the long-chain polyunsaturated fatty
acids (LC-PUFAs), which are abundant in fish oil but lacking
in most of the fats and oils commonly used to spare fish oil
in aquafeeds. Carnivorous fishes, especially marine carnivores,
appear particularly demanding in this context, some reportedly
requiring nearly 4% n-3 LC-PUFAs in the diet (Glencross 2009;
NRC 2011), though 22:6(n-3)1 may be the only limiting fatty
Downloaded by [Department Of Fisheries] at 19:57 08 April 2013
acid for some species (Trushenski et al. 2012).
Unless n-3 LC-PUFA requirements are met via other dietary
sources (e.g., fish meal), complete or near-complete fish oil
replacement generally impairs performance in marine species,
including sparids, salmonids, serranids, and pleuronectiforms,
and results in significantly modified tissue fatty acid profiles
reflecting diminished LC-PUFA intake (Glencross 2009;
Turchini et al. 2009). Recent studies have suggested that the
use of saturated fatty acid (SFA)–rich lipids instead of C18
polyunsaturated fatty acid (C18 PUFA)–rich lipids may be
advantageous in preventing the loss of LC-PUFAs from the
tissues (Laporte and Trushenski 2011; Trushenski et al. 2011c;
Ramezani-Fard et al. 2012) as well as in increasing tissue plas-
ticity and LC-PUFA augmentation during finishing (Trushenski
et al. 2008a, 2009, 2011b; Turchini et al. 2011). In these studies,
fish fed SFA-rich diets deposited or retained more LC-PUFAs
within their tissues than fish fed C18 PUFA–rich diets. It is pos-
sible that SFA-rich alternative lipids may also offer a strategic
advantage in terms of meeting the LC-PUFA requirements of
marine carnivores while minimizing dietary fish oil inclusion.
For example, it might be possible to meet LC-PUFA require-
ments at lower dietary concentrations if these nutrients are
provided in the context of SFA-rich feeds. Accordingly, we
assessed the survival, growth performance, and tissue fatty acid
composition of juvenile White Seabass fed graded levels of
C18 PUFA–rich soy oil or SFA-rich hydrogenated soy oil as
alternatives to fish oil.
METHODS
Feed preparation and analyses.—A practical feed (Table 1)
including approximately 400 g/kg menhaden fish meal and
71 g/kg menhaden fish oil (FISH ONLY; Virginia Gold, Omega
1In fatty acid designations of this nature, the number to the left of the colon
is the number of carbon atoms in the compound, the number immediately to
the right of the colon is the number of double bonds, and the number after the
hyphen indicates the position of the first double bond from the methyl end.
Protein, Inc., Houston, Texas) and previously verified in White
Seabass culture (Jirsa et al. 2010), served as the positive control
feed in the present work. Experimental feeds were derived from
the FISH ONLY formulation using standard (STD SO) or fully
hydrogenated (SFA SO) soybean oil (Archer Daniels Midland
Company, Decatur, Illinois) replacing 25, 50, 75, or 100% of the
fish oil included in the control formulation. Standard soybean oil
is comprised largely of C18 PUFAs, containing approximately
16% SFAs (primarily 16:0 and 18:0), 23% monounsaturated
fatty acids (MUFAs; primarily 18:1[n-7] and 18:1[n-9]), and
61% C18 PUFAs (primarily 18:2[n-6]) (J. T. Trushenski, unpub-
lished data). Hydrogenation progressively reduces the degree
of fatty acid unsaturation, transforming PUFAs to MUFAs and
MUFAs to SFAs. Consequently, the fully hydrogenated soybean
oil is comprised almost exclusively of SFAs, containing approx-
imately 98% SFAs (primarily 16:0 and 18:0; J. T. Trushenski,
unpublished data). Feeds were prepared at the Fisheries and
Illinois Aquaculture Center (Carbondale, Illinois) according to
standard in-house feed manufacturing practices. All feedstuffs
were incorporated using a cutter–mixer (Model CM450; Hobart
Corporation, Troy, Ohio), pelleted using a commercial-grade
food grinder (1.5-hp [1 hp = 746 W] electric grinder; Cabela’s,
Sidney, Nebraska), and dried in a commercial-grade food de-
hydrator (100◦ C; Harvest Saver R-5A; Commercial Dehydrator
Systems, Inc., Eugene, Oregon) to approximately 936 g/kg dry
matter. The hydrogenated soy lipid was a solid at room temper-
ature but was slowly melted using a consumer-grade microwave
prior to addition to the other feedstuffs. All feeds were packaged
in plastic bags and shipped to the Hubbs-SeaWorld Research
Institute (HSWRI; Carlsbad, California) and stored frozen
(–20◦ C) until needed. Prior to shipping, triplicate diet samples
were collected and analyzed to confirm their proximate (Table 1)
and fatty acid composition (Table 2). Samples were lyophilized
(Freezone 6; Labconco Corporation, Kansas City, Missouri) to
determine moisture content and prevent the changes to lipid
content or composition that can occur during oven drying. The
dried samples were then pulverized. Protein (LECO FP-528;
LECO Corporation, St. Joseph, Michigan) and ash (muffle fur-
nace; 600◦ C for 3 h) content were determined for each pulver-
ized sample, and lipid content was determined gravimetrically
following chloroform–methanol extraction (Folch et al. 1957).
Predictably, dietary fatty acid composition reflected the rela-
tive contributions of fish oil and the soy-derived lipids, with the
STD SO series containing levels of C18 PUFAs proportional to
the standard soy oil content and the SFA SO series containing
levels of SFAs proportional to hydrogenated soy oil content.
Minor differences in dietary macronutrient composition were
noted; these were attributed to small inaccuracies in feedstuff
weighing and/or incomplete mixing during feed manufacturing
or sample collection. In one diet, 100% STD SO, protein was
markedly lower (∼40% versus ∼45% in other diets); the cause
of this discrepancy is unclear, but the implications of reduced
protein content in this feed are addressed in the Discussion.
Reserved crude lipid samples were subjected to acid-catalyzed
 SOY OILS IN WHITE SEABASS FEEDS 307

TABLE 1. Formulation and proximate composition of diets fed to White Seabass to assess the effects of fish oil sparing with different soy-derived lipids.

100% 25% 50% 75% 100% 25% 50% 75% 100%

Ingredient FO STD SO STD SO STD SO STD SO SFA SO SFA SO SFA SO SFA SO
Formulation (g/kg; as-fed basis)
Menhaden fish meala 400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0 400.0
Soy protein concentrateb 190.0 190.0 190.0 190.0 190.0 190.0 190.0 190.0 190.0
Corn starch 118.8 118.8 118.8 118.8 118.8 118.8 118.8 118.8 118.8
Wheat flour 150.0 150.0 150.0 150.0 150.0 150.0 150.0 150.0 150.0
Menhaden fish oila 71.0 53.2 35.5 17.8 0.0 53.2 35.5 17.8 0.0
Standard soybean oilb 0.0 17.8 35.5 53.2 71.0 0.0 0.0 0.0 0.0
Hydrogenated soybean oilb 0.0 0.0 0.0 0.0 0.0 17.8 35.5 53.2 71.0
Soy lecithinb 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0
Carboxymethyl cellulose 20.0 20.0 20.0 20.0 20.0 20.0 20.0 20.0 20.0
Choline chloride 6.0 6.0 6.0 6.0 6.0 6.0 6.0 6.0 6.0
Stay-Cc 2.0 2.0 2.0 2.0 2.0 2.0 2.0 2.0 2.0
Vitamin premixd 1.2 1.2 1.2 1.2 1.2 1.2 1.2 1.2 1.2
Downloaded by [Department Of Fisheries] at 19:57 08 April 2013

Mineral premixe 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0
Corn gluten meal 30.0 30.0 30.0 30.0 30.0 30.0 30.0 30.0 30.0
Proximate compositionf (g/kg; dry-matter basis [except dry matter])
Dry matter 931 ± 1 935 ± 1 935 ± 1 932 ± 1 924 ± 1 941 ± 1 943 ± 1 942 ± 1 940 ± 1
Protein 456 ± 12 435 ± 12 461 ± 12 450 ± 12 409 ± 12 454 ± 12 459 ± 12 460 ± 12 441 ± 12
Lipid 122 ± 10 126 ± 10 124 ± 10 127 ± 10 112 ± 10 125 ± 10 140 ± 12 104 ± 10 115 ± 10
Ash 101 ± 6 103 ± 6 108 ± 6 89 ± 6 95 ± 6 104 ± 6 98 ± 6 97 ± 6 102 ± 6
a
Omega Protein, Inc., Houston, Texas.
b
Archer Daniels Midland Co., Decatur, Illinois.
c
Argent Laboratories, Redmond, Washington.
d
Formulated to contain 25.000% L-ascorbyl-2-polyphosphate, 14.000% RRR-alpha tocopheryl acetate, 13.160% vitamin K, 12.500% inositol, 12.500% nicotinic acid, 7.500%
riboflavin, 6.250% calcium pantothenate, 2.500% pyridoxine hydrochloride, 1.250% thiamine mononitrate, 1.000% vitamin A palmitate, 0.500% cyanocobalamin, 0.450% folic acid,
0.125% biotin, and 0.010% cholecalciferol in a cellulose base.
e
Formulated to contain 24.897% zinc oxide, 14.933% ferrous sulfate, 3.470% manganese oxide, 0.967% cupric carbonate, 0.262% potassium iodide, 0.060% sodium selenate, and
0.030% cobalt carbonate in a cellulose base.
f
Values are least-squares means ± pooled SEs of triplicate samples.

transmethylation performed overnight at 50◦ C (Christie 1982).
The resultant fatty acid methyl esters (FAMEs) were separated
using a gas chromatograph equipped with a flame-ionization
detector fitted with a permanently bonded polyethylene gly-
col, fused silica capillary column (Omegawax 250; 30 m ×
0.25 mm interior diameter, 0.25 µm film; Supelco, Bellefonte,
Pennsylvania). The injection volume was 1.0 µL, helium was
the carrier gas (30 cm/s; 205◦ C), and the injector temperature
was 250◦ C. A split injection technique (100:1) was used, and the
temperature program was as follows: 50◦ C for 2 min, increased
to 220◦ C at 4◦ C/min, and 220◦ C for 15 min. Individual FAMEs
were identified by reference to external standards (Supelco 37
Component FAME Mix, PUFA-1, and PUFA-3; Supelco).
Experimental design and feeding trial.—The feeding trial
was conducted at HSWRI’s marine fish hatchery in Carlsbad,
California. The culture system utilized for these trials was a
semiclosed recirculating system consisting of forty 60-L-square
culture tanks, a water pump, supplemental aeration (provided
using a central line, regenerative blower, and air diffusers) as
well as mechanical and biological filtration. A small amount
(<2 L/min) of ozonated seawater was continuously added to
the system for water exchange. Water temperature was con-
trolled with a heat exchanger (Aqua Logic, Inc., San Diego,
California). Tanks were siphoned daily to remove solids. Tem-
perature, dissolved oxygen, and salinity were measured daily,
and pH was measured biweekly using a YSI optical ODO multi-
probe meter (YSI, Inc., Yellow Springs, Ohio). Total ammonia-,
nitrite- and nitrate-nitrogen were quantified weekly by spec-
trophotometric analysis (HACH; Hach, Inc., Loveland, Col-
orado). Prior to stocking the system and beginning the trial,
all fish had been fed a commercially available marine fish feed
(Gemma Diamond; Skretting Global, Stavanger, Norway); all
fish were fed the 100% FO control feed for 2 weeks to ac-
climate them to the new feed. Fifteen juvenile fish (3.8 ±
0.2 g [mean ± SE]) were stocked into each of 36 tanks, and
each tank was randomly assigned one of the experimental feeds
(4 tanks/feed; N = 4) for the 8-week feeding trial. Fish were
fed by percent body weight at a rate slightly in excess of the
best-feeding tanks using Fish Mate F14 automatic feeders (Pet
Mate, Ltd., Hersham, Surrey, England), which divided the ra-
tions into five feedings per day. Throughout the trial, photope-
riod was kept on a 14.5 light : 9.5 dark cycle, and all water
quality parameters were maintained within ranges suitable for
White Seabass culture: temperature = 19.9 ± 0.4◦ C, dissolved
 308 TRUSHENSKI ET AL.

TABLE 2. Fatty acid composition (g/100 g fatty acid methyl esters) of diets fed to White Seabass to assess the effects of fish oil sparing with different soy-derived
lipids. Data are least-squares means of triplicate samples; pooled SE estimates < 0.1 are reported as 0.0.

100% 25% 50% 75% 100% 25% 50% 75% 100% Pooled
Fatty acid(s) FO STD SO STD SO STD SO STD SO SFA SO SFA SO SFA SO SFA SO SE
14:0 8.2 6.6 5.4 4.2 2.9 7.0 5.5 4.5 2.8 0.1
15:0 0.8 0.6 0.5 0.4 0.3 0.7 0.5 0.5 0.3 0.0
16:0 20.9 19.2 18.3 17.4 16.5 20.1 18.8 19.0 17.5 0.3
17:0 0.6 0.5 0.5 0.4 0.3 0.6 0.5 0.5 0.4 0.0
18:0 4.1 4.2 4.4 4.7 4.9 16.0 28.1 35.7 51.2 0.4
20:0 0.3 0.3 0.3 0.3 0.3 0.3 0.4 0.4 0.5 0.0
SFAsa 35.0 31.4 29.3 27.4 25.3 44.7 53.8 60.6 72.7 0.7
16:1(n-7) 10.3 8.2 6.6 5.0 3.4 8.6 6.7 5.4 3.2 0.1
18:1(n-7) 3.1 2.8 2.5 2.3 2.1 2.7 2.1 1.8 1.2 0.0
18:1(n-9) 7.9 10.2 12.9 15.2 17.6 6.9 5.7 5.1 3.9 0.2
20:1(n-9) 0.8 0.6 0.6 0.5 0.4 0.6 0.5 0.4 0.3 0.0
MUFAsb 22.1 21.9 22.6 23.0 23.4 18.8 15.1 12.8 8.6 0.3
Downloaded by [Department Of Fisheries] at 19:57 08 April 2013

16:2(n-4) 1.2 0.9 0.7 0.5 0.3 1.0 0.8 0.6 0.3 0.0
16:3(n-4) 1.1 1.1 0.8 0.6 0.3 1.1 0.8 0.7 0.4 0.1
18:2(n-6) 6.8 14.8 22.0 28.8 35.5 6.3 6.0 6.0 5.8 0.1
20:4(n-6) 1.2 1.0 0.8 0.6 0.4 1.0 0.8 0.7 0.5 0.0
n-6c 8.0 15.8 22.8 29.4 36.0 7.3 6.8 6.6 6.3 0.1
18:3(n-3) 1.8 2.8 3.5 4.3 5.0 1.6 1.4 1.3 1.0 0.1
18:4(n-3) 2.6 2.1 1.7 1.2 0.8 2.1 1.7 1.4 0.8 0.0
20:4(n-3) 1.3 1.1 0.9 0.6 0.4 1.1 0.9 0.7 0.4 0.0
20:5(n-3) 12.6 10.6 8.0 5.8 3.6 10.3 8.5 6.7 3.9 0.3
22:5(n-3) 2.3 2.0 1.5 1.0 0.6 1.9 1.6 1.3 0.7 0.1
22:6(n-3) 12.0 10.4 8.2 6.2 4.2 10.0 8.7 7.3 4.8 0.4
n-3d 32.7 28.9 23.7 19.1 14.6 27.1 22.8 18.7 11.7 0.9
PUFAse 42.9 46.7 48.1 49.6 51.2 36.4 31.2 26.6 18.7 0.9
C18 PUFAsf 11.3 19.7 27.2 34.3 41.3 10.1 9.1 8.6 7.6 0.1
LC-PUFAsg 29.4 25.0 19.3 14.2 9.2 24.4 20.4 16.7 10.3 0.8
a
Sum of all fatty acids without double bonds.
b
Sum of all fatty acids with a single double bond.
c
Sum of all n-6 fatty acids.
d
Sum of all n-3 fatty acids.
e
Sum of all fatty acids with ≥2 double bonds.
f
Sum of all PUFAs with chain lengths of 18 carbon atoms.
g
Sum of all fatty acids with chain length ≥20 carbon atoms and double bonds ≥3.

oxygen = 6.1 ± 0.5 mg/L, pH = 7.7 ± 0.2, total ammonia = Specific Growth Rate (SGR, % body weight/day) = 100
0.01 ± 0.02 mg/L, nitrite = 0.0 ± 0.0 mg/L, nitrate = 1.1 ± log average final weight − loge average initial weight
× e
2.8 mg/L, and salinity = 33.4 ± 0.2 g/L. days of feeding
Growth performance, sample collection, and analysis.—At
the end of the feeding trial, fish were group-weighed by tank
Three fish/tank were then euthanized by an overdose of tricaine
and counted in order to calculate the following standard metrics
methanesulfonate (MS-222; ∼200 mg/L in culture water, with
of production performance:
fish being exposed to the solution until cessation of opercular
movement [∼5 min]), frozen, and shipped to the Fisheries and
Weight Gain (%) = 100
Illinois Aquaculture Center for further processing. Upon deliv-
(average individual final weight − average individual initial weight)
× ery, the carcasses were dissected to remove liver, white muscle,
average individual initial weight
eye, and brain tissue samples. All tissue samples were stored at
Feed Conversion Ratio (FCR) –80◦ C prior to analysis.
average individual dry matter feed intake Tissue fatty acid analysis.—Muscle tissue samples were
=
average individual weight gain lyophilized, pulverized, and extracted following the procedures
 SOY OILS IN WHITE SEABASS FEEDS 309

described previously for feed samples. Intact brain, eye, and
liver samples were homogenized in the extraction solvent mix-
ture using a tissue homogenizer (PowerGen Model 1000; Fisher
Scientific, Waltham, Massachusetts) and then extracted in the
same manner as other samples. Fillet samples were processed
in the same manner as diet samples because incomplete sam-
ple homogenization and cross-contamination commonly occurs
when attempting to process intact muscle samples in solvent
using a tissue homogenizer. All other preparatory and analytical
procedures were as described previously. Coefficient of dis-
tance (Djh) values (Turchini et al. 2006) comparing the overall
fatty acid profiles between the experimental treatments and the
100% FO control group were calculated for each tissue type as
follows:
 1/2

n
Djh = (Pij − Pih) 2
,
Downloaded by [Department Of Fisheries] at 19:57 08 April 2013
oil replacement level as main effects and whether there was a
significant interactive effect of these factors on performance.
Among the parameters assessed, weight gain appeared to be the
most sensitive to treatment effects. Accordingly, weight gain and
dietary LC-PUFA content data were subjected to single-slope,
broken-line regression analyses (PROC NLIN) as described by
Robbins et al. (2006) in order to identify minimum LC-PUFA
inclusion levels. For all statistical procedures, differences were
considered significant at P < 0.05. Coefficient of distance data
were determined using mean fatty acid composition data, yield-
ing a single value for each dietary treatment within each tissue
type. Accordingly, these data were not analyzed using formal
statistical procedures, though linear regressions were used to
qualitatively describe the relationships between dietary fish oil
content and Djh values.

