(This is a sample cover image for this issue. The actual cover is not yet available at this time.)This article appeared in a journal published by Elsevier. The attachedcopy is furnished to the author for internal non-commercial researchand education use, including for instruction at the authors institution and sharing with colleagues. Other uses, including reproduction and distribution, or selling or licensing copies, or posting to personal, institutional or third party websites are prohibited. In most cases authors are permitted to post their version of the article (e.g. in Word or Tex form) to their personal website or institutional repository. Authors requiring further information regarding Elsevier’s archiving and manuscript policies are encouraged to visit: http://www.elsevier.com/copyright
Authors personal copy P.C. Pessoa et al. / Aquatic Toxicology 105 (2011) 312–320 313LC50 to different freshwater ﬁsh vary from 88 to 1990 g/L (USEPA, 26 (initial juvenile) according to (Fujimura and Okada, 2007). Fish2004), whereas crustaceans and insects are among the most sensi- were transported to the Ecotoxicology Laboratory at Universidadetive organisms, with 96 h LC50 s varying from 1.6 to 500 g/L (Dutra Federal de Pernambuco (UFPE) in plastic containers, and were keptet al., 2009). Environmental agencies in the state of Pernambuco in 70 L tanks until experiments begun. Water was renewed daily at ado not monitor organic contaminants in water bodies, and there rate of 100%, after passing through 5 and 1 m pressure ﬁlters. Tem-is no information available on carbofuran concentrations in water. perature was kept at 25.5 ± 0.5 ◦ C, pH was 7.5 ± 0.5, total hardnessIt has been shown that carbofuran concentration in irrigated rice was 100 mg CaCO3 L−1 and oxygen was 6 ± 0.5 mg L−1 . Fish wereﬁelds in the southeast of Brazil can reach maximum concentrations fed 45% protein commercial ﬁsh food (Alcon Artemia, Camboriu,of 233 g/L in laminar water (Plese, 2005). Additionally, carbofu- Brazil), and total ammonia, nitrite and nitrate were undetectable.ran concentrations of 26 g/L and 264 g/L have been detected Experiments have been carried out in accordance with The Code ofin a headwater stream, and in agricultural ﬁeld drains that ﬂow Ethics of the World Medical Association (Declaration of Helsinki)to this headwater stream in UK, respectively (Matthiessen et al., for animal experiments.1995). Estimated environmental concentrations of carbofuran insurface water for selected use patterns in the United States havebeen modeled and vary from 5.2 to 36 g/L using maximum rates of 2.2. Exposure proceduresapplication (USEPA, 2006). Adverse sublethal effects can result fromexposure to these concentrations of carbofuran. Goldﬁsh exposed A stock solution of carbofuran at 8.75 g/L was prepared by addingto carbofuran concentrations as low as 1 g/L showed reduced 2.5 mL of the commercial form of carbofuran as 350 g/L Furadanattraction to a chironomid extract (Saglio et al., 1996), and when 350-SC (FMC Corporation, Philadelphia, USA) into a 100 mL volu-the same species was exposed to 5 g/L, signiﬁcant alterations in metric ﬂask completed with distilled water. Different volumes ofsheltering and burst swimming were found (Bretaud et al., 2002). In this stock solution were added daily into mixing aquariums withanother study, the amplitude of the electro-olfactogram of paciﬁc 20 L of clean water to create nominal exposure concentrations ofsalmon (Oncorhynchus sp.) exposed to 10.4 g/L carbofuran was 10, 50, 100, 200, 300 and 400 g/L carbofuran plus a control group.reduced (Jarrard et al., 2004). After homogenization of stock solutions added into aquariums with Low concentrations of organophosphates and carbamates can 20 L volume, water ﬂowed by gravity into exposure aquariums withinhibit acetylcholinesterase, leading to accumulation of the neu- 15 L of water, providing a daily renewal rate of 133% for all treat-rotransmitter acetylcholine in the synaptic gap of cholinergic ments, including the control.synapses and neuromuscular junctions, effect that has become a A second stock solution was prepared from the dissolution ofclassic biomarker of exposure to these chemicals (Sturm et al., 2.5 g of carbofuran 98% pure (Sigma–Aldrich, St. Louis, USA) into1999). Acetylcholinesterase inhibition has been used as an indi- 100 mL of reagent grade ethanol in a volumetric ﬂask. Different vol-cator of potential ﬁsh exposure to these agrochemicals in Brazil umes of this stock solution were added daily into mixing aquariums(Oliveira et al., 2007). However, information regarding whether this with 20 L to create exposure concentrations of 10, 100, 200, 300,biomarker of exposure might be used to ultimately predict more 400 and 500 g carbofuran L−1 plus a control and a solvent controlecologically relevant endpoints related to this exposure are lacking. group. This solution was used only for mortality experiments.As pointed out by Scholz and Hopkins (2006), there are important Samples of water from these treatments were analyzed by liquiddata gaps that need to be addressed before these predictions can be chromatography for carbofuran after extraction of 1 L with methy-made, and these include an understanding of sublethal effects that lene chloride, and the extract was dried and concentrated to acan be linked to deﬁcits in individual ﬁtness. Acetylcholinesterase volume of 10 mL. The extract was cleaned up on a C-18 cartridge,inhibition in ﬁsh sublethaly exposed to these agrochemicals has ﬁltered, and eluted on a C-18 analytical column with a mobilebeen related to several measures of behavioral toxicity, includ- phase of 50% acetonitrile in water at a ﬂow rate of 2.0 mL/min,ing neuromotor effects on swimming activity (Brewer et al., 2001; with a retention time of 3.5 min and ultra-violet (UV) detectionLittle and Finger, 1990), effects on speciﬁc sensorial systems such as at 280 nm, according to (USEPA, 2010). Method detection limitvision (Dutta et al., 1992) and smell (Tierney et al., 2008), as well as was 3.2 g carbofuran L−1 . Chemical analysis results indicated thatbehavioral measures encompassing more complex situations like measured water concentrations were below nominal concentra-predator–prey interactions (Scholz et al., 2000). Proper functioning tions by a factor varying from 5% to 30%, and measured carbofuranof swimming skills and sensorial systems is essential for success- concentrations were subsequently used in all graphs and analysis.ful detection, attack and capture of prey, as well as for predator The effect of carbofuran concentrations prepared from 98% car-evasion (Fuiman et al., 2006). The joint assessment of traditional bofuran (Sigma–Aldrich, St. Louis, USA) and prepared from thebiomarkers like cholinesterase inhibition with more ecologically commercial form Furadan 350-SC on larval mortality was com-relevant behavioral biomarkers will provide important knowledge pared. On this experiment groups freely swimming larvae werethat can help in the development of predictive models of popu- exposed to each of the carbofuran concentrations described abovelation level effects for ﬁshery resources. Within this framework, in 15 L aquariums, a group of 100 ﬁsh for each concentration fromthe present study was designed to evaluate the effects of exposing each stock solution preparation method.larval tilapia Oreochromis niloticus to carbofuran on cholinesterase All behavioral parameters and cholinesterase activity measuredinhibition and behavioral parameters related to vision, swimming, were based on larvae exposed to carbofuran solutions preparedprey capture, predator evasion and growth. from the commercial form of carbofuran, Furadan 350-SC, for higher environmental realism. On these experiments 10 larvae were kept in each one of four 250 mL volume ﬂoating beakers2. Materials and methods adapted with 300 m windows to allow water ﬂow in and out while they ﬂoated in the 15 L exposure aquariums. A total of 402.1. Animals larvae (4 replicates of 10) were exposed to each concentration described before. Each group of 10 larvae was fed once a day with Nile tilapia (O. niloticus) larvae at 9 days post hatch, total length 0.3 g of 45% protein commercial ﬁsh food (Alcon Artemia, Cambo-9 mm, were obtained from the Fish Culture Facility Professor Johei riu, Brazil). Dead larvae were counted and removed daily. MortalityKoike at Universidade Federal Rural de Pernambuco (UFRPE). Fish rates were calculated by totaling all dead larvae from each exposureused in experiments were between stages 23 (advanced larvae) and concentration at the end of a 96 h exposure period.
