An endocrine disrupting chemical changes courtship and parental care in the sand goby

Aquatic Toxicology 97 (2010) 285–292

Contents lists available at ScienceDirect

Aquatic Toxicology journal homepage: www.elsevier.com/locate/aquatox

An endocrine disrupting chemical changes courtship and parental care in the sand goby Minna Saaristo a,∗ , John A. Craft b , Kari K. Lehtonen c , Kai Lindström d a

Department of Bio- and Environmental Sciences, University of Helsinki, P.O. Box 65, FI-00014 Helsinki, Finland Biological and Biomedical Sciences, Glasgow Caledonian University, Cowcaddens Road, Glasgow G4 0BA, Scotland, UK Finnish Environment Institute, Marine Research Centre, P.O. Box 140, FI-00251 Helsinki, Finland d Environmental and Marine Biology, Åbo Akademi University, FI-20520 Turku, Finland b c

a r t i c l e

i n f o

Article history: Received 24 September 2009 Received in revised form 26 November 2009 Accepted 12 December 2009 Keywords: Endocrine disrupting chemical EE2 Reproductive behaviour Courtship Parental care Sand goby

a b s t r a c t Endocrine disrupting chemicals (EDCs) are a diverse group of compounds that can mimic, block or modulate the synthesis of natural hormones. They are known to cause impairment of reproduction of aquatic organisms at very low concentrations. The aim of this study was to examine how exposure from 10 to 31 days to 17␣-ethinyl estradiol (EE2, 41 ng L−1 ) affects the courtship and parental care behaviour of male sand gobies (Pomatoschistus minutus). The sand goby exhibits a polygynous mating system, where males compete for females and provide paternal care. First, male courtship performance towards a stimulus female was recorded with video camera. Secondly, after the male had received eggs his parental care behaviour was video recorded. In addition to behavioural endpoints, we measured the expression of hepatic vitellogenin (Vtg) and zona radiata protein (Zrp) mRNA, as well as common somatic indices. Our study shows that exposure to EE2 affected male fanning behaviour during both courtship and parental care. Interestingly, small exposed males increased their courtship fanning to similar levels as larger control males. However, during parental care egg fanning was not related to male size, and all exposed males fanned more than control males. The EE2-exposure induced Vtg and Zrp mRNA expression in males and decreased hepatosomatic index (HSI), and increased gonadosomatic index (GSI). Females prefer males that fan more, which will favour the small EDC exposed males. This may lead to mating that favours males that are not strong enough to tend the eggs until they hatch, thus decreasing the reproductive success of individuals. © 2009 Elsevier B.V. All rights reserved.

1. Introduction Endocrine disrupting chemicals (EDCs) are a diverse group of compounds from persistent organic pollutants (DDT, PCB), dioxins, polychlorinated biphenyls, pesticides, phenols, and phthalates to natural and synthetic hormones (Caserta et al., 2008). EDCs may mimic, block or modulate hormone synthesis, with many acting as agonists of estrogen receptors or antagonizing androgen receptors (Scippo et al., 2004). Additionally, they can also act through nongenomic mechanisms altering steroid synthesis and metabolism (Waring and Harris, 2005; Tabb and Blumberg, 2006). Most known EDCs originate from paper, plastic and pharmaceutical industries, agriculture and households. They occur in the aquatic environment at very low concentrations (ng L−1 ), and often in mixtures, which can be more harmful than individual compounds (Aerni et al., 2004; Brian et al., 2007).

∗ Corresponding author. Tel.: +358 9 191 57826. E-mail address: Minna.Saaristo@helsinki.fi (M. Saaristo). 0166-445X/$ – see front matter © 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.aquatox.2009.12.015

The use of behaviour as a tool to measure effects of EDC exposure is fairly recent (Jones and Reynolds, 1997; Clotfelter et al., 2004; Zala and Penn, 2004). Even though it has been shown to be a sensitive marker of EDC exposure (Smith and Logan, 1997; Scott and Sloman, 2004) and provides a starting point for evaluating population level consequences, it is still quite rarely used in ecotoxicological studies. This may be due to behavioural experiments being time-consuming when the behaviour is assessed by a human observer, and difficulties with repetition. However, changes in behaviour bring us direct information on the consequences that exposure may have on individual fitness. Moreover, relaying on single molecular, biochemical or cellular biomarker responses alone can underestimate the impact of EDCs on aquatic organisms (Thorpe et al., 2009). Reproductive systems are especially sensitive to EDCs and reproduction is closely connected to individual fitness. The reproductive behaviour of fish exhibiting parental care, like the sand goby (Pomatoschistus minutus), is therefore a good subject for studying potential effects of EDCs. Courtship behaviour often consists of several individual behaviours, and active courtship is

286

M. Saaristo et al. / Aquatic Toxicology 97 (2010) 285–292

energetically demanding (Andersson, 1994; Lindström, 1998) and therefore potentially serving as an honest signal of male condition (Zahavi, 1975). Parental care behaviour protects and promotes the development of offspring (Keenleyside, 1979; Smith and Wootton, 1995). In many fish, it includes (1) egg fanning, where the parent moves fresh water over the eggs with his pectoral fins, (2) cleaning the nest from fungi and other harmful substances, (3) removing diseased eggs, and (4) protecting the nest against predators. Parental care has its costs since it can reduce the survival of a parent by increasing its exposure to predators (Svensson, 1988) or deplete its energy reserves to an extent that the parent is more susceptible to disease or starvation (Smith and Wootton, 1995). Therefore, parental care can reduce the chances for additional matings, although among fish with paternal care most male parents continue to court females while still caring for previous broods (Gross and Sargent, 1985). The sand goby is a small marine fish with a wide geographic distribution. It has a one-year life cycle and mating system with resource-defence and paternal care (Healey, 1971). Sand gobies live a major part of their life in shallow, sandy shores and feed on benthic invertebrates and zooplankton. They also breed in shallow water, and the male builds a nest under a shell or rock, attracts females with courtship and tends the eggs until they hatch. Females base their mate choice on male body size, colouration, courtship display, parental care, and nest size and quality. Males that court intensively (Forsgren, 1997) and show high level of parental care are particularly favoured (Lindström et al., 2006). A male usually receives eggs from several females (Jones et al., 2001) and females prefer males who already have eggs in their nest (Forsgren et al., 1996). Our model compound representing EDCs was the synthetic pharmaceutical 17␣-ethinyl estradiol (EE2). EE2 is used in oral contraceptives and has been detected in ecologically relevant concentrations (<1 to 15 ng L−1 ) in sewage effluents worldwide (Baronti et al., 2000; Johnson and Sumpter, 2001; Onda et al., 2002; Muller et al., 2008). EE2 was chosen because it is more persistent in the environment than natural steroids (Schweinfurth et al., 1996; Jürgens et al., 2002; Clouzot et al., 2008), has a high tendency for bioaccumulation (Lai et al., 2000, 2002), and can cause 35–50% of the surface water estrogenicity (Cargouet et al., 2004). A concentration as low as of 1 ng L−1 of EE2 can impair reproductive success of fish (Parrott and Blunt, 2005; Schäfers et al., 2007; Länge et al., 2009). Because sand gobies spend such a major part of their life in shallow water areas, it is likely that they will encounter EDCs during their lifetime. To confirm that EE2 exposure caused a physiological response in the experimental fish, two “classical” molecular biomarkers for estrogenic exposure were used in this study. Vitellogenin (Vtg) is the egg yolk precursor protein (Mommsen and Walsh, 1988) and zona radiata protein (Zrp) forms the inner core of the egg envelope (Oppen-Berntsen et al., 1992). Males have estrogen receptors, and therefore exposure to estrogenic EDCs can be detected as production of Vtg and Zrp by the liver. Our aim was to examine the effects of short-term EDC exposure on courtship and parental care behaviour of sand goby males. In addition, we compared the performance of traditional biomarkers to our behavioural measurements for assessing EDC exposure. 2. Material and methods 2.1. Exposure setup The study was carried out at the Tvärminne Zoological Station, southern Finland, during May–July 2005. The fish were caught with a hand trawl from their natural breeding area. Males were randomly placed into nine flow-through exposure tanks (80 cm × 80 cm × 40 cm), each receiving 45 males. Three tanks were