i=1
where Pij is the percent content of fatty acid i in the control
treatment (100% FO) and Pih is the percent content of fatty
acid i in an experimental treatment. The total number of fatty
acids included in each calculation (n) varied among tissue types,
but in all cases only major individual fatty acids (>2% of total
quantified FAMEs; no fatty acid groupings, e.g., SFAs, MUFAs)
were used in the calculation of Djh.
Statistical analysis.—Although each tank represented mul-
tiple individual fish, replicate tanks served as the experimen-
tal units for all statistical analyses (N = 4). All production
performance and fatty acid composition data were analyzed
by one-way analysis of variance (ANOVA; ANOVA, PROC
MIXED) using the Statistical Analysis System (version 9.1,
SAS Institute, Cary, North Carolina). For parameters exhibiting
significant treatment effects, Tukey’s honestly significant dif-
ference pairwise comparison tests were used to determine the
significance of differences among means. Additionally, produc-
tion data from the experimental groups (excluding the 100%
FO control group) were analyzed by two-way ANOVA (PROC
MIXED) to determine the significance of soy oil type and fish
RESULTS
Production performance of White Seabass was not impaired
by partial or complete replacement of dietary fish oil with hy-
drogenated soy oil; in the case of FCR, weight gain, and SGR,
performance was marginally improved among fish fed the SFA
SO feeds (Table 3). However, fish oil sparing with standard soy
oil was associated with declining growth and increasing FCR.
Survival was unaffected by dietary oil source. These trends were
supported by the results of the two-way ANOVA, which revealed
significant effects of soy oil type (P < 0.001; reduced perfor-
mance in the STD SO series compared with the SFA SO series
across fish oil replacement levels), but not fish oil replacement
level (P = 0.159–0.226; no consistent effect of fish oil replace-
ment level across soy oil types), on FCR, weight gain, and SGR.
Significant interaction effects were also observed for these pa-
rameters (P = 0.020–0.032), indicating decreasing performance
among fish fed increasing amounts of standard soybean oil but
not hydrogenated soybean oil.
Although the fillet lipid content did not vary, the fatty acid
composition of the fillets and other tissues was significantly in-
fluenced by dietary treatment (Table 4; Figure 1). Compared
with the FO feed, the STD SO diets contained higher levels of

TABLE 3. Production performance of White Seabass fed diets containing fish oil or soy-derived lipids. Values are least-squares means. The P-values resulting
from one-way ANOVA tests are also provided; within rows, means with common lowercase letters are not significantly different (P > 0.05).

100% 25% 50% 75% 100% 25% 50% 75% 100% Pooled
Parameter FO STD SO STD SO STD SO STD SO SFA SO SFA SO SFA SO SFA SO SE P-value
Survival (%) 93 98 100 98 98 100 100 95 97 3 0.147
Initial weight (g) 3.7 3.8 3.8 3.8 3.8 3.9 3.8 3.7 3.8 0.1 0.049a
Final weight (g) 25.5 xy 26.3 yz 25.0 xy 24.4 xy 23.0 x 28.7 z 28.7 z 28.7 z 28.4 z 0.9 <0.001
Weight gain (%) 582 xy 596 xyz 549 wx 544 wx 508 w 638 yz 649 yz 674 z 654 yz 24 <0.001
FCR 1.24 xyz 1.24 xyz 1.28 yz 1.31 z 1.33 z 1.18 xy 1.18 x 1.15 x 1.16 x 0.03 <0.001
SGR (% body
weight/day) 3.43 xy 3.46 xyz 3.34 wx 3.22 wx 3.22 w 3.56 yz 3.60 yz 3.66 z 3.60 yz 0.06 <0.001
a
Although one-way ANOVA indicated a significant treatment effect, more conservative Tukey’s honestly significant difference pairwise comparison tests indicated no significant
differences among means.
 310 TRUSHENSKI ET AL.

TABLE 4. Fillet lipid content and fatty acid composition (g/100 g) of White Seabass fed diets containing fish oil or soy-derived lipids. Values are least-squares
means of the major fatty acids (>1% of total fatty acid methyl esters). The P-values resulting from one-way ANOVA tests are also provided; within rows, means
with common lowercase letters are not significantly different (P > 0.05).

Fatty 100% 25% 50% 75% 100% 25% 50% 75% 100% Pooled
acid(s) FO STD SO STD SO STD SO STD SO SFA SO SFA SO SFA SO SFA SO SE P-value
14:0 2.3 z 1.9 yz 1.3 wx 1.2 vw 0.7 v 2.1 z 1.9 xyz 1.5 wxy 1.3 w 0.2 <0.001
16:0 22.6 z 21.7 yz 20.7 xy 19.7 wx 18.5 w 22.5 z 22.5 z 21.8 yz 20.8 xy 0.3 <0.001
18:0 8.9 9.0 9.2 9.2 9.5 9.1 9.5 9.8 9.1 0.3 0.047i
SFAsa 34.1 z 32.9 z 31.5 y 30.3 xy 29.0 x 34.0 z 34.2 z 33.2 z 31.4 y 0.5 <0.001
16:1(n-7) 4.9 z 3.8 wxy 2.6 uv 2.3 tu 1.5 t 4.5 yz 4.2 xyz 3.5 wx 3.3 vw 0.3 <0.001
18:1(n-7) 3.8 yz 3.4 wx 3.2 w 2.9 v 2.8 v 3.7 yz 3.8 z 3.8 yz 3.5 xy 0.1 <0.001
18:1(n-9) 8.7 w 9.9 wx 10.7 xy 11.9 yz 12.8 z 8.7 w 9.1 wx 9.5 wx 9.9 wx 0.5 <0.001
MUFAsb 18.1 17.9 17.1 17.6 17.7 17.5 17.7 17.5 17.3 0.7 0.951
18:2(n-6) 5.7 u 11.2 w 15.8 x 22.7 y 28.2 z 5.8 u 6.7 uv 7.9 v 10.9 w 0.4 <0.001
20:4(n-6) 2.3 yz 2.1 xy 1.9 x 1.5 w 1.4 w 2.4 yz 2.3 yz 2.4 z 2.3 yz 0.1 <0.001
n-6c 8.1 u 13.3 w 17.7 x 24.2 y 29.6 z 8.1 u 9.1 u 10.3 v 13.2 w 0.3 <0.001
Downloaded by [Department Of Fisheries] at 19:57 08 April 2013

18:3(n-3) 0.8 wx 1.1 xy 1.2 y 1.8 z 1.8 z 0.8 w 0.8 wx 0.8 wx 1.1 xy 0.1 <0.001
20:5(n-3) 8.1 z 6.8 y 5.4 x 4.8 x 3.2 w 8.0 z 7.8 z 7.6 yz 7.9 z 0.3 <0.001
22:5(n-3) 2.6 z 2.3 y 2.1 y 1.8 x 1.5 w 2.6 z 2.7 z 2.7 z 2.7 z 0.1 <0.001
22:6(n-3) 26.9 z 24.7 yz 24.1 y 18.9 x 16.8 x 27.7 z 26.6 yz 26.9 yz 25.4 yz 1.0 <0.001
n-3d 39.8 z 35.9 xy 33.6 x 27.9 w 23.7 v 40.4 z 39.1 yz 39.0 yz 38.1 yz 1.0 <0.001
PUFAse 47.8 x 49.2 xy 51.3 yz 52.1 yz 53.3 z 48.5 x 48.1 x 49.3 xy 51.3 yz 0.9 <0.001
C18 PUFAsf 7.2 u 12.8 w 17.4 x 24.8 y 30.2 z 7.2 u 8.1 uv 9.2 v 12.5 w 0.5 <0.001
LC-PUFAsg 40.7 z 36.4 xy 33.9 x 27.3 w 23.1 v 41.3 z 40.0 yz 40.1 yz 38.8 yz 1.2 <0.001
Lipidh 57 55 55 61 61 55 49 47 52 5 0.120
a
Sum of all fatty acids without double bonds; includes 20:0 in addition to the individually reported SFAs.
b
Sum of all fatty acids with a single double bond; includes 20:1(n-9) in addition to the individually reported MUFAs.
c
Sum of all n-6 fatty acids.
d
Sum of all n-3 fatty acids; includes 18:4(n-3) and 20:4(n-3) in addition to the individually reported n-3 fatty acids.
e
Sum of all fatty acids with ≥2 double bonds; includes 18:4(n-3) and 20:4(n-3) in addition to the individually reported PUFAs.
f
Sum of all PUFAs with chain lengths of 18 carbon atoms; includes 18:4(n-3) in addition to the individually reported C18 PUFAs.
g
Sum of all fatty acids with chain length ≥20 carbon atoms and double bonds ≥3; includes 20:4(n-3) in addition to the individually reported LC-PUFAs.
h
Units are g/kg tissue (dry-matter basis).
i
Although the one-way ANOVA indicated a significant treatment effect, more conservative Tukey’s honestly significant difference pairwise comparison tests indicated no significant
differences among means.

C18 PUFAs, particularly in the form of 18:2(n-6), which was
reflected in the fillets and other tissues of White Seabass fed
these diets. Conversely, the SFA SO feeds contained higher
levels of SFAs, particularly 18:0, than the 100% FO feed, but
these fatty acids were not observed to increase substantially
among the tissues of fish fed these diets. In general, the tis-
sues of fish fed the SFA SO diets contained very similar fatty
acid profiles to that of the control, particularly in terms of their
SFA, n-3, and LC-PUFA levels. However, feeding the STD SO
feeds resulted in a concomitant loss of fish oil–associated n-3
fatty acids and LC-PUFAs as the fish oil replacement rate in-
creased. Changes in tissue fatty acid profiles, whether overt (in
the case of the STD SO treatment groups) or more subtle (in the
case of the SFA SO groups), became increasingly apparent as
fish oil was progressively replaced by either soy-derived lipid.
Consequently, the shape of the radial diagrams illustrating tis-
sue fatty acid profiles became increasingly distorted, reflecting
greater deviation from the 100% FO profiles, particularly among
the STD SO groups (Figure 1). However, the magnitude of di-
etary effect, as indicated by increasing Djh values, was greater
among liver and fillet tissues than among brain and eye tissues
(Figure 2).
Broken-line regression analysis indicated a break point at
29.3 g LC-PUFAs/kg diet, suggesting that this dietary concen-
tration was necessary to maintain weight gain among White
Seabass fed the STD SO feeds (Figure 3). Conversely, no posi-
tive slope or break point was identified in the SFA SO data set
(Figure 3).
DISCUSSION
The fish performance observed in the present trial was
consistent with results previously reported for juvenile White
Seabass with regards to survival and FCR (Jirsa et al. 2010,
in press). Growth in the present experiment was markedly
better than that reported by Jirsa et al. (2010), but the average
water temperature in the previous experiment was 2◦ C cooler,
which readily explains the difference between the two trials.
 SOY OILS IN WHITE SEABASS FEEDS 311

In that study, a maximum weight gain of 402% was reported,
whereas in the present study weight gain exceeded 500% in all
treatment groups, with a maximum of 674%. Our observations
suggest that growth is reduced among fish fed standard soybean
oil–based diets containing less than 29.3 g LC-PUFAs/kg diet
but not among those fed equivalent, hydrogenated soybean
oil–based feeds (Figure 3). This trend culminated in signifi-
cantly impaired weight gain and SGR in the 100% STD SO
dietary treatment, and similar, albeit statistically nonsignificant,
effects were observed in FCR. Dietary protein content was also
reduced in the 100% STD SO feed, which likely contributed
to the reduced performance observed in this treatment group.
Thus, growth as a function of fish oil replacement with standard
soybean oil may have been disproportionately suppressed by
the lower protein content of this particular feed. However,
reduced performance is nonetheless consistent with the trend
observed among fish fed the other standard soybean oil–based
Downloaded by [Department Of Fisheries] at 19:57 08 April 2013
SO series, and though most of the treatments were statistically
similar to the 100% FO treatment, all were numerically superior
to the control group in terms of weight gain, SGR, and FCR. Al-
though it is unclear why growth performance was numerically
improved among fish fed hydrogenated soybean oil, similar
improvements in growth performance have been associated
with fish oil sparing in other species (Ng et al. 2003; Trushenski
et al. 2013, this issue). Together, these data suggest a limitation
associated with standard soybean oil, but not with hydrogenated
soybean oil, which became progressively problematic at higher
inclusion rates. We posit that this limitation is the high level of
18:2(n-6) associated with standard soybean oil, and the inferior
growth performance of White Seabass fed increasing amounts
of this lipid is essentially a function of a developing LC-PUFA
deficiency state exacerbated by increased 18:2(n-6) intake. A
fish fed an essential fatty acid–deficient diet “stops growing and
reproducing, develops various pathologies, and eventually dies”

feeds. These performance trends were not observed in the SFA (Sargent et al. 2002), but given the relatively short-term nature of

FIGURE 1. Fatty acid composition of White Seabass (A) fillet, (B) brain, (C) liver, and (D) eye total lipid for experimental treatment groups expressed as
fractions of the fatty acid composition observed in the positive control group (100% FO). Values were calculated from the fatty acid methyl ester composition
(g/100 g). Based on this calculation, a value of 1 represents equality between tissue profiles. Only fatty acids or fatty acid groupings representing >2% of the total
fatty acid methyl esters quantified for each tissue type are shown. All fatty acid abbreviations are as reported in Table 2. (Continued on next page)
 312 TRUSHENSKI ET AL.
Downloaded by [Department Of Fisheries] at 19:57 08 April 2013
FIGURE 1.
our experiment, it is not surprising that these ultimate signs of
deficiency were not observed. Nonetheless, our observations of
fish fed higher levels of standard soybean oil (e.g., 75% STD
SOand 100% STD SO) are consistent with the early stages of
deficiency, i.e., impaired growth performance and decreasing
levels of essential fatty acids in peripheral and central tissues.
Saturated fatty acid–rich feeds result in more efficient
utilization of available LC-PUFAs—at least in terms of tissue
deposition—than otherwise equivalent feeds containing higher
levels of C18 PUFAs. Several studies have demonstrated that
deposition of LC-PUFAs is greater among fish consuming
high levels of SFAs and MUFAs rather than C18 PUFA–rich
feeds. Largemouth Bass Micropterus salmoides (Laporte and
Trushenski 2011), hybrid Striped Bass (White Bass Morone
chrysops × Striped Bass M. saxatilis; Trushenski et al. 2008a),
Rainbow Trout Oncorhynchus mykiss (Trushenski et al. 2011b,
2011c), Nile Tilapia Oreochromis niloticus (Trushenski et al.
2009), Murray Cod Maccullochella peelii (Turchini et al. 2011),
and Cobia Rachycentron canadum (Trushenski et al. 2013,
this issue) fed diets rich in C18 PUFAs, particularly 18:2(n-6),
(Continued)
deposit less LC-PUFAs in their tissues than those fed SFA-rich
diets. Similar tissue fatty acid profile effects were observed in
the present study. Tissue profiles were influenced by dietary
fatty acid composition, though as has been reported elsewhere
(Noffs et al. 2009; Trushenski et al. 2011c), central tissues
(brain and eye) were less compositionally plastic than peripheral
tissues (fillet and liver). Although profiles varied significantly
among dietary treatment groups and changed to reflect fatty
acid intake, this effect was greatly attenuated among fish fed
the hydrogenated, SFA-rich soybean oil–based feeds. By some
mechanism(s)—preferential inclusion, retention, or improved
utilization of available LC-PUFAs—provision of SFAs effec-
tively yields more efficient utilization of dietary LC-PUFAs, an
effect Turchini et al. (2011) called the “omega-3 sparing effect.”
Several groups have argued that selective catabolism of SFAs
via β-oxidation is responsible for SFA intake being underrepre-
sented in the fatty acid profile of fish tissues (Lane et al. 2006;
Trushenski et al. 2008a; Turchini et al. 2011). Fatty acids ap-
pear to be catabolized to a greater or lesser extent based on their
degree of unsaturation and chain length as well their dietary
 SOY OILS IN WHITE SEABASS FEEDS 313

40 40
Eye Liver
Y = -5.500X + 39.539
r² = 0.998

30 30
Coefficient of Distance (Djh)

Coefficient of Distance (Djh)

Y = -2.351X + 17.522
r² = 0.968
20 20

Y = -0.990X + 7.389
10 10
r² = 0.985

Y = -1.700X + 13.444
r² = 0.866

0 0
0 1 2 3 4 5 6 7 8 0 1 2 3 4 5 6 7 8
Dietary Fish Oil Content (%) Dietary Fish Oil Content (%)

40
Brain 40
Fillet

STD SO SFA SO
Downloaded by [Department Of Fisheries] at 19:57 08 April 2013

30 30 Y = -3.711X + 25.961
Coefficient of Distance (Djh)

Coefficient of Distance (Djh)

r² = 0.995

20 20

Y = -0.910X + 7.016
r² = 0.922
10 10
Y = -0.113X + 1.504 Y = -0.822X + 5.272
r² = 0.235 r² = 0.905

0 0
0 1 2 3 4 5 6 7 8 0 1 2 3 4 5 6 7 8
Dietary Fish Oil Content (%) Dietary Fish Oil Content (%)