Authors personal copy314 P.C. Pessoa et al. / Aquatic Toxicology 105 (2011) 312–3202.3. Video recording system 2.7. Prey capture Behavioral tests for swimming activity, feeding and predator During prey capture tests run on the same aquariums used forevasion were based on digital video recordings obtained from a swimming tests an individual tilapia larva was used in each test.closed circuit television (CFTV) system based on cameras with Model prey used were Daphnia magna neonates (24–48 h old), total6–60 mm zoom lenses that allowed the capture of a full superior length between 3 mm and 5 mm. Each larval tilapia was acclimatedview of each experimental arena. A video encoder board (Geovision for 5 min inside a PVC tube in the test aquarium, while 5 D. magnamodel GV-800, Irvine, CA) received the images from the cameras neonates were also enclosed in another PVC tube inside the testmonitoring one experimental arena from each exposure concen- aquarium, and the experimental arena setup was enclosed by atration (a total of 7 cameras) simultaneously, and videos were black curtain to avoid disturbance. All PVC tubes holding the preda-recorded on a hard disk for later analysis. tor and prey were tied to a PVC rod that when lifted allowed the simultaneous release of all predator and prey in each of the 7 are-2.4. Cholinesterase activity nas monitored simultaneously by the video system. Predator–prey interactions were recorded for approximately 10 min, and 15 ﬁsh Each sample consisted of pools of 3 whole tilapia larvae homog- from each exposure concentration were individually tested. Pos-enized after being euthanized, due to their small size. For each terior video analysis generated data on the number of attackscarbofuran concentration, 5 samples were analyzed (n = 5 for each performed by each tilapia larva towards prey during the test.concentration). Larval pools were homogenized in Tris buffer (HCl50 mM, KCl 0.15 M, pH 7.4, PMSF 0.1 mM) in the proportion of 2.8. Predator evasion1 g of sample to 4 mL of buffer (1:4), using a Potter homoge-nizer (Glas-Col). Samples were then centrifuged at 9000 × g during Individual tilapia larvae originating from each of the different20 min at 4 ◦ C. The supernatant (S9 fraction) was separated in exposure concentrations and a control predator (Parachromis man-aliquots and stocked at a −80 ◦ C freezer for posterior measure- aguensis) were conﬁned inside PVC tubes at opposing sides of anment of cholinesterase activity according to the method described aquarium (14 cm length × 10.5 cm width × 13.5 cm height). All PVCby Ellman et al. (1961) adapted to 96 well microplates. Analy- tubes holding the prey (tilapia) and predator were tied to a PVC rodses were run on duplicates at 25 ◦ C using the microplate reader that when lifted allowed the simultaneous release of all prey andSpectramax 250 (Molecular Devices, Sunnyvale, CA). Reagents were predator in each of the 7 arenas monitored simultaneously by thepurchased from Sigma (St. Louis, MO, EUA). Total protein was mea- video system (Fig. 1). Prey–predator interactions were recorded forsured according to Peterson (1977), using bovine serum albumin approximately 10 min. A total of 15 ﬁsh from each treatment wereas standard. Cholinesterase activity was normalized to total pro- tested individually. Posterior video analysis generated data on thetein content of samples. We used only acetylthiocholine iodide number of predator attacks needed to capture each prey (tilapia).(ASCh) as substrate during our measurements in whole ﬁsh, butcholinesterases present in liver and muscle of Nile tilapia have 2.9. Weight gainproperties that resemble both AChE and BChE (Rodríguez-Fuentesand Gold-Bouchot, 2004). Therefore, our measurements might Before exposure initiated, 25 control tilapia larvae were sacri-relate to total cholinesterase activity, and will be referred to as ChE. ﬁced, dried on a paper towel, and weighted on an analytical balance with a precision of ±0.001 g (Toledo, Brazil, model AR1530). At the end of the 96 h exposure to each carbofuran concentration, 152.5. Swimming activity tilapia larvae from each treatment were weighted using the same procedure to evaluate weight gain during this period. Before the start of the experiment, larvae were individually putin aquariums (8 cm length × 6 cm width × 8.5 cm depth) with a col- 2.10. Reaction distance to prey and application of Blaxter’sumn of water 3 cm deep, and acclimated for 10 min inside the feeding modelexperimental arena setup enclosed by a black curtain to avoid dis-turbance by people moving in the lab. Each video recording was Reaction distance (RD) to an object of speciﬁc size is an alter-created in segments of 3 min. Recordings were processed by the native form of representation of the visual acuity of an animal, andsoftware Spyneurotracking (Bose, 2005), which identiﬁes the ani- is deﬁned as the maximum distance at which an object of certainmal coordinates x and y in each