Chemosphere 112 (2014) 18–25

Contents lists available at ScienceDirect

Chemosphere
journal homepage: www.elsevier.com/locate/chemosphere

Environmental effects and fate of the insecticide bifenthrin

in a salt-marsh mesocosm
Paul L. Pennington a,⇑, Heather Harper-Laux b, Yelena Sapozhnikova c,1, Michael H. Fulton a
a
NOAA’s National Centers for Coastal Ocean Science, Center for Coastal Environmental Health and Biomolecular Research, Charleston, SC, United States
b
College of Charleston, Graduate Program in Marine Biology, Grice Marine Laboratory, Charleston, SC, United States
c
JHT Incorporated, Contractor to NOAA, Charleston, SC, United States

h i g h l i g h t s

 We examined the lethal and sublethal effects of bifenthrin on shrimp and fish.
 Bifenthrin was toxic to shrimp at environmentally relevant levels.
 Bifenthrin had no effect on oxidative stress assays in shrimp or fish.
 Bifenthrin was rapidly eliminated from the water column.
 Bifenthrin readily partitioned to sediments.

a r t i c l e i n f o a b s t r a c t

Article history: Bifenthrin is a widely used synthetic pyrethroid insecticide that is often applied to crops, turf, and
Received 25 November 2013 residential structures for the control of insects. Like other insecticides, bifenthrin has the potential to con-
Received in revised form 25 February 2014 taminate bodies of water that are adjacent to the application site via spray drift and runoff during storm
Accepted 8 March 2014
events. The objective of this study was to examine the lethal and sublethal effects of bifenthrin on grass
shrimp, Palaemonetes pugio, and sheepshead minnow, Cyprinodon variegatus in a 28 d mesocosm exper-
Handling Editor: S. Jobling iment under estuarine conditions. Endpoints included mortality and growth and the oxidative stress
biomarkers of lipid peroxidation, glutathione, and catalase.
Keywords: In the mesocosm experiment, 24 h and 96 h caged shrimp LC50s were 0.061 and 0.051 lg L1, respec-
Mesocosm tively. The uncaged grass shrimp 28 d LC50 was 0.062 lg L1. Fifty percent mortality was not reached in
Bifenthrin the uncaged sheepshead minnow. Bifenthrin did not have a significant effect on the growth of the shrimp,
Grass shrimp but there was an increasing impact on fish growth. However, it is uncertain as to whether this pattern is a
Sheepshead minnows direct effect of the chemical or if it is due to increased food availability resulting from mortality in prey
Estuarine species. The oxidative stress assays were largely inconclusive. Bifenthrin was eliminated rapidly from the
water column and readily partitioned to sediments. The LC50s for adult and larval P. pugio were below
published Estimated Environmental Concentration (EEC) values and were within the range of bifenthrin
concentrations that have been measured in rivers, channels, and creeks.
Published by Elsevier Ltd.

1. Introduction
Bifenthrin ((2-methyl-1,1-biphenyl-3-yl)-methyl-3-(2-chloro-
3,3,3-trifluoro-1-propenyl)-2,2-dimethylcycloprpanecarboxylate)
is a widely used pyrethroid that is the active ingredient in the
products TalstarÒ, BrigadeÒ, OnyxÒ, CaptureÒ, and Ortho Home
DefenseÒ. Bifenthrin is used as an insecticide and acaricide
⇑ Corresponding author. Tel.: +1 843 762 8620; fax: +1 843 762 8700.
E-mail address: paul.pennington@noaa.gov (P.L. Pennington).
1
Current address: USDA Agricultural Research Service, Eastern Regional Research
Center, Wyndmoor, PA, United States.
(Briggs, 1992) and is applied to crops, turf, and around the home
to control pests. Pyrethroids, like bifenthrin, are nervous system
toxicants and act primarily upon sodium channels in nerve cell
membranes (Narahashi, 2002). It acts by slowing down the activa-
tion and inactivation of the sodium channel gates causing them to
be open for longer periods of time (Narahashi, 2002).
In general, pyrethroids are highly nonpolar, have low water sol-
ubility, low volatility, high octanol–water partition coefficients and
a high affinity for soil or sediment particles. Bifenthrin has a
molecular weight of 422.9 g mol1 (Laskowski, 2002) and its water
solubility at 25 °C is 0.1 mg L1 (Fecko, 1999). A low vapor pressure