FIGURE 2. Relationships between dietary fish oil content and coefficient of distance (Djh) values by tissue type for White Seabass fed diets containing standard
(STD SO) or fully hydrogenated (SFA SO) soybean oils. The Djh values are based on mean fatty acid composition using major individual fatty acids (>2% of
total quantified fatty acid methyl esters with no fatty acid groupings, e.g., SFAs, MUFAs). The solid lines represent linear regressions. Equivalency between the
profiles of fish fed the experimental diets and those of fish fed the control 100% FO diet would be represented by a Djh value of zero. Djh values representing the
100% FO group are denoted by gray diamonds.

abundance (Henderson and Sargent 1985; Kiessling and vided in the diet (Bell et al. 2001, 2002, 2004; Trushenski et al.
Kiessling 1993; Henderson 1996; Frøyland et al. 2000; 2008a, 2011a, 2011b, 2011c; Fountoulaki et al. 2009). This
Torstensen et al. 2004; Stubhaug et al. 2007); as Turchini et al. effect has been observed in a wide range of fish species fed
(2011) put it, “fatty acids provided in dietary surplus are the diets containing numerous oil types and blends, most notably
object of intensive β-oxidation.” Furthermore, the selective in- among species with limited capacity for de novo LC-PUFA
corporation of SFAs in newly synthesized lipid appears to be synthesis (Trushenski et al. 2008a, 2011b, 2011c). Excluding
somewhat limited in that only a certain SFA, i.e., 16:0, is the de novo LC-PUFA synthesis as a major contributing factor,
preferred substrate of the acyltransferases and fatty acid binding tissue enrichment of LC-PUFAs can be partially explained by
proteins which govern phospholipid synthesis (Trushenski et al. the affinity of acyltransferases and other elements involved in
2008b). For these reasons, it is reasonable to anticipate some tis- lipid synthesis for these fatty acids. But why is this effect ex-
sue accumulation of SFAs among fish fed increasing amounts of aggerated among fish fed higher levels of SFAs? It is possible
SFAs, but that the accumulation will be less than proportional to that LC-PUFA sparing by dietary SFAs is not a direct con-
dietary intake. These metabolic phenomena provide an adequate sequence of increased SFA intake but rather decreased C18
explanation for the underrepresentation of SFAs in the tissues of PUFA intake. In addition to LC-PUFAs, the enzymes involved
fish fed the SFA-rich, hydrogenated soy oil–based feeds. They in lipid synthesis also show considerably affinity for C18 PU-
are less satisfying, however, in explaining the conservation of FAs, especially 18:2(n-6). If 18:2(n-6) and other C18 PUFAs
LC-PUFAs in the tissues of fish fed the SFA-rich feeds. are abundant in the diet, they may effectively “outcompete”
Long-chain polyunsaturated fatty acids, particularly 22:6(n- LC-PUFAs for deposition during lipid synthesis–remodeling
3), are preferentially included and/or retained in fish tissues, (Trushenski et al. 2008a, 2011a; Laporte and Trushenski 2011).
typically becoming enriched beyond the concentrations pro- Eliminating or reducing competition from C18 PUFAs may be a
 314
800
750 Break Point = 29.3 g LC-PUFA/kg diet
Weight Gain (%)
700
650
600
550
500
450
400
0 10 20 30
800
750
Weight Gain (%)
700
650
600
550
500
450
400
Downloaded by [Department Of Fisheries] at 19:57 08 April 2013
0 10 20 30
Dietary LC-PUFA (g/kg diet)
FIGURE 3. Broken-line regressions depicting the weight gain of White
Seabass as a function of dietary long-chain polyunsaturated fatty acid (LC-
PUFA) content in standard soybean (upper panel) and fully hydrogenated soy-
bean (lower panel) oil-based feeds. Each data point represents the average
individual weight gain for a single replicate versus the dietary LC-PUFA con-
tent (reported as g LC-PUFA/kg diet on a dry-matter basis; calculated from the
data in Tables 1 and 2). Data from the 100% FO control group are included in
both panels. The lines represent best-fit regressions; an identifiable break point
was observed within the standard soybean oil data series only.
mechanism by which dietary SFAs improve the apparent effi-
ciency of LC-PUFA metabolism and allow for greater tissue
deposition. However, it has been suggested that the compet-
itive effect of C18 PUFAs is primarily associated with phos-
pholipid synthesis (Trushenski et al. 2008b, 2012); given the
greater abundance of neutral lipid in most fish tissues and the
less-selective processes which govern neutral lipid synthesis, it
seems unlikely that reduced competition for tissue deposition
is the primary means by which SFAs lead to greater tissue LC-
PUFA content. Turchini et al. (2011) also suggested that SFAs
(and MUFAs) literally spare LC-PUFAs by outcompeting LC-
PUFAs for catabolism as preferred substrates for β-oxidation.
If SFAs interfere with and limit LC-PUFA catabolism, they
may facilitate greater tissue deposition of LC-PUFAs simply by
increasing the total availability of LC-PUFAs. Generally, LC-
PUFAs are not considered ideal substrates for energy production
and are not shunted into β-oxidation in substantial amounts ex-
cept when consumed in considerable amounts (Torstensen et al.
2004). If these fatty acids are not typically oxidized to a great
extent, it seems somewhat implausible that improved catabolic
sparing is the dominant mechanism by which consuming SFAs
in lieu of C18 PUFAs leads to higher tissue levels of LC-PUFAs.
In other words, the argument is perfectly reasonable, but the
magnitude of the catabolic sparing effect seems insufficient to
fully explain the magnitude of LC-PUFA retention observed. It
seems like that the SFA “sparing effect” is the result of reduced
TRUSHENSKI ET AL.
competition from C18 PUFAs for deposition and catabolic spar-
ing of LC-PUFAs, though it is also possible that SFAs influence
LC-PUFA metabolism by another, as-yet-undetermined means.
Increased tissue deposition of LC-PUFAs likely results from
a greater abundance of LC-PUFAs; whether this surplus is the
result of reduced LC-PUFA demand for other physiological pur-
poses or reduced catabolic losses is unclear. The former suggests
a direct effect of SFAs on absolute LC-PUFA requirements,
40
whereas the latter suggests an indirect effect of SFAs that al-
lows the same absolute requirements to be met at lower dietary
inclusion rates. Neither of these mechanisms can be disregarded:
feeding different dietary lipid sources to Atlantic Salmon Salmo
salar influenced gene expression of the proteins involved in fatty
acid uptake, trafficking, and oxidation (Torstensen et al. 2009),
which could be interpreted as supporting evidence for either or
both of the proposed direct and indirect effects of SFAs. Regard-
less of the mechanism, our data do suggest that the efficiency
40 of LC-PUFA metabolism can be enhanced by feeding SFA-rich
alternative lipids and perhaps that the LC-PUFA requirements
of White Seabass can be effectively reduced by using this ap-
proach. If successfully implemented, this strategy could allow
for a greater level of fish oil sparing without growth impair-
ment or tissue profile modification than is possible with C18
PUFA–rich lipids.
ACKNOWLEDGMENTS
The authors sincerely thank Lauren Blazeck for her sub-
stantial assistance in completing the analytical portion of this
project. The authors sincerely thank Andrea Marino and Jose
Velazquez for their diligent hands-on care and management of
the feeding trial. We thank the United Soybean Board for pro-
viding funding and support for this research (Project SB1463).
We thank Archer Daniels Midland and Omega Protein for their
donation of the feedstuffs used in the experimental feeds.
REFERENCES
Bell, J. G., R. J. Henderson, D. R. Tocher, F. McGhee, J. R. Dick, A. Porter,
R. P. Smullen, and J. R. Sargent. 2002. Substituting fish oil with crude palm
oil in the diet of Atlantic Salmon (Salmo salar) affects muscle fatty acid
composition and hepatic fatty acid metabolism. Journal of Nutrition 132:222–
230.
Bell, J. G., R. J. Henderson, D. R. Tocher, and J. R. Sargent. 2004. Replacement
of dietary fish oil with increasing levels of linseed oil: modification of flesh
fatty acid compositions in Atlantic Salmon (Salmo salar) using a fish oil
finishing diet. Lipids 39:223–232.
Bell, J. G., J. McEvoy, D. R. Tocher, F. McGhee, P. J. Campbell, and J. R.
Sargent. 2001. Replacement of fish oil with rapeseed oil in diets of Atlantic
Salmon (Salmo salar) affects tissue lipid compositions and hepatocyte fatty
acid metabolism. Journal of Nutrition 131:1535–1543.
Christie, W. W. 1982. Lipid analysis, 2nd edition. Pergamon, Oxford, UK.
Durazo, E., A. C. Cruz, L. M. López, J. P. Lazo, M. Drawbridge, and M. T.
Viana. 2010. Effects of digestible protein levels in isonitrogenous diets on
growth performance and tissue composition of juvenile Atractoscion nobilis.
Aquaculture Nutrition 16:54–60.
 SOY OILS IN WHITE SEABASS FEEDS
Folch, J., M. Lees, and G. H. Sloane Stanley. 1957. A simple method for
the isolation and purification of total lipides from animal tissues. Journal of
Biological Chemistry 226:497–509.
Fountoulaki, E., A. Vasilaki, R. Hurtado, K. Grigorakis, I. Karacostas, I.
Nengas, G. Rigos, Y. Kotzamanis, B. Venou, and M. N. Alexis. 2009. Fish
oil substitution by vegetable oils in commercial diets for Gilthead Sea Bream
(Sparus aurata L.); effects on growth performance, flesh quality and fillet
fatty acid profile: recovery of fatty acid profiles by a fish oil finishing diet
under fluctuating water temperatures. Aquaculture 289:317–326.
Frøyland, L., Ø. Lie, and R. K. Berge. 2000. Mitochondrial and peroxisomal
β-oxidation capacities in various tissues from Atlantic Salmon Salmo salar.
Aquaculture Nutrition 6:85–89.
Glencross, B. D. 2009. Exploring the nutritional demand for essential fatty acids
by aquaculture species. Reviews in Aquaculture 1:71–124.
Henderson, R. J. 1996. Fatty acid metabolism in freshwater fish with particular
reference to polyunsaturated fatty acids. Archives of Animal Nutrition 49:5–
22.
Henderson, R. J., and J. R. Sargent. 1985. Chain-length specificities of mito-
chondrial and peroxisomal β-oxidation of fatty acids in livers of Rainbow
Trout (Salmo gairdneri). Comparative Biochemistry and Physiology 82B:
Downloaded by [Department Of Fisheries] at 19:57 08 April 2013
79–85.
Jirsa, D., D. A. Davis, and M. Drawbridge. 2010. Development of a practical
soy-based diet for White Seabass. North American Journal of Aquaculture
72:332–337.
Jirsa, D., D. F. Deng, D. A. Davis, W. F. Wang, S. S. O. Hung, and M.
Drawbridge. In press. The effects of dietary lipid levels on performance and
heat-shock protein response of juvenile White Seabass, Atractoscion nobilis.
Aquaculture Nutrition. DOI: 10.1111/j.1365-2095.2012.00965.x.
Kiessling, K. H., and A. Kiessling. 1993. Selective utilization of fatty acids in
Rainbow Trout (Oncorhynchus mykiss Walbaum) red muscle mitochondria.
Canadian Journal of Zoology 71:248–251.
Lane, R. L., J. T. Trushenski, and C. C. Kohler. 2006. Modification of fillet
composition and evidence of differential fatty acid turnover in Sunshine Bass
Morone chrysops × M. saxatilis following change in dietary lipid source.
Lipids 41:1029–1038.
Laporte, J., and J. Trushenski. 2011. Growth performance and tissue fatty acid
composition of Largemouth Bass fed diets containing fish oil or blends of fish
oil and soy-derived lipids. North American Journal of Aquaculture 73:435–
444.
López, L. M., E. Durazo, M. T. Viana, M. Drawbridge, and D. P. Bureau. 2009.
Effect of dietary lipid levels on performance, body composition and fatty
acid profile of juvenile White Seabass, Atractoscion nobilis. Aquaculture
289:101–105.
López, L. M., A. L. Torres, E. Durazo, M. Drawbridge, and D. P. Bureau. 2006.
Effects of lipid on growth and feed utilization of White Seabass (Atractoscion
nobilis) fingerlings. Aquaculture 253:557–563.
Ng, W. K., P. K. Lim, and P. L. Boey. 2003. Dietary lipid and palm oil source
affects growth, fatty acid composition and muscle α-tocopherol concentration
of African Catfish, Clarias gariepinus. Aquaculture 215:229–243.
Noffs, M. D., R. C. Martino, L. C. Trugo, E. C. Urbinati, J. B. K. Fernandes, and
L. S. Takahashi. 2009. Dietary fish oil replacement with lard and soybean oil
affect triacylglycerol and phospholipid muscle and liver docosahexaenoic acid
content but not in the brain and eyes of Surubim juveniles Pseudoplatystoma
sp. Fish Physiology and Biochemistry 35:399–412.
NRC (National Research Council). 2011. Nutrient requirements of fish and
shrimp. National Academies Press, Washington, D.C.
Ramezani-Fard, E., M. S. Kamarudin, S. A. Harmin, and C. R. Saad. 2012.
Dietary saturated and omega-3 fatty acids affect growth and fatty acid profiles
315
of Malaysian Mahseer. European Journal of Lipid Science and Technology
114:185–193.
Robbins, K. R., A. M. Saxton, and L. L. Southern. 2006. Estimation of nutri-
ent requirements using broken-line regression analysis. Journal of Animal
Science 84(Supplement):155–165.
Sargent, J. R., D. R. Tocher, and J. G. Bell. 2002. The lipids. Pages 181–257 in
J. E. Halver and R. W. Hardy, editors. Fish nutrition, 3rd edition. Academic
Press, San Diego, California.
Stubhaug, I., Ø. Lie, and B. E. Torstensen. 2007. Fatty acid productive value and
β-oxidation capacity in Atlantic Salmon (Salmo salar L.) fed on different lipid
sources along the whole growth period. Aquaculture Nutrition 13:145–155.
Torstensen, B. E., L. Frøland, R. Ørnsrud, and Ø. Lie. 2004. Tailoring of a
cardioprotective muscle fatty acid composition of Atlantic Salmon (Salmo
salar) fed vegetable oils. Food Chemistry 87:567–580.
Torstensen, B. E., D. A. Nanton, P. A. Olsvik, H. Sundvold, and I. Stubhaug.
2009. Gene expression of fatty acid–binding proteins, fatty acid transport
proteins (cd36 and FATP) and β-oxidation-related genes in Atlantic Salmon
(Salmo salar L.) fed fish oil or vegetable oil. Aquaculture Nutrition 15:440–
451.
Trushenski, J. T., P. Blaufuss, B. Mulligan, and J. Laporte. 2011a. Growth
performance and tissue fatty acid composition of Rainbow Trout reared on
feeds containing fish oil or equal blends of fish oil and traditional or novel
alternative lipids. North American Journal of Aquaculture 73:194–203.
Trushenski, J. T., J. Boesenberg, and C. C. Kohler. 2009. Influence of grow-
out feed fatty acid composition on finishing success in Nile Tilapia. North
American Journal of Aquaculture 71:242–251.
Trushenski, J. T., B. Gause, and H. A. Lewis. 2011b. Selective fatty acid
metabolism, not the sequence of dietary fish oil intake, prevails in fillet fatty
acid profile change in Sunshine Bass. North American Journal of Aquaculture
73:204–211.
Trushenski, J. T., H. A. Lewis, and C. C. Kohler. 2008a. Fatty acid profile of
Sunshine Bass: I. profile change is affected by initial composition and differs
among tissues. Lipids 43:629–641.
Trushenski, J. T., H. A. Lewis, and C. C. Kohler. 2008b. Fatty acid profile of
Sunshine Bass: II. profile change differs among fillet lipid classes. Lipids
43:643–653.
Trushenski, J. T., M. Schwarz, A. Bergman, A. Rombenso, and B. Delbos. 2012.
DHA is essential, EPA appears largely expendable, in meeting the n-3 long-
chain polyunsaturated fatty acid requirements of juvenile Cobia Rachycentron
canadum. Aquaculture 326–329:81–89.
Trushenski, J. T., M. Schwarz, H. Lewis, J. Laporte, B. Delbos, R. Takeuchi,
and L. A. Sampaio. 2011c. Effect of replacing dietary fish oil with soybean
oil on production performance and fillet lipid and fatty acid composition of
juvenile Cobia Rachycentron canadum. Aquaculture Nutrition 17:437–447.
Trushenski, J. T., M. Schwarz, F. Woitel, and F. Yamamoto. 2013. Saturated fatty
acids limit the effects of replacing fish oil with soybean oil with or without
phospholipid supplementation in feeds for juvenile Cobia. North American
Journal of Aquaculture 75:316–328.
Turchini, G. M., D. S. Francis, and S. S. De Silva. 2006. Modification of tissue
fatty acid composition in Murray Cod (Maccullochella peelii peelii, Mitchell)
resulting from a shift from vegetable oil diets to a fish oil diet. Aquaculture
Research 37:570–585.
Turchini, G. M., D. S. Francis, S. P. S. D. Senadheera, T. Thanuthong, and S. S.
De Silva. 2011. Fish oil replacement with different vegetable oils in Murray
Cod: evidence of an “omega-3 sparing effect” by other dietary fatty acids.
Aquaculture 315:250–259.
Turchini, G. M., B. E. Torstensen, and W. K. Ng. 2009. Fish oil replacement in
finfish nutrition. Reviews in Aquaculture 1:10–57.
This article was downloaded by: [Department Of Fisheries]
On: 08 April 2013, At: 19:58
Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,
37-41 Mortimer Street, London W1T 3JH, UK

North American Journal of Aquaculture

Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/unaj20

Saturated Fatty Acids Limit the Effects of Replacing

Fish Oil with Soybean Oil with or without Phospholipid
Supplementation in Feeds for Juvenile Cobia
a a b c
Jesse Trushenski , Franklin Woitel , Michael Schwarz & Fernando Yamamoto
a
Fisheries and Illinois Aquaculture Center and Department of Zoology, Southern Illinois
University–Carbondale, 1125 Lincoln Drive, Carbondale, Illinois, 62901-6511, USA
b
Virginia Seafood Agricultural Research and Extension Center, Virginia Tech, 102 South King
Street, Hampton, Virginia, 23669, USA
c
International Initiative for Sustainable and Biosecure Aquafarming, Norfolk, Virginia,
23503, USA
Version of record first published: 02 Apr 2013.