frame of the video recordings, and dimensions can be away from the observer, and still be resolved.calculates the average swimming speed of the animal in cm s−1 . A RD can be calculated from the visual acuity angle and the size (D)total of 15 ﬁsh were tested in each treatment. of either a prey or predator, from equation D2.6. Visual acuity 2 × tan( ˛ ) × 2 180 RD = 10 Visual acuity tests were based on the system and operational where: RD = reaction distance (cm); D = prey dimension (lengthprocedures previously described by Carvalho et al., 2002, and is in mm) ˛ = visual acuity angle (◦ ); = 3.14.related to the capacity of discriminating detail. During this test indi- RD and swimming speed can be used to calculate the volumevidual larvae were kept inside a glass chamber surrounded by black scanned by larval ﬁsh during their search for food. This informa-and white stripes of varying widths. Brieﬂy, both optomotor (swim- tion relates the visual and motor capacity of the ﬁsh with its abilityming) and optokinetic (eye movement) responses were monitored to search and detect food in a three dimensional environment. RDas ﬁsh were exposed to moving stripes of decreasing width until the is an important parameter used to estimate the reaction area, theoptokinetic response ceased. An acuity angle was then calculated transversal section of the visual ﬁeld of ﬁsh in the vertical and hori-based on the smallest stripe width to which the ﬁsh responded pos- zontal directions, according to the formula proposed by Blaxter anditively. Ten individual tilapia larvae from each exposure treatment Staines (1971):were tested for visual acuity after 4 and 5 days of exposure to thedifferent carbofuran concentrations. A total of 15 ﬁsh were tested 2 Reaction area (RA) (cm2 ) = × (RD)2in each treatment. 3
Authors personal copy P.C. Pessoa et al. / Aquatic Toxicology 105 (2011) 312–320 315 Fig. 1. Design of the arenas for predator–prey interaction monitoring, showing superior and lateral views of the system. where: = 3.14; RD = reaction distance in cm 1.0 From RA it is possible to estimate the search volume by multi- 0.9plying RA times swimming speed: 0.8 0.7 Search volume (SV) (L s−1 ) Mortality rate 0.6 RA (cm2 ) × swimming speed (cm s−1 ) 0.5 = 1000 0.4 This simple modeling approach relates the visual and motor 0.3capacity of the ﬁsh with its ability to search and detect food in a 0.2three dimensional environment. 0.1 0.0 1 10 100 10002.11. Statistical analyses Carbofuran concentration (µg.L-1) Mortality rates were corrected for control responses using Fig. 2. Mortality rates, probit adjusted lines and ﬁducial limits for Furadan (blackAbbott’s procedure and evaluated for ﬁtness to probit model with circles) and carbofuran (white circles) exposed tilapia larvae Oreochromis niloticus.Chi-square goodness of ﬁt tests (p < 0.05), using the software Toxs-tat version 3.5 (West, Inc. and Gulley, D.D., University of Wyoming, 3.2. Cholinesterase activityUSA). If the data ﬁt the probit model, 96 h LC50 and 95% ﬁduciallimits were also calculated. A dose dependent inhibition of cholinesterase activ- For behavioral and cholinesterase parameters, tests for nor- ity was detected (Fig. 3). Activity varied from 443.7 ±mality (Kolmogorov–Smirnov test) and homoscedasticity (Levene 162.4 nmol min−1 mg protein−1 in control larvae (0% inhibi-median test) of the data were applied. If the data were nor- tion) to 143.7 ± 27.3 nmol min−1 mg protein−1 in larvae frommal and homoscedastic, we used one-way ANOVA to compare 300 g/L carbofuran (68% inhibition). The lowest observed effectmeans from larvae exposed to the different carbofuran treatmentgroups, followed by a Dunnett’s test procedure to test for dif-ferences between the control and carbofuran treatments. A LOEL 700(lowest observable effect level) and NOEL (no observable effectlevel) were estimated following Dunnett’s tests. In case data failed 600normality or homoscedasticity a non-parametric Kruskal–WallisANOVA was used, followed by Dunn’s test to estimate the LOEL and (nmol.min-1 mg protein-1) 500NOEL for each parameter. All statistical analysis for cholinesterase, Cholinesterase activitybehavioral and growth data were based on the software Sigmaplotversion 11 (Jandel Scientiﬁc, Erkrath, Germany). 400 300 *3. Results * *3.1. Mortality rates 200 * Tilapia larval mortality rates exposed to both types of carbofuran 100solutions increased in a dose dependent manner (Fig. 2). Mortalityrates calculated from both exposure types ﬁt a probit model, and the 0LC50 -96 h for the commercial form of carbofuran was 220.7 g/L, 0 8.3 40.6 69.9 140 297 397with 95% ﬁducial limits of 191.6 and 254.3 g/L. The LC50 -96 h for Carbofuran concentration (µg.L-1)carbofuran at 98% purity form was 214.7 g/L, with 95% ﬁducial Fig. 3. Cholinesterase activity (nmol min−1 mg protein−1 ) of carbofuran exposedlimits of 200.1 and 229.2 g/L. tilapia larvae Oreochromis niloticus. (*): Different from control values (p < 0.05).