http://dx.doi.org/10.1016/j.chemosphere.2014.03.047
0045-6535/Published by Elsevier Ltd.
 P.L. Pennington et al. / Chemosphere 112 (2014) 18–25
(1.8E7 mmHg) and low Henry’s Constant (7.20E3 atm m3 -
mol1) suggest that bifenthrin is not easily volatilized into the
atmosphere and the high octanol–water partition coefficient
(1.0E+6) suggests that bifenthrin tends to partition into lipids
(Oros and Werner, 2005).
Bifenthrin is used both in agriculture and in urban areas for pest
control. Agriculturally, bifenthrin is used mainly for corn and cot-
ton (Thelin and Gianessi, 2000) and has been a major contributor
to improve cotton yields over the last two to three decades
(Solomon et al., 2001). In the US, between 1992 and 1995, a total
of 115 080 lb of the chemical were applied nationally to about
1 700 000 acres, with about 70 000 lb applied to cotton (Thelin
and Gianessi, 2000). In 2009, the agricultural use of bifenthrin
was estimated at 844 000 lb of chemical nationally (Thelin and
Stone, 2013). The trend in bifenthrin use has recently moved more
toward urban applications. In Central Valley California, 13 879 lb of
bifenthrin were used in urban applications in 2003, with 13 801 lb
used for structural pest control and 78 lb used for pest control in
landscaping (Oros and Werner, 2005). In 2005, it was reported that
45 000 lb of bifenthrin were applied in urban settings and 20 000 lb
were used for agriculture in California (Moran, 2005).
There does seem to be a seasonal trend in applications with peak
periods of agricultural application being in the spring and summer
months. For example, in California the majority of bifenthrin is
sprayed on cotton during the month of August and on alfalfa during
the month of July (Weston et al., 2004). In the North Atlantic region,
bifenthrin is applied between March and August, with the majority
being applied in May (Pait et al., 1992). This may increase the effects
of bifenthrin on aquatic organisms because data from previous
studies suggest that pyrethroid concentrations in aquatic habitats
tend to be greater shortly after their use rather than after heavy
rains (Weston et al., 2004) and these peak application times often
coincide with the spawning periods of aquatic organisms.
Bifenthrin has been detected in various aquatic settings includ-
ing agricultural drains, creeks, rivers, open wells, nursery runoff,
channels, and even golf course ponds (Kelley and Starner, 2004;
LeBlanc et al., 2004; Hunt et al., 2006; Smith Jr. et al., 2006). Surface
water concentrations of bifenthrin ranged from 0.005 to
3.79 lg L1 with the highest concentration measured in the Hines
Channel in California (Siepmann and Holm, 2000). Bottom and sus-
pended sediment concentrations ranged from 1.2 to 437 ng g1
(LeBlanc et al., 2004; Smalling et al., 2005; Weston et al., 2005)
with the highest concentration measured in bottom sediment from
a tributary of Kaseberg Creek in California (Weston et al., 2005).
Bifenthrin can potentially enter surface waters through a variety
of mechanisms depending on the product type used. These include
spray drift, particle transport or via storm water runoff.
It has been determined that bifenthrin is very highly toxic to
fish and aquatic invertebrates (Briggs, 1992) and therefore it is a
restricted use pesticide (USEPA, 2003). For example, the 96 h
LC50s (estimated concentration in which 50% of the test organisms
die after a 96 h exposure) for rainbow trout, bluegill sunfish,
sheepshead minnow, and mysids are 0.15, 0.35, 17.8, and
0.00397 lg L1, respectively.
Harper et al. (2008) indicated that there is an information gap
for the effects of bifenthrin in estuaries and that risks need to be
evaluated. That study demonstrated that bifenthrin is one of the
most toxic pesticides to the grass shrimp, Palaemonetes pugio, with
96 h LC50s for adult and larval shrimp at 0.020 lg L1 and
0.013 lg L1, respectively. Bifenthrin was found to be significantly
less toxic (> than an order of magnitude) to grass shrimp when
sediment was present.
The Estimated Environmental Concentration (EEC) for bifenth-
rin in surface water is 0.1 lg L1 for acute exposure and
0.032 lg L1 for chronic exposures (USEPA, 2000). Both of these
values are greater than median toxicity values determined by
19
Harper et al. (2008) indicating that the potential exists for impacts
to aquatic invertebrates at these EEC values.
The objective of this study was to further evaluate the fate and
toxicity of bifenthrin within the framework of an estuarine simula-
tion using a modular estuarine mesocosm (Lauth et al., 1996;
Pennington et al., 2004). Given that bifenthrin estuarine inverte-
brate toxicity values are lower than EEC values and that toxicity
in the presence of sediment is decreased, a more realistic exposure
scenario was devised. While using structural and functional com-
ponents of an estuarine system in a mesocosm setting, applications
of bifenthrin were applied to mesocosms weekly.
2. Materials and methods
2.1. Animal acquisition and holding
Juvenile sheepshead minnows (1–2 cm TL), Cyprinodon
variegatus, were purchased from Aquatic Biosystems in Fort Col-
lins, Colorado. The fish were held in the mesocosm building in a
340-L tank with 170 L of 20 ppt seawater at ambient temperature
and photoperiod. The fish were held for 28 d prior to testing to
allow them to acclimate and were fed daily with Tetramin Fish
FlakesÒ during the holding period.
Adult grass shrimp, P. pugio, were collected by dip net from
Leadenwah Creek on Wadmalaw Island, South Carolina
(32°38.8500 N, 080°13.3010 W). Only non-ovigerous shrimp were
used. They were held in the mesocosm building in a 340-L holding
tank with 170 L of 20 ppt seawater at ambient temperature and
photoperiod. The shrimp were held for 14 d prior to testing and
were fed daily with Tetramin Fish FlakesÒ.
2.2. Mesocosm setup
The mesocosm design was similar to that of Lauth et al. (1996)
with modifications following Pennington et al. (2004, 2007).
Briefly, the system consists of an upper tank containing the salt
marsh materials (sediment, flora, and fauna) and a lower tank (or
sump) that serves as a tidal seawater reservoir. The mesocosm
tanks were established in a greenhouse under ambient tempera-
ture and light. Approximately 294 L of 20 ppt seawater were added
to each tank. Sediment was collected from Leadenwah Creek
(32°38.8480 N, 080°13.2830 W), sieved with a 3-mm sieve, and
homogenized. It was then dispensed into sediment trays. Four sed-
iment trays were placed into each tank and elevated 5.08 cm to
allow for drainage. There was approximately 49 kg of sediment
per tank. The sediment to water ratio in the mesocosms was
approximately 165 g of sediment for every liter of water. This
was nearly the same ratio employed by Harper et al. (2008) in lab-
oratory scale assays with sediment (170 g L1). A semidiurnal tidal
cycle was set for each tank. The mesocosm tanks were setup one
month prior to dosing.
The mesocosms were dosed 4 times: on days 0, 7, 14, and 21 of
the 28 d exposure period. This weekly dosing matched a possible
application pattern for the bifenthrin containing product Brigade
2ECÒ (FMC Corporation, Philadelphia, PA) which, for many crops
such as brassicas, cucurbits, lettuce, spinach, cilantro, coriander,
beans, peas, okra, and turnip greens, can be applied weekly.
Depending on the maximum allowable amounts of bifenthrin
active ingredient per acre, these crops can receive anywhere from