To cite this article: Jesse Trushenski , Franklin Woitel , Michael Schwarz & Fernando Yamamoto (2013): Saturated Fatty Acids
Limit the Effects of Replacing Fish Oil with Soybean Oil with or without Phospholipid Supplementation in Feeds for Juvenile
Cobia, North American Journal of Aquaculture, 75:2, 316-328

To link to this article: http://dx.doi.org/10.1080/15222055.2012.713897

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions

This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to
anyone is expressly forbidden.
The publisher does not give any warranty express or implied or make any representation that the contents
will be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses should
be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims,
proceedings, demand, or costs or damages whatsoever or howsoever caused arising directly or indirectly in
connection with or arising out of the use of this material.
 North American Journal of Aquaculture 75:316–328, 2013

C American Fisheries Society 2013

ISSN: 1522-2055 print / 1548-8454 online

DOI: 10.1080/15222055.2012.713897

SPECIAL SECTION: LIPIDS IN AQUACULTURE NUTRITION AND PHYSIOLOGY II

Saturated Fatty Acids Limit the Effects of Replacing

Fish Oil with Soybean Oil with or without Phospholipid
Supplementation in Feeds for Juvenile Cobia

Jesse Trushenski* and Franklin Woitel

Fisheries and Illinois Aquaculture Center and Department of Zoology,
Southern Illinois University–Carbondale, 1125 Lincoln Drive, Carbondale, Illinois 62901-6511, USA

Michael Schwarz
Virginia Seafood Agricultural Research and Extension Center, Virginia Tech, 102 South King Street,
Downloaded by [Department Of Fisheries] at 19:58 08 April 2013

Hampton, Virginia 23669, USA

Fernando Yamamoto
International Initiative for Sustainable and Biosecure Aquafarming, Norfolk, Virginia 23503, USA

Abstract
The high cost and limited availability of fish oil makes plant-derived lipids attractive for aquafeed manufacturing,
but replacing fish oil with these lipids can result in long-chain polyunsaturated fatty acid (LC-PUFA) deficiencies.
Fatty acid metabolism, specifically the efficiency of LC-PUFA utilization, may be influenced by the dietary saturated
fatty acid (SFA) content versus that of C18 polyunsaturated fatty acids (PUFAs). We assessed the growth and tissue
composition of Cobia Rachycentron canadum (55.3 ± 0.2 g initial weight [mean ± SE]; 10 fish/tank, 3 tanks/diet) fed
diets (∼49% protein, ∼10% lipid) containing fish oil; 22:6(n-3)–amended standard, partially hydrogenated, or fully
hydrogenated soybean oil; and these same soybean oils supplemented with soybean lecithin for 8 weeks. Although
survival (range = 97–100%), final weight (160–189 g), and feed conversion ratio (1.40–1.52) were unaffected by diet,
differences were observed in weight gain (185–241%), specific growth rate (1.87–2.19% body weight/d), and feed
intake (2.94–3.44% body weight/d). Significant effects of soybean oil type on final weight, weight gain, feed conversion
ratio, specific growth rate, and feed intake were noted, with standard soybean oil generally outperforming the other
soybean lipids when oil types were pooled across phospholipid supplementation treatments, whereas phospholipid
supplementation had no significant effect on any of the performance measures. Differences in dietary fatty acid profile
yielded differences in tissue composition. Feeding standard soybean oil resulted in the most greatly modified profiles,
whereas the profiles of fish fed fully hydrogenated, completely saturated soybean oil were most similar to those of
the fish oil–fed fish. The magnitude of profile change was greatest in the liver and fillet tissues and smallest in the
eye and brain tissues. Although further research is necessary to demonstrate whether SFA-rich lipids can effectively
reduce the LC-PUFA requirements of Cobia, it is clear that SFA-rich oils offer a strategic advantage in minimizing
the effects of fish oil replacement on tissue fatty acid profile.

Cobia Rachycentron canadum is an emerging global aqua- Benetti et al. 2010). As is typical of marine carnivore feeds,
culture species because of its fast growth rate, high economic Cobia rations typically contain higher levels of lipids as the
value, adaptability to captive breeding, and resistance to dis- primary source of metabolic energy rather than carbohydrates.
eases (Liao et al. 2004; Sun et al. 2006; Liao and Leaño 2007; Traditionally, the primary lipid source used in aquafeeds was

*Corresponding author: saluski@siu.edu

Received April 15, 2012; accepted July 13, 2012
Published online April 2, 2013

316
 SOYBEAN OILS IN COBIA FEEDS
fish oil, and thiscommodity is still widely used in the produc-
tion of industrially compounded aquafeeds. Although dietary
inclusion rates have declined and are expected to decrease fur-
ther (Tacon and Metian 2008), most of the fish oil saved by
sparing with alternative lipids has been consumed by an overall
increase in aquafeed production. At this time, nearly 85% of
global annual fish oil production is consumed by the aquafeed
manufacturing sector (FAO 2010). As a result of this bottle-
neck and the rising price of fish oil, researchers continue to
pursue more judicious use of fish oil in aquafeeds through spar-
ing and replacement with alternative lipids. Terrestrial plant–
derived lipids are considered attractive alternatives because they
have lower prices and production volumes are sufficient to meet
expanding aquaculture demand (Naylor et al. 2009). However,
plant-derived lipids do not contain long-chain polyunsaturated
fatty acids (LC-PUFAs, i.e., fatty acids with a chain length of 20
or more carbon atoms and at least 3 double bonds), which are es-
Downloaded by [Department Of Fisheries] at 19:58 08 April 2013
sential for some fish species, including Cobia (Chou et al. 2004;
Trushenski et al. 2011e). Arguably, demand for LC-PUFAs may
be met through de novo synthesis from shorter-chain fatty acid
precursors, and Cobia reportedly express the elongase and de-
saturase genes involved in these biosynthetic pathways (Zheng
et al. 2009; Monroig et al. 2011). Despite their having the en-
zymatic capacity for de novo synthesis, in some species the
production of biologically relevant volumes of these critical nu-
trients is apparently limited. As a result, fish oil sparing with
plant oils can result in essential fatty acid deficiencies, reduc-
ing the growth and survival of a number of fishes (Glencross
2009; Turchini et al. 2009), including Cobia (Faulk and Holt
2005; Liu et al. 2007; Wang et al. 2007; Nhu et al. 2011). For
example, replacing fish oil with soybean oil impairs growth in
juvenile Cobia (Trushenski et al. 2011e, 2012), but these effects
are attenuated when the diet is amended with sufficient amounts
of docosahexaenoic acid (22:6[n-3]1) (Trushenski et al. 2012).
If essential fatty acid requirements are met (22:6[n-3] re-
quirements, in the case of Cobias), fish oil sparing or replace-
ment in feeds may be more successful than originally observed
or anticipated. However, previous research has suggested that
fatty acid requirements, and thus the necessary supplementa-
tion level, may be reduced under certain conditions. Nutrient
requirements are known to vary among taxa, life stages, and
culture conditions and are influenced by interactions among
nutrients (NRC 2011). Classic examples of these types of inter-
actions include the sparing effect of cysteine on methionine; the
interaction between vitamin C, vitamin E, and selenium related
to total antioxidant capacity; the relationship between dietary
oxidation potential and the demand for antioxidant nutrients;
and the relative potency of different molecular species of mi-
cronutrients (NRC 2011). Essential fatty acid requirements may
1In fatty acid designations of this nature, the number to the left of the colon
is the number of carbon atoms in the compound, the number immediately to
the right of the colon is the number of double bonds, and the number after the
hyphen indicates the position of the first double bond from the methyl end.
317
depend on the “potency” of different fatty acids: although the
n-3 fatty acid requirements of some species may be met exclu-
sively by 18:3(n-3), they may also be met by 20:5(n-3) and/or
22:6(n-3) at lower dietary concentrations (Ruyter et al. 2000).
Essential fatty acid requirements may also be influenced by in-
teractions with other fatty acids as well as dietary lipid content
(Glencross 2009). Our previous research takes this paradigm a
step further, suggesting that fatty acid metabolism, and therefore
absolute requirements, may be influenced by the relative pro-
portions of saturated fatty acids (SFAs, i.e., fatty acids with
no double bonds) and C18 polyunsaturated fatty acids (C18
PUFAs, i.e., fatty acids with chain lengths of 18 carbon atoms
and at least 2 double bonds) in the diet. Specifically, it has
been demonstrated that fish fed SFA-rich diets retain higher
levels of LC-PUFAs, including 22:6(n-3) than those fed C18
PUFA-rich diets (Trushenski et al. 2008a, 2008b, 2011a, 2011c;
Laporte and Trushenski 2011; Turchini et al. 2011; Ramezani-
Fard et al. 2012). Through various mechanisms, the efficiency
of LC-PUFA utilization may be increased through increased di-
etary provision of SFAs in lieu of C18 PUFAs. By extension,
it may be possible to meet LC-PUFA requirements at lower
dietary inclusion rates if the required fatty acids are provided
in the context of a SFA-rich alternative lipid base. In addition
to the importance of meeting essential fatty acid requirements
and the potential influence of dietary SFA versus C18 PUFA con-
tent, the availability of phospholipid may determine, in part, the
success of fish oil sparing. Phospholipid intake is known to be
critically important in successful completion of early ontogeny
in fish (Tocher et al. 2008; Cahu et al. 2009) and likely has a
continuing importance in the diet of rapidly growing fish (Seoka
et al. 2008; Sandel et al. 2010; Hansen et al. 2011), including Co-
bia (Trushenski et al., in press). As dietary fish oil and fish meal
are replaced with plant-derived lipids and lipid-extracted protein
meals, phospholipid may emerge as an essential element of lipid
nutrition of o(n-growing fish and an increasingly important fac-
tor in aquafeed manufacturing (Tocher et al. 2008). Accordingly,
we assessed the growth and tissue composition of juvenile Cobia
fed diets containing fish oil or 22:6(n-3)–amended soybean oils
containing different levels of SFAs versus C18 PUFAs with or
without supplemental phospholipid. Our objectives were to de-
termine (1) whether these diets yielded adequate and equivalent
growth performance and (2) by inferring from the differences
in tissue fatty acid composition, whether the apparent require-
ments for 22:6(n-3) vary depending on the lipid and fatty acid
composition of the diet.
METHODS
Feed preparation and analyses.—A practical feed (Table 1),
including ∼246 g/kg menhaden fish meal (Special Select)
and 53 g/kg menhaden fish oil (FO; Virginia Prime Gold,
Omega Protein, Inc., Houston, Texas) and previously ver-
ified in Cobia culture (Trushenski et al. 2011d, 2012, in
press), served as the positive control feed in the present work.
 318 TRUSHENSKI ET AL.

TABLE 1. Dietary formulations and proximate compositions of feeds fed to Cobias. The FO formulation was the control (see Methods). Other formulations
were derived from it using standard (STD SO), partially hydrogenated (HYD SO), and fully hydrogenated saturated fatty acid–rich (SFA SO) soybean oil with or
without the addition of soybean lecithin (PL).

STD STD HYD HYD SFA SFA

Ingredient FO SO SO + PL SO SO + PL SO SO + PL
Dietary formulation (g/kg on an as-fed basis)
Wheat bran 302.3 289.0 289.0 289.0 289.0 289.0 289.0
Menhaden fish meala 245.5 245.6 245.6 245.6 245.6 245.6 245.6
Corn gluten meal 150.1 150.1 150.1 150.1 150.1 150.1 150.1
Soybean protein concentrateb 137.0 137.2 137.2 137.2 137.2 137.2 137.2
Soybean protein isolateb 79.2 79.2 79.2 79.2 79.2 79.2 79.2
Menhaden fish oila 53.1 0.0 0.0 0.0 0.0 0.0 0.0
Soybean-derived lipidb 0.0 44.5 24.5 44.5 24.5 44.5 24.5
Carboxymethyl cellulose 20.2 20.2 20.2 20.2 20.2 20.2 20.2
Choline chloride 5.5 5.5 5.5 5.5 5.5 5.5 5.5
Methionine chloride 3.3 3.3 3.3 3.3 3.3 3.3 3.3
Downloaded by [Department Of Fisheries] at 19:58 08 April 2013

Stay-Cc 1.8 1.8 1.8 1.8 1.8 1.8 1.8

Vitamin premixd 1.1 1.1 1.1 1.1 1.1 1.1 1.1
Mineral premixe 0.9 0.9 0.9 0.9 0.9 0.9 0.9
Soybean lecithinb 0.0 0.0 20.0 0.0 20.0 0.0 20.0
Algal mealf 0.0 21.6 21.6 21.6 21.6 21.6 21.6
Proximate compositiong (g/kg on a dry-matter basis [except dry matter])
Dry matter 944 951 950 959 959 945 960
Protein 486 ± 2 486 ± 3 487 ± 6 491 ± 5 485 ± 7 487 ± 4 489 ± 5
Lipid 94 ± 1 100 ± 1 94 ± 1 100 ± 7 102 ± 3 98 ± 5 102 ± 2
Ash 87 ± 4 98 ± 6 82 ± 4 80 ± 5 95 ± 7 78 ± 6 83 ± 2
a
Omega Protein, Inc., Houston, Texas.
b
Archer Daniels Midland Co., Decatur, Illinois.
c
Argent Laboratories, Redmond, Washington.
d
Formulated to contain 25.000% L-ascorbyl-2-polyphosphate, 13.160% vitamin K, 12.500% inositol, 12.500% nicotinic acid, 7.500% riboflavin, 6.250% calcium pantothenate,
2.500% pyridoxine hydrochloride, 1.250% thiamine mononitrate, 1.000% vitamin A palmitate, 0.500% cyanocobalamin, 0.450% folic acid, 0.125% biotin, and 0.010% cholecalciferol
in a cellulose base.
e
Formulated to contain 24.897% zinc oxide, 14.933% ferrous sulfate, 3.470% manganese oxide, 0.967% cupric carbonate, 0.262% potassium iodide, 0.060% sodium selenate, and
0.030% cobalt carbonate in a cellulose base.
f
Aqua Grow Gold; Advanced BioNutrition Corporation, Columbia, Maryland.
g
Values represent least-square means ± SEs of triplicate samples, except for dry matter, which was determined from a single sample for each diet.

Experimental feeds were derived from the FO formulation using
standard (STD SO), partially hydrogenated (HYD SO), or fully
hydrogenated saturated fatty acid–rich (SFA SO) soybean oil
(Archer Daniels Midland Company, Decatur, Illinois) supple-
mented with ∼22 g/kg of an LC-PUFA concentrate (Aqua Grow
Gold; Advanced BioNutrition Corporation, Columbia, Mary-
land) in lieu of menhaden fish oil. The LC-PUFA concentrate
was used to ensure that the soybean oil–based feeds contained
∼75% of the 22:6(n-3) found in the FO feed and would thus
meet the 22:6(n-3) requirement of juvenile Cobia (Trushenski
et al. 2012). Although the LC-PUFA concentrate used was par-
ticularly rich in 22:6(n-3), it was not a purified source of this
fatty acid. Consequently, the levels of other fatty acids, such
as 20:5(n-3), in the LC-PUFA–supplemented feeds were also
slightly above the levels associated with the unsupplemented
oils and basal formulations. An additional series of experimental
feeds was prepared using the same 22:6(n-3)-supplemented soy-
bean oils but with soybean lecithin (Yelkin TS; Archer Daniels
Midland; typical phospholipid content: phosphatidylcholine =
15%, phosphatidylethanolamine = 13%, phosphatidylinositol
= 9%, and phosphatidic acid = 4%) replacing 20 g/kg of the
soybean oil base (STD SO + PL, HYD SO + PL, and SFA SO +
PL). Slight adjustments to the basal formulation (i.e., menhaden
fish meal and supplemental lipid inclusion rates) were made in
order to accommodate the macronutrient content of the LC-
PUFA concentrate while maintaining the intended proximate
composition of the experimental feeds, but these adjustments
were minor (∼1% decrease in fish meal and lipid inclusion) in
the context of the overall formulation.
All feeds were prepared at the Fisheries and Illinois Aquacul-
ture Center (Carbondale, Illinois). All feedstuffs were incorpo-
rated using a cutter–mixer (Model CM450; Hobart Corporation,
Troy, Ohio), pelleted using a commercial-grade food grinder
(1.5-hp electric grinder; Cabela’s, Sidney, Nebraska), and dried
 SOYBEAN OILS IN COBIA FEEDS 319

TABLE 2. Dietary fatty acid composition (g/100 g fatty acid methyl esters). Data are presented as least-squares means ± SEs of triplicate samples. See Table 1
for descriptions of the dietary formulations and the text for fatty acid nomenclature.