Authors personal copy316 P.C. Pessoa et al. / Aquatic Toxicology 105 (2011) 312–320 1,2 Number of tilapia larvae attacks on Daphnia 32 30 28 1,0 26 24 Swimming speed (cm.s-1) 0,8 22 20 18 0,6 16 * 14 0,4 12 * 10 8 0,2 6 4 2 0,0 0 0 8.3 40.6 69.9 140 297 397 0 8.3 40.6 69.9 140 297 397 Carbofuran concentration (µg.L-1) Carbofuran water concentration (µg.L-1) Fig. 6. Number of attacks on Daphnia performed by carbofuran exposed tilapia lar-Fig. 4. Swimming speed (cm s−1 ) of carbofuran exposed tilapia larvae Oreochromis vae Oreochromis niloticus. N = 15 ﬁsh for all treatments, (*): different from controlniloticus. N = 15 ﬁsh for all treatments, (*): different from control values (p < 0.05). values (p < 0.05).concentration (LOEC) was 69.9 g/L, and the no observed effectconcentration (NOEC) was 40.6 g/L. Cholinesterase inhibition dependent increase in median acuity angle was detected alongin larvae exposed to carbofuran concentrations of 69.9, 140, 297 exposure concentrations of 8.3, 40.6, 69.9, 140, 297 and 397 g/L,and 397 g/L were 59%, 66%, 68% e 46%, respectively, and were with median visual acuities of 1.4◦ , 1.9◦ , 1.9◦ , 2.4◦ , 3.6◦ and 3.4◦ ,signiﬁcantly different than control values (ANOVA F6, 34 = 6.07, respectively (Fig. 5). The LOEC for visual acuity was 40.6 g/L, andfollowed by Dunnet’s test, p < 0.05). the NOEC was 8.3 g/L (Kruskal–Wallis H6 = 56.8, p < 0.001, fol- lowed by Dunn’s test, p < 0.05).3.3. Swimming speed 3.5. Prey capture A tendency of dose dependent decrease in swimming speed ofexposed larvae was detected (Fig. 4), although only at the 397 g/L A dose dependent decrease in the number of attacks per-exposure group the swimming speed of 0.25 cm s−1 was signif- formed by carbofuran exposed tilapia towards 24 h old D. magnaicantly reduced when compared to the control larvae speed of was observed, although a statistically signiﬁcant difference was0.65 cm s−1 (ANOVA F6,55 = 3.61, followed by Dunnet’s test, p < 0.05). detected only at the 397 g/L exposure group. Median number of attacks varied from 17 in the control to 5 attacks at 397 g/L,3.4. Visual acuity the LOEC (Fig. 6) (Kruskal–Wallis, H6 = 14.3, p < 0.001, followed by Dunn’s test, p < 0.05). The median reductions in number of attacks Visual acuity was signiﬁcantly affected by carbofuran exposure. compared to controls were 17.6%, 23.5%, 11.7%, 29.4%, 35.2% andMedian visual acuity angle was 0.4◦ in control larvae, and a dose 70.5% for the exposure groups of 8.3, 40.6, 69.9, 140, 297 and 397 g/L. 7 * * 3.6. Predator evasion 6 Control tilapias were more successful in escaping the model predator P. managuensis, which needed a median of 5 attacks to 5 capture each individual tilapia tested. Carbofuran exposed tilapia Visual acuity (degrees) larvae were captured by the predator after a median number of 4 attacks of 4, 3, 1, 1 and 1 for exposure groups at 8.3, 40.6, 69.9, 140 and 297 g/L, respectively (Fig. 7), and the latter three groups were statistically different from control values (Kruskal–Wallis, 3 * H5 = 14.1, p = 0.015, followed by Dunn’s test, p < 0.05). The * * NOEC for predator evasion was 40.6 g/L and the LOEC was 2 69.9 g/L. 3.7. Weight gain 1 Weight gain in carbofuran exposed larvae was signiﬁcantly 0 reduced at all exposure concentrations (Fig. 8). Tilapia larvae had 0 8.3 40.6 69.9 140 297 397 an average initial weight at the beginning of the experiment of 10.7 ± 1.8 mg (mean ± standard deviation). Weight gain after the Carbofuran concentration (µg.L-1) exposure period was equal to 6.5 mg in the control group, and aFig. 5. Visual acuity (◦ ) of carbofuran exposed tilapia larvae Oreochromis niloticus. dose dependent decrease in weight gain was found. At the exposureN = 15 ﬁsh for all treatments, (*): different from control values (p < 0.05). groups of 8.3, 40.6, 69.9, 140, 297 and 397 g/L carbofuran, weight
Authors personal copy P.C. Pessoa et al. / Aquatic Toxicology 105 (2011) 312–320 317Table 1Reaction distances to prey, reaction areas and volume searched by carbofuran exposed Nile tilapia larvae hunting Daphnia magna neonates based on visual acuity andswimming speeds applied to Blaxter’s feeding model. Carbofuran concentration Visual Swimming Prey total Reaction Reaction Search volume ( g/L) acuity (◦ ) speed (cm s−1 ) length (mm) distance (cm) area (cm2 ) (L s−1 ) Control 0.4 0.65 2.0 28.6 1718.9 1.109 8.3 1.3 0.81 2.0 8.9 165.3 0.134 40.6 2.1 0.66 2.0 5.6 65.4 0.043 69.9 2.1 0.69 2.0 5.6 65.4 0.045 140 2.6 0.56 2.0 4.4 41.3 0.023 297 4.7 0.31 2.0 2.5 12.7 0.004 397 4.5 0.25 2.0 2.5 13.3 0.003 Number of predator attacks to capture tilapia larvae 3.8. Visual acuity and swimming speed applied to Blaxter’s feeding model 9 Control larval tilapia reaction distance, reaction area and search 8 volume towards prey represented by a 2 mm total length D. magna 7 neonate is equal to 28.6 cm, 1718 cm2 and 1.1 L s−1 , respectively (Table 1). A reduction in these parameters related to the capac- 6 ity of the free swimming larvae to encounter prey was calculated for carbofuran exposed larvae using measured data on visual acuity 5 and swimming speed applied to (Blaxter and Staines, 1971) feeding 4 model assumptions. Reaction distances were reduced from 31% to 9% of control RD from lowest (10 g/L) to highest (400 g/L) carbo- 3 furan exposure groups, respectively. Reaction areas were reduced 2 * from 10% to 0.8% of control RA from lowest (10 g/L) to high- * * est (400 g/L) carbofuran exposure groups, respectively. Search 1 volumes were reduced from 12% to 0.3% of control SV from low- est (10 g/L) to highest (400 g/L) carbofuran exposure groups, 0 respectively. 0 8.3 40.6 69.9 140 297 Carbofuran concentration (µg.L-1) 4. DiscussionFig. 