2 to 5 weekly applications during the growing season. Bifenthrin
(97.2% cis-isomer and 2.5% trans-isomer) was purchased from
ChemService (West Chester, PA, USA) and was dissolved into
100% acetone.
There was a control and three treatments (0.002, 0.02,
0.2 lg L1) with three replicates each. These concentrations were
chosen because they bracketed the LC50 values determined by
20 P.L. Pennington et al. / Chemosphere 112 (2014) 18–25
Harper et al. (2008). The lower two doses were below the afore-
mentioned EEC values and the high dose was above both the acute
and chronic EEC values. Acetone was used as the carrier and it was
normalized at 0.01% for each dose. The controls and treatments
were randomly assigned to tanks in the greenhouse to compensate
for sunlight and temperature variations. The bifenthrin and ace-
tone were added to the water in the saltwater reservoir or sump.
The seawater was then circulated between the upper tank and
the sump to allow for mixing.
One week before dosing, 50 uncaged shrimp and 25 uncaged fish
were randomly added to each tank. Prior to adding the organisms, a
subset of ten shrimp and ten fish were measured for length and wet
weight. The length measurement for the shrimp was from the end
of the tail to the tip of the rostrum and length for the fish was total
length. Immediately after each of the four dosing and mixing
events, previously unexposed shrimp were randomly placed into
cages with ten shrimp per cage. Cages were constructed of clear,
rectangular acrylic glass (25 cm  5.3 cm  5.3 cm) covered with
2 mm nylon screen). Each cage contained 10 compartments holding
one shrimp each. The cages were then placed into the sumps with
one cage per tank, monitored daily and removed after 96 h. The pur-
pose of this was to quantify the toxicity of each successive dose in
an incipient fashion. Water quality parameters were taken daily
both by a YSI 556 MPS (YSI Inc. Yellow Springs, OH, USA) discrete
handheld sampler and four (one per treatment) YSI6920 multi-
parameter sondes (YSI Inc. Yellow Springs, OH, USA). Temperature,
pH, salinity, dissolved oxygen, and conductivity were measured.
Deionized water was added daily in order to maintain salinity
(20 ppt).
2.3. Mesocosm endpoints
Ninety-six hours after dosing, the shrimp cages were removed
from each tank and mortality was assessed. Living shrimp from
the first dose were frozen at 20 °C and used for the oxidative
stress assays. After the second dose (day 7), ten unexposed shrimp
were placed into each cage and the cages were deployed back into
the sumps. Weekly during the 28 d exposure, the cages were
removed, mortality was assessed, and the cages were deployed
again.
At the end of the 28 d experiment, mortality was assessed on all
uncaged organisms. Length and wet weight were also measured.
Five surviving fish from each tank were dissected and their livers
were removed. The five livers were pooled into one sample from
each tank in order to obtain enough tissue for analysis. They were
frozen at 70 °C and later used for the lipid peroxidation assay. Ten
surviving shrimp from each tank were frozen whole and were used
for the oxidative stress assays.
2.4. Lipid peroxidation assay
The lipid peroxidation assay followed the procedure in
Ringwood et al. (2003). Tissue was homogenized in 50 mM potas-
sium phosphate buffer (K2PO4) at four times the tissue weight. The
homogenate was transferred to microcentrifuge tubes and centri-
fuged at 13 000g for 5 min at 4 °C. Standards were prepared by
making serial dilutions of malondialdehyde tetraethylacetal
(MDA) with K2PO4 buffer. In another set of microcentrifuge tubes,
100 lL of each standard, blank (K2PO4), and sample supernatant
was added to 1400 lL of 0.375% thiobarbituric acid and 14 lL of
2% butylated hydroxytoluene.
The samples were then heated in a 100 °C water bath for 15 min
and were centrifuged at 13 000g for 5 min at room temperature.
The supernatant was then transferred to cuvettes and placed into
a spectrophotometer (Ultraspec 4300 Pro Spectrophotometer)
and absorbance was read at 532 nm. MDA concentrations in the
samples were determined from the standard curve after account-
ing for any dilutions and were measured in nmol g1 of tissue.
2.5. Glutathione assay
The glutathione assay followed the procedure in Ringwood et al.
(2003). Tissue was homogenized in 5% sulfosalicylic acid (SSA) at
ten times the tissue weight. The homogenate was transferred to
microcentrifuge tubes and centrifuged at 13 000g for 5 min at
4 °C. Standards were prepared by making serial dilutions of gluta-
thione (GSH) with 5% SSA. In another set of microcentrifuge tubes,
25 lL of each standard, blank (5% SSA), and sample supernatant
was added to 700 lL of 0.238 mg mL1 NADPH buffer, 100 lL of
10 mM 5,50 -dithiobis(2-Nitrobenzoic acid), and 175 lL of deion-
ized water. The tubes were vortexed and 900 lL of the mixture
was transferred to cuvettes. Next, 15 lL of 50 U mL1 glutathione
reductase (GSSG) was added to each cuvette. The cuvettes were
shaken and quickly placed into the spectrophotometer. Absorbance
was read at 405 nm every 15 s for at least 90 s. GSH concentrations
were determined from the standard curve after accounting for any
dilutions and were measured in nmol g1 of tissue.
2.6. Catalase assay
The catalase assay utilized a kit from Cell Technology
(Fluoro:Catalase Kit™) with a few modifications. Tissue was
homogenized in TRIS buffer at ten times the tissue weight for
shrimp and 100 times for fish. The samples were diluted 1:100
for shrimp and 1:500 for fish with 1 rxn buffer. Standards were
prepared by making serial dilutions of catalase with 1 rxn buffer.
Next, 50 lL of each standard, blank (1 rxn buffer), and samples
were added to a 96 well plate in triplicate along with 50 lL of
40 lM hydrogen peroxide.
The plate was then incubated in the dark for 60 min. A mixture
was prepared, including 9.3 mL of 1 rxn buffer, 100 lL of horse-
radish peroxidase, 100 lL of the detection reagent (ADHP), and
500 lL of 40 U mL1 superoxide dismutase. After the incubation
period, 100 lL of the mixture was added to each well and the sam-
ples were incubated an additional 15 min. The plate was then
placed into the microplate reader (FLx 800 microplate fluorescence
reader) and absorbance was read at an excitation/emission of 560/
590. Catalase activity was determined from the standard curve and
was measured in mmol of H2O2 consumed1 min1 mg of protein.
Protein concentrations were determined following Lowry et al.
(1951). In a microcentrifuge tube, 50 lL of the homogenate was
added to 950 lL of deionized water. The standards were prepared
by adding the appropriate amount of protein standard, deionized
water, and 50 lL of TRIS buffer. Next, 1.0 mL of Lowry reagent solu-
tion was added to all standards, blanks, and samples. The tubes
were vortexed and allowed to sit for 20 min. Afterwards, 500 lL
of Folin and Ciocalteu’s Phenol reagent was added to all tubes.
The tubes were vortexed again and allowed to sit for 30 min. The
solutions were then transferred to cuvettes and absorbance was
read at 500 nm. Protein concentrations were determined from
the standard curve and measured in milligrams of protein.
2.7. Analytical chemistry
Water and sediment samples were collected to examine the
transport and fate of bifenthrin. In the mesocosm experiment, each
replicate of a given concentration was sampled prior to the initial
dose, at 1 h after initial dose, 24 h after initial dose and before and
