Contaminants reduce male contribution to ... - Wiley Online Library

2 downloads 0 Views 1002KB Size Report
Apr 24, 2015 - 2013, Lyons and. Lowe 2013) ... Bourgeon, S., E. K. H. Leat, R. W. Furness, K. Borga, ... Jobling, S., M. Nolan, C. R. Tyler, G. Brighty, and J. P..
Contaminants reduce male contribution to reproduction at the population scale JEFFREY J. HILLIS, JAMES E. GARVEY, 

AND

MICHAEL J. LYDY

Department of Zoology, Center for Fisheries, Aquaculture, and Aquatic Sciences, Center for Ecology, Southern Illinois University, Carbondale, Illinois 62901 USA Citation: Hillis, J. J., J. E. Garvey, and M. J. Lydy. 2015. Contaminants reduce male contribution to reproduction at the population scale. Ecosphere 6(4):66. http://dx.doi.org/10.1890/ES14-00391.1

Abstract. In many populations, males compete for mates and contribute to genetic diversity, but are assumed to have little impact on population growth rates and density, because one father is all that is necessary to fertilize a large number of eggs and sire a potential cohort. A factor affecting male reproductive output is exposure to contaminants that either disrupt endocrine function or mimic hormones. Polychlorinated biphenyls (PCBs) persist in aquatic ecosystems and may interfere with the reproductive contribution of male fish to populations. We quantified 28 PCB congeners in adult male bluegills across eight Illinois lakes. In spring 2011 and 2012, mature male bluegills from lakes with conspecifics with PCB whole-body residues ranging from undetectable to high (2,000 ng/g dry mass total PCBs) were placed in experimental ponds with mature females containing no detectable PCB body residues and allowed to reproduce in spring under natural conditions. The presence of PCBs in fish had no effect on the presence of intersex or other gonadal deformities in adult males nor was gonad size, lipid concentration, and survival affected. Males with high PCBs were heavier than those from lakes with low PCB concentrations. The relative density of juvenile bluegills produced in ponds by fall, a fitness endpoint in this species, was negatively related to contaminant concentration in fathers. Contaminant exposure and other factors that affect male reproduction may not only suppress the relative contribution of males to genetic diversity, but also the overall density of cohorts within populations. Key words: bluegills; contaminants; endocrine disruption; Lepomis macrochirus; male reproductive output; polychlorinated biphenyls (PCBs); reproductive success. Received 14 October 2014; revised 16 December 2014; accepted 23 December 2014; final version received 7 March 2015; published 24 April 2015. Corresponding Editor: D. P. C. Peters. Copyright: Ó 2015 Hillis et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. http://creativecommons.org/licenses/by/3.0/   E-mail: [email protected]

INTRODUCTION

duction and mortality (Wilson and Bossert 1971). In fact, many species switch to asexual reproduction by females to rapidly increase density, and only invest in the production of males when environmental variability makes it necessary to enhance genetic diversity to offset potential unanticipated variable effects on offspring survival (Brzyski et al. 2014, Innes and Ginn 2014). In this paper, we provide evidence that factors affecting male reproductive state and, thereby, individual fitness directly translates to the

A general tenant of ecology is that males matter for maintaining genetic diversity through sexual reproduction in a population, but it is the reproductive output of females that often determines rates of population growth (Monosson et al. 1994, Nagler and Daniel 1997). Females typically invest more energy in gametes than males, and most population models are sensitive to female demographics, such as age at reprov www.esajournals.org

1

April 2015 v Volume 6(4) v Article 66

HILLIS ET AL.

density of offspring transforming from the larval to juvenile stage (i.e., recruits) and ultimately to population density. Reproductive success of males may be affected by several factors, both natural and anthropogenic. In many populations, competition for mates among males leads to strong selection for complex physical and behavioral traits, such as large antlers, loud mating calls, or more successful sperm, that result in higher genetic contributions from males with successful characters (Wedell et al. 2006). Differential mortality of males may temper sexual selection, when these traits reduce survival (Kokko et al. 2002). Large body size is related to increased reproductive output in males within many species, because big males may be selected by females, are often superior competitors for mates, and may produce more sperm (Wedell et al. 2006). Poor growth, exposure to disease, selective harvest (Parkos et al. 2011), and contaminants (Baldigo et al. 2006) may reduce male reproductive output. Persistent organic pollutants, such as polychlorinated biphenyls (PCBs), may interfere with male reproduction due to their potential to bioaccumulate and disrupt endocrine processes associated with reproduction (Derocher et al. 2003). All life stages of animals in aquatic ecosystems are particularly vulnerable to PCB exposure, because these contaminants are hydrophobic and accumulate in the sediments, remaining in their stable state for decades and then becoming bioavailable and ultimately bioaccumulating in long-lived organisms, like fishes. Reproduction of male fish may be particularly sensitive to elevated PCB concentrations in muscle, gonad, and brain tissues (Gray 1998, Jobling et al. 1998, Koch et al. 2006, DeFalco et al. 2015), leading to intersexual characteristics in species that should be strictly diecious. We explored how PCB exposure affected the reproductive contribution of male bluegills (Lepomis macrochirus) to populations in small lakes. Male bluegills have complex, size-dependent reproductive behavior where individuals build benthic nests, compete for females, and participate in male parental care of eggs and larvae (Gross and Sargent 1985, Cargnelli and Gross 1996, Garvey et al. 2002). Large male fishes that reproduce early and are best able to protect nests are often the individuals with the highest fitness within a cohort (Parkos et v www.esajournals.org

