See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/319409799

The effects of ozonation on select waterborne steroid hormones in

recirculation aquaculture systems containing sexually mature Atlantic salmon
Salmo salar

Article  in  Aquacultural Engineering · August 2017

DOI: 10.1016/j.aquaeng.2017.08.004

CITATIONS READS

9 817

5 authors, including:

Christopher Good John Davidson

The Conservation Fund Freshwater Institute The Conservation Fund
51 PUBLICATIONS   1,240 CITATIONS    54 PUBLICATIONS   1,843 CITATIONS   

SEE PROFILE SEE PROFILE

Ryan L Earley Joseph Styga

University of Alabama Centre College
174 PUBLICATIONS   3,525 CITATIONS    9 PUBLICATIONS   34 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Fish Welfare and Performance in Recirculating Aquaculture Systems (RASALMO) View project

RASALMO View project

All content following this page was uploaded by Steven T. Summerfelt on 25 September 2017.

The user has requested enhancement of the downloaded file.

 Aquacultural Engineering 79 (2017) 9–16

Contents lists available at ScienceDirect

Aquacultural Engineering
journal homepage: www.elsevier.com/locate/aque

The effects of ozonation on select waterborne steroid hormones in MARK

recirculation aquaculture systems containing sexually mature Atlantic
salmon Salmo salar
⁎
Christopher Gooda, , John Davidsona, Ryan L. Earleyb, Joseph Stygab, Steven Summerfelta
a
The Conservation Fund’s Freshwater Institute, 1098 Turner Road, Shepherdstown, WV 25443, United States
b
Department of Biological Sciences, University of Alabama, 300 Hackberry Lane, Tuscaloosa, AL 35401, United States

A R T I C L E I N F O A B S T R A C T

Keywords: Steroid hormones have been shown to accumulate in recirculation aquaculture system (RAS) water over time;
RAS however, their influence on the reproductive physiology of fish within RAS remains unknown. Whether ozo-
Ozone nation impacts waterborne hormone levels in RAS has likewise not been fully evaluated. To this end, a controlled
Estradiol 3-month study was conducted in 6 replicated RAS containing a mixture of sexually mature and immature
11-ketotestosterone
Atlantic salmon Salmo salar to determine whether ozone, as typically applied in RAS to improve water quality, is
Testosterone
associated with a reduction in waterborne hormones. Post-smolt Atlantic salmon (1253 ± 15 g) were stocked
Atlantic salmon
into each RAS; 109 of 264 fish placed in each system were sexually mature males, and 5 were mature females.
Water ozonation, controlled using an ORP set-point of 290–300 mV, was applied with the pure oxygen feed gas
within the low-head oxygenators of 3 randomly selected RAS, while the remaining 3 RAS did not receive ozone.
The RAS hydraulic retention time was 6.9 ± 0.3 days. Study fish were raised under these conditions for 12
weeks; during weeks 10 and 12, triplicate water samples were collected from the following locations in each
RAS: i) culture tank, ii) makeup water, iii) pre-biofilter, iv) post-biofilter, and v) post-gas conditioning.
Concentrations of 3 waterborne hormones – testosterone, 11-ketotestosterone (11-KT), and estradiol (17β-es-
tradiol) – were quantified using enzyme immunoassays (EIA). Estradiol was significantly reduced by ozonation;
testosterone and 11-KT were also reduced by ozonation, although these reductions were not observed across all
sampling locations and events. Testosterone and 11-KT concentrations, however, were significantly reduced
following water passage through the biofilters of both ozonated and non-ozonated RAS. The results of this study
demonstrate the potential for ozone to be used in RAS as a means of preventing the accumulation of steroid
hormones. Further research is required to assess whether reducing hormones in this manner impacts precocious
sexual maturation in RAS-produced Atlantic salmon.

1. Introduction
Land-based, closed containment facilities utilizing recirculation
aquaculture system (RAS) technologies are becoming more common-
place in the Atlantic salmon Salmo salar farming industry, not only to
raise smolts and post-smolts prior to sea transfer but also, in a growing
number of locations, to raise salmon to market size entirely on land
(Summerfelt and Christianson, 2014). While the latter approach to
salmon culture is still in its infancy, a major issue encountered thus far
in land-based salmon growout is precocious sexual maturation, parti-
cularly in males (Davidson et al., 2016; Good and Davidson, 2016).
Early maturation is detrimental to farm profitability due to the asso-
ciated decrease in growth performance and feed conversion efficiency
(McClure et al., 2007) and reduced product quality (Aksnes et al.,
1986). Unfortunately, sexual maturation in Atlantic salmon is a com-
plicated and highly flexible process (Taylor, 1991; Fjelldal et al., 2011),
influenced by numerous factors including photoperiod (Taranger et al.,
1998; Bromage et al., 2001), temperature (Taranger et al., 2003;
Vikingstad et al., 2008), growth rate (Friedland and Haas, 1996; Duston
and Saunders, 1999), and genetics (Wolters, 2010; Barson et al., 2015).
The accumulation of steroid hormones in recirculating water has been
suggested as a potential, additional instigator of sexual maturation in
RAS facilities (Good and Davidson, 2016). Previous research has in-
dicated that steroid hormones do indeed accumulate in RAS water
(Good et al., 2014; Mota et al., 2014); however, their influence on
salmon maturation has not been adequately characterized given the

⁎
Corresponding author.
E-mail address: cgood@conservationfund.org (C. Good).