assigned to each treatment level: (1) EE2 exposure, with the measured concentration 41 ng L−1 (SD = 24.4 ng L−1 , n = 12), (2) solvent control (2-propanol, <0.002%, v/v), and (3) natural seawater control. A stock solution of EE2 was prepared every second day by dissolving EE2 (Sigma–Aldrich) in 2-propanol (Merck), and added to flow-through mixing chambers using peristaltic pumps (Watson Marlow). From the mixing chambers, water was channelled into the exposure tanks using silicon tubing and flow rate was set to 9.6 L h−1 using glass flow meters equipped with adjustable valves (Kyrömäki, Finland). The water temperature ranged between 11 and 19 ◦ C, salinity 5.2–6.6‰ and pH 7.7–8.3 following natural variation in the sea. During the exposure fish were fed ad libitum with live Mysis spp. and frozen chironomid larvae. Males were exposed from 10 to 31 days before they were either used in behavioural experiments and dissected for Vtg and Zrp mRNA analysis or only dissected. A set of samples (EE2: n = 6, solvent: n = 6, seawater control: n = 6) was dissected every 5th day from the exposure tanks to follow the effect of EE2 exposure duration on Vtg and Zrp mRNA expression. The first behavioural experiments were performed after 10 days of exposure, and thereafter a new set of behavioural experiments was started every 4th day, always using new fish. This gave us a total sample size of 223 males that had gone through the behavioural tests and another sample of 81 males that only had gone through the exposure treatment. 2.2. Behavioural experiments We used 30 individual tanks (35 cm × 40 cm × 16 cm) for the behavioural experiments, 10 tanks for each treatment. These tanks had no flow-through of seawater. Each tank had a 3-cm layer of fine sand on the bottom and a large halved clay flowerpot (10 cm in diameter) that served as a nest site. Inside the flowerpot, a transparent film lined the ceiling of the pot. The film served as a substrate for the eggs laid by the female. It could be then removed and returned to the nest, making it possible to photograph and count the number of eggs. The water temperature in each tank was checked daily. 2.2.1. Nest building A total of 30 males (10 from each treatment group) were randomly picked from the exposure tanks, and measured for total length (to the nearest mm) and body mass (to the nearest g). One male was placed in each behavioural tank, and males were left to build a nest overnight. During nest building, the tanks were exposed according to the treatment from which the male originated. The mean measured concentration in the EE2 treatment was 50 ng EE2 L−1 , (SD = 18.4 ng L−1 , n = 20). In the other two treatments (solvent and seawater controls), EE2 was below the detection limit (<1 ng L−1 ). Males build their nests by excavating the inside and piling sand on top of the nest substrate. As a measure of nest building we used the amount of sand piled on the nest. This was scored as follows: (1) no sand had been piled on the nest; (2) some sand was piled to the top and front of the nest, but the clay pot was still visible; (3) the male had totally hidden the clay pot by piling plenty of sand to the top of the nest. In most cases, the entrance to the nest was as small as the male’s head. 2.2.2. Courtship behaviour Male courtship involves a repertoire of behaviours, which he uses to attract the female’s attention and to lead her into his nest to spawn. An important part of this repertoire is courtship fanning, a behaviour where the male moves his pectoral fins back and forth to create a flow of water. This behaviour is usually used to fan fresh water over the developing eggs, but is frequently used among fish even when there are no eggs in the nest. This behaviour was first described for the three-spined stickleback by Sevenster (1961). It is important to note the difference between fanning during courtship

M. Saaristo et al. / Aquatic Toxicology 97 (2010) 285–292

and parental care. From here onwards, we will call fanning during courtship ‘courtship fanning’. When a male had successfully built a nest (score 2), a gravid female was introduced to the tank. In order not to affect the female by exposing her to EE2, we first replaced the water in the tank with clean seawater (this was done irrespective of treatment). The female was first enclosed inside a glass container, containing clean seawater before she was introduced into the male tank. The fish were allowed to settle for 30 min before starting video recording. Males willingly court enclosed females and females respond to this courtship in a natural way. Male courtship behaviour was recorded with a digital video camera for 20 min and after filming the female was released from the container and left overnight to spawn with the male. The nest was checked with at regular intervals using a flashlight, and if eggs were found, the film was carefully removed and the eggs were photographed using a digital camera. Thus, we could count the total number of eggs spawned. The film with the eggs was then carefully returned to the nest and female was released back to the sea. 2.2.3. Parental care behaviour After the eggs had been photographed and returned to the nest, the male was allowed 12 h to settle and rebuild his nest, before the parental care observations were begun. We used a video camera that faced the nest opening to record the male’s parental care behaviour. One of the recorded behaviours was fanning. Male uses fanning during parental care to fan fresh water over the eggs and possibly to remove debris and waste products. From here on we will call fanning during the parental phase ‘egg fanning’ to distinguish it from ‘courtship fanning’. Parental care behaviour was recorded using a digital video camera for 20 min, and after recordings the male was taken to laboratory for dissection. The videotapes from the courtship and parental care experiments were analysed by the same person (M.S.) in a blind control manner using a simple event recorder computer program written specifically for this purpose. The program calculated duration (s) and frequency of the recorded behaviours. 2.3. Post-experiment morphometric measurements The nuptial colouration of each experimental male was scored according to the following scale: (1) very weak colouration, (2) anal and ventral fins light blue, and (3) anal and ventral fins dark blue with a bright blue spot on the first dorsal fin. The individuals were then anaesthetized using benzocaine (0.6 g in 200 ml seawater). Their body mass (to the closest g) and total length (to the closest mm) were measured. The fish was killed by cutting the spinal cord, and liver, gonads and accessory glands were excised and weighed. The liver was snap-frozen in liquid nitrogen and stored in an ultra freezer at −80 ◦ C for Vtg and Zrp mRNA expression measurements (see below). Gonadosomatic index (GSI), hepatosomatic index (HSI) and sperm-duct gland somatic index (SDGSI) were calculated as specific tissue mass/total body mass × 100. 2.4. Vtg and Zrp mRNA expression Vtg and Zrp expression were determined from total RNA on slot blots hybridised with [32 P]-labelled cDNA fragments and subsequent quantization by phosphoimager using the method described previously (Kirby et al., 2003; Saaristo et al., 2009a). Every 5th day, 18 fish (EE2: n = 6, solvent: n = 6, seawater: n = 6) were sampled from the exposure tanks for the Vtg and Zrp analysis, starting after 10 days of exposure (n = 137). Males that went through the behavioural experiments were also dissected, and analysed for Vtg and Zrp mRNA expression, and thus the total sample size for Vtg and Zrp mRNA expression was 218.