al. 2011). In species where complex male traits are impaired by contaminant exposure, relying on complex reproductive features as a life history tactic may hinder reproductive output rather than enhance it. Our objective was to determine whether mature male bluegills collected from different lakes with varying PCB contaminant levels influenced the production of offspring in small, replicated ponds (i.e., the natural habitat for bluegills) at our research facility. We hypothesized that high PCB contaminant loads in fathers reduce individual fitness by increasing the occurrence of intersex, reducing testis size, reducing body condition and survival, and perhaps interfering with reproductive behavior. A strong relationship between population density, especially of adult males, and the production of the next generation of recruits is elusive, especially in fish populations or other species with Type III survivorship, because recruitment of vulnerable offspring is often driven by density-independent factors (Myers 2001, Garvey et al. 2009). Juvenile bluegills only persist in lakes after going through months of multiple, complex and dangerous life stages, transforming from embryos in spring, to larvae in nests, to freeswimming larvae, where survival varies greatly by fall (Werner and Hall 1988, Aday et al. 2002, Garvey et al. 2002). No strong evidence exists that the density of adult males guarding nests, embryo density, and larval density are correlated with future population density, because early survival varies independent of population density (Garvey et al. 2009). However, the density of juveniles produced by fall (i.e., cohort recruits) correlates positively with future adult density, and therefore is a population-level fitness endpoint (Werner and Hall 1988, Cargnelli and Gross 1997, Garvey et al. 2002).

METHODS Study system Lakes.—In August 2010, adult bluegills were collected from eight Midwestern USA lakes (Table 1) to determine whether total PCB body residues were elevated and to determine the incidence of intersex and gonadosomatic index (GSI). These data were used to select fish for subsequent research pond reproduction experi2

April 2015 v Volume 6(4) v Article 66

HILLIS ET AL. Table 1. Summary of sites in Illinois selected for sampling based on Illinois Department of Public Health fish consumption advisories. CCF ¼ Channel Catfish, CRP ¼ Common Carp, LMB ¼ Largemouth Bass, BLG ¼ Bluegill. Lake

County

PCB Advisory

Cedar Lake Frank Holden State Lake Herrin Lake Horseshoe Lake Lake of Egypt Marion City Reservoir Crab Orchard Lake  SIUC Campus Lake 

Jackson St. Clair Williamson Madison Williamson Williamson Williamson Jackson

No advisory CCF and CRP 1 meal per week No advisory CCF and CRP 1 meal per week CRP 1 meal per week BLG 1 meal per week LMB 1 meal per week; 1 meal per month CCF and CRP BLG 1 meal per week

  Listed as a 303d impaired waterway due to PCBs.

ments in 2011 and 2012. All fish were captured with standard, direct current, boat-mounted electrofishing gear. Fifty adult bluegills (including both males and females .150 mm total length) were selected from each of eight lakes (n ¼ 400). Fish were placed on ice and transported back to the laboratory for processing. Fish length (total length, mm), mass (blotted wet mass, g), and gonadosomatic index (GSI ¼ 100 3 gonad mass=whole body mass) data were obtained from each fish. Each carcass was individually packaged in zipper closure plastic bags and frozen at 258C until further analysis for lipid and PCB concentrations. Ponds.—Experiments were conducted in rectangular, 0.05-ha ponds (maximum 2.5 m depth; natural habitat for bluegill) during 2011 and 2012, which allowed bluegills to reproduce, and offspring to successfully transform from eggs, embryos, and larvae in spring to juveniles by fall (i.e., recruits). Before experiments, each pond was drained and completely dried, then filled with water from a source reservoir with a large pump. When filling or topping off the ponds, each water source pipe was covered with two pond socks (81.3 cm long 3 17.8 cm diameter) with a mesh size of 800 lm inner sock and 300 lm outer sock (Pentair Aquatic Ecosystems, Apopka, Florida, USA) to assure that fish larvae from the source reservoir did not enter the experimental ponds. Bluegills consumed naturally occurring zooplankton and macroinvertebrates. Also, a 3-mm mesh was installed on the top of each pond drainage stand-pipe to prevent larvae, juveniles, and adults from escaping the ponds. Water quality parameters (temperature, dissolved oxygen, and pH) were measured weekly (Hach Hydromet, Loveland, Colorado, USA). v www.esajournals.org

Histology In 2010, histology samples were taken to evaluate whether male bluegills from the field survey lakes had intersex characteristics. All males collected during the field study were histologically evaluated (;25 fish from each lake; total n ¼ 200). Testes were removed, weighed, and preserved in 10% neutral buffered formalin. Samples were then sectioned and stained. Three cross sections of gonadal tissue were taken equidistant from each other down the length of each gonad lobe. A total of six sections (two on each slide) of each gonad sample were stained with hematoxylin and eosin. The slides were examined with a compound microscope at 400 times to identify gonadal abnormalities, such as oocytes in testicular tissue and to assess whether mature spermatocytes were present.