7. Number of attacks of the model predator Parachromis managuensis performed This study reveals sublethal effects of the exposure of larvalto capture each individually tested carbofuran exposed tilapia larvae Oreochromis tilapia to waterborne carbofuran on cholinesterase activity and aniloticus. N = 15 ﬁsh for all treatments, (*): different from control values (p < 0.05). suite of behavioral endpoints that are directly related to impor- tant ecological mechanisms relevant for their survival, growth andgains of 3.5, 2.8, 2.7, 0.2, −1.3 and −1.1 mg were found, respec- recruitment to the adult population.tively, all means statistically different than the control (ANOVA I,F6,49 = 27.7, p < 0.001, followed by Dunnett’s test, p < 0.05). The LOEC 4.1. Mortality ratesfor weight gain was 8.3 g/L. Carbofuran 96 h LC50 in freshwater ﬁshes range from 88 g/L in bluegill sunﬁsh Lepomis macrochirus to 1990 g/L in fathead min- 0,012 now Pimephales promelas, a 20 fold difference in sensitivity (USEPA, Tilapia weight gain after exposure (g) 2004). A carbofuran 96 h LC50 of 480 g/L has been determined for 0,010 juvenile (0.6–3 g wet weight) O. niloticus (Stephenson et al., 1984). Our results for carbofuran and Furadan 96 h LC50 of 214 g/L and 0,008 220 g/L, respectively, indicate that O. niloticus larvae (0.01 g wet weight) are more sensitive than larger juveniles and also that the 0,006 * species is among the most sensitive ﬁshes to carbofuran exposure. This increased sensitivity of larval tilapia could be explained by * * the smaller amount of AChE present in smaller ﬁsh that can be 0,004 rapidly affected by pesticides when compared to larger ﬁsh (Dutta 0,002 and Arends, 2003). * Additionally, the similarity of the 96 h LC50 for carbofuran in its pure form compared to its agricultural formulation Furadan 0,000 * * indicate that the toxic potency of the active ingredient in the agri- cultural formulation is not affected by the inert substances added -0,002 (Fig. 2). 0 8.3 40.6 69.9 140 297 397 4.2. Cholinesterase activity Carbofuran concentration (µg.L-1) The measurement of acetylcholinesterase inhibition in feralFig. 8. Weight gain (g) at the end of 96 h of carbofuran exposed tilapia larvae Ore-ochromis niloticus. N = 15 ﬁsh for all treatments, (*): different from control values ﬁsh is widely established as a biomarker to diagnose exposure to(p < 0.05). organophosphate and carbamate pesticides (Sturm et al., 1999).
Authors personal copy318 P.C. Pessoa et al. / Aquatic Toxicology 105 (2011) 312–320Nevertheless, the potential use of this biomarker as part of an tion decline (Weis et al., 1999, 2001). Fish during early life stagesearly warning system of impending ecologically relevant effects need to feed frequently to supply its energetic demands, and an efﬁ-is a central challenge in ﬁsh ecotoxicology, due to gaps in knowl- cient prey capture skill is essential for growth and survival (Zhouedge about whether this inhibition might affect the ﬁtness of the et al., 2001). Contaminants can affect the motivation to feed as wellexposed individuals (Scholz and Hopkins, 2006). This potential as the ability to capture prey. A decrease in feeding rates mea-effect in the overall ﬁtness can be assessed by the quantiﬁcation sured as a reduction in attacks or capture of live prey as well asof behavioral parameters relevant for the survival and growth of in reduced consumption of artiﬁcial food is a common and ecolog-exposed ﬁsh. Furthermore, they can provide the basis for ecological ically relevant effect of contaminants (Sandheinrich and Atchison,mechanisms involved in the propagation of these effects towards 1990; Weis et al., 2003). Examples related to organophosphatesthe population level, where they are a matter of societal concern. include decreased ability to capture Artemia after exposure ofRecent advances in this direction have been made, and an elegant atlantic salmon Salmo salar to fenitrothion (Morgan and Kiceniuk,modeling approach has been used to relate sublethal reductions 1990), and decreased ability to capture fathead minnows P. prome-in acetylcholinesterase activity to reductions in wild salmon pop- las by hybrid striped bass exposed to diazinon (Gaworecki et al.,ulations’ productivity and growth rates (Baldwin et al., 2009). A 2009). In this study prey capture skills of exposed tilapia were20% AChE inhibition is commonly used as a threshold to deter- affected by carbofuran, as control larvae attacked Daphnia moremine exposure to organophosphate and carbamates (Bretaud et al., frequently than exposed tilapia (Fig. 6), although the high vari-2000). We found a LOEC of 69.9 g/L carbofuran for ChE inhibi- ability of the data limited statistically signiﬁcant differences totion in tilapias exhibiting 59.4% inhibition relative to controls, and the 397 g/L treatment. The complexity of predator prey rela-a non statistically signiﬁcant ChE inhibition of 22.1% was detected tions might have inﬂuenced this variability. In spite of that, resultsat the lowest concentration tested of 8.3 g/L (Fig. 3). In addition, of both endpoints (swimming and attacks on prey) indicate awe found statistically signiﬁcant differences in weight gain at this clear tendency of decrease with increasing carbofuran concentra-lowest concentration of 8.3 g/L (Fig. 8), which opens up the pos- tion.sibility that this threshold discussed above by Bretaud et al. (2000) Vision is essential for several important behavioral activitiesmight also be correlated with a threshold for more ecologically like prey detection, orientation towards prey, search for sexualrelevant effects. Growth of early life stages is considered an impor- partners, and detection and escape from predators. An inhibitiontant endpoint for risk assessment of effects of contaminants on ﬁsh of Ache activity in diazinon exposed Indian carp Labeo rohita hasrecruitment. The lack of statistical signiﬁcance in ChE inhibition at been correlated with deﬁcits in the optomotor response (Dutta8.3 g/L might also be related to a relatively large variability in the et al., 1992), which is considered a measure of the visual abil-results (Fig. 3), as pools of 3 whole larvae had to be homogenized. ity of ﬁsh. However, Dutta et al. (1992) analyzed the optomotorIt was unfeasible to use tissues like brain where the variability of response using a method that does not quantify the actual visualthe results could have been lower. acuity of the ﬁsh being tested, but rather only whether they respond Nevertheless, it is essential