287

2.5. EE2 measurements The concentration of EE2 in the exposure aquaria was measured by liquid chromatograph–mass spectrometer (LC–MS; HS 1100-Waters Quattro II) using monitoring techniques adapted from tandem-spectrophotometric reactions (MRM). A 1-L sample was taken from each EE2 exposure aquarium and from one solvent and from one control aquarium every week. Samples were acidified to pH 2 with formic acid immediately after sampling and stored in at −4 ◦ C. Three control samples were prepared from the stock solution (20, 40 and 75 ng L−1 EE2). The samples were vacuum-filtered using glass fibre filters (GF/C, 1.2 ␮m, Whatman) and cleaned using the solid-phase extraction (SPE) method. The 1-L samples were loaded on Oasis HLB cartridges (5 ml) preconditioned with 10 ml methanol followed by 10 ml Milli-Q water (MQ). After loading, the cartridges were washed with 10 ml MQ. Samples were dried using an electronic vacuum pump for 1 h and then eluted with 8 ml ethinyl acetate, which was then evaporated with a slow nitrogen flow. To the pure sample 250 ␮l methanol and 50 ␮l of MQ were added to prepare the sample for LC–MS analysis. The LC–MS was operated in the negative electrospray ionisation mode using multiple reactions monitoring (MRM). EE2 was fractionated from the matrix with the liquid chromatograph, ionised by atmospheric pressure chemical ionisation (APCI) and analysed following two reactions of the molecule-ion of the proton-EE2 ([M + H]+ ; m/z 279): m/z 279 > 159 (0.2 s) and m/z 279 > 133 (0.2 s). External standards were used for the analysis and three-point calibration curves were made from the three control samples. Limit of detection was 1 ng L−1 and limit of quantification 5 ng L−1 . 2.6. Statistical analyses Parametric statistical methods were applied whenever the variables fulfilled or could be transformed to fulfil the requirements of these analyses. Initially we included both treatment and exposure time in the models. However, if exposure time or the interaction between treatment and exposure time had no significant effect on the response variable, exposure time was omitted from the final model. To test the effect of treatment and exposure time on Vtg and Zrp mRNA expression we used the Scheirer-Ray-Hare two factor non-parametric ANOVA (Sokal and Rohlf, 1995). All statistical analyses were performed using SPSS 12.0 software. 2.7. Ethical notes During exposure, fish were kept in flow-through aquaria which had a layer of sand on the bottom to maintain aquarium conditions as close to natural conditions as possible. All fish were fed twice a day during the exposure period. Non-exposed females were immediately returned to their natural habitat after the trials. We minimized the stress to animals during the experiment by gentle handling and keeping the aquaria in peaceful surroundings. Males were anesthetized with benzocaine before dissection. The study was approved and permit granted by the Finnish National Board for Laboratory Animals. 3. Results 3.1. Courtship behaviour Courtship fanning, i.e. fanning that male performs before having received any eggs, differed among treatments. The time spent in courtship fanning was a significant positive function of male size in the water control (linear regression on arcsine transformed fanning time b = −0.012, F1,26 = 6.256, p = 0.019), nearly so in the solvent (b = 0.010, F1,19 = 3.406, p = 0.081), but there was no relation-

288

M. Saaristo et al. / Aquatic Toxicology 97 (2010) 285–292

Fig. 1. The proportion of time sand goby male displayed courtship fanning in comparison to his size in different treatments (n = 81). (A) 17␣-Ethinyl estradiol (41 ng L−1 EE2), (B) solvent (<0.002%, v/v 2-propoanol), (C) seawater control. In the x-axis, male length is presented in mm.

ship for the EE2-exposed males (b = −0.004, F1,30 = 0.831, p = 0.369) (Fig. 1). These slopes differ significantly from each other suggesting that the relationship between size and courtship fanning is different under the different treatments (ANCOVA with male length as a covariate; treatment and covariate interaction, F2,75 = 3.511, p = 0.035, n = 81). The largest male in the solvent treatment did not fan at all (Fig. 1), suggesting that it is an outlier. If this male is removed from the analysis the relationship between male fanning and male length becomes similar to that in the water control (b = 0.017, F1,18 = 9.356, p = 0.007). Hence, small EE2 exposed males increased their courtship fanning to a level comparable to that of large males in the control treatment.

an effect on the index levels on exposed males (F3,211 = 17.111, p < 0.0001, n = 218) (Fig. 4a). There was no significant interaction between treatment and exposure length (F6,211 = 1.266, p = 0.274, n = 218). EE2 exposure affected GSI. After 18 days, GSI of exposed males started to increase, while among control males (seawater and solvent control) the index levels decreased (Fig. 4b). There was a significant interaction effect between the treatment and exposure length (ANOVA: F6,211 = 3.912, p = 0.001, n = 218). The other somatic and morphometric indices (SDGSI, UGP and nuptial colouration) showed no clear concentration- or time-dependent effects during the experiment.

3.2. Parental care behaviour

3.5. EE2 measurements

Egg fanning was the only behaviour during the parental care phase that was affected by EE2 exposure (ANOVA on arcsine transformed fanning time, treatment effect: F2,50 = 3.833, p = 0.028, n = 53) (Fig. 2). Exposed males spent more time egg fanning than males in the other treatments (Fisher’s LSD, Fig. 2). Unlike during the courtship phase, male length had no effect on egg fanning time, which is why we left it out of the above analysis.