http://dx.doi.org/10.1016/j.aquaeng.2017.08.004
Received 20 April 2017; Received in revised form 25 August 2017; Accepted 28 August 2017
0144-8609/ © 2017 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/BY/4.0/).
C. Good et al.
complexity of sexual development in salmon. Whether water ozonation
at levels typically applied in RAS is sufficient to reduce accumulating
hormones has likewise not been satisfactorily investigated.
All fish release conjugated (sulfonated or glucuronidated) or un-
conjugated (“free”) steroid hormones into the aquatic environment
through urine and feces (Vermeirssen and Scott, 1996) or through the
gills (Ellis et al., 2005; Sorensen et al., 2005), respectively. Approaches
to quantify hormones in water have been developed, such as radio-
immunoassay (RIA) (e.g. Ellis et al., 2004) and enzyme immunoassay
(EIA) (e.g. Kidd et al., 2010; Friesen et al., 2012), and have been va-
lidated in cultured species such as rainbow trout (Oncorhynchus mykiss)
(Ellis et al., 2004). Various non-aquaculture studies have demonstrated
that ozonation has the capacity to reduce or eliminate specific water-
borne steroid hormones (e.g. Westerhoff et al., 2005; Snyder et al.,
2008; Kawasaki et al., 2009); however, such studies have pre-
dominantly utilized relatively high ozone doses (e.g. 8 mg/L), and as
such are difficult to compare to ozone levels typically observed in RAS
operations (usually < 1 mg/L). The possibility that ozone, as typically
applied in RAS (i.e., at low-level, non-disinfecting dosages), can reduce
waterborne hormones has thus far not been adequately investigated.
In the present study, we focused on three hormones, estradiol (i.e.,
17β-estradiol), 11-ketotestosterone (11-KT), and testosterone, based on
their major role(s) in the sexual maturation of fish, as well as on the
availability of non-invasive techniques for their measurement in water
samples. The major androgen produced by the testes of teleost fish is
11-KT, which is a derivative of testosterone (Taranger et al., 2010), and
is associated with the onset of male sexual maturation in numerous fish
species (e.g. Cavaco et al., 2001; Schulz and Miura, 2002; Campbell
et al., 2003; Rodriguez et al., 2005). Circulating plasma levels of both
11-KT and testosterone are usually low in immature fish and rise
markedly during spermatogenesis, declining shortly before the devel-
opment of mature spermatozoa (Liley and Stacey, 1983; Fostier et al.,
1987). Both of these androgens are also associated with reproductive
behavior and the development of secondary sexual characteristics
(Lofts, 1987), which in the case of male Atlantic salmon include bronze
coloration and the development of a prominent kype (i.e., hooked jaw).
In female Atlantic salmon, rising plasma estradiol levels are observed
during commencement of secondary oocyte growth (Chadwick et al.,
1987; King and Pankhurst, 2003). In general, estradiol is associated
with female sexual maturation in many teleost species through, among
other things, its role in promoting the synthesis and release of vitello-
genin (the precursor of egg yolk, which is taken up by oocytes) (Skipper
and Hamilton, 1977). Testosterone is a precursor for estradiol
(Nagahama, 1987), and therefore this hormone plays a key role in both
male and female Atlantic salmon development.
The following describes a 3-month study carried out to determine
whether water ozonation, at levels typically applied in RAS, impacts the
concentration of waterborne steroid hormones. A secondary objective
was to determine the impact, if any, of biofiltration and/or gas con-
ditioning on steroid hormone concentrations as recirculating water
passes through these unit processes.
2. Materials & methods
2.1. Water recirculation aquaculture systems
The replicated (n = 6) experimental RAS used in this study have
previously been described in detail (Davidson et al., 2011; Good et al.,
2011; Davidson et al., 2013); the components of an individual RAS are
shown in Fig. 1. To summarize, each system consisted of a fluidized-
sand biofilter, CO2 stripping column, low-head oxygenator (LHO), cir-
cular dual-drain culture tank (5.3 m3), radial flow settler, microscreen
drum filter (60 μm), heat exchanger, and a 1-HP centrifugal pump. The
total RAS water volume was 9.5 m3; water was recirculated at a rate of
380 L/min (100 gpm) through all processes, except only 60% was
passed through the biofilter and 40% was passed through the heat
10
Aquacultural Engineering 79 (2017) 9–16
exchanger. Makeup water originated from a freshwater spring source.
All RAS were operated at an exchange rate of 99.7% on a flow basis, i.e.
each unit of recirculating water volume consisted of 0.3% fresh makeup
water. The RAS hydraulic retention time was 6.9 ± 0.3 days. A con-
stant photoperiod (LD24:0; i.e., 24 h light, zero hours dark) was pro-
vided throughout the study, and fish were fed once an hour using au-
tomated feeders (T-drum 2000CE, Arvo-Tec, Finland) and a standard
commercially available salmonid diet (Bio-Oregon, Westbrook, ME
USA). Fish were fed to satiation based on standardized feeding charts,
supplemented by observations of feeding activity and wasted feed. The
mean feed loading was 1.5 ± 0.1 kg/day per m3/day of makeup water
flow.
2.2. Atlantic salmon
Fertilized, eyed Atlantic salmon eggs (Salmobreed AS, Bergen,
Norway) were hatched on-site and raised in a flow-through system at
12°C under LD24:0 photoperiod until 40 g, at which point they received
a 6-week LD12:12 artificial winter followed by a return to LD24:0 to
induce smoltification. Post-smolt Atlantic salmon (102 ± 1 g) were
then stocked into the replicated RAS, and raised for 8 months under
high NO3-N (99 ± 1 mg/L) versus low NO3-N (10 ± 0 mg/L) condi-
tions, described separately (Good et al., 2016). Following the Good
et al. (2016) study, all fish were removed from the replicated RAS and
combined into a holding tank to mix previously treated populations into
a single source population for the present study. From this source po-
pulation, salmon were selected for restocking each RAS (264 fish per
RAS) with focus on distributing equal numbers of sexually mature fish;
109 mature males and 5 mature females were stocked into each RAS,
such that the percentage of mature fish at study commencement was
43.2% in each culture tank. Sexually mature male and female Atlantic
salmon were identified by characteristic morphological features (i.e.,
bronze coloration and prominent kype in males and ovipositor in fe-
males). Mean starting weight was 1253 ± 15 g, and mean starting
density was 62 ± 1 kg/m3. Water ozonation treatment (see below)
began shortly afterward, and continued for 12 weeks; final mean
salmon weight and density were 1853 ± 24 g and 91 ± 1 kg/m3,
respectively.
2.3. Ozonation
Three RAS were randomly selected to be ozonated using generators
(Model G22, Pacific Ozone Technology, Benecia, CA, USA) that con-
verted pure oxygen feed gas into ozone, which was then combined with
the primary oxygen gas flow to the LHOs. To prevent ozone levels from
reaching unsafe levels in the culture tank water, oxidation-reduction
potential (ORP) was monitored in each tank via an ORP digital sensor
(Model DRD1R5, Hach Company, Loveland, CO, USA) placed directly in
front of the inlet flow. An ORP set-point of 290–300 mV was used, and
SC100 Universal Controllers (Hach) provided proportional-integral-
derivative control of generator output to maintain target ozone levels
and prevent exposure of fish to toxic ozone residuals. An average of
1.79% ozone was contained in the feed gas produced by the ozone
generators, and this was continuously monitored by Teledyne
Instruments Ozone Monitor – Model 465H (Teledyne Instruments, City
of Industry, CA, USA); data were recorded from the unit once daily.
2.4. Water sampling
For routine water quality assessments, water samples were collected
weekly from culture tank side drains and tested on-site. Specific para-
meters, methodologies, and frequencies of testing are summarized in
Table 1. To assess the effects of ozonation on waterborne hormones,
250 mL water samples were collected in triplicate in high density
polyethylene bottles at the following locations in each RAS: i) makeup
water, ii) pre-biofilter, iii) post-biofilter, iv) post-gas conditioning, and
C. Good et al. Aquacultural Engineering 79 (2017) 9–16