Fish lipid and contaminant concentrations Five male fish from each lake for lake surveys and before the pond experiments were randomly selected for quantification of body residues of 28 PCB congeners summed as ng/g per fish. Wholebody fish tissue samples (minus gonadal tissue, which was removed for GSI and histological processing, hereafter referred to as whole-body fish tissue) were homogenized using an acetonerinsed stainless steel tissue grinder (Waring, South Shelton, Connnecticut, USA). Whole-body homogenates were stored in solvent rinsed glass jars at 258C until solvent extraction. Approximately 3 g of freeze-dried whole-body homogenate was extracted using a Dionex 200 accelerated solvent extractor (ASE) (Sunnyvale, California, USA) and processed for lipid and PCB concentrations at the SIU aquatic toxicology laboratory (see Appendix). Additional toxicolog3

April 2015 v Volume 6(4) v Article 66

HILLIS ET AL.

ical analyses for other chemicals such as DDE and DDT were done on a subset of fish for another study.

Crab Orchard Lake (intermediate PCBs), and Campus Lake (high PCBs) were added to the ponds, respectively. A subset (n ¼ 3) of adult male bluegills from each treatment was processed for PCB body residues using the same methods as described above. Juvenile bluegills were retrieved in each pond (see below), and lengths and masses of adults were recorded. Both experiments ended in fall when nightly pond temperatures were consistently below 208C, which was determined by using continuous temperature monitors with hourly data points (Onset, Bourne, Massachusetts, USA). At the end of each experiment, water levels were drained to approximately 1 m and three seine hauls per pond were performed with a 10 3 2.5 m beach seine with a 2.5 m bag and a 3 mm mesh size, which covered the entire length of each pond. Survival (i.e., % remaining), total length (mm), and mass (g) were recorded for adult bluegills. The total mass (g) of juvenile bluegills obtained from each seine haul was quantified for each pond. Comparisons of juvenile production among lakes were made using abundance estimates. Abundance estimates were calculated by randomly subsampling the masses of 150 juveniles. The average individual fish mass and the total mass of juveniles were used to calculate a relative abundance estimate for each pond (total mass of three seine hauls (g)/average mass of individual (g) ¼ juvenile abundance estimate). The number of juvenile fish caught declined exponentially with each pass of the seine haul, with at least 90% of bluegill removed by the third pass. Subsequent seine hauls produced ,3% additional fish biomass, suggesting that three seine hauls was appropriate for capturing the relative density of juvenile bluegill produced among ponds.

Pond experiments Two pond experiments were conducted, where mature females from reference lakes (i.e., nondetectable PCB body residues) were placed in each pond. Males collected from lakes containing varying PCB body residues were added to ponds to determine the impact of male PCB body burdens on reproductive output. A pilot experiment was conducted in 2011 to identify logistical difficulties and determine whether differences existed in juvenile abundance among lakes. Fish were collected in the same manner as the bluegill field survey. Lakes were selected based on Illinois fish consumption advisories (IDPH 2010) and were verified using the results from the 2010 PCB field survey. Four ponds were used in year one of the study. Each pond represented a single treatment. First, 10 female bluegills from Cedar Lake (non-detectable PCB body residues, reference lake) were added to each of the four ponds on 18 June 2011. Ten male bluegills from Lake of Egypt (low PCBs), and Crab Orchard Lake (intermediate PCBs) were then added. Ten males from Campus Lake (high PCBs) and Cedar Lake (non-detectable PCBs) were added the next day. Juvenile abundance estimates (see below) and adult length and mass were measured at the end of the 2011 trial. The study ended on 2 November 2011. The pond experiment in 2012 was conducted to expand on the preliminary trial and included 12 ponds and four treatments. The experiment commenced 7 June 2012 and ended 21 September 2012. Lakes used for the reproductive trial were selected based on 2010 field survey male bluegill PCB body residues. Field collected bluegill data indicated that Cedar Lake contained detectable concentrations of another contaminant (DDE); therefore, Herrin Lake was selected as the reference lake for the 2012 trial. First, female reference site bluegills (no detectable PCB body residues) were collected from Herrin Lake and distributed randomly into each pond (n ¼ 120; 10 per pond). Herrin Lake (non-detectable PCB body residues) males were collected and added to three randomly selected ponds (n ¼ 30). Then 10 males from Marion Reservoir (low PCBs), v www.esajournals.org

Statistics Loge(x þ 1) transformed adult PCB whole-body residues of contaminants (ng/g dry mass) and proportion lipids (g/kg dry mass) were compared across the Illinois lakes using a one-way ANOVA with pair-wise comparisons (Tukey’s HSD). Because no replication occurred during the 2011 pond experiment, only mean differences were evaluated and post-study adult PCB concentrations were not verified, although they likely were similar to those from the 2010 lake 4

April 2015 v Volume 6(4) v Article 66

HILLIS ET AL.