In the exposure aquaria, the measured average concentration of 41 ng EE2 L−1 (SD = 24.4 ng L−1 , n = 12) was 55% of the intended

3.3. Vtg and Zrp mRNA expression EE2 exposure induced Vtg and Zrp mRNA expression. The expression levels differed significantly between the treatments (Scheirer-Ray-Hare, Vtg: treatment effect, H = 95.89, df = 2, p < 0.001; n = 218; Zrp: H = 43.53, df = 2, p < 0.001, n = 218, and were highest among EE2-exposed males. Exposure length had a significant effect on the expression levels of both markers (Vtg: Scheirer-Ray-Hare, exposure length effect, H = 10.47, df = 3, p = 0.015, n = 218; Zrp: H = 14.68, df = 3, p = 0.518, n = 218) (Fig. 3a, b). 3.4. Somatic indices HSI was significantly smaller in EE2-treated males as compared to the solvent and seawater control males (ANOVA, treatment: F2,211 = 6.539, p = 0.002, n = 218). Moreover, exposure length had

Fig. 2. The proportion of time sand goby males spent egg fanning during parental care (n = 53), White bars represent seawater control males, grey bars solvent control (<0.002%, v/v 2-propanol) males, and black bars males exposed to 41 ng L−1 of 17␣-ethinyl estradiol (EE2). Error bars represent one S.E.M. Same letters on the graphs indicate that groups do not differ from each other (Fisher’s LSD; EE2 to solvent, diff = 0.093, p = 0.037; EE2 to water, diff = 0.116, p = 0.012; solvent to water, diff = 0.023, p = 0.597).

M. Saaristo et al. / Aquatic Toxicology 97 (2010) 285–292

289

Fig. 4. Effects of 17␣-ethinyl estradiol (EE2) exposure on (A) hepatosomatic index (HSI), and (B) gonadosomatic index (GSI). Males were exposed from 10 to 31 days to 41 ng L−1 of EE2. White bars represent seawater control males, grey bars solvent control (<0.002%, v/v 2-propoanol) males, and black bars males exposed to EE2. Error bars represent one S.E.M and total n = 218 (n = 6 in each treatment, and in each sampling time).

Fig. 3. Effects on male (A) vitellogenin (Vtg), and (B) zona radiata protein (Zrp) mRNA expression in the seawater control, solvent control (< 0.002%, v/v 2-propanol) and 17␣-ethinyl estradiol (41 ng L−1 ) treatment. Sand goby males were exposed from 10 to 31 days (total n = 218, n = 6 in each treatment, and in each sampling time). White bars represent seawater males, grey bars represent males exposed to solvent and black bars males exposed to EE2. Bars represent mean expression levels in arbitrary units + one S.E.M.

nominal concentration (75 ng L−1 ). In the behavioural tanks, the EE2 concentration was 50 ng L−1 (SD = 18.4 ng L−1 , n = 20). EE2 in the control and solvent tanks was below the detection limit (<1 ng L−1 ) throughout the experiment. 4. Discussion 4.1. Courtship behaviour EE2 exposure changed courtship fanning in a complex manner. Among control males, courtship fanning was an increasing function of male body size. In EE2-exposed males, however, this relationship disappeared and all males fanned at a level typical of large control males. Sand goby females, when given a choice, prefer to mate with males that fan more (Lindström et al., 2006). Males also use

fanning as part of their courtship repertoire, as shown by the fact that they increase their courtship fanning when perceived mating opportunities are good (Pampoulie et al., 2004). Another important mate choice cue is male body size. When given a choice between a large and small male, females prefer to mate with the bigger male (Forsgren, 1992). Previous studies have shown that mating with a good caretaker may be more important than mating with a large and dominant male (Forsgren, 1997; Wong, 2004). According to the present study, a female who used courtship fanning as a signal to find the best mate among control males would also end up mating with a large male. However, a female mating with a high-fanning EE2-exposed male could end up mating with either a small or a large male. In the present population competition for nest sites is intense and it could be equally important for a female to find a good caretaker as to find a male who can successfully defend the nest (Lindström and Pampoulie, 2005). Our finding here could also, at least partly, explain the changes in the sand goby mating system that we observed in our previous study (Saaristo et al., 2009a). In that study, we found that the size difference between mated and unmated males was smaller among exposed males than among control, non-exposed fish. The reason for this difference was that among exposed fish females more often mated with small males whereas in the control tanks almost only the biggest males usually received matings. If females primarily used courtship fanning as a mate choice cue it could explain the weaker relationship between male mating status and male size among the exposed males.

290

M. Saaristo et al. / Aquatic Toxicology 97 (2010) 285–292

Even if small males may benefit from increased courtship fanning, it could be harmful to them in two ways: (1) their courtship will last a shorter time because they may run out of energy, and (2) the quality and quantity of parental care will be impaired since the male has to compensate for the loss of energy reserves during courtship. Impaired parental care is likely to lead to reduced hatching rates and reduced reproductive success. Hence, EDCs might change the reproductive output of a population. Alternatively, a change in courtship fanning could be an adaptive response to increase current reproductive success, if EE2 exposure indicated a reduced life expectancy for smaller males. In threespine sticklebacks, females use the red nuptial colouration of males as a mate choice cue. Redness is an honest signal of male quality as only males in good condition are able to maintain high red intensity (Milinski and Bakker, 1990). However, some males in very poor condition increase the intensity of their nuptial colouration possibly as a terminal fitness investment (Candolin, 1999). If exposure to EE2 was associated with a lower life expectancy, then the increased fanning of small males could have similarly represented an increased investment in current reproduction, as sand goby females prefer males that fan more (Lindström et al., 2006). In the present study, EE2 exposure did not affect any other measured courtship behaviours except for courtship fanning. This is contrary to previous studies, which have reported decreased courtship after exposure to EDCs (Bayley et al., 1999; Bjerselius et al., 2001; Larsen et al., 2008; Colman et al., 2009; Saaristo et al., 2009b). This is most likely due to the difference in exposure concentrations as our two previous studies (Saaristo et al., 2009b, in preparation) used exposure concentrations less than 5 ng L−1 . 4.2. Parental care behaviour Exposure to EE2 increased the amount of time males spent for egg fanning during the parental care phase. Egg fanning was independent of male body size in all treatments. Fanning is important for egg development rate (Zoran and Ward, 1983) and hatching success (Sargent and Gebler, 1980), and therefore crucial for offspring survival. Egg fanning is also costly (Coleman and Fischer, 1991; Lindström and Hellström, 1993; Lissåker et al., 2003). Hence, it is likely that the increase in egg fanning rate exhibited by EE2- exposed males also increased their parental care costs. This could result in a shortened reproductive season and/or decreased survival. At the same time, one would expect egg survival to improve due to increased egg fanning (Hale et al., 2003; Green and McCormick, 2005; Karino and Arai, 2006), which will be a clear benefit to both males and females. Unfortunately, we did not measure egg survival and development rate in this study. Parental care has been shown to suppress androgen levels in some teleost species (Wingfield et al., 1990; Hirschenhauser and Oliveira, 2006), but not in others (Knapp et al., 1999; Ros et al., 2004; Rodgers et al., 2006). Knowledge so far suggests that 11-ketotestosterone (KT) facilitates parental care to some extent (Rodgers et al., 2006). If exogenous estrogen does down-regulate androgen production (Bell, 2001), a decreased androgen level could have caused increased parental care, such as egg fanning. Egg fanning behaviour was shown to be negatively correlated with free and total KT levels in the Azorean rock-pool blenny (Parablennius sanguinolentus parvicornis) (Oliveira et al., 2001). However, a study by Ros et al. (2004) using the same blenny species as found that 11-KT implants did not suppress the egg fanning by nest holders. In the three-spined sticklebacks (Gasterosteus aculeatus), egg fanning behaviour was found to be similar in castrated males, castrated males with 11-KT implants and sham-operated males (Pall et al., 2002a,b). Alternatively, high androgen levels may even facilitate effective parental care (Desjardins et al., 2008). Several recent studies on fish, avian and mammalian species found that androgen