after each subsequent dose. While not ideal, but necessary due to
unavoidable laboratory constraints, water and sediment samples
were pooled across replicates to make composite samples for each
treatment at each time point.
 P.L. Pennington et al. / Chemosphere 112 (2014) 18–25
Separatory funnel liquid/liquid extraction was employed for
water samples, where 0.9 L was extracted with 100 mL of ethyl
acetate two times for 1 min and then passed through anhydrous
sodium sulfate. The extracts were reduced in volume using an
automated evaporation system (TurboVap II) to 0.5 mL. To verify
bifenthrin stock solution concentration, the stock solution under-
went serial dilution, and internal standard was added prior to
quantification using the gas chromatography–mass spectrometry.
Each batch of samples contained 1 reagent blank, 1 matrix spike,
and 1 matrix spike duplicate samples.
For sediment samples, 15 g of sediment was placed into an
Accelerated Solvent Extraction (ASE) 33 mL cell with 20 g of anhy-
drous sodium sulfate and 20 g of activated copper granules. After
adding the internal standard, the samples were extracted with
hexane:ethyl acetate (1:1, v:v) at 120 °C and 2000 psi. The result-
ing extracts were filtered through anhydrous sodium sulfate, evap-
orated to 0.5 mL at 40 °C and 15 psi with TurboVap II, solvent
exchanged to hexane, and evaporated again tp 0.5 mL.
The extracts were cleaned using Solid Phase Extraction (SPE)
cartridges (20 mL) containing Florisil (5 g), conditioned with
10 mL of ethyl acetate, and followed by 20 mL of hexane. After
the extract was passed through the SPE cartridge, they were eluted
with 15 mL of hexane:ethyl acetate (1:1, v:v) with flow rate of
2 mL min1. The extract was evaporated to 0.5 mL, solvent
exchange with hexane, and evaporated again until 0.5 mL, and
transferred to an autosampler vial for GC analysis.
Quantification of bifenthrin was based on internal standard cal-
culations using permethrin-phenoxy-13C6. The internal standard
was added to each sample prior to extraction and d-HCH was added
as a recovery compound before instrument analysis to enable the
calculation of internal standard recoveries. Analysis was performed
on a capillary gas chromatograph/mass spectrometer (Agilent
6890N/5973) using electron impact ionization operating in selec-
tive ion monitoring (SIM) mode. RH-5MS (Restek Corporation) cap-
illary column (30 m  0.25 mm  0.25 lm) was used. The method
detection limit (MDL) was calculated to be 0.4 ng L1 for water
samples and 0.3 ng g1 for sediment samples. Bifenthrin recoveries
were 112 ± 2% for water samples and 102 ± 7% for sediments.
2.8. Data analysis
The LC50 values and confidence intervals were calculated by the
Probit procedure [PROC Probit – SAS Institute Cary, NC, USA and
Newman (1995)] using either a normal, logistic, or a Wiebull distri-
bution. Nominal concentrations were used in the analysis rather
than measured concentrations due to the limited amount of bif-
enthrin residue sampling that occurred. If the data did not fit any
of the probit models (p < 0.05), then the Trimmed Spearman–Karber
approach was used (Hamilton et al., 1977; Newman, 1995). To look
for significant differences between treatments, Analysis of
Variance (ANOVA) was used. For this, normality and homogeneity
of variances were checked by looking at the Shapiro–Wilk test and
the Levene’s test respectively in addition to graphical examination
of residual plots. If the assumptions were met, then the one-way
ANOVA with Dunnett’s test for multiple comparisons was used
(Zar 1999, PROC GLM, SAS). If these assumptions were not met,
then a transformation was done. If the transformation did not
improve the data, then the Kruskal–Wallace with Dunnett’s test
was used. The no observable effects concentration (NOEC) was
the highest concentration in which there was no significant effect,
the lowest observable effects concentration (LOEC) was the lowest
concentration in which there was a significant effect, and the
threshold concentrations (the estimated concentration above
which an effect would be expected to occur) were determined by
taking the geometric mean on the NOEC and LOEC (DeLorenzo
et al., 2006).
21
The growth parameters were analyzed using one-way ANOVA
with Dunnett’s test to compare length and wet weight in the treat-
ments to the controls. Percent coefficient of variation (% CV) was
also calculated to show within treatment variation. The results
from the oxidative stress assays were also analyzed using one-
way ANOVA with Dunnett’s test. Log transformations were used
for a few of the data sets. Alpha (a) for all statistical tests was
set to 0.05 a priori.
3. Results
3.1. Caged shrimp mortality
After 24 h, there was 100% mortality of caged P. pugio in the
highest treatment (0.2 lg L1) for all weekly doses except for dose
four, which had 90% mortality (Supplemental Fig. 1). The 24 h LC50
ranged from 0.054 to 0.078 lg L1 with the highest LC50 being from
dose four (Table 1). At each dose, the NOEC, LOEC, and TC were
0.02, 0.2, and 0.063 lg L1, respectively. The results at the 96 h
time point showed similar results. The LC50 ranged from 0.054 to
0.059 lg L1 and the NOEC, LOEC, and TC were the same as the
24 h time point (Table 1, Supplemental Fig. 2).
3.2. Uncaged shrimp and fish mortality
As with the caged shrimp, there was 100% mortality in the
uncaged P. pugio at the highest mesocosm treatment (Supplemen-
tal Fig. 3). The 28 d LC50 was determined to be 0.062 lg L1 and the
NOEC, LOEC, and TC was 0.02, 0.2, and 0.063 lg L1, respectively.
Bifenthrin had a significant effect on the mortality of the uncaged
shrimp (p < 0.0001). The uncaged C. variegatus results showed that
there was only slight mortality in the control and the lowest treat-
ment of 0.002 lg L1. The NOEC, LOEC, and TC were not able to be
determined and the LC50 was greater than the highest treatment
(0.2 lg L1). Bifenthrin did not have a significant effect on the mor-
tality of the uncaged sheepshead minnows in the mesocosm study
(p = 0.5296).
3.3. Growth
The lengths of adult P. pugio ranged from 2.0 to 3.8 cm and were
consistent within treatments (8.99–12.80% CV). There was no over-
all effect of bifenthrin on length (p = 0.5673) and there were no sig-
nificant differences between treatments (Supplemental Fig. 4). The
wet weights of the shrimp ranged from 0.056 to 0.289 g and were
not as consistent within treatments as the lengths (21.05–39.29%
CV). As with the lengths, there was no effect of bifenthrin on
weight (p = 0.9862) and there was no significant difference
between treatments (Supplemental Fig. 4).
The lengths of juvenile C. variegatus ranged from 1.6 to 3.5 cm
and were fairly consistent within treatments (9.27–16.94% CV).
There was an overall effect of bifenthrin on length (p = 0.0419)
and there was a slight increasing trend with increasing concentra-
tion (Supplemental Fig. 5). The wet weights ranged from 0.042 to
0.847 g and they were not as consistent within treatments
(33.17–65.81% CV). Bifenthrin did have an effect on weight
(p = 0.028) and there was an increasing trend with increasing con-
centration (Supplemental Fig. 5).
3.4. Lipid peroxidation assay
The caged P. pugio (96 h, 1 dose) and the uncaged P. pugio (28 d,
4 doses) from the mesocosm showed that the controls had slightly
higher malondialdehyde (MDA) levels than the treatments
(Supplemental Fig. 6). There was no significant effect for either
22 P.L. Pennington et al. / Chemosphere 112 (2014) 18–25