Fatty acid(s) FO STD SO STD SO + PL HYD SO HYD SO + PL SFA SO SFA SO + PL

14:0 7.0 ± 0.1 3.0 ± 0.0 3.0 ± 0.0 2.9 ± 0.1 2.9 ± 0.1 2.9 ± 0.1 3.0 ± 0.0
15:0 0.7 ± 0.0 0.3 ± 0.0 0.3 ± 0.0 0.3 ± 0.0 0.3 ± 0.0 0.3 ± 0.0 0.3 ± 0.0
16:0 19.7 ± 0.0 17.1 ± 0.1 18.3 ± 0.0 17.6 ± 0.0 18.4 ± 0.1 18.1 ± 0.2 18.8 ± 0.0
17:0 0.6 ± 0.0 0.3 ± 0.0 0.3 ± 0.0 0.3 ± 0.0 0.3 ± 0.0 0.3 ± 0.0 0.3 ± 0.0
18:0 3.9 ± 0.0 3.9 ± 0.0 3.9 ± 0.0 6.1 ± 0.0 5.1 ± 0.0 43.9 ± 0.5 27.3 ± 1.0
20:0 0.3 ± 0.0 0.3 ± 0.0 0.3 ± 0.0 0.3 ± 0.0 0.3 ± 0.0 0.4 ± 0.0 0.4 ± 0.0
SFAsa 32.2 ± 0.1 24.9 ± 0.1 26.1 ± 0.1 27.5 ± 0.1 27.3 ± 0.2 66.0 ± 0.2 50.2 ± 1.0
16:1(n-7) 8.9 ± 0.1 2.6 ± 0.0 2.8 ± 0.0 2.6 ± 0.0 2.7 ± 0.1 2.5 ± 0.1 2.7 ± 0.0
18:1(n-7) 2.8 ± 0.0 1.7 ± 0.0 1.7 ± 0.0 14.4 ± 0.2 8.9 ± 0.1 1.0 ± 0.0 1.3 ± 0.0
18:1(n-9) 9.2 ± 0.1 15.8 ± 0.0 14.5 ± 0.0 27.5 ± 0.1 20.9 ± 0.1 5.2 ± 0.1 8.3 ± 0.4
20:1(n-9) 0.8 ± 0.0 0.3 ± 0.0 0.3 ± 0.0 0.3 ± 0.0 0.3 ± 0.0 0.3 ± 0.0 0.3 ± 0.0
MUFAsb 21.8 ± 0.0 20.4 ± 0.0 19.3 ± 0.0 44.8 ± 0.3 32.8 ± 0.1 9.0 ± 0.0 12.6 ± 0.3
16:2(n-4) 1.0 ± 0.0 0.3 ± 0.0 0.3 ± 0.0 0.3 ± 0.0 0.3 ± 0.0 0.3 ± 0.0 0.3 ± 0.0
Downloaded by [Department Of Fisheries] at 19:58 08 April 2013

16:3(n-4) 1.1 ± 0.0 0.3 ± 0.0 0.3 ± 0.0 0.3 ± 0.0 0.3 ± 0.0 0.3 ± 0.0 0.3 ± 0.0
18:2(n-6) 10.9 ± 0.4 34.9 ± 0.1 34.3 ± 0.1 11.5 ± 0.5 22.1 ± 0.3 9.3 ± 0.3 19.7 ± 0.9
20:4(n-6) 1.0 ± 0.0 0.8 ± 0.0 0.8 ± 0.0 0.8 ± 0.0 0.8 ± 0.0 0.7 ± 0.0 0.6 ± 0.1
n-6c 12.0 ± 0.4 35.7 ± 0.0 35.1 ± 0.1 12.2 ± 0.5 22.9 ± 0.3 10.1 ± 0.2 20.3 ± 0.8
18:3(n-3) 1.9 ± 0.0 4.5 ± 0.0 4.3 ± 0.0 0.9 ± 0.0 2.5 ± 0.0 0.9 ± 0.0 2.1 ± 0.0
18:4(n-3) 2.6 ± 0.0 0.7 ± 0.0 0.8 ± 0.0 0.7 ± 0.0 0.7 ± 0.0 0.7 ± 0.0 0.7 ± 0.0
20:4(n-3) 1.4 ± 0.0 0.5 ± 0.0 0.5 ± 0.0 0.5 ± 0.0 0.5 ± 0.0 0.5 ± 0.0 0.5 ± 0.0
20:5(n-3) 11.5 ± 0.1 3.6 ± 0.0 3.8 ± 0.0 3.6 ± 0.0 3.6 ± 0.1 3.5 ± 0.1 3.7 ± 0.0
22:5(n-3) 2.2 ± 0.0 0.6 ± 0.0 0.7 ± 0.0 0.7 ± 0.0 0.7 ± 0.0 0.6 ± 0.0 0.7 ± 0.0
22:6(n-3) 12.3 ± 0.1 8.5 ± 0.0 9.0 ± 0.0 8.5 ± 0.1 8.5 ± 0.1 8.3 ± 0.2 8.7 ± 0.1
n-3d 31.9 ± 0.2 18.5 ± 0.1 19.0 ± 0.1 14.9 ± 0.1 16.5 ± 0.2 14.4 ± 0.4 16.4 ± 0.1
PUFAse 46.0 ± 0.1 54.7 ± 0.1 54.6 ± 0.0 27.7 ± 0.4 39.9 ± 0.2 25.0 ± 0.1 37.2 ± 0.7
C18 PUFAsf 15.4 ± 0.4 40.2 ± 0.1 39.3 ± 0.1 13.1 ± 0.5 25.3 ± 0.3 10.9 ± 0.2 22.5 ± 0.9
LC-PUFAsg 28.4 0.2 13.9 ± 0.0 14.8 ± 0.0 14.0 ± 0.1 14.0 ± 0.2 13.6 ± 0.4 14.1 ± 0.2
(n-3):(n-6) 2.7 ± 0.1 0.5 ± 0.0 0.5 ± 0.0 1.2 ± 0.1 0.7 ± 0.0 1.4 ± 0.1 0.8 ± 0.0
a
Sum of all fatty acids without double bonds.
b
Sum of all fatty acids with a single double bond.
c
Sum of all n-6 fatty acids.
d
Sum of all n-3 fatty acids.
e
Sum of all fatty acids with ≥2 double bonds.
f
Sum of all PUFAs with chain lengths of 18 carbon atoms.
g
Sum of all fatty acids with chain lengths ≥20 carbon atoms and double bonds ≥3.

in a commercial-grade food dehydrator (100◦ C; Harvest Saver
R-5A; Commercial Dehydrator Systems, Inc., Eugene, Ore-
gon) to approximately 953 g/kg dry matter. All feeds were then
packaged in plastic bags and shipped to the Virginia Seafood
Agricultural Research and Extension Center (VSAREC; Hamp-
ton, Virginia) and stored frozen (–20◦ C) throughout the study.
Prior to shipping, triplicate diet samples were collected and ana-
lyzed to confirm proximate (Table 1) and fatty acid composition
(Table 2). Samples were lyophilized (Freezone 6; Labconco
Corporation, Kansas City, Missouri) to determine moisture
content and then pulverized. Protein (LECO FP-528; LECO
Corporation, St. Joseph, Michigan) and ash (muffle furnace;
600◦ C for 3 h) content were determined for each pulverized
sample, and lipid content was determined gravimetrically fol-
lowing chloroform–methanol extraction (Folch et al. 1957).
Minor differences in dietary composition were noted; these
were attributed to small inaccuracies in feedstuff weighing
and/or incomplete mixing during feed manufacturing or sam-
ple collection. Reserved crude lipid samples were subjected to
acid-catalyzed transmethylation performed overnight at 50◦ C
(Christie 1982). The resultant fatty acid methyl esters (FAMEs)
were separated using a gas chromatograph equipped with a
flame-ionization detector fitted with a permanently bonded
polyethylene glycol, fused silica capillary column (Omegawax
250; 30 m × 0.25 mm [interior diameter], 0.25-µm film; Su-
pelco, Bellefonte, Pennsylvania). The injection volume was
1.0 µL, helium was the carrier gas (30 cm/s, 205◦ C), and the
injector temperature was 250◦ C. A split injection technique
(100:1) was used, and the temperature program was as follows:
50◦ C for 2 min, increased to 220◦ C at 4◦ C/min, and 220◦ C for
 320 TRUSHENSKI ET AL.
15 min. Individual FAMEs were identified by reference to ex-
ternal standards (Supelco 37 Component FAME Mix, PUFA-1,
and PUFA-3; Supelco).
Experimental design and feeding trial.—The feeding trial
was conducted at the VSAREC in a recirculating aquacul-
ture system consisting of 300-L fiberglass tanks equipped with
fluidized-bed biofilters, bubble bead filters, protein skimmers,
UV sterilization, immersion titanium heaters, and a diffusion
aeration system. Water temperature and dissolved oxygen were
monitored daily using a YSI-85 series dissolved oxygen me-
ter (YSI, Inc., Yellow Springs, Ohio) and pH with a YSI-
pH100 meter. Total ammonia-, nitrite- and nitrate-nitrogen were
quantified daily by spectrophotometric analysis (Hach, Inc.,
Loveland, Colorado). Prior to stocking the system and beginning
the trial, all fish had been fed a commercially available carniv-
orous fish feed (Otohime; Marubeni Nisshin Feed Company,
Tokyo, Japan); all fish were fed the FO control feed for 1 week
Downloaded by [Department Of Fisheries] at 19:58 08 April 2013
to acclimate them to the new feed. Ten juvenile fish (55.3 ±
0.2 g [mean ± SE]) were stocked into each of 21 tanks, and
each tank was randomly assigned one of the experimental feeds
(3 tanks/feed, N = 3) for the 8-week feeding trial. Fish in all
tanks were fed to apparent satiation, twice daily. Throughout the
trial, photoperiod was kept on a 12 light : 12 dark cycle, and all
water quality parameters were maintained within ranges suit-
able for Cobia culture: temperature = 27.3 ± 0.2◦ C, dissolved
oxygen = 6.42 ± 0.12 mg/L, pH = 7.75 ± 0.19, total ammo-
nia nitrogen = 0.20 ± 0.06 mg/L, nitrite = 0.04 ± 0.01 mg/L,
nitrate = 16.77 ± 3.59 mg/L, alkalinity = 144 ± 21 mg/L,
salinity = 24.7 ± 0.5 g/L.
Growth performance, sample collection, and analysis.—At
the end of the feeding trial, fish were group-weighed by tank
and counted in order to calculate standard metrics of production
performance as follows:
Weight Gain (%)
(average final weight – average initial weight)
= 100 ×
average initial weight
Feed Conversion Ratio (FCR)
average individual dry matter feed intake
=
average individual weight gain
Specific Growth Rate (SGR, % body weight/d) = 100
loge average final weight − loge average initial weight
× soybean oil–based feeds were analyzed by two-way ANOVA
days of feeding
Feed Intake (% body weight/d) = 100
average individual dry matter feed intake
×
(initial individual weight × final individual weight)0.5 /days of feeding
Five fish per tank were then euthanized by an overdose of
tricaine methanesulfonate (∼200 mg/L in culture water, with
fish being exposed to the solution until cessation of opercular
movement [∼5 min]) and dissected to remove liver, white mus-
cle, eye, and brain tissue samples. Liver samples were weighed
in order to calculate the hepatosomatic index
HSI = 100 × (liver weight/whole body weight),
and then all tissue samples were frozen, shipped overnight to the
Fisheries and Illinois Aquaculture Center, and stored at –80◦ C
prior to analysis.
Tissue fatty acid analysis.—Muscle tissue samples were
lyophilized, pulverized, and extracted following the procedures
described previously for feed samples. Intact brain, eye, and
liver samples were homogenized in the extraction solvent mix-
ture using a tissue homogenizer (PowerGen Model 1000; Fisher
Scientific, Waltham, Massachusetts) and then extracted in the
same manner as other samples. All other preparatory and ana-
lytical procedures were as described previously. Coefficient of
distance (Djh) values (Turchini et al. 2006) comparing overall
fatty acid profiles between the experimental treatments and the
FO positive control treatment were calculated for each tissue
type as follows:
 1/2

n
Djh = (Pij − Pih) 2
i=1
where Pij is the percent content of fatty acid i in the control
treatment (FO) and Pih is the percent content of fatty acid i in an
experimental treatment. The total number of fatty acids included
in each calculation (n) varied among tissue types, but in all
cases only major individual fatty acids (>2% of total quantified
FAMEs; no fatty acid groupings, e.g., SFA, LC-PUFA) were
used in the calculation of Djh.
Statistical analysis.—Although multiple fish were sampled
from each tank, replicate tanks served as the experimental units
for all statistical analyses (N = 3). All production performance
and fatty acid composition data were analyzed by one-way anal-
ysis of variance (ANOVA; PROC MIXED) using the Statisti-
cal Analysis System (version 9.1, SAS Institute, Cary, North
Carolina). For parameters exhibiting significant treatment ef-
fects, Tukey’s honestly significantly different (HSD) tests were
used to determine the significance of differences among means.
Additionally, production data associated with the experimental,
(PROC MIXED) to assess the significance of soybean oil type
and phospholipid supplementation as main effects and to test
for a significant interaction between these two treatment fac-
tors. For all statistical procedures, differences were considered
significant at P < 0.05. Coefficient of distance data were de-
termined using mean fatty acid composition data, yielding a
single value for each dietary treatment within each tissue type.
Accordingly, these data were not analyzed using a formal sta-
tistical procedure but are provided for the purpose of qualitative
comparison only.
 SOYBEAN OILS IN COBIA FEEDS 321

TABLE 3. Production performance by dietary treatment. The values are the means ± SEs associated with triplicate tanks of fish (N = 3). The P-values resulting
from one-way ANOVA tests are also provided. Within rows, means with common lowercase letters are not significantly different (P > 0.05).

STD HYD SFA

Parameter FO STD SO SO + PL HYD SO SO + PL SFA SO SO + PL P Value

Survival (%) 97 ± 3 100 ± 0 100 ± 0 100 ± 0 100 ± 0 100 ± 0 100 ± 0 0.463

Initial weight (g) 55 ± 1 55 ± 0 55 ± 0 56 ± 0 56 ± 1 54 ± 1 55 ± 1 0.379
Final weight (g) 172 ± 9 177 ± 4 189 ± 9 169 ± 2 160 ± 3 176 ± 7 167 ± 1 0.078
Weight gain (%) 211 ± 17 yz 220 ± 8 yz 241 ± 16 z 201 ± 7 yz 185 ± 4y 224 ± 8 yz 206 ± 3 yz 0.039
FCR 1.40 ± 0.05 1.42 ± 0.03 1.45 ± 0.02 1.45 ± 0.01 1.51 ± 0.01 1.55 ± 0.05 1.53 ± 0.01 0.036a
SGR (% body
weight/d) 2.02 ± 0.10 yz 2.08 ± 0.04 yz 2.19 ± 0.08 z 1.97 ± 0.04 yz 1.87 ± 0.03 y 2.10 ± 0.04 yz 2.00 ± 0.01 yz 0.034
Feed intake
(% body weight/d) 2.98 ± 0.07 x 3.12 ± 0.02 xy 3.37 ± 0.10 yz 3.00 ± 0.10 x 2.94 ± 0.03 x 3.44 ± 0.04 z 3.21 ± 0.04 xyz <0.001
a
Although the one-way ANOVA indicated a significant treatment effect, the more conservative Tukey’s HSD pairwise comparison tests indicated no significant differences among
means.

RESULTS fatty acid profiles more similar to those observed in the STD
According to the one-way ANOVA, production performance SO group. In general, fatty acid profile change was more overt,
Downloaded by [Department Of Fisheries] at 19:58 08 April 2013

varied significantly among juvenile Cobias fed diets with dif-
ferent lipid and fatty acid compositions (Table 3). Although
survival (range = 97–100%), final weight (160–189 g), and
feed conversion ratio (FCR) (1.40–1.52) were unaffected by di-
etary treatment, significant differences were observed in weight
gain (185–241%), specific growth rate (SGR) (1.87–2.19% body
weight/d), and feed intake (2.94–3.44% body weight/d). Only
the HYD SO + PL group exhibited significantly reduced growth
in comparison with the FO control group, and none of the exper-
imental groups was significantly different from the control with
respect to SGR. Feed intake was significantly increased rela-
tive to the control only among fish fed the STD SO + PL and
SFA SO feeds. Although statistically significant results were
observed for some parameters, in several cases the magnitude
of these differences was relatively small. The results of the two-
way ANOVA revealed significant effects of soybean oil type
on final weight, weight gain, FCR, SGR, and feed intake, with
standard soybean oil generally outperforming the other soybean
lipids when oil types were pooled across phospholipid sup-
plementation treatments. Phospholipid supplementation had no
significant effect on any of the performance measures, and the
only significant interaction between soybean oil type and PL
supplementation observed was for feed intake.
Differences in dietary lipid composition and fatty acid profile
gave rise to significant differences in tissue fatty acid composi-
tion with respect to nearly all of the major fatty acids and fatty
acid groupings quantified in fillet (Table 4), brain (Table 5), liver
(Table 6), and eye (Table 7) total lipid. Generally, tissues of fish
fed the FO feed contained higher levels of fish oil–associated
fatty acids, such as 20:5(n-3) and 22:6(n-3). In contrast, the tis-
sues of fish fed the STD SO–, HYD SO–, and SFA SO–based
feeds generally contained higher levels of the fatty acids associ-
ated with these lipids, i.e., 18:2(n-6), 18:1(n-9), and, to a lesser
extent, 18:0. In terms of fatty acid composition, soybean lecithin
is most similar to standard soybean oil, so phospholipid supple-
mentation had the general effect of making dietary and tissue
in terms of the number and magnitude of significant differences
observed, among the fillet and liver tissues than the brain and
eye tissues. Despite these general trends, some important devi-
ations were observed. LC-PUFAs were conserved, to a greater
or lesser extent, among the tissue types and dietary treatment
groups. Of all dietary treatments, FO yielded the highest levels
of 20:5(n-3) in all tissues. Among other treatments, the 20:5(n-3)
levels in the brain and eye samples did not significantly differ
regardless of dietary treatment. However, in fillets, the SFA SO
treatment resulted in significantly greater 20:5(n-3) deposition
than any other non–fish oil treatment. SFA SO also resulted
in the second-highest 20:5(n-3) deposition in livers, though
these levels were not significantly different from those asso-
ciated with the HYD SO treatment (Table 6). Fillet 22:6(n-3)
was highest in the SFA SO treatment group, and fillets from SFA
SO + PL–treated fish did not differ significantly from those of
FISH ONLY–treated fish in 22:6(n-3) content. STD SO– and
STD SO + PL–treated fish had the lowest levels of 22:6(n-3)
in eye tissue, while other treatments did not significantly differ
in this respect. HYD SO–treated fish had the most 22:6(n-3) in
their livers, followed by SFA SO–treated fish. In fact, the total
LC-PUFA content in fillets did not significantly differ between
the FO and SFA SO treatments, and the SFA SO + PL treatment
fell only slightly below the FO treatment in this respect (Ta-
ble 4). Additionally, dietary increases in monounsaturated fatty
acids (MUFAs; i.e., fatty acids with one double bond) and SFAs
in the HYD SO– and SFA SO–based feeds, respectively, were
not proportionally reflected in the tissues: tissue levels of SFAs,
particularly 18:0, were significantly higher in some cases but to
a lesser extent than would be anticipated based on proportional
enrichment based on dietary levels.
Coefficient of distance information (Figure 1) indicated that
overall the STD SO and STD SO + PL treatments resulted in
the most greatly modified tissue fatty acid profiles, whereas the
tissue profiles of SFA SO– and SFA SO + PL–fed fish were most
similar to those of the FO–fed fish. The profiles of tissues from
 322 TRUSHENSKI ET AL.