Fig. 1. Total whole-body (minus gonadal tissue) male PCB concentrations observed in bluegills collected in August 2010 from selected Illinois lakes (mean þ SD). Five male fish were randomly selected per lake and analyzed using gas chromatography coupled with electron capture detection. Significant differences among the different lakes are indicated by different letters as determined by Tukey’s honestly significant differences test (F7,32 ¼ 9.59, P , 0.0001).

survey. Adult retrieval, which may have reflected survival among treatments during the 2012 reproductive trial, was compared using a oneway ANOVA. For the purpose of regressions, non-detectable PCB body residues were assigned a value of half of the reporting limit. We combined juvenile estimates for both years to assess whether patterns were consistent between years. The total mass of juvenile bluegill collected across the three seine hauls was regressed against average mass of juvenile bluegill to determine whether density-dependent processes affected growth of young fish. A negative relationship would suggest that ponds with high densities of juvenile bluegill had poor growth and that resources may have been limited for survival. All statistical analysis was accomplished using the SAS statistical software package (SAS Institute, Cary, North Carolina, USA).

0.0001) (Fig. 1). Histology showed no intersex fish or abnormal gonadal tissue in any of the male bluegills. Males across all lakes contained spermatocytes.

Pond experiments The 2011 and 2012 experiments revealed similar patterns. At the outset of each experiment, male bluegills collected from Campus Lake had significantly greater lengths (F3, 116 ¼ 99.44, P , 0.0001) and masses (F3, 116 ¼ 103.48, P , 0.0001) than those from the other lakes. Female sizes did not differ among treatments. Female bluegill across both experiments averaged 188 6 8 SD mm total length. Lipid concentrations of bluegill males differed among lakes (F7, 39 ¼ 3.1, P ¼ 0.01; Table 2), with those from Horseshoe Lake having the highest concentrations and those from Lake of Egypt having the lowest. In 2011, the Cedar Lake (reference) pond was inadvertently drained and could not be sampled. Recovered adults from ponds, which may have been related to survival, ranged from 40 to 70% and was unrelated to contaminant concentration (Table 3). Campus Lake (high PCBs) adult males remained significantly longer (F2,7 ¼ 6.14, P ¼ 0.03) and heavier (F2,7 ¼ 10.43, P ¼ 0.008) among recovered males in all treatments (Fig. 2) at the end of the test. Recovered female mass and

RESULTS Lake survey Lake-specific male bluegill whole-body total PCB body residues in 2010 ranged from below reporting limit to 2,000 ng/g dry mass (non-lipid normalized basis). Male bluegill populations contained PCB body residues that were significantly different from each other (F7,32 ¼ 9.59, P , v www.esajournals.org

5

April 2015 v Volume 6(4) v Article 66

HILLIS ET AL. Table 2. Proportion of lipids in whole bodies (without gonads) of adult male bluegill from Illinois lakes in 2011. Lakes not sharing a common letter are significantly different (Tukeys, P , 0.05). Lake

Mean

SD

Difference

Cedar Lake Frank Holden State Lake Herrin Lake Horseshoe Lake Lake of Egypt Marion City Reservoir Crab Orchard Lake SIUC Campus Lake

0.09 0.12 0.09 0.18 0.06 0.09 0.15 0.10

0.04 0.05 0.02 0.07 0.03 0.06 0.06 0.05

ab ab ab b a ab ab ab

had significantly greater masses and lengths than the other treatments (Fig. 2). Recovered females were not significantly different from each other in terms of both mass and length among treatments. The GSI (males: F3,50 ¼ 1.70, P ¼ 0.18; females: F3,10 ¼ 0.45, P ¼ 0.73) of both males and females were similar among treatments (Fig. 3). When 2011 and 2012 reproductive trial juvenile abundance estimates were correlated against adult mean male body residues in each pond, the relationship was negative (F1,9 ¼ 0.89, P ¼ 0.002, R2 ¼ 0.71; Fig. 4). Mean total mass of juvenile bluegill was unrelated to the average mass of individuals across ponds (R2 ¼ 0.03), suggesting that there were no density-dependent effects on juvenile growth.

length were not significantly different among treatments at the end of the test. Dissolved oxygen (DO) concentrations ranged from 4.2 to 10.8 mg/L, temperatures ranged from 208 to 348C and pH ranged from 5.8 to 8.4. In 2012, adult bluegill recovery, which may reflect survival in each pond, ranged from 7% to 50%, again unrelated to pond treatment (Table 3). Due to excessive mortality or inability to sample juveniles due to excess draining of the pond and high vegetation, one replicate from each treatment was lost. Survival of adults was not significantly different among treatments (F3,8 ¼ 1.54, P ¼ 0.28). Dissolved oxygen, temperature, and pH ranged from 2.6 to10.5 mg/L, 178 to 328C, and 8.6 to 10.2, respectively. No significant differences were noted in mass or length of female bluegills stocked into the various treatments (Fig. 2). At the end of the test, male bluegills originally collected from Campus Lake

DISCUSSION This study suggests that the reproductive state of males in a population contributes directly to population density, and thereby impacts population size. A similar study was recently conducted by Parkos et al. (2011), showing that nesting male largemouth bass (Micropterus salmoides) that were able to protect larvae from predation contributed proportionately more to young production, meaning that these fish had higher fitness. We expand on this finding. Parkos et al. (2011) did not translate nest success to the density of recruits at the population scale, while the current study showed this. Males matter both in terms of genetic diversity within cohorts and density of populations, even in species with highly variable