Fig. 1. Process flow diagram of an individual experimental-scale RAS, illustrating direction of water flow and all water treatment unit processes, and with the five sampling locations for
waterborne hormones indicated.

Table 1
Water quality parameters evaluated and methodologies for testing.

Parameter Method of Analysis Testing Frequency

Biochemical oxygen demand Hach Method 8043 Weekly

Dissolved carbon dioxide Hach Method 8233–Buret titration Weekly
Dissolved oxygen Hach SC100 Universal Controller & LDO® Probe Continuous
Nitrite nitrogen Hach Method 8507–Diazotization Weekly
Nitrate nitrogen Hach Method 8171–Cadmium reduction Weekly
ORP Hach SC100 Universal Controller & Differential ORP Sensor Continuous
pH Hach Method 8156 – pH electrode 4–5x/week
Temperature Hach SC100 Universal Controller & Differential ORP Sensor Continuous
Total alkalinity HACH Method 8203 4–5x/week
Total ammonia nitrogen Hach Method 8038 – Nessler Weekly
Heterotrophic bacteria Standard Methods 9215D – Membrane Filtration and Agar Plate Counts Weekly during final 3 weeks
Total suspended solids Standard Methods 2540D – Dried at 103–105 °C Weekly
True color Hach Method 8025 – Platinum-Cobalt Weekly
Ultraviolet transmittance Standard Methods 5910B–Ultraviolet absorption Weekly