levels even increase during parental care (e.g. Van et al., 2000; Ros et al., 2004; Rodgers et al., 2006; Magee et al., 2006; Desjardins et al., 2008; Hanson et al., 2009). This might be due to the need for males to defend the nest during parental care (Ros et al., 2004) or because the male is courting new females while looking after the brood (Knapp et al., 1999; Rodgers et al., 2006). The mating system, degree of paternal care and the phase of paternal care are factors that all modulate hormonal levels (Hirschenhauser and Oliveira, 2006). Therefore, more studies should be focused on the effects of exogenous exposure to estrogenic and androgenic hormones and hormone-like contaminants at the reproductive behaviour level. 4.3. Traditional biomarkers: Vtg mRNA, Zrp mRNA, and somatic indices Induction of Vtg and Zrp mRNA was significantly higher in EE2exposed males compared to controls (seawater and solvent). This finding is consistent with our previous work, where we measured increasing expression of Vtg and Zrp mRNA among males exposed to 20 ng EE2 L−1 for 1–4 weeks (Saaristo et al., 2009a). Vtg and Zrp transcripts in the present study followed a delay (Craft et al., 2004) before reaching a maximum and then declining, as observed previously (Robinson et al., 2003; Brown et al., 2004; Saaristo et al., 2009a). This study strengthens our expectations that sand goby males express Vtg and Zrp mRNA when the exposure concentration is above 5 ng L−1 (Saaristo et al., 2009a,b). Hence, reliance on responses of only one biomarker should be avoided, because a single tool can underestimate the impacts of EDC contamination. Furthermore, absence of expression or response of the studied marker does not necessarily mean that the EDC compound has no effect on the reproduction or physiology of the individual. The HSI was significantly smaller in the EE2-treated males compared to control males, and the index level was dependent on length of EE2 exposure. The liver mass of exposed males increased slightly during the study, but was still significantly less than in the seawater control and solvent males. Decreased liver mass after EE2 exposure was also observed in our previous studies (Saaristo et al., 2009a,b) and in mummichog (Peters et al., 2007) and zebrafish (Versonnen and Janssen, 2004) after exposure to xenoestrogens. This is however, contradictory to several previous studies where authors have measured increased liver size (Herman and Kincaid, 1988; Sheahan et al., 2002; Zha et al., 2007) or no effect (Versonnen and Janssen, 2004; Andersson et al., 2007) after exposure. Since the HSI index often correlates with the degree of pollution (Adams and McLean, 1985), our study implies that estrogenic EDC exposure does have a toxic effect on the sand goby. However, the reason for the lower HSI may also be detoxification, because glycogen reserves are located in the liver and an increased usage of energy reserves to support increased detoxification reactions could result in decreased liver size. EE2-exposure increased male GSI. This result is not consistent with our previous studies (Saaristo et al., 2009a,b) where we observed no change in GSI of exposed males. This is likely due to different exposure concentrations as in the present work males were exposed to 41 ng L−1 EE2; while in our previous work the exposure concentration was 4 and 5 ng L−1 , respectively (Saaristo et al., 2009a,b). The GSI was also observed to increase after exposure to EE2 in male mummichog (10ng L−1 , Peters et al., 2007) and rare minnow (1 ng L−1 , Zha et al., 2007). Possible reasons for the increased GSI index are inhibited maturation cycle, lack of sperm release, accelerated gonad growth (Gill et al., 2002) and feminized vas deferens (Gimeno et al., 1996). Feminized vas deferens could block the release of viable gametes and increase GSI, if gametes continued to develop, but could not be released (Gill et al., 2002). E2 is known to simulate spermatogonial stem cell renewal (Schulz et al., 2009), and therefore it is possible that spermatogenesis was