Table 3. Summary of recovered adults out of 10 possible observed in both reproductive trials and replicates omitted from analysis. Potential survival among treatments was not significantly different among treatments. Year

Lake

Replicate

Males

Females

2011 2011 2011 2012

Lake of Egypt (low PCBs) Crab Orchard Lake (intermediate PCBs) Campus Lake (high PCBs) Herrin Lake (non-detectable PCBs)

2012

Marion Reservoir (low PCBs)

2012

Crab Orchard Lake (intermediate PCBs)

2012

Campus Lake (high PCBs)

1 1 1 1 2 3  1 2 3  1 2  3 1  2 3

9 3 3 7 6 6 6 13 1 3 0 11 1 4 3

4 5 11 2 1 3 1 2 0 0 1 1 3 1 2

  Excluded from contaminant analysis due to loss of males.

v www.esajournals.org

6

April 2015 v Volume 6(4) v Article 66

HILLIS ET AL.

Fig. 2. Recovered male and female length (mm total length; dark bars) and mass (g; light bars) measurements of the 2011 and 2012 reproductive trials at the end of the test (mean þ SD). A difference was observed among treatments in both reproductive trials for male length (2011: F2,7 ¼ 6.14, P ¼ 0.03; 2012: F3,50 ¼ 5.23, P ¼ 0.003) and mass (2011: F2,7 ¼ 10.43, P ¼ 0.008; F3,50 ¼ 4.54, P ¼ 0.007) as denoted with the bars with different letters. No differences in female length and mass were observed among treatments in 2011 and 2012.

Fig. 3. Gonadosomatic index (mean 6 SE) of recovered adults in the 2012 reproductive trial at the end of the test. No statistically significant differences were noted among treatments for both males (F3,50 ¼ 1.70, P ¼ 0.18) and females (F3,10 ¼ 0.45, P ¼ 0.73). Number of males measured were 19, 20, 13 and 8 for Herrin, Marion, Crab Orchard, and Campus, respectively. Number of females measured were 6, 3, 5, and 6 for Herrin, Marion, Crab Orchard, and Campus, respectively.

v www.esajournals.org

7

April 2015 v Volume 6(4) v Article 66

HILLIS ET AL.

Fig. 4. Juvenile bluegill abundance on a log scale for both 2011 and 2012 reproductive trials regressed against adult male PCB concentration. The 2011 PCB values are mean values from the 2010 field survey coupled with the 2011 young of year abundance estimates. Data for each adult bluegill, pond, and year are plotted. However, for statistical analyses, mean concentrations for bluegills within each ponds were used (F1,9 ¼ 0.89, P ¼ 0.002, R2 ¼ 0.71). Each symbol represents a lake and year combination: EGY ¼ Lake of Egypt; H ¼ Herrin Lake; CRAB ¼ Crab Orchard Lake; M ¼ Marion.

first-year survival. Many processes affect male reproductive success, including the characteristics that arise from male-male competition for mates. From an applied perspective, males are often the target of removal by harvest or eradication, such as bucks in deer populations. In most cases, this is not expected to have much of an impact on population growth. However, in organisms with males that provide some parental care, there is a debate whether this is important to population growth (Steinhart et al. 2004). This is particularly relevant to fish ecology, where there are numerous examples of males providing parental care in marine and freshwater systems (Alonzo 2012). Thus, the effects of contaminants on male parental care could be quite detrimental to this particularly diverse and abundant taxonomic group. In the bluegill populations, the mechanism underlying reduced reproductive output at the population level in ponds is unknown. Exposure to contaminants in the current study did not cause intersexuality in males or cause notable v www.esajournals.org

abnormalities in sperm production. Apparent survival of adults varied across ponds, but was unrelated to treatment. Terrestrial predation by birds and mammals is common in these ponds, and some adults may have been missed during pond draining. The finding of no direct impact of male physical characteristics was unexpected. A positive relationship has been shown between body size, GSI and reproductive output (number of eggs or sperm produced) in several fish species (Kang et al. 2003, Green et al. 2010). In most fish, an exponential increase in reproductive output is expected as fish mass increases (Tarnada and Iwata 2005). Male fish with larger gonads should participate in more spawning events and produce more sperm per event, because they have more gametes for fertilization (Wedell et al. 2006). Previous research indicates that a nesting male bluegill that is 6% larger than its cohorts can produce 50% more fry (Gross 1982). The bigger bodies, large gonads, and supposed better condition of Campus Lake contaminated males should have increased rather than reduced cohort size produced. Thus, reduced reproduc8

April 2015 v Volume 6(4) v Article 66

HILLIS ET AL.