to better understand how sublethal or not to a certain ﬁxed width of the black and white stripes withindose-dependent ChE inhibition relates to concomitant alterations their ﬁeld of view during testing. Using a different methodology,in behavioral parameters, if we want to improve our capacity to (Carvalho and Tillitt, 2004) exposed rainbow trout to 2,3,7,8-TCDDpropose models of contaminant effects from the suborganismal to and quantiﬁed the visual acuity angle of the tested ﬁsh also ana-the population level of biological organization. Deleterious effects lyzing optomotor and optokinetic responses. This difference inof ChE inhibiting pesticides on ﬁsh exposed during early life stages methodology has important ecological implications because thehave been detected in individual performance behavioral parame- visual acuity angle can be expressed in terms of a reaction dis-ters like spontaneous swimming speed (Brewer et al., 2001), swim- tance to prey, and further into reaction areas and search volumesming stamina (Van Dolah et al., 1997), vision dependent behaviors towards prey, important parameters used in ﬁsh foraging model-like the optomotor response (Dutta et al., 1992), or olfactory depen- ing (Blaxter, 1986; Breck and Gitter, 1983). Deleterious effects ondent behaviors like attraction to a food extract (Saglio et al., 1996). visual acuity of a ﬁsh involve a decrease in its reaction distanceAdditionally, ChE inhibition has also been correlated with maladap- to other subjects, either prey or predators. In the ﬁrst situation,tive behavioral effects on more complex situations involving the deﬁcient prey detection skills can lead to decreased energy intake,interaction of exposed ﬁsh either with their potential prey (Morgan and potential growth deﬁcits. The reaction distance of fenitroth-and Kiceniuk, 1990) or predators (Sandahl et al., 2005). Our results ion exposed atlantic salmon towards adult Artemia quantiﬁed inindicate a signiﬁcant dose-dependent ChE inhibition clearly cor- experiments with live prey was a sensitive biomarker indicatingrelated with several measures of behavioral performance that are signiﬁcant effects at 0.3% the LC50 (Morgan and Kiceniuk, 1990). Inusually evaluated separately in different studies. this study with larval tilapia, visual acuity was the second most sensitive parameter, with a LOEC of 40.6 g/L (Fig. 5). Further-4.3. Swimming more, combining the average swimming speeds and visual acuity angles in a modeling approach, we predicted that the reaction area A decrease in swimming speed after exposure to organophos- and search volume towards prey for the exposed larvae wouldphates and carbamates has been detected in several ﬁsh, as be reduced to less than 12% of the prey search volume in con-in carbamate exposed golﬁsh Carassius auratus, rainbow trout trol ﬁsh at the lowest carbofuran concentration tested, 8.3 g/LOncorhynchus mykiss and medaka Oryzias latipes (Brewer et al., (Table 1). According to this approach, control ﬁsh would be able2001; Heath et al., 1993; Zinkl et al., 1991), and also in organophos- to search 1.1 L s−1 of water for prey per second, and this search vol-phate exposed seabass Dicentrarchus labrax (Almeida et al., 2010). ume would be reduced to 0.13 L s−1 after exposure to the lowestOur results with larval tilapia support this tendency as we found a concentration mentioned above, and further reduced to 0.043 L s−1signiﬁcant tendency of hypo activity after a 96 h exposure period at 40.6 g/L. These concentrations are within a range that can be(Fig. 4), although a statistically signiﬁcant effect was observed only detected in various aquatic environments, as reported previously.at 397 g/L. Interestingly, this prediction of a modeled reduced prey detection capability based on swimming speed and visual acuity is conﬁrmed4.4. Vision, feeding and weight gain by a clear tendency of reduced attacks on prey (Fig. 6) and by a It is recognized that alterations in feeding behavior by aquatic signiﬁcantly decreased growth rate (Fig. 8), which was the mostcontaminants can be related to deﬁcits in growth and to popula- sensitive parameter, with a LOEC of 8.3 g/L.
Authors personal copy P.C. Pessoa et al. / Aquatic Toxicology 105 (2011) 312–320 3194.5. Predator evasion References Visual acuity is also important in the detection of predators, Almeida, J., Oliveira, C., Gravato, C., Guilhermino, L., 2010. Linking behavioural alterations with biomarkers responses in the European seabass Dicentrarchuswhich is the ﬁrst step in the sequence that can lead to successfully labrax L. exposed to the organophosphate pesticide fenitrothion. Ecotoxicology,escaping predator attacks, followed by proper timing, direction 1–13.and locomotor performance at the time of escape (Domenici and Baldwin, D.H., Spromberg, J.A., Collier, T.K., Scholz, N.L., 2009. A ﬁsh of many scales: extrapolating sublethal pesticide exposures to the productivity of wild salmonBlake, 1997; Webb, 1986). Escape at the time of attack can be chal- populations. Ecological Applications 19, 2004–2015.lenging and energetically expensive, as an early escape response Barros, A.C.B.M., de Moura, R.M., Pedrosa, E.M., 2005. Estudo de Interacão ¸can actually attract the predator, whereas a late response can Variedade-Nematicida em Cana-de-Acúcar, em Solo Naturalmente Infestado por ¸ Meloidogyne incognita, M. javanica e Pratylenchus zeae. Nematologia Brasileira 29,lead to death. Measured behavioral parameters directly related 39–46.to predator evasion skills of ﬁsh exposed to the organophosphate Blaxter, J.H.S., 1986. Development of sense organs and behaviour of teleost larvaemalathion include a visual startle response based on the reaction with special reference to feeding and predator avoidance. Transactions of the American Fisheries Society 115, 98–114.of ﬁsh to a stimulus followed by the quantiﬁcation of respon- Blaxter, J.H.S., Staines, M., 1971. Food searching potential in marine ﬁsh larvae.siveness, latency to respond and maximum response speed (Del In: Crisp, D.J. (Ed.), Fourth European Marine Biology Symposium. CambridgeCarmen Alvarez and Fuiman, 2006). However, the measurement of University Press, Cambridge, pp. 467–485.behavioral parameters in a realistic predator–prey interaction in Bose, R., 2005. A system for tracking and movement analysis of laboratory ani- mals based on image processing. Revista Brasileira de Engenharia Biomédicaa predation experiment between exposed ﬁsh and their potential 21, 61–72.predator is rare in the literature. Examples of increased vulnera- Breck, J.E., Gitter, M.J., 1983. Effect of ﬁsh size on the reactive distance of bluegill (Lep-bility to predator capture include carbaryl and pentachlorophenol omis macrochirus) sunﬁsh. Canadian Journal of Fisheries and Aquatic Sciences 40, 162–167.exposed rainbow trout Oncorhynchus mykiss (Little et al., 1990), Bretaud, S., Toutant, J.