M. Saaristo et al. / Aquatic Toxicology 97 (2010) 285–292

maintained in the EE2-treated males while control males showed normal seasonal decline of GSI. 5. Conclusions Our study showed that male reproductive behaviour was affected by EE2 exposure. Courtship fanning increased with increasing male size in non-exposed males, while in EE2- exposed males courtship fanning was high and independent of body size. During the parental care phase, EE2-exposed males spent more time egg fanning than did control males. Egg fanning is crucial for offspring survival, and could explain why females prefer males that fan more during courtship. The increase in courtship fanning by small EE2-expsed may lead to mating that favours males that do not have the resources to successfully complete their brood cycles, thus decreasing the reproductive success of individuals. Finally, our study illustrates that reproductive behaviour is a sensitive marker of EDC exposure and should be considered as a complementary tool to traditional physiological and molecular markers in ecotoxicological studies. Acknowledgements Authors thank Tvärminne Zoological Station for excellent facilities and support, Dr. Lorna Hanley for her irreplaceable guidance and training during mRNA expression work in Glasgow Caledonian University, and Heikki Björk for performing the LC–MS analyses. This work was funded by Onni Talas Foundation. References Adams, S.M., McLean, R.B., 1985. Estimation of largemouth bass, Micropterussalmoides Lacepede, growth using the liver somatic index and physiological variables. J. Fish Biol. 26, 111–126. Aerni, H.R., Kobler, B., Rutishauser, B.V., Wettstein, F.E., Fischer, R., Giger, W., Hungerbuhler, A., Marazuela, M.D., Peter, A., Schonenberger, R., Vogeli, A.C., Suter, M.J.F., Eggen, R.I.L., 2004. Combined biological and chemical assessment of estrogenic activities in wastewater treatment plant effluents. Anal. Bioanal. Chem. 378, 688–696. Andersson, C., Katsiadaki, I., Lundstedt-Enkel, K., Orberg, J., 2007. Effects of 17 alphaethynylestradiol on EROD activity, spiggin and vitellogenin in three-spined stickleback (Gasterosteus aculeatus). Aquat. Toxicol. 83, 33–42. Andersson, M., 1994. Sexual Selection. Princeton, New Jersey. Baronti, C., Curini, R., D’Ascenzo, G., Di Corcia, A., Gentili, A., Samperi, R., 2000. Monitoring natural and synthetic estrogens at activated sludge sewage treatment plants and in a receiving river water. Environ. Sci. Technol. 34, 5059–5066. Bayley, M., Nielsen, J.R., Baatrup, E., 1999. Guppy sexual behavior as an effect biomarker of estrogen mimics. Ecotoxicol. Environ. Saf. 43, 68–73. Bell, A.M., 2001. Effects of an endocrine disrupter on courtship and aggressive behavior of male three-spined stickleback, Gasterosteus aculeatus. Anim. Behav. 62, 775–780. Bjerselius, R., Lundstedt-Enkel, K., Olsen, H., Mayer, I., Dimberg, K., 2001. Male goldfish reproductive behavior and physiology are severely affected by exogenous exposure to 17 beta-estradiol. Aquat. Toxicol. 53, 139–152. Brian, J.V., Harris, C.A., Scholze, M., Kortenkamp, A., Booy, P., Lamoree, M., Pojana, G., Jonkers, N., Marcomini, A., Sumpter, J.P., 2007. Evidence of estrogenic mixture effects on the reproductive performance of fish. Environ. Sci. Tehcnol. 41, 337–344. Brown, M., Robinson, C., Davies, I.M., Moffat, C.F., Redshaw, J., Craft, J.A., 2004. Temporal changes in gene expression in the liver of male plaice (Pleuronectes platessa) in response to exposure to ethynyl oestradiol analysed by macroarray and Real-Time PCR. Mutat. Res. 18, 35–49. Candolin, U., 1999. The relationship between signal quality and physical condition: is sexual signalling honest in the three spine stickleback? Anim. Behav. 58, 1261–1267. Cargouet, M., Perdiz, D., Mouatassim-Souali, A., Tamisier-Karolak, S., Levi, Y., 2004. Assessment of river contamination by estrogenic compounds in Paris area (France). Sci. Total Environ. 324, 55–66. Caserta, D., Maranghi, L., Mantovani, A., Marci, R., Maranghi, F., Moscarini, M., 2008. Impact of endocrine disruptor chemicals in gynaecology. Hum. Reprod. Update 14, 59–72. Coleman, R.M., Fischer, R.U., 1991. Brood size, male fanning effort and the energetics of a nonshareable parental investment in bluegill sunfish, Lepomis macrochirus (Teleostei: Centrarchidae). Ethology 87, 177–188. Colman, J.R., Baldwin, D., Johnson, L.L., Scholz, N.L., 2009. Effects of the synthetic estrogen, 17␣-ethinylestradiol, on aggression and courtship behavior in male zebrafish (Danio rerio). Aquat. Toxicol. 91, 346–354.