tive output via contaminant loading was likely not caused by the morphological characteristics we quantified. Polychlorinated biphenyls could affect the complex mating behavior of male bluegills due to disruption of male-male competition and sexual selection by females. Although we did not quantify PCB concentrations in the brains of the male bluegill, Koch et al. (2006) found that presence of chemicals in the hypothalamus and pituitary correlated with endocrine disruption in male sturgeon. We hypothesize that male bluegills collected from PCB contaminated lakes may have altered reproductive activities, such as impaired nest building, parental care, male courtship of females, and nest defense associated with impairments in the brain. Recent research has demonstrated that parental care can be altered by endocrine disruptive contaminants (McCarty and Secord 1999, Bustnes et al. 2001, Saaristo et al. 2010). Parental male bluegills that invest in nest building activities may be at a disadvantage due to high PCB body residues or other contaminants that were not quantified but co-varied with PCB loading. To reach the juvenile stage, bluegill offspring must develop from embryos and fry in nests in spring, to free-swimming larvae in the open water, and then transform into juveniles by fall. This spring-summer gauntlet during early life is when most mortality occurs during the lifespan of most fish (i.e., Type III Survivorship) and varies dramatically among populations. Transfer of contaminants to offspring is usually maternal not paternal (Bourgeon et al. 2013, Lyons and Lowe 2013), so it is unlikely that offspring survival was affected by residual contaminants from their fathers. Other processes that may have negatively affected survival of offspring in ponds could be sperm-duct impairment, abnormal sperm production, nest abandonment, cannibalism, or some yet unknown effect of male exposure on offspring development and growth. Given that many other variables may have interfered with reproductive results in ponds, the negative relationship between whole-body male PCB burden and juvenile success was surprising. Other contaminants may have varied among individuals as we found that our ‘‘clean’’ females in 2011 contained low concentrations of DDE (i.e., many contaminants covary; Baldigo et al. v www.esajournals.org

2006). Most studies focus on contaminants in target tissues that influence endocrine function such as liver, gonad, and brain, whereas we quantified total body burden. Males from Campus Lake had the highest average concentration of PCBs, but also had the greatest variation among individuals. This individual variation may have translated to high variability in response in ponds. In our view, our experimental results provide a conservative relationship between contaminants and male success. Our survey revealed that contaminant concentrations varied in males across lakes, likely due to differences in historical spills of industrial waste that date back decades. The Campus Lake contaminants likely arise from a transformer spilled in the lake in the 1980s. Lakes vary widely in fish population and ecosystem dynamics and a ‘‘hidden’’ factor affecting these may be legacy contaminants affecting reproductive behavior. Bluegills consume zooplankton, suppressing their density and reducing their grazing pressure on phytoplankton (Turner and Mittelbach 1990). Bluegills also are important prey for apex predators in many lakes (Aday et al. 2005). Impaired reproductive output due to contaminants may scale up to ecosystem processes and also affect the ability of fisheries to respond to harvest. Given that environmental endocrine disruptors and estrogen mimics are a persistent problem in aquatic ecosystems, these hidden effects on reproduction need to be incorporated into predictive models, and male contributions to reproduction, especially in populations where males rely on complex mechanisms through sexual selection including attracting females (Alonzo 2012), should not be discounted.

ACKNOWLEDGMENTS The present research has been funded in part by the U.S. Environmental Protection Agency under the Greater Research Opportunities Graduate Program (EPA STAR to J. J. Hillis).

LITERATURE CITED Aday, D. D., C. M. Kush, D. H. Wahl, and D. P. Philipp. 2002. The influence of stunted body size on the reproductive ecology of bluegill Lepomis macrochirus. Ecology of Freshwater Fish 11:190–195.

9

April 2015 v Volume 6(4) v Article 66

HILLIS ET AL. Aday, D. D., D. E. Shoup, J. A. Neviackas, J. L. Kline, and D. H. Wahl. 2005. Prey community responses to bluegill and gizzard shad foraging: Implications for growth of juvenile largemouth bass. Transactions of the American Fisheries Society 134:1091– 1102. Alonzo, S. H. 2012. Sexual selection favours male parental care, when females can choose. Proceedings of the Royal Society B 279:1784–1790. Baldigo, B. P., R. J. Sloan, S. B. Smith, N. D. Denslow, V. S. Blazer, and T. S. Gross. 2006. Polychlorinated biphenyls, mercury, and potential endocrine disruption in fish from the Hudson River, New York, USA. Aquatic Sciences 68:206–228. Bourgeon, S., E. K. H. Leat, R. W. Furness, K. Borga, S. A. Hanssen, and J. O. Bustnes. 2013. Dietary versus maternal sources of organochlorines in top predator seabird chicks: an experimental approach. Environmental Science and Technology 47:5963– 5970. Brzyski, J. R., W. Taylor, and D. N. McLetchie. 2014. Reproductive allocation between the sexes, across natural and novel habitats, and its impact on genetic diversity. Evolutionary Ecology 28:247–261. Bustnes, J. O., V. Bakken, K. E. Erikstad, F. Mehlum, and J. U. Skaare. 2001. Patterns of incubation and nest-site attentiveness in relation to organochlorine (PCB) concentrations in glaucous gulls. Journal of Applied Ecology 4:791–801. Cargnelli, L. M., and M. R. Gross. 1996. The temporal dimension in fish recruitment: birth date, body size, and size-dependent survival in a sunfish (bluegill: Lepomis macrochirus). Canadian Journal of Fisheries and Aquatic Sciences 53:360–367. Cargnelli, L. M., and M. R. Gross. 1997. Fish energetics: larger individuals emerge from winter in better condition. Transactions of the American Fisheries Society 126:153–156. DeFalco, M., M. Forte, and V. Laforgia. 2015. Estrogenic and anti-androgenic endocrine disrupting chemical and their impact on the male reproductive system. Frontiers in Environmental Science 3:3. Derocher, A. E., H. Wolkers, T. Colborn, M. Schlabach, T. S. Larsen, and O. Wiig. 2003. Contaminants in Svalbard polar bear samples archived since 1967 and possible population level effects. Science of the Total Environment 301:163–174. Garvey, J. E., T. P. Herra, and W. C. Leggett. 2002. Protracted reproduction in sunfish: The temporal dimension in fish recruitment revisited. Ecological Applications 12:194–205. Garvey, J. E., R. A. Wright, and E. A. Marschall. 2009. Searching for threshold shifts in spawner-recruit data. Canadian Journal of Fisheries and Aquatic Sciences 66:312–320. Gray, L. E. 1998. Xenoendocrine disrupters: laboratory studies on male reproductive effects. Toxicology