Table 1
Toxicity of bifenthrin to caged grass shrimp.

Test Week Hours of exposure (h) LC50 (lg L1) 95% CI (lg L1)
Initial dose (dose 1) Week 1 24 0.054 0.044–0.067
96 0.054 0.044–0.067
Redose (dose 2) Week 2 24 0.059 0.050–0.068
96 0.059 0.050–0.068
Redose (dose 3) Week 3 24 0.063 NA
96 0.059 0.050–0.068
Redose (dose 4) Week 4 24 0.078 0.051–0.113
96 0.056 0.039–0.085

the caged and uncaged shrimp (p = 0.7947 and p = 0.2139, respec-
tively). In the uncaged C. variegatus, there were elevated MDA lev-
els in the 0.002 lg L1 treatments but then it dropped back down
to levels similar to the controls (Supplemental Fig. 7) and there
was not a statistically significant effect (p = 0.2375).
3.5. Glutathione assay
In the caged and uncaged shrimp there were slightly higher
GSH levels in the higher concentrations but once again they were
not significant (Supplemental Fig. 6). The overall p-values for the
caged and uncaged shrimp were 0.2466 and 0.4981, respectively.
3.6. Catalase assay
In the caged shrimp, there was a slight increasing catalase activ-
ity trend with increasing bifenthrin concentrations (Supplemental
Fig. 6), but it was not statistically significant (p = 0.6265). Catalase
activity in the uncaged shrimp was higher in the controls than in
the treatments with an overall p-value of 0.2788.
the nominal dose) for TRTs B and C followed a single-ordered expo-
nential elimination model during the first seven days (Fig. 1).
Bifenthrin steadily accumulated in sediments (Table 3)
throughout the 28 d (Fig. 2).
3.8. Mesocosm water quality parameters
The water quality parameters for the mesocosm experiment are
summarized in Supplemental Fig. 9 and overall there were no dif-
ferences between treatments. Conductivity ranged from 26.00 to
35.35 mS cm1 and remained fairly constant throughout the 28 d
with a slight decline at dose 4 (10/25). Salinity ranged from
20.14 to 22.00 ppt and pH ranged from 7.80 to 8.54. Temperature
also remained fairly constant except for a decrease at dose 4, in
which it dropped to 13.42 °C. The maximum temperature mea-
sured was 26.9 °C. Dissolved oxygen ranged from 4.61 to
13.53 mg L1, with the lowest levels measured early in the morn-
ing (0800).
4. Discussion