291

Clotfelter, E.D., Bell, A.M., Levering, K.R., 2004. The role of animal behaviour in the study of endocrine-disrupting chemicals. Anim. Behav. 68, 665–676. Clouzot, L., Marrot, B., Doumenq, P., Roche, N., 2008. 17␣-Ethinylestradiol: an endocrine disrupter of great concern. Analytical methods and removal processes applied to water purification. A review. Environ. Prog. 27, 383–396. Craft, J.A., Brown, M., Dempsey, K., Francey, J., Kirby, M.F., Scott, A.P., Katsiadaki, I., Robinson, C.D., Davies, I.M., Bradac, P., Moffat, C.F., 2004. Kinetics of vitellogenin protein and mRNA induction and depuration in fish following laboratory and environmental exposure to oestrogens. Marin. Environ. Res. 58, 419–423. Desjardins, J.K., Stiver, K.A., Fitzpatrick, J.L., Milligan, N., Van der Kraak, G.J., Balshine, S., 2008. Sex and status in a cooperative breeding fish: behavior and androgens. Behav. Ecol. Sociobiol. 62, 785–794. Forsgren, E., 1992. Predation risk affects mate choice in a Gobiid fish. Am. Nat. 140, 1041–1049. Forsgren, E., 1997. Female sand gobies prefer good fathers over dominant males. Proc. R. Soc. Lond. B 264, 1283–1286. Forsgren, E., Kvarnemo, C., Lindström, K., 1996. Mode of sexual selection determined by resource abundance in two sand goby populations. Evolution 50, 646–654. Gill, M.E., Spiropoulos, J., Moss, C., 2002. Testicular structure and sperm production in flounders from a polluted estuary: a preliminary study. J. Exp. Mar. Biol. Ecol. 281, 41–51. Gimeno, S., Gerritsen, A., Bowner, T., Komen, H., 1996. Feminisation of male carp. Nature 384, 221–222. Green, B.S., McCormick, M.I., 2005. O2 replenishment to fish nests: males adjust brood care to ambient conditions and brood development. Behav. Ecol. 16, 389–397. Gross, M., Sargent, R.C., 1985. The evolution of male and female parental care in fishes. Am. Zool. 25, 807–822. Hale, R.E., St. Mary, C.M., Lindström, K., 2003. Parental responses to changes in costs and benefits along an environmental gradient. Env. Biol. Fish. 67, 107–116. Hanson, K.C., O’Connor, C.M., Van Der Kraak, G., Cooke, S.J., 2009. Paternal aggression towards a brood predator during parental care in wild smallmouth bass is not correlated with circulating testosterone and cortisol concentrations. Horm. Behav. 55, 495–499. Healey, M.C., 1971. The distribution and abundance of sand gobies Gobius minutus, in the Ythan Estuary. J. Zool. 163, 177–229. Herman, R.L., Kincaid, H.L., 1988. Pathological effects of orally administered estradiol to rainbow trout. Aquaculture 72, 165–172. Hirschenhauser, K., Oliveira, R.F., 2006. A review on social modulation of androgens in male vertebrates: meta-analyses of the challenge hypothesis. Anim. Behav. 71, 265–277. Johnson, A.C., Sumpter, J.P., 2001. Removal of endocrine-disrupting chemicals in activated sludge treatment works. Environ. Sci. Technol. 35, 4697–4703. Jones, J.C., Reynolds, J.D., 1997. Effects of pollution on reproductive behaviour of fishes. Rev. Fish. Biol. Fisher 7, 463–491. Jones, A.G., Walker, D., Kvarnemo, C., Lindström, K., Avise, J.C., 2001. How cuckoldry can decrease the opportunity for sexual selection: Data and theory from a genetic parentage analysis of the sand goby, Pomatoschistus minutus. Proc. Natl. Acad. Sci. U.S.A. 98, 9151–9156. Jürgens, M.D., Holthaus, K.I., Johnson, A.C., Smith, J.L., Hetheridge, M., Williams, R.J., 2002. The potential for estradiol and ethinylestradiol degradation in English rivers. Environ. Toxicol. Chem. 21, 480–488. Karino, K., Arai, A., 2006. Effect of clutch size on male egg-fanning behavior and hatching success in the goby, Eviota prasina (Klunzinger). J. Exp. Mar. Biol. Ecol. 334, 43–50. Keenleyside, M.H.A., 1979. Diversity and Adaptation in Fish Behaviour. SpringerVerlag, Berlin, p. 208. Kirby, M.F., Bignell, J., Brown, E., Craft, J.A., Davies, I., Dyer, R.A., Feist, S.W., Jones, G., Matthiessen, P., Megginson, C., Robertson, F.E., Robinson, C., 2003. The presence of morphologically intermediate papilla syndrome in United Kingdom populations of sand goby (Pomatoschistus spp.): endocrine disruption? Environ. Toxicol. Chem. 22, 239–251. Knapp, R., Wingfield, J.C., Bass, A.H., 1999. Steroid hormones and paternal care in the plainfin midshipman fish (Porichthys notatus). Horm. Behav. 35, 81–89. Lai, K.M., Scrimshaw, M.D., Lester, J.N., 2002. Prediction of the bioaccumulation factors and body burden of natural and synthetic estrogens inaquatic organisms in the river systems. Sci. Total Environ. 289, 159–168. Lai, K.M., Johnson, K.L., Scrimshaw, M.D., Lester, J.N., 2000. Binding of waterborne steroid estrogens to solid phases in river and estuarine systems. Environ. Sci. Technol. 34, 3890–3894. Larsen, M.G., Hansen, K.B., Henriksen, P.G., Baatrup, E., 2008. Male zebrafish (Danio rerio) courtship behaviour resists the feminising effects of 17alphaethinyloestradiol – morphological sexual characteristics do not. Aquat. Toxicol. 87, 234–244. Lindström, K., 1998. Energetic constraints on mating performance in the sand goby. Behav. Ecol. 9, 297–300. Lindström, K., Hellström, M., 1993. Male size and parental care in the sand goby, Pomatoschistus minutus. Ethol. Ecol. Evol. 5, 97–106. Lindström, K., Pampoulie, C., 2005. Effects of resource holding potential and resource value on tenure at nest sites in sand gobies. Behav. Ecol. 16, 70–74. Lindström, K., St.Mary, C.M., Pampoulie, C., 2006. Sexual selection for male parental care in the sand goby, Pomatoschistus minutus. Behav. Ecol. Sociobiol. 60, 46–51. Lissåker, M., Kvarnemo, C., Svensson, O., 2003. Effects of a low oxygen environment on parental effort and filial cannibalism in the male sand goby, Pomatoschistus minutus. Behav. Ecol. 14, 374–381.

292

M. Saaristo et al. / Aquatic Toxicology 97 (2010) 285–292

Länge, A., Paull, G.C., Coe, T.S., Katsu, Y., Urushitani, H., Iguchi, T., Tyler, C.R., 2009. Sexual reprogramming and estrogenic sensitization in wild fish exposed to ethinylestradiol. Environ. Sci. Technol. 43, 1219–1225. Magee, S.E., Neff, B.D., Knapp, R., 2006. Plasma levels of androgens and cortisol in relation to breeding behavior in parental male bluegill sunfish, Lepomis macrochirus. Horm. Behav. 49, 598–609. Milinski, M., Bakker, T.C.M., 1990. Female sticklebacks use male coloration in mate choice and hence avoid parasitized males. Nature 344, 330–333. Mommsen, T.P., Walsh, P.J., 1988. Fish Physiology, vol. XIA. Academic Press, New York, USA, pp. 347–406. Muller, M., Rabenoelina, F., Balaguer, P., Patureau, D., Lemenach, K., Budzinski, H., Barcelo, D., De Alda, M.L., Kuster, M., Delgenes, J.P., Hernandez-Raquet, G., 2008. Chemical and biological analysis of endocrine-disrupting hormones and estrogenic activity in an advanced sewage treatment plant. Environ. Toxicol. Chem. 27, 1649–1658. Oliveira, R.F., Canario, A.V.M., Grober, M.S., Santos, R.S., 2001. Endocrine correlates of male polymorphism and alternative reproductive tactics in the Azorean rockpool blenny, Parablennius sanguinolentus parvicornis. Gen. Comp. Endocrinol. 121, 278–288. Onda, S., Hiraishi, A., Matsuo, Y., Takii, S., 2002. Polyphasic approaches to the identification of predominant polyphosphate-accumulating organisms in a laboratory-scale anaerobic/aerobic activated sludge system. J. Gen. Appl. Microbiol. 48, 43–54. Oppen-Berntsen, D.O., Gramjensen, E., Walther, B.T., 1992. Zona radiata proteins are synthesized by rainbow-trout (Oncorhynchus mykiss) hepatocytes in response to estradiol-17-beta. J. Endocrinol. 135, 293–302. Pall, M.K., Mayer, I., Borg, B., 2002a. Androgen and behavior in the male three-spined stickleback, Gasterosteus aculeatus. I. Changes in 11-ketotestosterone levels during the nesting cycle. Horm. Behav. 41, 377–383. Pall, M.K., Mayer, I., Borg, B., 2002b. Androgen and behavior in the male three-spined stickleback, Gasterosteus aculeatus. II. Castration and 11-Ketoandrostenedione effects on courtship and parental care during the nesting cycle. Horm. Behav. 42, 337–344. Pampoulie, C., Lindström, K., St.Mary, C.M., 2004. Have your cake and eat it too: male sand gobies show more parental care in the presence of female partners. Behav. Ecol. 15, 199–204. Parrott, J.L., Blunt, B.R., 2005. Life-cycle exposure of fathead minnows (Pimephales promelas) to an ethinylestradiol concentration below 1 ng/L reduces egg fertilization success and demasculinizes males. Environ. Toxicol. 20, 131– 141. Peters, R.E.M., Courtenay, S.C., Cagampan, S., Hewitt, M.L., MacLatchy, D.L., 2007. Effects on reproductive potential and endocrine status in the mummichog (Fundulus heteroclitus) after exposure to 17alpha-ethynylestradiol in a short-term reproductive bioassay. Aquat. Toxicol. 85, 154–166. Robinson, C.D., Brown, E., Craft, J.A., Davies, I.M., Moffat, C.F., Pirie, D., Robertson, F., Stagg, R.M., Struthers, S., 2003. Effects of sewage effluent and ethynyl oestradiol upon molecular markers of oestrogenic exposure, maturation and reproductive success in the sand goby (Pomatoschistus minutus, Pallas). Aquat. Toxicol. 62, 119–134. Rodgers, E.W., Earley, R.L., Grober, M.S., 2006. Elevated 11-ketotestosterone during paternal behavior in the bluebanded goby (Lythrypnus dalli). Horm. Behav. 49, 610–614. Ros, A.F.H., Bruintjes, R., Santos, R.S., Canario, A.V.M., Oliveira, R.F., 2004. The role of androgens in the trade-off between territorial and parental behavior in the Azorean rock-pool blenny, Parablennius parvicornis. Horm. Behav. 46, 491–497. Saaristo, M., Craft, J.A., Lehtonen, K.K., Björk, H., Lindström, K., 2009a. Disruption of sexual selection in sand gobies (Pomatoschistus minutus) by 17␣-ethinyl estradiol, an endocrine disruptor. Horm. Behav. 55, 530–537. Saaristo, M., Craft, J.A., Lehtonen, K.K., Lindström, K., 2009b. Sand goby (Pomatoschistus minutus) males exposed to an endocrine disrupting chemical fail in nest and mate competition. Horm. Behav. 56, 315–321.