-P., Saglio, P., 2000. Effects of carbofuran, diuron, andand mummichog Fundulus heteroclitus larvae from a polluted creek nicosulfuron on acetylcholinesterase activity in goldﬁsh (Carassius auratus). Eco-(Zhou and Weis, 1998, 1999). toxicology and Environmental Safety 47, 117–124. Bretaud, S., Saglio, P., Saligaut, C., Auperin, B., 2002. Biochemical and behavioral In this study the skills of exposed tilapia to escape predator effects of carbofuran in goldﬁsh (Carassius auratus). Environmental and Toxicol-attacks was affected by carbofuran exposure, increasing preda- ogy Chemistry 21, 175–181.tion vulnerability as measured by the number of predator attacks Brewer, S.K., Little, E.E., DeLonay, A.J., Beauvais, S.L., Jones, S.B., Ellersieck, M.R., 2001. Behavioral dysfunctions correlate to altered physiology in rainbow troutrequire to capture the exposed prey (Fig. 7). The escape behavior (Oncorynchus mykiss) exposed to cholinesterase-inhibiting chemicals. Archivescan be related to vision and locomotor ability of larvae, among other of Environmental Contamination and Toxicology 40, 70–76.factors (Fuiman et al., 2006). It is reasonable to suppose that the Carvalho, P.S.M., Noltie, D.B., Tillitt, D.E., 2002. Ontogenetic improvement of visual function in the medaka Oryzias latipes based on an optomotor testing system fordetected visual acuity deﬁcit and the hipoactivity pattern exhibited larval and adult ﬁshes. Animal Behaviour 64, 1–10.by the exposed tilapia (Fig. 4) may have favored a dose dependent Carvalho, P.S.M., Tillitt, D.E., 2004. 2,3,7,8-TCDD effects on visual structure andincrease in vulnerability to the predator attacks. function in swim-up rainbow trout. Environmental Science and Technology 38, Although biochemical responses are deemed sensitive with 6300–6306. Del Carmen Alvarez, M., Fuiman, L.A., 2006. Ecological performance of red drumlower statistically signiﬁcant effect concentrations, behavioral (Sciaenops ocellatus) larvae exposed to environmental levels of the insecticideeffects of carbofuran could be detected at lower exposure concen- malathion. Environmental and Toxicology Chemistry 25, 1426–1432.trations than biochemical responses. In our study, visual acuity and Domenici, P., Blake, R., 1997. The kinematics and performance of ﬁsh fast-start swimming. The Journal of Experimental Biology 200, 1165–1178.growth rates were the most sensitive biomarkers, with lower LOECs Dutra, B.K., Fernandes, F.A., Lauffer, A.L., Oliveira, G.T., 2009. Carbofuran-inducedthan ChE inhibition, while predator evasion showed a LOEC equal to alterations in the energy metabolism and reproductive behaviors of HyalellaChE inhibition. Examples of higher sensitivity of behavioral param- castroi (Crustacea, Amphipoda). Comparative Biochemistry and Physiology Part C: Toxicology and Pharmacology 149, 640–646.eters include signiﬁcant effects on swimming of Carassius auratus Dutta, H.M., Arends, D.A., 2003. Effects of endosulfan on brain acetylcholinesteraseafter a 4 h exposure to 1 g/L carbofuran, while no effect on AChE activity in juvenile bluegill sunﬁsh. Environmental Research 91, 157–162.activity was detected after an exposure of 6 h to 5 g/L (Bretaud Dutta, H., Marcelino, J., Richmonds, C., 1992. Brain acetylcholinesterase activity and optomotor behavior in bluegills, Lepomis macrochirus exposed to different con-et al., 2000). centrations of diazinon. Archives of Physiology and Biochemistry 100, 331–334. Ellman, G.L., Courtney, K.D., Andres, V.J., Feather-Stone, R.M., 1961. A new and rapid colorimetric determination of acetylcholinesterase activity. Biochemical Phar-5. Conclusion macology 7, 88–95. Fuiman, L.A., Rose, K.A., Cowan Jr., J.H., Smith, E.P., 2006. Survival skills required for predator evasion by ﬁsh larvae and their relation to laboratory measures of These results show that exposure of sensitive early life stages performance. Animal Behaviour 71, 1389–1399.of tilapia O. niloticus to sublethal concentrations of carbofuran can Gaworecki, K.M., Roberts, A.P., Ellis, N., Sowers, A.D., Klaine, S.J., 2009. Biochemicalaffect vision, growth and predator avoidance skills at environmen- and behavioral effects of diazinon exposure in hybrid striped bass. Environmen-tally signiﬁcant concentrations, while effects on swimming and tal Toxicology and Chemistry 28, 105–112. Heath, A.G., Cech, J.J., Zinkl, J.G., Steele, M.D., 1993. Sublethal effects of three pes-prey capture were noted only at relatively high concentrations. ticides on Japanese medaka. Archives of Environmental Contamination andThese behavioral effects followed inhibition of cholinesterase, and Toxicology 25, 485–491.provided relevant information about the propagation of effects Jarrard, H.E., Delaney, K.R., Kennedy, C.J., 2004. Impacts of carbamate pesti- cides on olfactory neurophysiology and cholinesterase activity in coho salmonfrom the suborganismal to the organismal level of organization that (Oncorhynchus kisutch). Aquatic Toxicology 69, 133–148.are key to our understanding of ecological mechanisms of carbofu- Little, E.E., Finger, S.E., 1990. Swimming behavior as an indicator of sublethal toxicityran toxicity to ﬁsh early life stages. in ﬁsh. Environmental Toxicology and Chemistry 9, 13–19. Little, E.E., Archeski, R.D., Flerov, B.A., Kozlovskaya, V.I., 1990. Behavioral indicators of sublethal toxicity in rainbow trout. Archives of Environmental Contamination and Toxicology 19, 380–385.Role of the funding source Matthiessen, P., Sheahan, D., Harrison, R., Kirby, M., Rycroft, R., Turnbull, A., Volkner, C., Williams, R., 1995. Use of a gammarus pulex bioassay to measure the effects We would like to thank International Foundation for Science of transient carbofuran runoff from farmland. Ecotoxicology and Environmental Safety 30, 111–119.