3.7. Analytical chemistry
The nominal stock concentration was 14.6 lg L1 and the actual
measured concentration was 15.4 lg L1 (105.35% of nominal).
Bifenthrin concentrations in the mesocosm are shown in Table 2.
Bifenthrin was not detected in the control mesocosms and treat-
ments ranged from 98.60% to 159.30% of the nominal 1 h after each
dose. There was a slight additive effect between the four doses and
although the concentrations increased slightly they were still fairly
similar to the nominal (Supplemental Fig. 8).
Bifenthrin was quickly eliminated from the water column
(Table 2). Seven days after each dose water column bifenthrin con-
centrations were less than the limit of detection in the lowest dose
(TRT A). TRT B and TRT C 7 d concentrations averaged 8.1% (±0.97)
and 10.9% (±1.8) of the 1 h measured concentrations, respectively.
Water column bifenthrin concentrations (expressed as a percent of
The caged shrimp in the mesocosm exposure showed similar
responses throughout the doses, except for dose four. The 24 h
LC50 for dose four was slightly higher (Table 1). This was initially
believed to be a result of the drop in water temperature down to
13 °C at the time of dose four (Supplemental Fig. 9). Although pyre-
throids usually have a negative temperature coefficient (Narahashi,
1972, 1989, 1992), it is possible that temperature sensitivity can
vary between species.
When the 24 h LC50 for the mesocosm-caged shrimp (expressed
as a composite value from all four doses) was compared to the 24 h
LC50s for laboratory exposed shrimp (Supplemental Table 1) from
Harper et al. (2008), significant differences were found (Supple-
mental Table 2). Using the LC50 Ratio test by Wheeler et al.
(2006), the caged shrimp (0.061 lg L1) were significantly less sen-
sitive at 24 h (p < 0.0001) than the laboratory exposed shrimp from
the aqueous assay (0.038 lg L1). These results were expected

Table 2
Measured bifenthrin concentrations (lg L1) from water column samples.

Test Week Duration (post dose) Control 0 lg L1 TRT A 0.002 lg L1 TRT B 0.02 lg L1 TRT C 0.2 lg L1
Initial dose (dose 1) Week 1 1h <LOD 0.00284 0.0198 0.204
24 h <LOD 0.00208 0.00862 0.0810
7d <LOD <LOD 0.00158 0.0189
Redose (dose 2) Week 2 1h <LOD 0.00274 0.0237 0.245
7d <LOD <LOD 0.00262 0.023
Redose (dose 3) Week 3 1h <LOD 0.00289 0.0252 0.255
7d <LOD <LOD 0.00409 0.0133
Redose (dose 4) Week 4 1h <LOD 0.00311 0.0319 0.273
7d <LOD <LOD 0.00271 0.0239

LOD = limit of detection.