Sargent, R.C., Gebler, J.B., 1980. Effects of nest site concealment of hatching success, reproductive success, and paternal behaviour of the threespine stickleback, Gasterosteus aculeatus. Behav. Ecol. Sociobiol. 7, 137–142. Schulz., R.W., Renato de Franc¸a, L., Lareyre,J., LeGac, F., Chiarini-Garcia, H., Nobrega, R.H., Miura, T., 2009. Spermatogenesis of fish. Gen. Comp. Endocrinol. doi:10.1016/j.ygcen.2009.02.013. Schweinfurth, H., Länge, R., Günzal, P., 9–10 May 1996. Environmental fate and ecological effects of steroidal estrogens. In: Presented at IBC Conference, London, UK. Schäfers, C., Teigeler, M., Wenzel, A., Maach, G., Fenske, M., Segner, H., 2007. Concentration- and time-dependent effects of the synthetic estrogen, 17␣ethinylestradiol, on reproductive capabilities of the zebrafish, Danio rerio. J. Toxicol. Environ. Health, Part A 70, 768–779. Scippo, M., Argiris, C., Van de Weerdt, C., Muller, M., Willemsen, P., Martial, J., Maghuin-Rogister, G., 2004. Recombinant human estrogen, androgen and progesterone receptors for detection of potential endocrine disruptors. Anal. Bioanal. Chem. 378, 664–669. Scott, G.R., Sloman, K.A., 2004. The effects of environmental pollutants on complex fish behaviour: integrating behavioural and physiological indicators of toxicity. Aquat. Toxicol. 68, 369–392. Sheahan, D.A., Brigghty, D.C., Daniel, M., Kirby, S.J., Hurst, M.R., Kennedy, J., Morris, S., Routledge, E.J., Sumpter, J.P., Waldock, M.J., 2002. Estrogenic activity measured in a sewage treatment works treating industrial inputs containing high concentrations of alkylphenolic compounds—a case study. Environ. Toxicol. Chem. 21, 507–514. Smith, C., Wootton, R.J., 1995. The costs of parental care in teleosts fishes. Rev. Fish Biol. Fish. 5, 7–22. Smith, E.H., Logan, D.T., 1997. Linking environmental toxicology, ethology, and conservation. In: Clemmonds, J.R., Buchholz, R. (Eds.), Behavioral Approaches to Conservation in the Wild. Cambridge University Press, Cambridge, pp. 277–302. Sokal, R.R., Rohlf, F.J., 1995. Biometry. W.H. Freeman and Company, New York, p. 885. Sevenster, P., 1961. A causal analysis of a displacement activity (fanning in Gasterosteus aculeatus). Behav. Suppl. 9, 1–170. Svensson, I., 1988. Reproductive costs in two sex-role reversed pipefish species (Syngnathidae). J. Anim. Ecol. 57, 929–942. Tabb, M.M., Blumberg, B., 2006. New models of action for endocrine-disrupting chemicals. Mol. Endocrinol. 20, 475–482. Thorpe, K.L., Maack, G., Benstead, R., Tyler, C.R., 2009. Estrogenic wastewater treatment works effluents reduce egg production in fish. Environ. Sci. Technol. 43, 2976–2982. Waring, R.H., Harris, R.M., 2005. Endocrine disruptors: a human risk? Mol. Cell. Endocrinol. 244, 2–9. Wingfield, J.C., Hegner, R.E., Dufty, A.M., Ball, G.F., 1990. The challenge hypothesis – theoretical implications for patterns of testosterone secretion, mating systems, and breeding strategies. Am. Nat. 136, 829–846. Wong, B.B.M., 2004. Superior fighters make mediocre fathers: male competition, female choice and egg hatching success in the Pacific blue eye fish. Anim. Behav. 67, 583–590. Van, D.E., Pinxten, R., Eens, M., 2000. Does testosterone affect the trade-off between investment in sexual/territorial behaviour and parental care in male great tits? Behavior 137, 1503–1515. Versonnen, B.J., Janssen, C.R., 2004. Xenoestrogenic effects of ethinylestradiol in zebrafish (Danio rerio). Environ. Toxicol. 19, 198–206. Zahavi, A., 1975. Mate selection: a selection for a handicap. J. Theor. Biol. 53, 205–214. Zala, S.M., Penn, D.J., 2004. Abnormal behaviours induced by chemical pollution: a review of the evidence and new challenges. Anim. Behav. 68, 649–664. Zha, J., Wang, Z., Wang, N., Ingersoll, C., 2007. Histological alternation and vitellogenin induction in adult rare minnow (Gobiocypris rarus) after exposure to ethynylestradiol and nonylphenol. Chemosphere 66, 488–495. Zoran, M.J., Ward, J.A., 1983. Parental egg care behavior and fanning activity for the orange chromide, Etroplus maculatus. Environ. Biol. Fish. 8, 301–310.