(Grant W/3632), and FACEPE-CNPq for the Masters fellowship pro- Morgan, M.J., Kiceniuk, J.W., 1990. Effect of fenitrothion on the foraging behavior ofvided. This work was also funded in part by INCT-TA (CNPq), and juvenile atlantic salmon. Environmental Toxicology and Chemistry 9, 489–495.Petrobras. No funding source had any participation in study design; Oliveira, M.M., Silva Filho, M.V., Cunha Bastos, V.L., Fernandes, F.C., Cunha Bastos, J.,in the collection, analysis and interpretation of data; in the writing 2007. Brain acetylcholinesterase as a marine pesticide biomarker using Brazilian ﬁshes. Marine Environmental Research 63, 303–312.of the report; and in the decision to submit the paper for publica- Otieno, P., Lalah, J., Virani, M., Jondiko, I., Schramm, K.-W., 2010. Carbofuran andtion. its toxic metabolites provide forensic evidence for furadan exposure in vul-
Authors personal copy320 P.C. Pessoa et al. / Aquatic Toxicology 105 (2011) 312–320 tures (Gyps africanus) in Kenya. Bulletin of Environmental Contamination and epa.gov/oppfead1/endanger/litstatus/effects/carbofuran/riskanalysis.pdf. on Toxicology 84, 536–544. 12/12/2010, 69 pp.Peterson, G.L., 1977. A simpliﬁcation of the protein assay method of Lowry et al. USEPA, 2006. Interim Reregistration Eligibility Decision - Carbofuran. U.S. Environ- which is more generally applicable. Analytical Biochemistry 83, 346–356. mental Protection Agency, Washington, DC, EPA-738-R-706-031.Plese, L.P.M., 2005. Kinetics of carbosulfan hydrolysis to carbofuran and the subse- USEPA, 2010. Method 632 The Determination of Carbamate and Urea quent degradation of this last compound in irrigated rice ﬁelds. Chemosphere Pesticides in Municipal and Industrial Wastewater. Accessed at 60, 149–156. http://www.epa.gov/waterscience/methods/method/ﬁles/632.pdf. onRodríguez-Fuentes, G., Gold-Bouchot, G., 2004. Characterization of cholinesterase 07/01/2010 19 pp. activity from different tissues of Nile tilapia (Oreochromis niloticus). Marine Van Dolah, R.F., Maier, P.P., Fulton, M.H., Scott, G.I., 1997. Comparison of azin- Environmental Research 58, 505–509. phosmethyl toxicity to juvenile red drum (Sciaenops ocellatus) and theSaglio, P., Trijasse, S., Azam, D., 1996. Behavioral effects of waterborne carbofuran in mummichog (Fundulus heteroclitus). Environmental Toxicology and Chemistry goldﬁsh. Archives of Environmental Contamination and Toxicology 31, 232–238. 16, 1488–1493.Sandahl, J.F., Baldwin, D.H., Jenkins, J.J., Scholz, N.L., 2005. Comparative thresholds Webb, P.W., 1986. Functional locomotor morphology of early life history stages of for acetylcholinesterase inhibition and behavioral impairment in coho salmon ﬁshes. Transactions of the American Fisheries Society 115, 115–127. exposed to chlorpyrifos. Environmental Toxicology and Chemistry 24, 136–145. Weis, J.S., Smith, G.M., Zhou, T., 1999. Altered predator/prey behavior in pollutedSandheinrich, M.B., Atchison, G.J., 1990. Sublethal toxicant effects on ﬁsh foraging environments: implications for ﬁsh conservation. Environmental Biology of behavior: empirical vs. mechanistic approaches. Environmental Toxicology and Fishes 55, 43–51. Chemistry 9, 107–119. Weis, J.S., Samson, J., Zhou, T., Skurnick, J., Weis, P., 2001. Prey capture ability ofScholz, N.L., Hopkins, W.A., 2006. Ecotoxicology of anticholinesterase pesticides: mummichogs (Fundulus heteroclitus) as a behavioral biomarker for contami- data gaps and research challenges. Environmental Toxicology and Chemistry nants in estuarine systems. Canadian Journal of Fisheries and Aquatic Sciences 25, 1185–1186. 58, 1442–1452.Scholz, N.L., Truelove, N.K., French, B.L., Berejikian, B.A., Quinn, T.P., Casillas, E., Col- Weis, J.S., Samson, J., Zhou, T., Skurnick, J., Weis, P., 2003. Evaluating prey cap- lier, T.K., 2000. Diazinon disrupts antipredator and homing behaviors in chinook ture by larval mummichogs (Fundulus heteroclitus) as a potential biomarker for salmon (Oncorhynchus tshawytscha). Canadian Journal of Fisheries and Aquatic contaminants. Marine Environmental Research 55, 27–38. Sciences 57, 1911–1918. Zhou, T., Weis, J.S., 1998. Swimming behavior and predator avoidance in three pop-Stephenson, R.R., Choi, S.Y., Olmos-Jerez, A., 1984. Determining the toxicity and ulations of Fundulus heteroclitus larvae after embryonic and/or larval exposure hazard to ﬁsh of a rice insecticide. Crop Protection 3, 151–165. to methylmercury. Aquatic Toxicology 43, 131–148.Sturm, A., da Silva de Assis, H.C., Hansen, P.D., 1999. Cholinesterases of marine Zhou, T., Weis, J.S., 1999. Predator avoidance in mummichog larvae from a polluted teleost ﬁsh: enzymological characterization and potential use in the monitoring habitat. Journal of Fish Biology 54, 44–57. of neurotoxic contamination. Marine Environmental Research 47, 389–398. Zhou, T., Scali, R., Weis, J.S., 2001. Effects of methylmercury on ontogeny of preyTierney, K.B., Sampson, J.L., Ross, P.S., Sekela, M.A., Kennedy, C.J., 2008. Salmon capture ability and growth in three populations of larval Fundulus heteroclitus. olfaction is impaired by an environmentally realistic pesticide mixture. Envi- Archives of Environmental Contamination and Toxicology 41, 47–54. ronmental Science and Technology 42, 4996–5001. Zinkl, J.G., Lockhart, W.L., Kenny, S.A., Ward, F.J., 1991. The effects of cholinesteraseUSEPA, 2004. Analysis of Risks Endangered and Threatened Salmon and Steel- inhibiting insecticides on ﬁsh. In: Mineau, P. (Ed.), Cholinesterase-Inhibiting head. U. S. Environment Protection Agency (EPA) Accessed at http://www. Insecticides. Elsevier, New York, pp. 233–254.