11
C. Good et al.
v) culture tank sidewall box (Fig. 1), during two separate sampling
events which took place during weeks 10 and 12 of the 12-week study
period. A total of 180 water samples were collected (i.e., 90 samples
from each sampling event). All water samples were frozen and stored
on-site until overnight shipment to the University of Alabama for hor-
mone quantifications.
2.5. Waterborne hormones sampling and quantification
Waterborne hormones were assessed using enzyme-immunoassay
(EIA) kits (Cayman Chemicals Inc., Ann Arbor, Michigan, USA) for T,
11-KT, and E2. Upon arrival at the University of Alabama, the frozen
water samples were immediately thawed at 4 °C overnight, and then
filtered through Whatman Grade 1 filter paper (GE Healthcare Life
Sciences, Piscataway, New Jersey, USA) into 400 mL glass beakers,
which had been pre-cleaned with 95% ethanol and distilled water.
Tygon tubing was then fitted to Waters Sep-Pak C18 columns (Milford,
Massachusetts, USA) on a vacuum manifold; the other ends of the
tubing placed into the 400 mL glass beakers containing the fish hor-
mone collection water. The tubing used was of special formulation
2275, which does not adsorb hormones or leach hormone-mimics (e.g.,
plasticizers) into the water. Prior to fastening the tubing to columns, the
columns were primed with 2 × 2 mL HPLC grade methanol and
2 × 2 mL distilled water (the last 0.5 mL of distilled water was retained
to keep the columns moist). Water samples were then vacuumed
through the C18 columns, trapping the steroid hormones, and 2 mL
distilled water was passed over the columns to remove any lingering
salts. Subsequently, 2 × 2 mL ethyl acetate was passed over the col-
umns into 13 × 100 mm borosilicate vials to elute the “free” hormone
fraction, and 2 × 2 mL methanol was passed over the columns into new
13 × 100 mm borosilicate vials to elute combined “conjugated” hor-
mone fractions (sulphated, glucuronidated). These fractions were
stored at −20 °C. All free hormone samples were then dried under a
gentle stream of ultrapure nitrogen at 37 °C in a water bath, and were
resuspended in 500 μL of 5% EtOH: 95% EIA buffer (provided with the
EIA kits); i.e., 25 μL EtOH followed by 1 min vortexing + 475 μL EIA
buffer followed by 20 min vortexing. Based on initial serial dilution
curves generated from pooled samples (see below), it was determined
that the T and KT samples should be diluted 1:16 prior to assaying and
that the E2 samples should not be diluted (i.e., remain 1:1). To conduct
the 1:16 dilution for each of T and 11-KT, 50 μL of the original re-
suspension (i.e., the 500 μL resuspension described above) were added
to 750 μL of EIA buffer, followed by vortexing. Enzyme immunoassays
were then performed following kit protocols. To validate the assays,
35 μL was taken off the top of each of the original 180 samples to
generate a pool of 6300 μL. This pool was serially diluted from 1:1 to
1:128 for T and 11-KT and from 1:1 to 1:16 for E2 (further dilutions
were undetectable for E2). S. salar serial dilution curves were parallel to
the kit standard curves for all hormones (comparison of slopes test (Zar,
1996); T: t12 = 0.03, p = 0.98; 11-KT: t12 = −0.05, p = 0.96; E2:
t9 = 0.22, p = 0.83). Quantitative recovery was then conducted: each
pool was spiked with a known standard, extracted through C18 col-
umns, dried and resuspended as above, and assayed. For this, 75 μL of
the 1:1 pool was mixed with 75 μL of standards 1, 3, 5, and 7 for E2
(and processed as described), and 75 μL of 1:16 pool was mixed with
75 μL of standards 1, 3, 5, and 7 for 11-KT and T (and processed as
described). The relationship between observed and expected con-
centrations was analyzed via linear regression, and the slopes for T, 11-
KT, and E2 were 0.860, 0.874, 0.875, respectively, indicating good
recovery. Minimum observed recovery for T, 11-KT, and E2 were
87.5%, 87.1%, and 89.9%, respectively. Each hormone was assayed on
6, 96-well plates; on each plate, a pooled S. salar control was assayed in
the first and last two sample wells. Intra-assay coefficients of variation
for these controls were: T (6.9%, 4.5%, 4.8%, 1.6%, 4.9%, & 2.9%); 11-
KT (1.7%, 4.3%, 10.0%, 2.7%, 4.0%, & 5.1%); E2 (4.1%, 3.1%, 6.5%,
2.4%, 9.6%, & 2.4%). Inter-assay coefficients of variation were: 6.0%
12
Aquacultural Engineering 79 (2017) 9–16
(T), 7.7% (11-KT), and 12.9% (E2).
2.6. Mass balance for hormone production in RAS
Hormone production rates were estimated using a mass balance
approach, assuming quasi-steady state conditions, as carried out pre-
viously (Good et al., 2014). The production of each hormone (P, in μg/
day per kilogram of fish biomass) was estimated from water flow rates
(Qm = makeup water flow ≈ 1 L/min; Qr = recirculating water
flow ≈ 380 L/min) and hormone concentrations (Ctank = hormone
concentrations exiting the culture tanks; Cm = hormone concentrations
in makeup water samples; and Cr = hormone concentrations returning
to the culture tanks). The following equation was used for this esti-
mation:
P = Qm (Ctank − Cm) + Qr (Ctank − Cr )
Hormone removal efficiencies across the i) the biofilter, and ii)
across the gas conditioning processes, were also calculated based on the
differences in inlet and outlet hormone concentrations, divided by the
inlet hormone concentrations, at each site of interest; these removal
efficiencies were expressed as a percent.
2.7. Statistical analysis
Water quality parameters were assessed for normality using the
Shapiro-Wilk test; Student’s t-test was utilized to compare normally-
distributed variables between treatments, while the Mann-Whitney U
test was utilized if data were not normally distributed. All assessments
of water quality data were conducted in SYSTAT (Systat Software, San
Jose, CA, USA). Differences in waterborne hormone concentrations
between ozonated and non-ozonated RAS were assessed for each sam-
pling event dataset, and for each hormone in two ways: i) at each
sampling location, and ii) among sampling locations within each
treatment. These assessments were carried out using analysis of var-
iance (ANOVA), with hormone concentration as the dependent variable
and ozone treatment as the dichotomous independent variable; a sig-
nificance level of p < 0.05 was used. For significant ANOVA models,
post-hoc pairwise comparisons were performed using Tukey’s proce-
dure. Statistical analyses of the water quality data were carried out
using SYSTAT (Systat Software, Inc., San Jose, CA, USA); all other
procedures were conducted using STATA software (StataCorp LP,
College Station, Texas, USA).
3. Results & discussion
3.1. Water quality and hormone production mass balances
Numerous studies have demonstrated that water ozonation can ef-
fectively reduce carbonaceous biochemical oxygen demand (cBOD5),
chemical oxygen demand, dissolved organic carbon, color, nitrite ni-
trogen (NO2-N), turbidity, total organic carbon, and total suspended
solids (TSS) (Rosenthal and Kruner, 1985; Hozalski et al., 1995; Brazil,
1996; Summerfelt and Hochheimer, 1997; Summerfelt et al., 1997;
Tango and Gagnon, 2003; Summerfelt et al., 2009). In the same re-
plicated RAS, Davidson et al. (2011) demonstrated that ozonation im-
proved water quality through the reduction of, among other things,
TSS, cBOD5, and heterotrophic bacterial counts, and through the in-
crease in dissolved oxygen, water clarity and UV transmittance. During
the present study, there were overall significant decreases in cBOD5,
TAN, NO2-N, and true color associated with water ozonation, as well as
increases in ORP, dissolved oxygen, and UV transmittance (Table 2).
Mass balances used to estimate hormone production determined
that each kilogram of fish biomass produced, on average, 26, 270, and
190 μg/d of estradiol, 11-KT, and testosterone, respectively. Minimum
and maximum production values of these hormones, respectively, were
12.8 & 41.1, 226 & 344, and 49.1 & 372 μg/d. The reason for the
C. Good et al. Aquacultural Engineering 79 (2017) 9–16