v www.esajournals.org

Letters 102:331–335. Green, C. C., C. T. Gothreaux, and C. G. Lutz. 2010. Reproductive output of Gulf killifish at different stocking densities in static outdoor tanks. North American Journal of Aquaculture 72:321–331. Gross, M. R. 1982. Sneakers, satellites and parentals: polymorphic mating strategies in North American sunfishes. Zeitschrift fur Tierpsychologie 60:1–26. Gross, M. R., and R. C. Sargent. 1985. The evolution of male and female parental care in fishes. American Zoologist 24:807–822. IDPH [Illinois Department of Public Health]. 2010. Illinois Department of Public Health fish advisory. http://www.idph.state.il.us/envhealth/fishadv/ 2010_fish_advisories.pdf Innes, D. J., and M. Ginn. 2014. A population of sexual Daphnia pulex resists invasion by asexual clones. Proceedings of the Royal Society B 281. Jobling, S., M. Nolan, C. R. Tyler, G. Brighty, and J. P. Sumpter. 1998. Widespread sexual disruption in wild fish. Environmental Science and Technology 32:2498–2506. Kang, S. G., K. S. Choi, A. A. Bulgakov, Y. Kim, and S. Y. Sung-Kim. 2003. Enzyme-linked immunosorbent assay (ELISA) used in quantification of reproductive output in the pacific oyster, (Crasso streagigas), in Korea. Journal of Experimental Marine Biology and Ecology 8:1–21. Koch, B. T., J. E. Garvey, J. You, and M. J. Lydy. 2006. Elevated organochlorines in the brain-hypothalamic-pituitary complex of intersexual shovelnose sturgeon. Environmental Toxicology and Chemistry 25:1689–1697. Kokko, H., R. Brooks, J. M. McNamara, and A. I. Houston. 2002. The sexual selection continuum. Proceedings of the Royal Society B 269:1331–1340. Lyons, K., and C. G. Lowe. 2013. Mechanisms of maternal transfer of organochlorine contaminants and mercury in the common thresher shark (Alopias vulpinus). Canadian Journal of Fisheries and Aquatic Sciences 70:1667–1672. McCarty, J. P., and A. L. Secord. 1999. Nest-building behavior in PCB-contaminated tree swallows. Auk 153:55–63. Monosson, E., W. J. Fleming, and C. V. Sullivan. 1994. Effects of the planar PCB 3,3 0 ,4,4 0 -tetrachlorobiphenyl (TCB) on ovarian development, plasma levels of sex steroid hormones and vitellogenin, and progeny survival in the white perch (Morone americana). Aquatic Toxicology 29:1–19. Myers, R. A. 2001. Stock and recruitment: generalizations about maximum reproductive rate, density dependence, and variability using meta-analytic approaches. Ices Journal of Marine Science 58:937– 951. Nagler, J. J., and C. G. Daniel. 1997. Exposure of male American Plaice (Hippoglossoides platessoides) to

10

April 2015 v Volume 6(4) v Article 66

HILLIS ET AL. contaminated marine sediments decreases the hatching success of the progeny. Environmental Toxicology and Chemistry 6:1733–1738. Parkos, J. J., D. H. Wahl, and D. P. Philipp. 2011. Influence of behavior and mating success on broodspecific contribution to fish recruitment in ponds. Ecological Applications 21:2576–2586. Saaristo, M., J. A. Craft, K. K. Lehtonen, and K. Lindstrom. 2010. An endocrine disrupting chemical changes the courtship and parental care in the sand goby. Aquatic Toxicology 97:285–292. Steinhart, G. B., E. A. Marschall, and R. A. Stein. 2004. Round goby predation on smallmouth bass offspring in nests during simulated catch-and-release angling. Transactions of the American Fisheries Society 133:121–131. Tarnada, K., and K. Iwata. 2005. Intra-specific variations of egg size, clutch size and larval survival related to maternal size in amphidromous Rhinogobius goby. Environmental Biology of Fishes 73:379–389. Turner, A. M., and G. G. Mittelbach. 1990. Predator