 P.L. Pennington et al. / Chemosphere 112 (2014) 18–25
Fig. 1. Water column bifenthrin concentrations (expressed as a percent of the
nominal dose) for TRTs B and C followed a single-ordered exponential elimination
curve during the first seven days.
because the bifenthrin most likely partitioned to the sediment par-
ticles in the mesocosm units, therefore reducing availability for
uptake or exposure. The caged shrimp (0.061 lg L1) were signifi-
cantly more sensitive at 24 h (p < 0.0001) than the laboratory
exposed shrimp from the sediment assay (0.339 lg L1) of Harper
et al. (2008). This result was not expected. Although there was an
equal ratio of sediment to water (170 g L1) in both the sediment
assay and the mesocosm tanks, the mesocosm tanks also had rooted
vegetation as well as other organisms. It was expected that the sed-
iment, plants, and animals in the mesocosm tanks would have
reduced the availability of bifenthrin to caged shrimp thereby
resulting in a higher LC50 for these shrimp but this was not the case.
The cages may have caused additional stress to the shrimp by
restricting mobility making them more susceptible to the toxin.
Another plausible reason for the lower LC50 for the caged
shrimp is that because the mesocosm is a more dynamic system
and there may have been mixing issues resulting in areas of higher
bifenthrin concentrations. The caged shrimp were deployed into
the sumps, where the dosing occurred. Therefore, it is possible that
the shrimp were exposed to higher than nominal concentrations
for a short period of time.
In P. pugio, there were no significant differences in either length
or wet weight. In the C. variegatus, there was an increasing trend in
length and wet weight with increasing concentrations with the
highest concentration being significantly greater than the control
(p = 0.031). It was unclear as to whether this was a direct or indi-
rect effect of the chemical on growth. It is possible that there
was an increase in food availability due to the bifenthrin mediated
increased mortality in amphipods, polychaetes, shrimp, and other
aquatic organisms that resided in the mesocosm tanks. It is also
possible that hormesis may be one of the causes of the increased
Table 3
Sediment bifenthrin concentrations for TRT C (0.2 lg L1 nominal dose).
Test Week Duration (post dose)
Initial dose (dose 1) Week 1 1h
7d
Redose (dose 2) Week 2 1h
7d
Redose (dose 3) Week 3 1h
7d
Redose (dose 4) Week 4 1h
7d
a
These values are less than the reported MDL (0.3 ng g1).
23
Fig. 2. Sediment bifenthrin concentration accumulation plots at 24 h and 7 d after
each dose for TRT C. The 7 d sampling events occurred just prior to next dose. Each
data point represents a composite sample across the three replicates from each
treatment.
growth in C. variegatus exposed to bifenthrin. Hormesis is a
dose–response relationship in which there is a stimulatory
response at low levels of potentially toxic agents (Stebbing,
1982; Calabrese and Baldwin, 2001).
Similar results were found in a study done with the pyrethroid
esfenvalerate on larval crimson-spotted rainbow fish (Barry et al.,
1995). They observed larger larvae in treatments with significant
mortality than those in the controls. They were also unable to
determine if this was a direct effect or if it was due to density-
dependent interactions. In another study (Wirth et al., 2004)
which utilized the same mesocosm design as the present study,
C. variegatus exhibited a trend of increased fish production with
increased fipronil concentration after 64 d. As in the study by
Barry et al. (1995) and Wirth et al. (2004) was not able to determine
if the observed effect was a result of direct contaminant exposure.
Like the results of Harper et al. (2008), the oxidative stress
assays (lipid peroxidation, glutathione, and catalase) indicated that
there was no evidence of an oxidative stress response as a result of
sublethal bifenthrin expose for P. pugio. The sheepshead minnow,
C. variegatus, also failed to show a significant response in lipid per-
oxidation as a result of bifenthrin exposure at the concentrations
tested.
While our study had limited bifenthrin residue sampling, it does
appear that bifenthrin in the water column follows, at least, a single
exponential elimination model (Fig. 1). A greater frequency of sam-
pling might have revealed a higher ordered elimination model from
the water column, but the basic result would be the same; that the
residence time of bifenthrin in the water column is very short. It is
clear from Fig. 1 that the nominal concentration was reduced below
50% well before the 24 h time point. Given the amount of sediment,
the other sites of potential of partitioning (biota), and bifenthrin’s
TRT C ng g1 wet weight TRT C ng g1 dry weight
a
0.09 0.12a
0.14a 0.24a
0.31 0.52
0.32 0.54
0.68 1.15
0.57 0.98
0.86 1.55
0.58 0.99
24 P.L. Pennington et al. / Chemosphere 112 (2014) 18–25
high KOW (1.0  106), this finding was expected. A greater sampling
frequency would have also allowed for the use of time-weighted
measured concentrations to be used for the LC50 determinations
rather than the nominal concentrations.
Bifenthrin rapidly bound to sediments as indicated in Fig. 2 for
TRT C. This graph shows that 1 h after each of the four weekly
doses, the bifenthrin concentration in the sediments rose sharply.
These measures showed a linear increase over time (r2 = 0.99).
Seven days after each dose (and immediately prior to the next
dose), bifenthrin concentration also rose in a linear fashion
(r2 = 0.91), but the concentrations were less than what was mea-
sured at the 1 h post-dose samples. This difference between the
1 h samples and the 7 d samples after each dose could be due to
bioaccumulation into biota (which was not measured), the bifenth-
rin equilibrium shifting back towards the water column, or degra-
dation of the compound into metabolites that were not accounted
for in the analysis. It is not clear as to how persistent the bifenthrin
was in the sediment after the 28 d time point.
Water quality parameters remained quite consistent during the
study as evidenced by visual observation of the data in Supplemen-
tal Fig. 9. There does not appear to be an effect of the bifenthrin on
water quality conditions during the study, therefore it can be con-
cluded that these parameters did not contribute to the difference
observed in effects or bifenthrin’s partitioning in the system.
In this realistically designed mesocosm study where dose tim-
ings followed typical application timings for an actual product, bif-
enthrin was found to be lethal to the grass shrimp, P. pugio, at
concentrations below the identified acute EEC value (0.1 lg L1 –
(USEPA, 2000)). While bifenthrin was found to rapidly disappear
from the water column after dosing, we can also conclude that bif-
enthrin readily and rapidly accumulates in sediments and thus
presents a significant hazard to infaunal invertebrate species.
Future studies should focus the effects of bifenthrin in sediments
to estuarine infauna. While the presence of sediment significantly
lowered the toxicity of bifenthrin as opposed to assays with no
sediment, lethality remains an issue for crustaceans due to mixing
patterns and the potential for repeated applications.
This study stresses the importance of following product labels.
In the United States, most formulated products require buffer
zones of at least 10–25 feet between treated areas and aquatic hab-
itats. There are also strict requirements regarding applications and
wind speeds to prevent against spray drift. It is important for pes-