Table 2 immediate reduction in estradiol concentration, i.e. estradiol was not

Mean tank water quality concentrations (mg/L, unless otherwise noted) collected at the significantly reduced between the post-biofilter and post-gas con-
culture tank sidewall drain from ozonated and non-ozonated RAS.
ditioning sampling locations (as occurred with 11-ketotestosterone and
Parameter Ozone No Ozone p-value testosterone – see below), but rather was associated with RAS-wide
reduction in estradiol levels. The reason for this observation is likely
Alkalinity 196 ± 1.0 197 ± 0.2 0.653 due to the low dose (0.3 mg/L) of ozone that was applied, which clearly
Carbon Dioxide 7.5 ± 0.5 8.2 ± 0.5 0.464
had an impact on overall estradiol concentrations but was not sufficient
cBOD5 1.07 ± 0.03 1.22 ± 0.03 0.030
Dissolved Oxygen 10.6 ± 0.3 10.2 ± 0.1 0.046 to demonstrate estradiol reduction immediately following exposure to
Heterotrophic Bacteria (CFU/mL) 466 ± 127 490 ± 80 0.907 ozone as water passed through the LHO. Previously, ozone has been
Nitrite Nitrogen 0.006 ± 0.001 0.011 ± 0.001 0.043 shown to reduce estradiol in a nearly linear dose-response relationship
Nitrate Nitrogen 50 ± 3 50 ± 2 0.825 (Kawasaki et al., 2009), a process that is further accelerated when
Oxidative Reduction Potential 293 ± 4 270 ± 2 0.008
combined with activated carbon. The study by Kawasaki et al. (2009),
(mV)
pH 7.81 ± 0.01 7.79 ± 0.02 0.399 however, utilized high concentrations of both estradiol and ozone
Temperature (°C) 15.1 ± 0.01 15.2 ± 0.01 0.456 under laboratory conditions; therefore, it is difficult to directly compare
Total Ammonia Nitrogen 0.165 ± 0.005 0.203 ± 0.005 0.046 their findings with those observed in the present study. Hormones, such
Total Suspended Solids 1.5 ± 0.2 1.7 ± 0.1 0.436
as estradiol, containing phenolic moieties have been shown to be more
True Color (Pt-Co Units) 3.7 ± 0.3 20 ± 1 0.002
UV Transmittance (%) 91 ± 0.4 79 ± 0.8 0.001 susceptible to oxidation through ozonation compared to steroids
without phenolic moieties (Westerhoff et al., 2005), and therefore the
specific molecular structure of estradiol could have rendered it more
observed high variability in testosterone production, compared to es- susceptible to ozone reduction compared to the other hormones ex-
tradiol and 11-KT, cannot be adequately explained by the present scope amined in this study.
of investigation. Whether testosterone production might, for example,
be periodic in nature (i.e., non-steady state) compared to the other 3.3. 11-ketotestosterone
hormones, is a hypothesis requiring further study. Passage of the re-
circulating water flow through the fluidized-sand biofilters removed an Results for 11-KT were largely consistent between the two sampling
average of 85% and 37% of the 11-KT and testosterone, respectively, events, the major observation being that biofiltration, under study
but effectively no estradiol. Passage of the recirculating water flow conditions, significantly reduced 11-KT concentrations in both ozone
through the gas conditioning column and LHO did not appear to re- and non-ozonated RAS. This confirms previous observations showing
move estradiol, 11-KT, or testosterone, and these results can be viewed reduction of waterborne 11-KT following passage through RAS unit
as a refinement of previous findings presented by Good et al. (2014). processes (Good et al., 2014), and identifies the site of this reduction as
The ozone dosage applied to the recirculating water flows in the present the biofilters (vs. the gas conditioning processes). Concentrations of 11-
study, however, was effective at reducing waterborne estradiol. KT were significantly higher in post-gas conditioning water samples
compared to post-biofilter samples (Fig. 2); however, this is likely the
result of biofilter-bypassed water (relatively high in 11-KT concentra-
3.2. Estradiol tion) combining with the biofiltered water prior to the post-gas con-
ditioning sampling location (Fig. 1). As was postulated in Good et al.
As observed through both sampling events, waterborne estradiol (2014), the relative susceptibility of 11-KT to reduction through bio-
concentrations were significantly reduced by ozonation at all four filtration is likely related to its individual physicochemical properties,
sampling locations within the RAS recirculation loop (i.e., pre-biofilter, specifically its relatively low octanol-water partition coefficient
post-biofilter, post-gas conditioning, and culture tank; no significant (log Kow). Log Kow is a measure of the tendency of an organic compound
differences observed at the influent, non-ozonated makeup water lo- to adsorb to sediment, such that higher log Kow values indicate a ten-
cation) (Fig. 2). A significant reduction in estradiol concentration was dency to bind to organics instead of remaining in an aqueous phase. In
observed following biofiltration (in non-ozonated RAS) during the first the case of 11-KT (log Kow = 1.92) (Yang et al., 2012), its low log Kow
sampling; however, this observation was not repeated during the indicates a greater degree of partitioning into the aqueous phase, which
second sampling, where no significant differences were detected be- may have facilitated biodegradation during transit through the study
tween the four within-RAS sampling locations in either treatment RAS biofilters, regardless of ozone treatment. In contrast, the relatively
group. Therefore, it cannot be definitively stated that biofiltration, as high estradiol log Kow (3.94; Yang et al., 2012) indicates that estradiol
applied in the present study, consistently reduces estradiol at the ob- is more adsorbed to organics, which might hinder its degradation by
served waterborne concentrations and over the actual contact time with biofilter microbes. The relationship between RAS biofiltration and
the biofilter media (< 5 min; see Davidson et al., 2008). Much pub- steroid hormone log Kow values requires further investigation in order
lished research on water treatment removal of estrogenic compounds to be sufficiently characterized. With respect to the effects of ozonation
(and other endocrine disrupting chemicals) has focused on wastewater on waterborne 11-KT concentrations, results were inconsistent between
or sewage treatment facilities (e.g. Onda et al., 2003; Chimchirian et al., sampling events; the first sampling demonstrated, predominantly, that
2007; Cicek et al., 2007); however, such facilities typically have much ozonation did not significantly reduce 11-KT concentration when
longer retention times (e.g. 10–15 days) than RAS treatment processes. compared with samples from the non-ozonated RAS, whereas in the
Overall, no specific sampling location demonstrated evidence of sig- second sampling 11-KT was significantly reduced in all four within-RAS
nificant estradiol production, suggesting that estradiol was being pro- sampling locations by water ozonation. The reason for this incon-
duced and/or introduced in relatively small amounts and that ozona- sistency is unclear, although concentration values obtained during the
tion prevented significant accumulation of this hormone, compared to first sampling exhibited greater variation than those measured in
the higher concentration observed in systems not receiving ozonation samples from the later sampling event, which could have obscured the
that otherwise gradually accumulated. Previous research has demon- observed relationship in the second sampling.
strated that estradiol concentrations in RAS are higher than those in
influent makeup water, and that passage through unit processes (bio- 3.4. Testosterone
filter and gas-conditioning) did not significantly impact waterborne
levels of this hormone (Good et al., 2014); these findings were largely In comparison to estradiol and 11-KT observations, the results ob-
confirmed in the present study. Ozone did not, however, induce an tained for testosterone are cryptic and ultimately require further