TABLE 4. Fillet fatty acid composition (g/100 g) with respect to major fatty acids (>2% of total fatty acid methyl esters quantified). Data are presented as
least-squares means ± SEs of multiple tissue subsamples from triplicate tanks (N = 3); SEs < 0.1 are reported as 0.0. The P-values resulting from one-way
ANOVA tests are also provided. Within rows, means with common lowercase letters are not significantly different (P > 0.05).

STD HYD SFA

Fatty acid(s) FO STD SO SO + PL HYD SO SO + PL SFA SO SO + PL P-value

14:0 5.6 ± 0.2 z 2.7 ± 0.2 y 2.8 ± 0.2 y 3.4 ± 0.2 y 3.1 ± 0.2 y 3.6 ± 0.2 y 3.5 ± 0.2 y <0.001
16:0 21.9 ± 0.2 z 19.1 ± 0.2 x 20.2 ± 0.2 y 20.9 ± 0.2 y 20.9 ± 0.2 y 22.8 ± 0.2 z 22.6 ± 0.2 z <0.001
18:0 6.6 ± 0.1 wx 6.5 ± 0.2 w 6.8 ± 0.2 wx 7.2 ± 0.1 xy 7.0 ± 0.2 wxy 8.5 ± 0.2 z 7.7 ± 0.2 y <0.001
SFAsa 34.4 ± 0.3 z 28.6 ± 0.3 w 30.0 ± 0.3 wx 31.7 ± 0.3 y 31.2 ± 0.3 xy 35.2 ± 0.3 z 34.0 ± 0.3 z <0.001
16:1(n-7) 8.4 ± 0.2 z 3.0 ± 0.2 v 3.2 ± 0.2 vw 3.9 ± 0.1 wx 3.5 ± 0.2 vw 5.0 ± 0.2 y 4.4 ± 0.2 xy <0.001
18:1(n-7) 3.6 ± 0.0 y 2.3 ± 0.0 v 2.3 ± 0.0 v 4.2 ± 0.0 z 3.5 ± 0.0 y 3.1 ± 0.0 x 2.8 ± 0.0 w <0.001
18:1(n-9) 11.8 ± 0.2 u 16.3 ± 0.2 x 15.2 ± 0.2 w 25.6 ± 0.2 z 20.0 ± 0.2 y 13.3 ± 0.2 v 13.5 ± 0.2 v <0.001
MUFAsb 24.5 ± 0.3 x 22.0 ± 0.3 w 21.1 ± 0.3 w 34.1 ± 0.3 z 27.5 ± 0.3 y 21.9 ± 0.3 w 21.2 ± 0.3 w <0.001
18:2(n-6) 8.6 ± 0.3 w 29.1 ± 0.3 z 27.2 ± 0.3 y 10.8 ± 0.3 v 19.1 ± 0.3 x 12.1 ± 0.3 v 18.5 ± 0.3 x <0.001
20:4(n-6) 1.4 ± 0.0 wx 1.1 ± 0.1 w 1.3 ± 0.1 wx 1.5 ± 0.0 xy 1.4 ± 0.1 wx 2.0 ± 0.1 z 1.7 ± 0.1 y <0.001
n-6c 10.0 ± 0.3 u 30.3 ± 0.3 z 28.5 ± 0.3 y 12.4 ± 0.3 v 20.5 ± 0.3 x 14.1 ± 0.3 w 20.2 ± 0.3 x <0.001
18:3(n-3) 1.4 ± 0.1 x 3.2 ± 0.1 z 2.9 ± 0.1 z 0.8 ± 0.1 w 1.8 ± 0.1 y 1.1 ± 0.1 w 1.8 ± 0.1 y <0.001
Downloaded by [Department Of Fisheries] at 19:58 08 April 2013

20:5(n-3) 8.5 ± 0.1 z 2.7 ± 0.1 v 3.0 ± 0.1 wv 3.9 ± 0.1 x 3.3 ± 0.1 w 4.9 ± 0.1 y 4.1 ± 0.1 x <0.001
22:5(n-3) 2.7 ± 0.0 z 1.0 ± 0.0 u 1.1 ± 0.0 v 1.3 ± 0.0 w 1.2 ± 0.0 v 1.9 ± 0.0 y 1.6 ± 0.0 x <0.001
22:6(n-3) 15.6 ± 0.6 y 11.2 ± 0.6 w 12.3 ± 0.6 wx 14.8 ± 0.6 xy 13.3 ± 0.6 wxy 19.4 ± 0.7 z 15.9 ± 0.6 y <0.001
n-3d 31.2 ± 0.6 z 19.1 ± 0.6 x 20.3 ± 0.6 x 21.9 ± 0.6 xy 20.8 ± 0.6 x 28.8 ± 0.7 z 24.7 ± 0.6 y <0.001
PUFAse 41.1 ± 0.6 x 49.4 ± 0.6 z 48.8 ± 0.6 z 34.2 ± 0.6 w 41.3 ± 0.6 x 42.9 ± 0.6 xy 44.9 ± 0.6 y <0.001
C18 PUFAsf 11.8 ± 0.4 v 32.9 ± 0.4 z 30.7 ± 0.4 y 12.3 ± 0.4 vw 21.6 ± 0.4 x 14.0 ± 0.4 w 21.0 ± 0.4 x <0.001
LC-PUFAsg 29.3 ± 0.7 z 16.5 ± 0.7 w 18.1 ± 0.7 w 21.9 ± 0.7 xy 19.7 ± 0.7 wx 28.9 ± 0.8 z 23.8 ± 0.7 y <0.001
a
Sum of all fatty acids without double bonds; includes 20:0 in addition to the individually reported SFAs.
b
Sum of all fatty acids with a single double bond; includes 20:1(n-9) in addition to the individually reported MUFAs.
c
Sum of all n-6 fatty acids.
d
Sum of all n-3 fatty acids; includes 18:4(n-3) and 20:4(n-3) in addition to the individually reported n-3 fatty acids.
e
Sum of all fatty acids with ≥2 double bonds; includes 18:4(n-3) and 20:4(n-3) in addition to the individually reported PUFAs.
f
Sum of all PUFAs with chain lengths of 18 carbon atoms; includes 18:4(n-3) in addition to the individually reported C18 PUFAs.
g
Sum of all fatty acids with chain lengths ≥20 carbon atoms and double bonds ≥3; includes 20:4(n-3) in addition to the individually reported LC-PUFAs.

fish fed the HYD SO and HYD SO + PL feeds were intermediate
between those of the fish fed the other pairs of soybean oil–
based feeds. Supplementing the STD SO treatment with PL
decreased its Djh values, whereas supplementing either the HYD
SO or SFA SO treatments with PL generally increased their
Djh values. Regardless, PL-supplemented diets tended to yield
tissue profiles broadly consistent with their nonsupplemented
counterparts. As observed for individual fatty acids, Djh values
suggested that the magnitude of diet-related fatty acid profile
change was greatest among liver (range of Djh values = 7–27)
and fillet tissues (8–23) and smallest in the eye (6–17) and brain
tissues (2–3).
DISCUSSION
As long as essential fatty acid requirements are met, fish
performance is typically not affected by changes in dietary lipid
sources. Our results are consistent with this generalization:
previous research demonstrated that Cobias require 0.8–1.2%
22:6(n-3) (dry matter basis) in diets containing approximately
10–11% lipid (Trushenski et al. 2012) and, in the present
work, comparable diets (∼10% lipid, no PL supplementation)
containing 0.7–1.1% 22:6(n-3) yielded equivalent performance.
It is interesting to note that PL supplementation appeared to
improve the STD SO formulation (weight gain increased from
220% to 241%) but had a somewhat negative effect on growth
performance when added to the HYD SO (weight gain = 201%
versus 185%) and SFA SO (224% versus 206%) formulations.
However, these effects were not statistically significant (i.e.,
one-way ANOVA–Tukey’s HSD tests indicated no differences
between unsupplemented and PL-supplemented treatment
pairs; two-way ANOVA indicated no significant effect of phos-
pholipid supplementation) and may be more related to minor
differences in dietary proximate composition and feed intake.
Numerically, the HYD SO + PL feed yielded the lowest weight
gain and the highest FCR. However, dietary protein content and
feed intake were both lowest in this treatment. Alone, these mi-
nor differences in protein content and feed intake are unlikely to
have substantially affected growth, but perhaps the combination
of these effects—not partially hydrogenated soybean oil or PL
supplementation—is to blame for the slightly impaired per-
formance observed. Although previous research has suggested
that the provision of marine-origin phospholipid is a means of
improving the performance and physiological competence of
Cobias fed reduced–fish meal feeds, previous and current results
suggest that the benefits of soybean lecithin supplementation in
juvenile Cobia feeds are limited (Trushenski et al. 2013). Niu
et al. (2008a, 2008b) reported that PL supplementation (8%
 SOYBEAN OILS IN COBIA FEEDS 323

TABLE 5. Brain fatty acid composition (g/100 g) with respect to major fatty acids (>2% of total fatty acid methyl esters quantified). Data are presented as
least-squares means ± SEs of multiple tissue subsamples from triplicate tanks (N = 3); SEs < 0.1 are reported as 0.0. The P-values resulting from one-way
ANOVA tests are also provided. Within rows, means with common lowercase letters are not significantly different (P > 0.05).

STD HYD SFA

Fatty acid(s) FO STD SO SO + PL HYD SO SO + PL SFA SO SO + PL P-value
16:0 18.7 ± 0.3 18.5 ± 0.3 18.8 ± 0.3 19.3 ± 0.3 18.7 ± 0.3 19.1 ± 0.3 18.9 ± 0.3 0.600
18:0 16.4 ± 0.4 16.3 ± 0.3 16.3 ± 0.3 16.2 ± 0.3 16.2 ± 0.3 17.1 ± 0.3 16.1 ± 0.3 0.428
24:0 2.7 ± 0.1 2.7 ± 0.1 2.6 ± 0.1 2.6 ± 0.1 2.9 ± 0.1 2.6 ± 0.1 2.6 ± 0.1 0.705
SFAsa 39.6 ± 0.3 39.3 ± 0.3 39.4 ± 0.3 39.8 ± 0.3 39.5 ± 0.3 40.6 ± 0.3 39.5 ± 0.3 0.132
18:1(n-7) 2.1 ± 0.0 z 1.8 ± 0.0 y 1.8 ± 0.0 y 1.8 ± 0.0 y 1.8 ± 0.0 y 1.9 ± 0.0 yz 1.9 ± 0.0 yz <0.001
18:1(n-9) 20.8 ± 0.4 21.1 ± 0.3 20.5 ± 0.3 21.8 ± 0.3 21.9 ± 0.4 20.8 ± 0.4 20.8 ± 0.4 0.074
MUFAsb 25.0 ± 0.4 24.6 ± 0.4 24.0 ± 0.4 25.6 ± 0.4 25.5 ± 0.4 24.4 ± 0.4 24.7 ± 0.4 0.097
18:2(n-6) 1.1 ± 0.5 y 3.9 ± 0.4 z 3.6 ± 0.4 z 1.6 ± 0.4 y 2.1 ± 0.4 yz 1.3 ± 0.4 y 2.8 ± 0.4 yz 0.002
20:4(n-6) 1.7 ± 0.0 x 2.0 ± 0.0 y 2.0 ± 0.0 y 2.2 ± 0.0 z 2.1 ± 0.0 yz 2.2 ± 0.0 z 2.1 ± 0.0 yz <0.001
n-6c 3.1 ± 0.4 x 6.6 ± 0.4 z 6.4 ± 0.4 z 4.0 ± 0.4 xy 4.8 ± 0.4 xyz 3.8 ± 0.4 xy 5.3 ± 0.4 yz <0.001
20:5(n-3) 2.2 ± 0.1 z 1.3 ± 0.1 y 1.3 ± 0.1 y 1.5 ± 0.1 y 1.3 ± 0.1 y 1.4 ± 0.1 y 1.6 ± 0.1 y <0.001
22:5(n-3) 2.4 ± 0.0 z 1.6 ± 0.0 w 1.7 ± 0.0 xw 1.9 ± 0.0 y 1.8 ± 0.0 xy 1.8 ± 0.0 wxy 1.8 ± 0.0 wxy <0.001
Downloaded by [Department Of Fisheries] at 19:58 08 April 2013

22:6(n-3) 27.4 ± 0.6 26.3 ± 0.5 26.9 ± 0.5 27.1 ± 0.5 27.1 ± 0.5 27.9 ± 0.5 26.8 ± 0.5 0.521
n-3d 32.4 ± 0.4 z 29.5 ± 0.4 y 30.2 ± 0.4 y 30.6 ± 0.4 yz 30.2 ± 0.4 y 31.2 ± 0.4 yz 30.5 ± 0.4 yz 0.009
PUFAse 35.5 ± 0.4 yz 36.1 ± 0.3 yz 36.6 ± 0.3 z 34.6 ± 0.3 y 35.0 ± 0.4 yz 35.0 ± 0.4 yz 35.8 ± 0.4 yz 0.016
C18 PUFAsf 1.2 ± 0.6 y 4.1 ± 0.5 z 3.9 ± 0.5 z 1.7 ± 0.5 yz 2.1 ± 0.5 yz 1.3 ± 0.5 y 3.1 ± 0.5 yz 0.004
LC-PUFAsg 34.2 ± 0.5 z 31.6 ± 0.5 y 32.3 ± 0.5 yz 32.8 ± 0.5 yz 32.6 ± 0.5 yz 33.6 ± 0.5 yz 32.6 ± 0.5 yz 0.047
a
Sum of all fatty acids without double bonds; includes 14:0, 20:0, and 22:0 in addition to the individually reported SFAs.
b
Sum of all fatty acids with a single double bond; includes 20:1(n-9) in addition to the individually MUFAs.
c
Sum of all n-6 fatty acids; includes 18:3(n-6), 20:2(n-6), and 20:3(n-6) in addition to the individually reported n-6 fatty acids.
d
Sum of all n-3 fatty acids; includes 18:3(n-3), 18:4(n-3), 20:3(n-3), and 20:4(n-3) in addition to individually reported (n-3) fatty acids.
e
Sum of all fatty acids with ≥2 double bonds; includes 18:3(n-6), 18:3(n-3), 18:4(n-3), 20:2(n-6), 20:3(n-6), 20:3(n-3), and 20:4(n-3) in addition to the individually reported PUFAs.
f
Sum of all PUFAs with chain lengths of 18 carbon atoms; includes 18:3(n-6), 18:3(n-3), and 18:4(n-3) in addition to the individually reported C18 PUFAs.
g
Sum of all fatty acids with chain lengths ≥20 carbon atoms and double bonds ≥3; includes 20:3(n-6), 20:3(n-3), and 20:4(n-3) in addition to the individually reported LC-PUFAs.

supplemental lecithin) was critical for ensuring survival and
adequate growth among larval Cobias. The feeds used by these
authors contained more than 50% fish meal, and thus a consid-
erable amount of marine-origin PL prior to supplementation.
Nonetheless, significant improvement was associated with addi-
tional PL supplementation. In light of these data, it is somewhat
surprising that the addition of 2% PL to our feeds, which con-
tained less fish meal and therefore less marine-origin PL, had no
effect on performance as our diets certainly contained less than
the levels recommended by Niu et al. (2008a, 2008b) for larval
Cobias. However, one must consider that absolute nutrient
requirements may change from one life stage to another. While
fish may require PL throughout their life, it may be at lower lev-
els than necessary during early life history (Tocher et al. 2008;
Cahu et al. 2009; Hansen et al. 2011), and dietary levels of native
phospholipid (i.e., marine-origin PL associated with fish meal)
may be reduced much further in juvenile Cobia feeds before PL
insufficiency (and corresponding positive effects of soybean
lecithin supplementation) are observed (Trushenski et al.,
in press).
As anticipated, fatty acid intake was largely reflected in Cobia
tissues, particularly the fillet and liver tissues, which are known
to be more responsive to changes in fatty acid intake (Noffs
et al. 2009; Trushenski et al. 2012): the tissues of Cobias fed the
FO feed contained more LC-PUFAs, the tissues of those fed the
STD SO and STD SO + PL feeds contained more C18 PUFAs,
the tissues of those fed the HYD SO and HYD SO + PL feeds
contained more MUFAs, and the tissues of those fed the SFA SO
and SFA SO + PL feeds contained more SFAs, at least in com-
parison with fish fed the other soybean lipid–based feeds. How-
ever, as has been observed in several other species, LC-PUFAs
were disproportionately enriched within the tissues (Bell et al.
2001, 2002, 2003; Trushenski et al. 2008a; Fountoulaki et al.
2009; Trushenski et al. 2011b), whereas SFAs tended to be un-
derrepresented when dietary intake increased (Lane et al. 2006;
Trushenski et al. 2008a; Turchini et al. 2011). Although the
brain, eye, liver, and fillet tissue levels of LC-PUFAs exceeded
dietary levels in all dietary treatments, SFAs were only slightly
enriched in the fillets and livers of fish in the SFA SO and SFA
SO + PL treatment groups even though SFAs represented more
than half of the fatty acid content of these feeds.
The disproportionately low tissue levels of SFAs are likely
the result of intensive catabolism of these fatty acids. Generally,
catabolism of fatty acids is a function of the molecular structure
of the fatty acids themselves as well as their relative abundance.
Saturated fatty acids tend to be directed to catabolic pathways
more than unsaturated fatty acids, but surplus unsaturates, in-
cluding LC-PUFAs, can also be shunted into beta oxidation
(Henderson and Sargent 1985; Kiessling and Kiessling 1993;
Henderson 1996; Frøyland et al. 2000; Torstensen et al. 2004;
 324 TRUSHENSKI ET AL.