ticide applicators, both licensed and unlicensed (home and garden
use) to read and follow the label instructions very carefully to pro-
tect sensitive aquatic life.
Acknowledgements
We would like to thank Jennifer Hoguet with her assistance in
performing the oxidative stress biomarker assays. This publication
does not constitute an endorsement of any commercial product or
intend to be an opinion beyond scientific or other results obtained
by the National Oceanic and Atmospheric Administration (NOAA).
No reference shall be made to NOAA, or this publication furnished
by NOAA, to any advertising or sales promotion which would indi-
cate or imply that NOAA recommends or endorses any proprietary
product mentioned herein, or which has as its purpose an interest
to cause the advertised product to be used or purchased because of
this publication.
Appendix A. Supplementary material
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.chemosphere.
2014.03.047.
References
Barry, M.J., Logan, D.C., Ahokas, J.T., Holdway, D.A., 1995. Effects of esfenvalerate
pulse-exposure on the survival and growth of larval Australian crimson-spotted
rainbow fish (Melanotaenia-Fluviatilis). Environ. Toxicol. Water 10, 267–274.
Briggs, S.A., 1992. Basic Guide to Pesticides: Their Characteristics and Hazards. CRC.
Calabrese, E.J., Baldwin, L.A., 2001. The frequency of U-shaped dose responses in the
toxicological literature. Toxicol. Sci. 62, 330–338.
DeLorenzo, M.E., Serrano, L., Chung, K.W., Hoguet, J., Key, P.B., 2006. Effects of the
insecticide permethrin on three life stages of the grass shrimp, Palaemonetes
pugio. Ecotoxicol. Environ. Saf. 64, 122–127.
Fecko, A., 1999. Environmental Fate of Bifenthrin. Environmental Monitoring and
Pest Management Branch, Department of Pesticide Regulation, Sacramento.
Hamilton, M.A., Russo, R.C., Thurston, R.V., 1977. Trimmed Spearman–Karber
method for estimating median lethal concentrations in toxicity bioassays.
Environ. Sci. Technol. 11, 714–719.
Harper, H.E., Pennington, P.L., Hoguet, J., Fulton, M.H., 2008. Lethal and sublethal
effects of the pyrethroid, bifenthrin, on grass shrimp (Palaemonetes pugio) and
sheepshead minnow (Cyprinodon variegatus). J. Environ. Sci. Health, Part B 43,
476–483.
Hunt, J., Anderson, B., Phillips, B., Tjeerdema, R., Richard, N., Connor, V., Worcester,
K., Angelo, M., Bern, A., Fulfrost, B., Mulvaney, D., 2006. Spatial relationships
between water quality and pesticide application rates in agricultural
watersheds. Environ. Monit. Assess. 121, 243–260.
Kelley, K., Starner, K., 2004. Preliminary Results for Study 219: Monitoring Surface
Waters and Sediments of the Salina and San Joaquin River Basins for
Organophosphate and Pyrethroid Pesticides. California Environmental
Protection Agency.
Laskowski, D.A., 2002. Physical and chemical properties of pyrethroids. Rev.
Environ. Contam. Toxicol. 174, 49.
Lauth, J.R., Scott, G.I., Cherry, D.S., Buikema, A.L.J., 1996. A modular estuarine
mesocosm. Environ. Toxicol. Chem. 15, 630–637.
LeBlanc, L.A., Orlando, J.L., Kuivila, K.M., 2004. Pesticide Concentrations in
Water and in Suspended and Bottom Sediments in the New and Alamo
Rivers, Salton Sea, Watershed California, April 2003. US Geological Survey
Reston, Virginia.
Lowry, O.H., Rosebrough, N.J., Farr, A.L., Randall, R.J., 1951. Protein measurement
with the Folin phenol reagent. J. Biol. Chem. 193, 265–275.
Moran, K.D., 2005. Pesticides in Urban Surface Water: Urban Pesticides Use Trends
Annual Report. TDC Environmental, San Mateo, CA.
Narahashi, T., 1972. Effects of insecticides on excitable tissues. Adv. Insect Physiol.
8, 1. APL.
Narahashi, T., 1989. The role of ion channels in insecticide action. In: Narahashi, T.,
Chambers, J.E. (Eds.), Insecticide Action: From Molecule to Organism. Plenum
Publishing Corporation, New York, pp. 55–84.
Narahashi, T., 1992. Nerve membrane Na+ channels as targets of insecticides.
Trends Pharmacol. Sci. 13, 236–241.
Narahashi, T., 2002. Nerve membrane ion channels as the target site of insecticides.
Mini-Rev. Med. Chem. 2, 419–432.
Newman, M.C., 1995. Quantitative Methods in Aquatic Ecotoxicology. CRC.
Oros, D.R., Werner, I., 2005. Pyrethroid Insecticides: An Analysis of Use Patterns,
Distributions, Potential Toxicity and Fate in the Sacramento-San Joaquin Delta
and Central Valley. San Francisco Estuary Institute Oakland, California.
Pait, A.S., De Souza, A.E., Farrow, D.R.G., 1992. Agricultural Pesticide Use in Coastal
Areas: A National Summary. USDOC NOAA, Rockville, Maryland.
Pennington, P.L., DeLorenzo, M.E., Lawton, J.C., Strozier, E.D., Fulton, M.H., Scott, G.I.,
2004. Modular estuarine mesocosm validation: ecotoxicological assessment of
direct effects with the model compound endosulfan. J. Exp. Mar. Biol. Ecol. 298,
369–387.
Pennington, P.L., DeLorenzo, M.E., Key, P.B., Wirth, E.F., Fulton, M.H., Scott, G.I., 2007.
The Design, Construction, Operation, and Maintenance of the Replicated
Modular Estuarine Mesocosm, Chalreston, SC, p. 77.
Ringwood, A., Hoguet, J., Keppler, C., Gielazyn, M., Ward, B., Rourk, A., 2003. Cellular
Biomarkers (Lysosomal Destabilization, Glutathione and Lipid Peroxidation) in
Three Common Estuarine Species: A Methods Handbook. Marine
Resources Research Institute, South Carolina Department of Natural
Resources, Charleston.
Siepmann, S., Holm, S., 2000. Hazard Assessment of the Synthetic Pyrethroid
Insecticides Bifenthrin, Cypermethrin, Esfenvalerate, and Permethrin to Aquatic
Organisms in the Sacramento-San Joaquin River system. Resources Agency,
Department of Fish and Game.
Smalling, K.L., Orlando, J.L., Kuivila, K.M., 2005. Analysis of Pesticides in Surface
Water and Sediment From Yolo Bypass, California, 2004–2005. US Geological
Survey. p. 20.
Smith Jr, S., Cooper, C.M., Lizotte Jr., R.E., Shields Jr., F.D., 2006. Storm pesticide
concentrations in Little Topashaw Creek, USA. Int. J. Ecol. Environ. Sci. 32, 173–
182.
Solomon, K.R., Giddings, J.M., Maund, S.J., 2001. Probabilistic risk assessment of
cotton pyrethroids: I. Distributional analyses of laboratory aquatic toxicity data.
Environ. Toxicol. Chem. 20, 652–659.
Stebbing, A.R.D., 1982. Hormesis—the stimulation of growth by low levels of
inhibitors. Sci. Total Environ. 22, 213–234.
Thelin, G.P., Gianessi, L.P., 2000. Method for Estimating Pesticide Use for County
Areas of the Conterminous United States. US Geological Survey US Department
of Interior, Sacramento, CA. p. 62.
 P.L. Pennington et al. / Chemosphere 112 (2014) 18–25
Thelin, G.P., Stone, W.W., 2013. Estimation of Annual Agricultural Pesticide Use for
Counties of the Conterminous United States, 1992–2009. US Department of the
Interior, US Geological Survey.
USEPA, 2000. Bifenthrin; Pesticide Tolerances for Emergency Exemptions. Federal
Register 65 FR 57972, 57972–57980.
USEPA, 2003. Updated Restricted Use Products Report. USEPA, Washington, DC.
Weston, D.P., You, J., Lydy, M.J., 2004. Distribution and toxicity of sediment-
associated pesticides in agriculture-dominated water bodies of California’s
central valley. Environ. Sci. Technol. 38, 2752–2759.
25
Weston, D.P., Holmes, R.W., You, J., Lydy, M.J., 2005. Aquatic toxicity due to
residential use of pyrethroid insecticides. Environ. Sci. Technol. 39, 9778–
9784.
Wheeler, M.W., Park, R.M., Bailer, A.J., 2006. Comparing median lethal concentration
values using confidence interval overlap or ratio tests. Environ. Toxicol. Chem.
25, 1441–1444.
Wirth, E.F., Pennington, P.L., Lawton, J.C., DeLorenzo, M.E., Bearden, D., Shaddrix, B.,
Sivertsen, S., Fulton, M.H., 2004. The effects of the contemporary-use insecticide
(fipronil) in an estuarine mesocosm. Environ. Pollut. 131, 365–371.