13
C. Good et al. Aquacultural Engineering 79 (2017) 9–16

Fig. 2. Waterborne concentrations of estradiol, 11-ketotestosterone, and testosterone, as measured by enzyme immunoassays, during the first (left; 10-week) and second (right; 12-week)
RAS water sampling events. Asterisks over individual water sample locations indicate significant (p < 0.05) differences between ozone and no ozone treatment groups. Differing letters
between sample locations indicate significant differences within ozone (lower case) and no ozone (upper case) treatments. ns = not significant.

investigation to adequately explain. One consistent finding was that in
both ozonated and non-ozonated RAS, testosterone was significantly
reduced following passage through the biofilters (Fig. 2), which is in
agreement with previous findings demonstrating the susceptibility of
testosterone to biodegradation (Vanderford et al., 2003). Previous re-
search in RAS (Good et al., 2014) did not demonstrate reduction of
testosterone across the RAS unit processes; however, it is likely that
these previous findings are not in agreement to the present results due
to their measurement across all unit processes, i.e. that biofilter-asso-
ciated hormone reduction was likely obscured due to the narrower
scope of RAS sampling locations. An increase in testosterone con-
centration from post-biofilter to post-gas conditioning sample locations,
due to the addition of biofilter-bypassed water, was not shown, unlike
the increase in 11-KT observed between these locations. The reason for
this inconsistency is unclear and warrants further study. The effect of
ozonation on waterborne testosterone concentrations is even less clear,
based on the data obtained in this experiment. In both sampling events,
post-gas conditioning and culture tank testosterone concentrations did
not appear to be affected by ozonation, whereas significant hormone
reduction through ozonation was observed in pre-biofilter samples (and
in the case of the first sampling, in post-biofilter samples). Why tes-
tosterone appeared to be significantly reduced by ozonation in only
these sampling locations is unknown, and additional investigation is
required to determine if this is a repeatable finding, and if so, to explain
its occurrence. As mentioned previously, ozone has been shown to
oxidize steroid hormones containing phenolic moieties (such as