avoidance and community structure: interactions among piscivores, planktivores, and plankton. Ecology 71:2241–2254. USEPA [US Environmental Protection Agency]. 2000. United States Environmental Protection Agency guidelines for assessing chemical contaminant data for use in fish advisories. Fish Sampling and Analysis. Volume 1. Third edition. Office of Science and Technology, Washington, D.C., USA. Wedell, N., C. Kvarnemo, C. M. Lessells, and T. Tregenza. 2006. Sexual conflict and life histories. Animal Behaviour 71:999–1011. Werner, E. E., and D. J. Hall. 1988. Ontogenetic habitat shifts in bluegill: the foraging rate-predation risk trade-off. Ecology 69:1352–1366. Wilson, E. O. and W. H. Bossert. 1971. A primer of population biology. Sinauer, Sunderland, Massachusetts, USA. You, J., P. F. Landrum, T. A. Trimble, and M. J. Lydy. 2007. Availability of polychlorinated biphenyls in field-contaminated sediment. Environmental Toxicology and Chemistry 9:1940 –1948.

SUPPLEMENTAL MATERIAL APPENDIX

(ISCO, Lincoln, Nebraska, USA) was used to collect PCB congener time fractions. Time fractions were collected from 8’00’’ to 10’00’’ and from 10’01’’ to 12’00’’ (Koch et al. 2006). Fractions were then solvent exchanged to hexane and reduced in volume to 2 ml using a Zymark Turbovap II. Cleanup methods followed those described by the United States Environmental Protection Agency (USEPA 2000). Concentrated sulfuric acid (1 ml) was added to each 1 ml extract and shaken for 5 min, followed by 5 min in a centrifuge at 2300 revolutions per minute. The hexane layer was removed and passed through 10 g of anhydrous Na2SO4 to remove any remaining water. This process was repeated twice more with 2-mL volumes of hexane (2 mL). The combined extracts were evaporated to 500 ll under a steady nitrogen stream and transferred with hexane rinses to a gas chromatography vial with a 1 ml final volume. Once at

Extraction surrogates 4,4 0 -dibromooctafluorobiphenyl (DBOFB) and PCB 203 were added to the accelerated solvent exchange cells before the extraction process. The 50 ml solvent extract (1:1 [v/v] acetone and methylene chloride) was evaporated with nitrogen and solvent exchanged to methylene chloride in a Zymark Turbovap II (Zymark, Hopkinton, Massachusetts, USA) to a volume of 6 ml. One ml of fish tissue extract was filtered through a 0.2 lm Whatman filter (Whatman, Piscataway, New Jersey, USA). Separation of lipids from the sample extract was performed using gel permeation chromatography. A 300 mm 3 19 mm Envirogel gel permeation chromatography cleanup column with 5 mm 3 19 mm guard column (Waters, Milford, Massachusetts, USA) was used. High-pressure liquid chromatography (Agilent Technologies, Palo Alto, California, USA) and a Foxy Jr. Fraction collector v www.esajournals.org

11

April 2015 v Volume 6(4) v Article 66

HILLIS ET AL.

final volume, samples were stored at 208C until analysis. Polychlorinated biphenyls were quantified using an Agilent 6890 gas chromatograph (Agilent Technologies, Palo Alto, California, USA) equipped with a 7683 auto-sampler and a microelectron capture detector (GC/ECD) using dualcolumn confirmation. The columns DB-608 (30 m 3 0.32 mm 3 0.25 lm) and RTX-1614 (30 m 3 0.250 mm 3 0.10 lm) were used for analysis with a carrier gas flow rate of 1 ml/min. The injector temperature was 2508C and the ECD temperature was 3508C. The oven temperature ramp for the DB-608 column consisted of an initial temperature of 1008C, a ramp of 108C/min. to 2408C, held at 2408C for 2 min., then 38C/min. up to 2808C and held for 5.67 min. at 2808C. The temperature ramp for the RTX-1614 column was set at 1008C, heated to 1808C at 158C/min., then heated to 2008C at 68C/min., held at 2008C for 4 min., then heated to 2208C at 58C/min., held at 2208C for 4 min, and heated to 2808C at 38C/min (You et al. 2007). The carrier gas was helium and the make-up gas was nitrogen. All gases used for GC analysis were 99.9999% ultra pure (Airgas USA, Bowling Green, Kentucky, USA). External

v www.esajournals.org

standards at concentrations of 5, 10, 50, 100, 250, and 500 ng/mL of each PCB and surrogate standard were prepared for calibration. Analysis followed methods specified by the Environmental Protection Agency method 8080a (USEPA 1994). Values for each column were averaged if within 20% of each other. For congeners that exceeded 20% difference, the lowest recorded values for each congener between the columns was used for data analysis. A mid-range (50 ng/mL) calibration standard was analyzed every eight samples to assure that instrument conditions did not significantly change. To verify performance of the analytical method, two surrogates (DBOFB and PCB 203) were added to each sample prior to extraction. Blanks for each matrix, in addition to matrix spike and matrix spike duplicates for fish tissue analysis, were included with every 20 samples. Reporting limits (the lowest concentration of an analyte that could be reliably measured and reported with the analytical method) for PCB congeners were set at 5 ng/g for PCBs. Total PCB concentrations were calculated by summing the 28 individual congeners.

12

April 2015 v Volume 6(4) v Article 66