TABLE 6. Liver fatty acid composition (g/100 g) with respect to major fatty acids (>2% of total fatty acid methyl esters quantified). Data are presented as
least-squares means ± SEs of multiple tissue subsamples from triplicate tanks (N = 3); SEs < 0.1 are reported as 0.0. The P-values resulting from one-way
ANOVA tests are also provided. Within rows, means with common lowercase letters are not significantly different (P > 0.05).

STD HYD SFA

Fatty acid(s) FO STD SO SO + PL HYD SO SO + PL SFA SO SO + PL P-value

16:0 17.1 ± 0.4 xy 16.4 ± 0.4 x 16.7 ± 0.4 xy 16.3 ± 0.4 x 16.2 ± 0.4 x 19.4 ± 0.4 z 18.3 ± 0.4 yz <0.001
18:0 6.6 ± 0.4 z 4.1 ± 0.4 y 4.5 ± 0.4 y 7.3 ± 0.5 z 7.1 ± 0.5 z 7.0 ± 0.4 z 6.0 ± 0.4 yz <0.001
SFAsa 24.8 ± 0.5 y 21.3 ± 0.5 w 22.0 ± 0.4 xw 24.2 ± 0.5 xy 23.7 ± 0.5 xy 27.5 ± 0.5 z 25.3 ± 0.5 yz <0.001
16:1(n-7) 5.5 ± 0.2 z 2.5 ± 0.2 wx 2.5 ± 0.2 wx 2.7 ± 0.2 wxy 2.1 ± 0.2 w 3.5 ± 0.2 y 3.3 ± 0.2 xy <0.001
18:1(n-7) 4.7 ± 0.1 z 2.5 ± 0.1 vw 2.6 ± 0.1 w 2.1 ± 0.1 uv 1.9 ± 0.1 u 3.7 ± 0.1 y 3.3 ± 0.1 x <0.001
18:1(n-9) 11.9 ± 0.4 wx 15.2 ± 0.4 yz 13.9 ± 0.4 xy 17.4 ± 0.5 z 15.2 ± 0.5 yz 11.3 ± 0.4 w 12.4 ± 0.5 wx <0.001
MUFAsb 22.5 ± 0.7 z 20.6 ± 0.7 yz 19.4 ± 0.6 y 22.8 ± 0.7 z 19.6 ± 0.7 yz 19.1 ± 0.7 y 19.3 ± 0.7 y 0.005
18:2(n-6) 9.6 ± 0.6 w 33.3 ± 0.6 z 31.7 ± 0.6 z 12.2 ± 0.6 wx 21.1 ± 0.6 y 13.9 ± 0.6 x 21.4 ± 0.6 y <0.001
20:4(n-6) 2.8 ± 0.2 vwx 1.8 ± 0.2 v 2.1 ± 0.2 vw 5.0 ± 0.2 z 4.1 ± 0.2 yz 3.9 ± 0.2 xy 3.0 ± 0.2 wxy <0.001
n-6c 12.4 ± 0.4 w 35.1 ± 0.4 z 33.8 ± 0.4 z 17.3 ± 0.5 x 25.3 ± 0.5 y 17.8 ± 0.4 x 24.5 ± 0.4 y <0.001
18:3(n-3) 1.4 ± 0.1 x 3.5 ± 0.1 z 3.3 ± 0.1 z 0.7 ± 0.1 w 1.6 ± 0.1 xy 1.2 ± 0.1 x 2.0 ± 0.1 y <0.001
20:5(n-3) 8.0 ± 0.1 z 2.6 ± 0.1 w 2.7 ± 0.1 w 4.6 ± 0.1 y 3.7 ± 0.1 x 4.3 ± 0.1 y 3.5 ± 0.1 x <0.001
Downloaded by [Department Of Fisheries] at 19:58 08 April 2013

22:5(n-3) 4.2 ± 0.1 z 1.7 ± 0.1 w 1.9 ± 0.1 w 2.3 ± 0.1 x 2.1 ± 0.1 wx 2.8 ± 0.1 y 2.5 ± 0.1 y <0.001
22:6(n-3) 25.2 ± 0.7 xyz 14.6 ± 0.7 w 16.3 ± 0.7 w 27.1 ± 0.7 z 23.2 ± 0.7 xy 26.2 ± 0.7 yz 22.0 ± 0.7 x <0.001
n-3d 40.2 ± 0.6 z 23.0 ± 0.7 w 24.7 ± 0.6 w 35.7 ± 0.7 y 31.4 ± 0.7 x 35.5 ± 0.6 y 30.9 ± 0.7 x <0.001
PUFAse 52.6 ± 0.5 w 58.1 ± 0.5 z 58.6 ± 0.5 z 53.0 ± 0.5 wx 56.6 ± 0.5 yz 53.3 ± 0.5 wx 55.4 ± 0.5 xy <0.001
C18 PUFAsf 11.5 ± 0.7 w 37.0 ± 0.7 z 35.3 ± 0.6 z 13.5 ± 0.7 wx 23.1 ± 0.7 y 15.6 ± 0.7 x 23.8 ± 0.7 y <0.001
LC-PUFAsg 41.1 ± 0.9 z 21.1 ± 0.9 w 23.3 ± 0.9 w 39.5 ± 1.0 z 33.5 ± 1.0 xy 37.7 ± 0.9 yz 31.6 ± 0.9 x <0.001
a
Sum of all fatty acids without double bonds; includes 14:0 and 20:0 in addition to the individually reported SFAs.
b
Sum of all fatty acids with a single double bond; includes 20:1(n-9) in addition to the individually reported MUFAs.
c
Sum of all n-6 fatty acids; includes 18:3(n-6), 20:2(n-6), and 20:3(n-6) in addition to the individually reported (n-6) fatty acids.
d
Sum of all n-3 fatty acids; includes 18:3(n-3), 18:4(n-3), 20:3(n-3), and 20:4(n-3) in addition to the individually reported n-3 fatty acids.
e
Sum of all fatty acids with ≥2 double bonds; includes 18:3(n-6), 18:3(n-3), 18:4(n-3), 20:2(n-6), 20:3(n-6), 20:3(n-3), and 20:4(n-3) in addition to the individually reported PUFAs.
f
All PUFAs with chain lengths of 18 carbon atoms; includes 18:3(n-6), 18:3(n-3), and 18:4(n-3) in addition to the individually reported C18 PUFAs.
g
Sum of all fatty acids with chain lengths ≥20 carbon atoms and double bonds ≥3; includes 20:3(n-6), 20:3(n-3), and 20:4(n-3) in addition to the individually reported LC-PUFAs.

Stubhaug et al. 2007). Increased catabolism of SFAs, coupled
with the relatively poor affinity of biosynthetic enzymes (e.g.,
acyltransferases) for most SFAs (Trushenski et al. 2008b), read-
ily explains the minor tissue enrichment noted for these fatty
acids, even among Cobias fed the SFA-rich SFA soy oilSO
feeds.
The disproportionately high tissue levels of LC-PUFAs can
be explained, in part, by the physiological relevance of these
fatty acids in some tissues (e.g., 22:6[n-3] in brain and eye mem-
branes; Glencross 2009) and preferential inclusion of polyunsat-
urated fatty acids in certain lipids, particularly some phospho-
lipids (Tocher et al. 2008; Trushenski et al. 2008b). Although it
has been reported that Cobias express the elongase and desat-
urase genes involved in the synthesis of LC-PUFAs from C18
precursors (Zheng et al. 2009; Monroig et al. 2011), perhaps
as early as 12–18 h postfertilization, the functional importance
of de novo LC-PUFA synthesis in this species is negligible,
as it is apparently insufficient to compensate for dietary LC-
PUFA deficiencies to maintain growth or tissue levels of these
fatty acids (Faulk and Holt 2005; Liu et al. 2007; Wang et al.
2007; Nhu et al. 2011; Trushenski et al. 2011e, 2012). It is
relatively easy to explain tissue LC-PUFA enrichment in Co-
bias as a function of selective retention of these fatty acids,
though de novo synthesis may also contribute in a more minor
sense. However, it is more difficult to explain differences in
the magnitude of LC-PUFA tissue enrichment among groups of
fish with essentially equivalent LC-PUFA intake. For example,
the dietary 22:6(n-3) content was very similar among the soy-
bean lipid–based feeds (0.7–0.8% of the diet, dry matter basis),
and minor differences in intake were largely attenuated by the
slight differences in feed intake (i.e., intake tended to be ele-
vated among diets containing slightly lower levels of 22:6[n-3]).
Assuming broadly equivalent intake of 22:6(n-3), one might an-
ticipate broadly equivalent tissue levels, but the opposite was
observed. Despite equivalent intake, fillet levels of 22:6(n-3)
varied widely (values ranged from 11.2% FAMEs in the STD
SO group to 19.4% in the SFA SO group), as did liver levels
(values ranged from 14.6% in the STD SO group to 27.1% in
the HYD SO group). In general, the magnitude of 22:6(n-3)
and total LC-PUFA enrichment was positively correlated with
dietary SFA content and negatively correlated with C18 PUFA
intake. Similar responses have been reported for other fishes, in-
cluding Largemouth Bass Micropterus salmoides (Laporte and
Trushenski 2011), hybrid Striped Bass (White Bass Morone
chrysops × Striped Bass M. saxatilis (Trushenski et al. 2008a),
Rainbow Trout Oncorhynchus mykiss (Trushenski et al. 2011a,
2011c), Nile Tilapia Oreochromis niloticus (Trushenski et al.
2009), Malaysian Mahseer (also known as Thai Mahseer) Tor
tambroides (Ramezani-Fard et al. 2012), and Murray Cod Mac-
cullochella peelii (Turchini et al. 2011).
 SOYBEAN OILS IN COBIA FEEDS 325

TABLE 7. Eye fatty acid composition (g/100 g) with respect to major fatty acids (>2% of total fatty acid methyl esters quantified). Data are presented as
least-squares means ± SEs of multiple tissue subsamples from triplicate tanks (N = 3); SEs < 0.1 are reported as 0.0. The P-values resulting from one-way
ANOVA tests are also provided. Within rows, means with common lowercase letters are not significantly different (P > 0.05).

STD HYD SFA

Fatty acid(s) FO STD SO SO + PL HYD SO SO + PL SFA SO SO + PL P-value

14:0 4.8 ± 0.2 z 2.9 ± 0.2 y 3.0 ± 0.3 y 3.3 ± 0.2 y 3.4 ± 0.2 y 3.8 ± 0.3 yz 3.7 ± 0.2 y 0.001
16:0 21.7 ± 0.4 yz 20.2 ± 0.4 y 21.0 ± 0.4 yz 22.0 ± 0.4 yz 21.8 ± 0.4 yz 22.5 ± 0.5 z 22.2 ± 0.4 z 0.023
18:0 8.0 ± 0.5 8.0 ± 0.5 8.1 ± 0.5 9.7 ± 0.5 8.4 ± 0.5 10.2 ± 0.6 9.2 ± 0.5 0.057
SFAsa 34.5 ± 0.5 xyz 31.1 ± 0.5 v 32.0 ± 0.5 vx 35.0 ± 0.5 yz 33.5 ± 0.5 vxy 36.6 ± 0.6 z 35.1 ± 0.5 yz <0.001
16:1(n-7) 7.2 ± 0.3 z 3.5 ± 0.3 y 3.6 ± 0.3 y 4.2 ± 0.3 y 4.2 ± 0.3 y 4.8 ± 0.3 y 4.6 ± 0.3 y <0.001
18:1(n-7) 3.7 ± 0.1 z 2.5 ± 0.1 x 2.6 ± 0.1 x 2.7 ± 0.1 x 2.5 ± 0.1 x 3.3 ± 0.1 y 3.2 ± 0.1 y <0.001
18:1(n-9) 15.9 ± 0.6 x 18.0 ± 0.6 xy 17.2 ± 0.6 x 22.7 ± 0.6 z 20.2 ± 0.6 yz 16.0 ± 0.7 x 17.3 ± 0.6 x <0.001
MUFAsb 27.7 ± 0.7 yz 24.6 ± 0.7 xy 24.0 ± 0.8 x 30.3 ± 0.7 z 27.6 ± 0.7 yz 25.0 ± 0.9 xy 25.9 ± 0.7 xy <0.001
18:2(n-6) 6.6 ± 0.4 x 22.3 ± 0.5 z 21.0 ± 0.5 z 8.0 ± 0.5 x 14.6 ± 0.4 y 8.2 ± 0.6 x 12.6 ± 0.4 y <0.001
20:4(n-6) 1.5 ± 0.1 xy 1.3 ± 0.1 x 1.4 ± 0.1 xy 1.8 ± 0.1 yz 1.5 ± 0.1 xy 2.0 ± 0.1 z 1.7 ± 0.1 xyz 0.002
n-6c 8.1 ± 0.4 x 23.6 ± 0.4 z 22.4 ± 0.4 z 9.8 ± 0.4 x 16.1 ± 0.4 y 10.3 ± 0.5 x 14.3 ± 0.4 y <0.001
18:3(n-3) 1.1 ± 0.1 wx 2.5 ± 0.1 z 2.3 ± 0.1 z 0.7 ± 0.1 v 1.5 ± 0.1 y 0.8 ± 0.1 vw 1.3 ± 0.1 xy <0.001
Downloaded by [Department Of Fisheries] at 19:58 08 April 2013

20:5(n-3) 7.3 ± 0.3 z 3.0 ± 0.3 y 3.2 ± 0.3 y 3.7 ± 0.3 y 3.7 ± 0.3 y 4.3 ± 0.3 y 3.9 ± 0.3 y <0.001
22:5(n-3) 2.8 ± 0.1 z 1.1 ± 0.1 x 1.3 ± 0.1 x 1.8 ± 0.1 y 1.4 ± 0.1 xy 1.8 ± 0.1 y 1.6 ± 0.1 y <0.001
22:6(n-3) 17.1 ± 0.9 yz 13.5 ± 1.0 y 14.2 ± 1.0 y 18.1 ± 1.0 yz 15.4 ± 0.9 yz 20.3 ± 1.2 yz 17.2 ± 0.9 yz 0.005
n-3d 29.7 ± 0.6 z 20.7 ± 0.7 w 21.5 ± 0.7 w 24.9 ± 0.7 y 22.7 ± 0.6 wx 28.1 ± 0.8 yz 24.8 ± 0.6 xy <0.001
PUFAse 37.8 ± 0.5 y 44.3 ± 0.5 z 43.9 ± 0.6 z 34.7 ± 0.5 x 38.8 ± 0.5 y 38.4 ± 0.6 y 39.0 ± 0.5 y <0.001
C18 PUFAsf 9.1 ± 0.6 x 25.4 ± 0.6 z 23.9 ± 0.6 z 9.3 ± 0.6 x 16.8 ± 0.6 y 9.8 ± 0.7 x 14.6 ± 0.6 y <0.001
LC-PUFAg 28.6 ± 0.8 z 18.9 ± 0.8 w 20.0 ± 0.9 w 25.4 ± 0.8 xyz 22.0 ± 0.8 wx 28.5 ± 1.0 yz 24.4 ± 0.8 xy <0.001
a
Sum of all fatty acids without double bonds.
b
Sum of all fatty acids with a single double bond; includes 20:1(n-9) in addition to the individually reported MUFAs.
c
Sum of all n-6 fatty acids.
d
Sum of all n-3 fatty acids; includes 18:4(n-3) in addition to the individually reported n-3 fatty acids.
e
Sum of all fatty acids with ≥2 double bonds; includes 18:4(n-3) in addition to the individually reported PUFAs.
f
Sum of all PUFAs with chain lengths of 18 carbon atoms; includes 18:4(n-3) in addition to individually reported C18 PUFAs.
g
Sum of all fatty acids with chain lengths ≥20 carbon atoms and double bonds ≥3.

Two complementary hypotheses have emerged regarding the
“omega-3 sparing effect” (Turchini et al. 2011), one related to
reduced competition for tissue deposition between LC-PUFAs
and C18 PUFAs, and the other related to catabolic sparing of
LC-PUFAs by SFAs and/or MUFAs. The enzymes involved in
lipid synthesis show considerable affinity for LC-PUFAs as well
as C18 PUFAs, especially 18:2(n-6). Thus, at high intake lev-
els, C18 PUFAs may “outcompete” LC-PUFAs for deposition
during lipid synthesis and remodeling (Trushenski et al. 2008a,
2011b; Laporte and Trushenski 2011). Thus, SFAs may indi-
rectly improve the apparent efficiency of LC-PUFA metabolism
by limiting the competitive exclusion of LC-PUFAs by C18 PU-
FAs. However, as noted above, preferential inclusion/exclusion
of fatty acids is primarily associated with polar lipid synthesis
and remodeling. For tissues with high levels of neutral lipid,
which is synthesized with less specificity for certain fatty acids,
reduced competition from C18 PUFAs may not be the primary
mechanism by which SFA-rich feeds yield higher tissue levels
of LC-PUFAs. As previously mentioned, SFAs are preferred
substrates for beta oxidation and may literally spare LC-PUFAs
from catabolism (Turchini et al. 2011). Arguably, if fewer LC-
PUFAs are catabolized, more are available for tissue deposition.
However, LC-PUFAs are not typically catabolized unless con-
sumed in excess (Torstensen et al. 2004). Consequently, it seems
unlikely that catabolic sparing is the primary means by which
SFA-rich feeds result in greater LC-PUFA tissue enrichment.
Although both of these hypotheses provide plausible explana-
tions, neither seems wholly sufficient to explain the magnitude
of LC-PUFA sparing by SFAs. Likely, both hypotheses are cor-
rect in some respects, and tissue LC-PUFA levels may become
elevated among fish fed SFA-rich feeds as a result of both re-
duced competition from C18 PUFAs and catabolic sparing of
LC-PUFAs by SFAs.
The question remains whether feeding Cobias SFA-rich
lipids effectively lowers their requirements for 22:6(n-3) or other
LC-PUFAs. Production performance was not equivalent among
fish fed the fish oil–free feeds, but none of these feeds yielded
inferior performance relative to the fish oil–based control feed
and, as explained before, the minor discrepancies among treat-
ments are perhaps best explained as experimental artifacts. The
general absence of production performance effects is unsurpris-
ing, given that each of these feeds was formulated to meet the
22:6(n-3) and other nutrient requirements of juvenile Cobias.
Higher 22:6(n-3) content in the tissues of fish fed the partially
and fully hydrogenated soybean oils suggests greater efficiency
in 22:6(n-3) metabolism and the possibility that production per-
formance could have been maintained at even lower dietary
inclusion rates. Unfortunately, we cannot answer this question
 326 TRUSHENSKI ET AL.