14
C. Good et al.
estradiol) more effectively than hormones without phenolic moieties
(e.g., testosterone) (Westerhoff et al., 2005), and therefore the clear
reduction of estradiol by ozone, as observed in this study, should not be
expected for a more resilient hormone such as testosterone. In previous
benchtop studies, both estradiol and testosterone (spiked in river water)
have been shown to be reduced 100% by ozonation following five
minutes of contact time (Snyder et al., 2008); however, the ozone do-
sage used was approximately eight times greater than that applied the
present study, and therefore it is difficult to extrapolate these findings
to the present observations in low-dose ozone RAS.
3.5. Conclusions
Ozonation, as applied in the present study, resulted in a significant
reduction in waterborne estradiol concentrations in RAS; this re-
lationship was not observed for 11-KT, while results for testosterone
remain ambiguous and require further examination. Biofiltration re-
duced concentrations of 11-KT and testosterone in both ozonated and
non-ozonated RAS. Further studies should focus on the application of
ozone in RAS to reduce steroid hormone accumulation and whether this
reduction has the potential to reduce early sexual maturation in im-
portant cultured species, such as Atlantic salmon.
Acknowledgements
Special thanks to Karen Schroyer, Susan Glenn, Natalie Redman,
and Christina Russell for water chemistry analysis; and to Jeremy
Ruffner and Ryan Snader for fish husbandry and technical assistance.
This research was supported by the USDA Agricultural Research Service
under Agreement Nos. 59-1930-0-046 and 59-8082-5-001. All experi-
mental protocols and methods were in compliance with the Animal
Welfare Act (9CFR) requirements and were approved by The
Freshwater Institute’s Institutional Animal Care and Use Committee.
Use of trade names does not imply endorsement by the U.S.
Government.
References
Aksnes, A., Gjerde, B., Roald, S.O., 1986. Biological, chemical and organoleptic changes
during maturation of farmed Atlantic salmon, Salmo salar. Aquaculture 53, 7–20.
Barson, N.J., Aykanat, T., Hindar, K., Baranski, M., Bolstad, G.H., Fiske, P., Jacq, C.,
Jensen, A.J., Johnston, S.E., Karlsson, S., Kent, M., Moen, T., Niemelä, E., Nome, T.,
Næsje, T.F., Orell, P., Romakkaniemi, A., Sægrov, H., Urdal, K., Erkinaro, J., Lien, S.,
Primmer, C.R., 2015. Sex-dependent dominance at a single locus maintains variation
in age at maturity in salmon. Nature 528, 405–408.
Brazil, B.L., 1996. Impact of Ozonation on System Performance and Growth
Characteristics of Hybrid Striped Bass (Morone chrysops x M. saxatilis) and Tilapia
Hybrids (Sarotherodon sp.) Reared in Recirculating Aquaculture Systems. Virginia
Polytechnic Institute and State University, Blacksburg, VA, USA Ph.D. Dissertation.
Bromage, N., Porter, M., Randall, C., 2001. The environmental regulation of maturation
in farmed finfish with special reference to the role of photoperiod and melatonin.
Aquaculture 197, 63–98.
Campbell, B., Dickey, J., Beckman, B., Young, G., Pierce, A., Swanson, P., 2003.
Endocrine changes associated with the growth of pre-vitellogenic oocytes in coho
salmon, Oncorhynchus kisutch. Fish Physiol. Biochem. 28, 287–289.
Cavaco, J.E.B., Bogerd, J., Goos, H., Schulz, R.W., 2001. Testosterone inhibits 11-keto-
testosterone-induced spermatogenesis in African catfish (Clarias gariepinus). Biol.
Reprod. 65, 1807–1812.
Chadwick, E.M.P., Claytor, R.R., Leger, C.E., Saunders, R.L., 1987. Inverse correlation
between ovarian development of Atlantic salmon (Salmo salar) smolts and sea age.
Can. J. Fish. Aquat. Sci. 44, 1320–1325.
Chimchirian, R.F., Suri, P.S.S., Fu, H., 2007. Free synthetic and natural estrogen hor-
mones in influent and effluent of three municipal wastewater treatment plants. Water
Environ. Res. 9, 969–974.
Cicek, N., Londry, K., Oleszkiewicz, J.A., Wong, D., Lee, Y., 2007. Removal of selected
natural and synthetic estrogenic compounds in a Canadian full-scale municipal
wastewater treatment plant. Water Environ. Res. 79, 795–800.
Davidson, J., Helwig, N., Summerfelt, S.T., 2008. Fluidized sand biofilters used to remove
ammonia, biochemical oxygen demand, total coliform bacteria, and suspended solids
from an intensive aquaculture effluent. Aquac. Eng. 39, 6–15.
Davidson, J., Good, C., Welsh, C., Summerfelt, S.T., 2011. The effects of ozone and water
exchange rates on water quality and rainbow trout Oncorhynchus mykiss performance
in replicated water recirculating systems. Aquac. Eng. 44, 80–96.
Davidson, J., Good, C., Barrows, F.T., Welsh, C., Kenney, P.B., Summerfelt, S.T., 2013.
15
Aquacultural Engineering 79 (2017) 9–16
Comparing the effects of feeding a grain- or a fish meal-based diet on water quality,
waste production, and rainbow trout Oncorhynchus mykiss performance within low
exchange water recirculating aquaculture systems. Aquac. Eng. 52, 45–57.
Davidson, J., May, T., Good, C., Waldrop, T., Kenney, B., Terjesen, B.F., Summerfelt, S.,
2016. Production of market-size North American strain Atlantic salmon Salmo salar in
a land-based recirculation aquaculture system using freshwater. Aquac. Eng. 74,
1–16.
Duston, J., Saunders, R.L., 1999. Effect of winter food deprivation on growth and sexual
maturity of Atlantic salmon (Salmo salar) in sea water. Can. J. Fish. Aquat. Sci. 56,
201–207.
Ellis, T., James, J.D., Stewart, C., Scott, A.P., 2004. A non-invasive stress assay based
upon measurement of free cortisol released into the water by rainbow trout. J. Fish
Biol. 65, 1233–1252.
Ellis, T., James, J.D., Scott, A.P., 2005. Branchial release of free cortisol and melatonin by
rainbow trout. J. Fish Biol. 67, 535–540.
Fjelldal, P.G., Hansen, T., Huang, T.-S., 2011. Continuous light and elevated temperature
can trigger maturation both during and immediately after smoltification in male
Atlantic salmon (Salmo salar). Aquaculture 321, 93–100.
Fostier, A., Le Grac, F., Loir, M., 1987. In: Idler, D.R., Crim, L.W., Walsh, J.M. (Eds.),
Proceedings of the 3rd International Symposium on the Reproductive Physiology of
Fish. Memorial University, St. Johns, Newfoundland. pp. 239–241.
Friedland, K.D., Haas, R.E., 1996. Marine post-smolt growth and age at maturity of
Atlantic salmon. J. Fish Biol. 48, 1–15.
Friesen, C.N., Chapman, L.J., Aubin-Horth, N., 2012. Holding water steroid hormones in
the African cichlid fish Pseudocrenilabrus multicolor victoriae. Gen. Comp. Endocr. 179,
400–405.
Good, C., Davidson, J., 2016. A review of factors influencing maturation of Atlantic
salmon Salmo salar with focus on water recirculation aquaculture system environ-
ments. J. World Aquac. Soc. 47 (5), 605–632. Online at: http://onlinelibrary.wiley.
com/doi/10.1111/jwas.12342/full.
Good, C., Davidson, J., Welsh, C., Snekvik, K., Summerfelt, S., 2011. The effects of ozo-
nation on performance, health, and welfare of rainbow trout Oncorhynchus mykiss in
low-exchange water recirculation aquaculture systems. Aquac. Eng. 44, 97–102.
Good, C., Davidson, J., Earley, R.L., Lee, E., Summerfelt, S., 2014. The impact of water
exchange rate and biofiltration on water-borne hormones in recirculation aqua-
culture systems containing sexually maturing Atlantic salmon Salmo salar. Aquac.
Res. Dev. 5 7 pp.
Good, C., Davidson, J., Iwanowicz, L., Meyer, M., Dietze, J., Kolpin, D.W., Marancik, D.,
Birkett, J., Russell, C., Summerfelt, S., 2016. Investigating the influence of nitrate
nitrogen on post-smolt Atlantic salmon Salmo salar reproductive physiology in
freshwater recirculation aquaculture systems. Aquac. Eng Available online at: http://
www.sciencedirect.com/science/article/pii/S0144860916301431.
Hozalski, R.M., Goel, S., Bouwer, E.J., 1995. TOC removal in biological filters. J. Am.
Waste Water Assoc. 87 (12), 40–54.
Kawasaki, N., Ogata, F., Yamaguchi, I., Tominaga, H., 2009. Degradation characteristics
of 17β-estradiol by ozone treatment with activated carbon. J. Oleo Sci. 58, 261–266.
Kidd, C.E., Kidd, M.R., Hofmann, H.A., 2010. Measuring multiple hormones from a single
water sample using enzyme immunoassays. Gen. Comp. Endocr. 165, 277–285.
King, H.R., Pankhurst, N.W., 2003. Ovarian growth and plasma sex steroid and vitello-
genin profiles during vitellogenesis in Tasmanian female Atlantic salmon (Salmo
salar). Aquaculture 219, 797–813.
Liley, N.R., Stacey, N.E., 1983. Hormones, pheromones, and reproductive behavior in
fish. In: Hoar, W.S. (Ed.), Fish Physiology IX Part B. Academic Press Inc., New York,
NY, pp. 1–63.
Lofts, B., 1987. In: Norris, D.O., Jones, R.E. (Eds.), Hormones and Reproduction in Fishes,
Amphibians and Reptiles. Plenium Press, New York, pp. 283–326.
McClure, C.A., Hammell, K.L., Moore, M., Dohoo, I.R., Burnley, H., 2007. Risk factors for
early sexual maturation in Atlantic salmon in seawater farms in New Brunswick and
Nova Scotia, Canada. Aquaculture 272, 370–379.
Mota, V.C., Martins, C.I.M., Eding, E.H., Canario, A.V.M., Verreth, J.A.J., 2014. Steroids
accumulate in the rearing water of commercial recirculating aquaculture systems.
Aquac. Eng. 62, 9–16.
Nagahama, Y., 1987. Gonadotropin action on gametogenesis and steroidogenesis in tel-
eost gonads. Zool. Sci. 4, 209–222.
Onda, K., Nakamura, Y., Takatoh, C., Miya, A., Katsu, Y., 2003. The behavior of estrogenic
substances in the biological treatment process of sewage. Water Sci. Technol. 47,
109–116.
Rodriguez, L., Begtashi, I., Zanuy, S., Carrillo, M., 2005. Long-term exposure to con-
tinuous light inhibits precocity in European male sea bass (Dicentrarchus labrax L.):
hormonal aspects. Gen. Comp. Endocr. 140, 116–125.
Rosenthal, H., Kruner, G., 1985. Treatment efficiency of an improved ozonation unit
applied to fish culture situations. Ozone-Sci. Eng. 7, 179–190.
Schulz, R.W., Miura, T., 2002. Spermatogenesis and its endocrine regulation. Fish Physiol.
Biochem. 26, 43–56.
Skipper, J.K., Hamilton, T.H., 1977. Regulation by estrogen of vitellogenin gene. Proc.
Natl. Acad. Sci. U. S. A. 74, 2384–2388.
Snyder, S.A., Lei, H., Wert, E.C., 2008. Removal of endocrine disruptors and pharma-
ceuticals during water treatment. In: Aga, Diana S. (Ed.), Fate of Pharmaceutical in
the Environment and in Water Treatment Systems. Taylor & Francis Group, Boca
Raton, Florida, pp. 229–260.
Sorensen, P.W., Pinillos, M., Scott, A.P., 2005. Sexually mature male goldfish release large
quantities of androstenedione into the water where it functions as a pheromone. Gen.
Comp. Endocr. 140, 164–175.
Summerfelt, S., Christianson, L., 2014. Fish farming in land-based closed containment
systems. World Aquac. Mag. 2014 (March), 18–21.
Summerfelt, S.T., Hochheimer, J.N., 1997. Review of ozone processes and applications as
C. Good et al.
an oxidizing agent in aquaculture. Prog. Fish Cult. 59, 94–105.
Summerfelt, S.T., Hankins, J.A., Weber, A.W., Durant, M.D., 1997. Ozonation of a re-
circulating rainbow trout culture system. II. Effects on microscreen filtration and
water quality. Aquaculture 158, 57–67.
Summerfelt, S.T., Sharrer, M.J., Tsukuda, S.M., Gearheart, M., 2009. Process require-
ments for achieving full-flow disinfection of recirculating water using ozonation and
UV irradiation. Aquac. Eng. 40, 17–27.
Tango, M.S., Gagnon, G.A., 2003. Impact of ozonation on water quality within marine
recirculating systems. Aquac. Eng. 29, 125–137.
Taranger, G.L., Haux, C., Stefansson, S.O., Björnsson, B.T., Walther, B.T., Hansen, T.,
1998. Abrupt changes in photoperiod affect age at maturity, timing of ovulation and
plasma testosterone and oestradiol-17β profiles in Atlantic salmon, Salmo salar.
Aquaculture 162, 85–98.
Taranger, G.L., Vikingstad, E., Klenke, U., Mayer, I., Stefansson, S.O., Norberg, B.,
Hansen, T., Zohar, Y., Andersson, E., 2003. Effects of photoperiod: temperature and
GnRHa treatment on the reproductive physiology of Atlantic salmon (Salmo salar L.)
broodstock. Fish Physiol. Biochem. 28, 403–406.
Taranger, G.L., Carillo, M., Schulz, R.W., Fontaine, P., Zanuy, S., Felip, A., Weltzien, F.-A.,
Dufour, S., Karlsen, O., Norberg, B., Andersson, E., Hansen, T., 2010. Control of
puberty in farmed fish. Gen. Comp. Endocr. 165, 483–515.
Taylor, E.B., 1991. A review of local adaptation in Salmonidae, with special reference to
16
View publication stats
Aquacultural Engineering 79 (2017) 9–16
Pacific and Atlantic salmon. Aquaculture 98, 185–207.
Vanderford, B.J., Pearson, R.A., Rexing, D.J., Snyder, S.A., 2003. Analysis of endocrine
disruptors, pharmaceuticals, and personal care products in water using liquid chro-
matography/tandem mass spectrometry. Anal. Chem. 75, 6265–6274.
Vermeirssen, L.M., Scott, A.P., 1996. Excretion of free and conjugated steroids in Rainbow
trout (Oncorhynchus mykiss): evidence of branchial excretion of the maturation-in-
ducing steroid 17, 20β-dihydroxy-4-pregnen-3-one. Gen. Comp. Endocr. 101,
180–194.
Vikingstad, E., Andersson, E., Norberg, B., Mayer, I., Klenke, U., Zohar, Y., Stefansson,
S.O., Taranger, G.L., 2008. The combined effects of temperature and GnRHa treat-
ment on the final stages of sexual maturation in Atlantic salmon (Salmo salar L.)
females. Fish Physiol. Biochem. 34, 289–298.
Westerhoff, P., Yoon, Y., Snyder, S., Wert, E., 2005. Fate of endocrine-disruptor, phar-
maceutical, and personal care product chemicals during simulated drinking water
treatment processes. Environ. Sci. Technol. 39, 6649–6663.
Wolters, W.R., 2010. Sources of phenotypic and genetic variation for seawater growth in
five North American Atlantic salmon stocks. J. World Aquac. Soc. 41, 421–429.
Yang, Y.-Y., Gray, J.L., Furlong, E.T., Davis, J.G., ReVello, R.C., Borch, T., 2012. Steroid
hormone runoff from agricultural test plots applied with municipal biosolids.
Environ. Sci. Technol. 46, 2746–2754.
Zar, J.H., 1996. Biostatistical Analysis, 3rd ed. Prentice Hall, New Jersey (662 pp.).