Liver Fay Acid Profile Similarity Fillet Fay Acid Profile Similarity
30 30
Coefficient of Distance (Djh)

Coefficient of Distance (Djh)

25 25
20 20
15 15
10 10
5 5
0 0
STD SO STD SO + HYD SO HYD SO + SFA SO SFA SO + STD SO STD SO + HYD SO HYD SO + SFA SO SFA SO +
PL PL PL PL PL PL

Eye Fay Acid Profile Similarity Brain Fay Acid Profile Similarity
30 30
Coefficient of Distance (Djh)

Coefficient of Distance (Djh)

25 25
Downloaded by [Department Of Fisheries] at 19:58 08 April 2013

20 20
15 15
10 10
5 5
0 0
STD SO STD SO + HYD SO HYD SO + SFA SO SFA SO + STD SO STD SO + HYD SO HYD SO + SFA SO SFA SO +
PL PL PL PL PL PL

FIGURE 1. Coefficient of distance (Djh) values by dietary treatment and tissue type. The values compare the overall fatty acid profiles between the experimental
treatments and the FO positive control treatment, with smaller values indicating greater profile similarity and zero indicating perfect congruence between the
profiles.

unequivocally based on the present results, but they are nonethe-
less compelling and suggestive. Although further research is
necessary to demonstrate whether SFA-rich lipids can effec-
tively reduce the LC-PUFA requirements of juvenile Cobias, it
is clear that partially and fully hydrogenated soybean oils offer
strategic advantages over standard soybean oil in minimizing
the effects of fish oil replacement on tissue fatty acid profiles.
ACKNOWLEDGMENTS
We thank the Illinois Soybean Association, Virginia Sea
Grant, Commercial Fisheries and Shellfish Technologies, and
the International Initiative for Sustainable and Biosecure Aqua-
farming for providing funding and support for this research.
We also thank Brendan Delbos, Steve Urick, and Brian Gause
for their assistance in completing the feeding trial, harvesting
tissue samples, and processing feed samples, respectively. We
thank Archer Daniels Midland and Omega Protein, Inc., for the
donation of the feedstuffs used in the experimental feeds. We
also thank Luı́s A. Sampaio for his assistance in identifying and
evaluating Yamamoto as a participant through the International
Initiative for Sustainable and Biosecure Aquafarming student
exchange program.
REFERENCES
Bell, J. G., R. J. Henderson, D. R. Tocher, F. McGhee, J. R. Dick, A. Porter, R. P.
Smullen, and J. R. Sargent. 2002. Substituting fish oil with crude palm oil in
the diet of Atlantic Salmon (Salmo salar) affect muscle fatty acid composition
and hepatic fatty acid metabolism. Journal of Nutrition 132:222–230.
Bell, J. R., R. J. Henderson, D. R. Tocher, and J. R. Sargent. 2003. Replacement
of dietary fish oil with increasing levels of linseed oil: modification of flesh
fatty acid compositions in Atlantic Salmon (Salmo salar) using a fish oil
finishing diet. Lipids 39:223–232.
Bell, J. G., J. McEvoy, D. R. Tocher, F. McGhee, P. J. Campbell, and J. R.
Sargent. 2001. Replacement of fish oil with rapeseed oil in diets of Atlantic
Salmon (Salmo salar) affects tissue lipid compositions and hepatocyte fatty
acid metabolism. Journal of Nutrition 131:1535–1543.
Benetti, D. D., B. O’Hanlon, J. A. Rivera, A. W. Welch, C. Maxey, and M.
R. Orhun. 2010. Growth rate of Cobia (Rachycentron canadum) cultured in
open ocean submerged cages in the Caribbean. Aquaculture 302:195–201.
Cahu, C. L., E. Gisbert, L. A. N. Villeneuve, S. Morais, N. Hamza, P. A. Wold,
and J. L. Zambonio-Infante. 2009. Influence of dietary phospholipids on early
ontogenesis of fish. Aquaculture Research 40:989–999.
Chou, R. L., B. Y. Her, M. S. Su, G. Hwang, Y. H. Wu, and H. Y. Chen.
2004. Substituting fish meal with soybean meal in diets of juvenile Cobia
(Rachycentron canadum). Aquaculture 229:325–333.
Christie, W. W. 1982. Lipid analysis, 2nd edition. Pergamon, Oxford, UK.
FAO (Food and Agriculture Organization of the United Nations). 2010. The
state of world fisheries and aquaculture. FAO, Rome.
Faulk, C. K., and G. J. Holt. 2005. Advanced in rearing Cobia Rachycentron
canadum larvae in recirculating aquaculture systems: live pretty enrichment
and greenwater culture. Aquaculture 249:231–243.
 SOYBEAN OILS IN COBIA FEEDS
Folch, J., M. Lees, and G. H. Sloane-Stanley. 1957. A simple method for
the isolation and purification of total lipids from animal tissues. Journal of
Biological Chemistry 276:497–507.
Fountoulaki, E., A. Vasilaki, R. Hurtado, K. Grigorakis, I. Karacostas, I.
Nengas, G. Rigos, Y. Kotzamanis, B. Venou, and M. N. Alexis. 2009.
Fish oil substitution by vegetable oils in commercial diets for Gilthead
Sea Bream (Sparus aurata L.); effects on growth performance, flesh qual-
ity and fillet fatty acid profile: recovery of fatty acid profiles by a fish
oil finishing diet under fluctuating water temperatures. Aquaculture 289:
317–326.
Frøyland, L., Ø. Lie, and R. K. Berge. 2000. Mitochondrial and peroxisomal
b-oxidation capacities in various tissues from Atlantic Salmon (Salmo salar).
Aquaculture Nutrition 6:85–89.
Glencross, B. D. 2009. Exploring the nutritional demand for essential fatty acids
by aquaculture species. Reviews in Aquaculture 1:71–124.
Hansen, O. J., V. Puvanendran, J. P. Jostenson, and C. Ous. 2011. Effects of
dietary levels and ratio of phosphatidylcholine and phosphatidylinositol on
the growth, survival and deformity levels of Atlantic Cod larvae and early
juveniles. Aquaculture Research 42: 1026–1033.
Henderson, R. J. 1996. Fatty acid metabolism in freshwater fish with particular
Downloaded by [Department Of Fisheries] at 19:58 08 April 2013
reference to polyunsaturated fatty acids. Archives in Animal Nutrition 49:5–
22.
Henderson, R. J., and J. R. Sargent. 1985. Chain length specificities of mitochon-
drial and peroxisomal b-oxidation of fatty acids in livers of Rainbow Trout
(Salmo gairdneri). Comparative Biochemistry and Physiology B 82:79–85.
Kiessling, K.-H., and A. Kiessling A. 1993. Selective utilization of fatty acids in
Rainbow Trout (Oncorhynchus mykiss, Walbaum) red muscle mitochondria.
Canadian Journal of Zoology 71:248–251.
Lane, R. L., J. T. Trushenski, and C. C. Kohler. 2006. Modification of fillet
composition and evidence of differential fatty acid turnover in Sunshine Bass
Morone chrysops × M. saxatilis following change in dietary lipid source.
Lipids 41:1029–1038.
Laporte, J., and J. T. Trushenski. 2011. Growth performance and tissue fatty
acid composition of Largemouth Bass fed diets containing fish oil or blends
of fish oil and soy-derived lipids. North American Journal of Aquaculture
73:435–444.
Liao, I. C., T.-S. Huang, W.-S. Tsai, C.-M. Hsueh, S.-L. Chang, and E. M. Leano.
2004. Cobia culture in Taiwan: current status and problems. Aquaculture
237:155–165.
Liao, I. C., and E. M. Leaño. 2007. Cobia aquaculture: research, developments
and commercial production. Asian Fisheries Society, Fisheries Society of
Taiwan, World Aquaculture Society, and National Taiwan Ocean University.
Liu, X.-W., B.-P. Tank, K.-S. Mai, Q.-H. Ai, and Q.-C. Zhou. 2007. Ef-
fects of dietary (n-3 highly unsaturated fatty acids on growth and fatty
acid composition of juvenile Cobia (Rachycentron canadum). Acta Hy-
drobiologica Sinica 31:190–195. (In Chinese with English abstract.) Avail-
able: http://en.cnki.com.cn/Article en/CJFDTOTAL-SSWX200702006.htm.
(September 2012).
Monroig, Ó., K. Webb, L. Ibarra-Castro, G. J. Holt, and D. R. Tocher. 2011.
Biosynthesis of long-chain polyunsaturated fatty acids in marine fish: charac-
terization of an Elovl4-like elongase from Cobia Rachycentron canadum and
activation of the pathway during early life stages. Aquaculture 312:145–153.
Naylor, R. L., R. W. Hardy, D. P. Bureau, A. Chiu, M. Elliott, A. P. Farrell,
I. Foster, D. M. Gatlin, R. J. Goldberg, K. Hua, and P. D. Nichols. 2009.
Feeding aquaculture in an era of finite resources. Proceedings of the National
Academy of Sciences of the USA 106:15103–15110.
Nhu, V. C., H. Q. Nguyen, T. L. Le, M. T. Tran, P. Sorgeloos, K. Dierckens,
H. Reinertsen, E. Kjørsvik, and N. Svennevig. 2011. Cobia Rachycentron
canadum aquaculture in Vietnam: recent developments and prospects. Aqua-
culture 315:20–25.
Niu, J., Y.-J. Liu, L.-X. Tian, K.-S. Mai, H.-J. Yang, C.-X. Ye, and Y. Zhu. 2008a.
The effect of different levels of dietary phospholipid on growth, survival,
and nutrient composition of early juvenile Cobia (Rachycentron canadum).
Aquaculture Nutrition 14:249–256.
327
Niu, J., Y.-J. Liu, L.-X. Tian, K.-S. Mai, H.-J. Yang, C.-X. Ye, and Y. Zhu. 2008b.
Effects of dietary phospholipid level in Cobia (Rachycentron canadum) lar-
vae: growth, survival, plasma lipids and enzymes of lipid metabolism. Fish
Physiology and Biochemistry 34:9–17.
Noffs, M. D., R. C. Martino, L. C. Trugo, E. C. Urbinati, J. B. K. Fernandes, and
L. S. Takahashi. 2009. Dietary fish oil replacement with lard and soybean oil
affect triacylglycerol and phospholipid muscle and liver docosahexaenoic acid
content but not in the brain and eyes of Surubim juveniles Pseudoplatystoma
sp. Fish Physiology and Biochemistry 35:399–412.
NRC (National Research Council). 2011. Nutrient requirements of fish and
shrimp. National Academies Press, Washington, D.C.
Ramezani-Fard, E., M. S. Kamarudin, S. A. Harmin, and C. R. Saad. 2012.
Dietary saturated and omega-3 fatty acids affect growth and fatty acid profiles
of Malaysian Mahseer. European Journal of Lipid Science and Technology
114:185–193.
Ruyter, R., C. Røsjø, O. Einen, and M. S. Thomassen. 2000. Essential fatty
acids in Atlantic Salmon: effects of increasing dietary doses of (n-6 and (n-3
fatty acids on growth, survival and fatty acid composition of liver, blood and
carcass. Aquaculture Nutrition 6:119–127.
Sandel, E., O. Nixon, S. Lutzky, B. Ginsbourg, A. Tandler, Z. Uni, and W.
Koven. 2010. The effect of dietary phosphatidylcholine/phosphatidylinositol
ratio on malformation in larvae and juvenile Gilthead Sea Bream (Sparus
aurata). Aquaculture 304:42–48.
Seoka, M., M. Kurata, R. Tamagawa, A. K. Biswas, B. K. Biswas, A. S. K. Yong,
Y. S. Kim, S. C. Ji, K. Takii, and H. Kumai. 2008. Dietary supplementation
of salmon roe phospholipid enhances the growth and survival of Pacific
Bluefin Tuna Thunnus orientalis larvae and juveniles. Aquaculture 275:225–
234.
Stubhaug, I., Ø. Lie, and B. E. Torstensen. 2007. Fatty acid productive value and
beta-oxidation capacity in Atlantic Salmon (Salmo salar L.) fed on different
lipid sources along the whole growth period. Aquaculture Nutrition 13:145–
155.
Sun, L., H. Chen, L. Huang, Z. Wang, and Y. Yan. 2006. Growth and energy bud-
get of juvenile Cobia (Rachycentron canadum) relative to ration. Aquaculture
257:214–220.
Tacon, A. G. J., and M. Metian. 2008. Global overview on the use of fish
meal and fish oil in industrially compounded aquafeeds: trends and future
prospects. Aquaculture 285:146–158.
Tocher, D. R., E. A. Bendiksen, P. J. Campbell, and J. G. Bell. 2008. The role of
phospholipids in nutrition and metabolism of teleost fish. Aquaculture 280:
21–34.
Torstensen, B. E., L. Frøyland, R. Ørnsrud, and Ø. Lie. 2004. Tailoring of a
cardioprotective muscle fatty acid composition of Atlantic Salmon (Salmo
salar) fed vegetable oils. Food Chemistry 87:567–580.
Trushenski, J. T., P. Blaufuss, B. Mulligan, and J. Laporte. 2011a. Growth
performance and tissue fatty acid composition of Rainbow Trout reared
on feeds containing fish oil or equal blends of fish oil and traditional or
novel alternative lipids. North American Journal of Aquaculture 73:194–
203.
Trushenski, J. T., J. Boesenberg, and C. C. Kohler. 2009. Influence of grow-
out feed fatty acid composition on finishing success in Nile Tilapia. North
American Journal of Aquaculture 71:242–251.
Trushenski, J. T., B. Gause, and H. A. Lewis. 2011b. Selective fatty acid
metabolism, not sequence of dietary fish oil intake, prevails in fillet fatty
acid profile change in Sunshine Bass. North American Journal of Aquacul-
ture 73:204–211.
Trushenski, J. T., H. A. Lewis, and C. C. Kohler. 2008a. Fatty acid profile of
Sunshine Bass: I. Profile change is affected by initial composition and differs
among tissues. Lipids 43:629–641.
Trushenski, J. T., H. A. Lewis, and C. C. Kohler. 2008b. Fatty acid profile of
Sunshine Bass: II. Profile change differs among fillet lipid classes. Lipids
43:643–653.
Trushenski, J. T., J. Rosenquist, and B. R. Gause. 2011c. Growth performance,
tissue fatty acid composition, and consumer appeal of Rainbow Trout reared
 328 TRUSHENSKI ET AL.
on feeds containing terrestrially derived rendered fats. North American Jour-
nal of Aquaculture 73:468–478.
Trushenski, J., M. Schwarz, A. Bergman, A. Rombenso, and B. Delbos. 2012.
DHA is essential, EPA appears largely expendable, in meeting the (n-3 long-
chain polyunsaturated fatty acid requirements of juvenile Cobia (Rachycen-
tron canadum). Aquaculture 326–329:81–89.
Trushenski, J. T., M. Schwarz, J. Laporte, H. Lewis, B. Delbos, R. Takeuchi,
and L. A. Sampaio. 2011d. Fish meal replacement with soy-derived protein
in feeds for juvenile Cobia: influence of replacement level and attractant
supplementation. Journal of the World Aquaculture Society 42:435–443.
Trushenski, J. T., M. Schwarz, H. Lewis, J. Laporte, B. Delbos, R. Takeuchi,
and L. A. Sampaio. 2011e. Effect of replacing dietary fish oil with soybean
oil on production performance and fillet lipid and fatty acid composition of
juvenile Cobia (Rachycentron canadum). Aquaculture Nutrition 17:437–447.
Trushenski, J., M. Schwarz, W. Vila Nova Pessoa, B. Mulligan, C. Crouse, B.
Gause, F. Yamamoto, and B. Delbos. 2013. Amending reduced fish meal
feeds with marine lecithin, but not soy lecithin, improves growth of ju-
venile Cobia and may attenuate heightened responses to stress challenge.
Journal of Animal Physiology and Animal Nutrition 97(1):170–180. DOI:
10.1111/j.1439-0396.2011.01255.x.
Downloaded by [Department Of Fisheries] at 19:58 08 April 2013
Turchini, G. M., D. S. Francis, and S. S. De Silva. 2006. Modification of tissue
fatty acid composition in Murray Cod (Mccullochella peelii peelii, Mitchell)
resulting from a shift from vegetable oil diets to a fish oil diet. Aquaculture
Research 37:570–585.
Turchini, G. M., D. S. Francis, S. P. S. D. Senadheera, T. Thanuthong, and S. S.
De Silva. 2011. Fish oil replacement with different vegetable oils in Murray
Cod: evidence of an “omega-3 sparing effect” by other dietary fatty acids.
Aquaculture 315:250–259.
Turchini, G. M., B. E. Torstensen, and W.-K. Ng. 2009. Fish oil replacement in
finfish nutrition. Reviews in Aquaculture 1:10–57.
Wang, Z., T. Han, L. Tian, Y. Wang, G. Liang, and Y. Liu. 2007. Impact of
three vegetable oil sources on growth, body composition, and tissue fatty
acid composition of juvenile Cobia (Rachycentron canadum). Journal of
Zhejiang Ocean University (Natural Science) 26:237–245. (In Chinese with
English abstract.) Available: http://en.cnki.com.cn/Article en/CJFDTOTAL-
REEF200703000.htm. (September 2012).
Zheng, X., Z. Ding, Y. Xu, Ó. Monroig, S. Morais, and D. R. Tocher. 2009.
Physiological roles of fatty acyl desaturases and elongases in marine fish:
characterization of cDNAs of fatty acyl 6 desaturase and elovl5 elongase
of Cobia (Rachycentron canadum). Aquaculture 290:122–131.