Journ31 of ijnderr.3dua.3- R-ese-arch
..Oluine , is ue S - [- .3, .. ' une -i.,,.
The Effect of Sewage Treatment Plant Effluent on the Reproductive
Behavior of Sailfin Mollies (Poecilia latipinna)
Endocrine disrupting chemicals (EDC) that are commonly found in the environment have been shown to have to
have deleterious effects on the morphology, physiology, and behavior of wildlife. Water effluent of industrial
and municipal sewage treatment plants is a frequent source of EDCs. Estrogens are a common type of EDC found
in domestic sewage treatment plants (STP), which is thought to enter the sewage effluent through human
excrement containing pharmaceutical chemicals such as the synthetic estrogen ethinyl estradiol in birth control
pills. While the effects of estrogens on morphology and physiology of wildlife, such as fish, have been
well documented, the effects on behavior, specifically sexual behavior, are less well known. The present
study explored the effect of sewage effluent on the sexual behavior of sailfin mollies. It was hypothesized
that female sailfin mollies will choose males from sites that do not receive sewage effluent over males from sites
that do receive effluent. It was also hypothesized that males from polluted sites will display decreased
sexual behavior. Sailfin mollies (Poecilia latipinna) were collected from a reference site receiving no sewage
effluent and an influenced site that contained sewage effluent. Female sailfin mollies were presented a male
molly from both sites and behavior indicative of female preference and male courtship were recorded.
Gonopodium length and standard length were also measured for each male fish. Females did not show a
mate preference. Male behavior, the total time and duration the male spent courting the female, was also
not significantly different between the two sites. Gonopodium length did not differ between sites, contrary
to expectations of an elongated gonopodium in influenced males. I believe that the lack of significant differences
in behavior is mainly attributed to the small sample size, supported by the high amount of variation. Future
studies should increase the sample size and expand the study to include evaluation of more sexual behavior
and quantification of estrogens in the sewage effluent.
Sewage treatment plants (STP) release a concoction of chemicals, natural and synthetic into the
aquatic environment. Many of these chemicals are known to be endocrine disrupting chemicals (EDCs).
The endocrine system is responsible for regulating various functions within the body and does so by
secreting hormones that change the actions of particular tissues or organs. EDCs act to disrupt normal
endocrine functions, resulting in potential abnormalities in the morphology, physiology, and behavior of
animals. Aquatic animals, such as fish, are especially at risk because of constant exposure to low concentrations
of EDCs (Jobling et al., 1996) acquired through aquatic respiration, osmoregulation, and maternal transfer
though uptake in the egg yolk (Van der Kraak et al., 2001). Additionally, because many EDCs are lipid soluble,
they can accumulate in animal tissues, and through biomagnification adversely affect species higher in the food
chain (Tyler et al, 1998).
One class of endocrine disruptor that is quite common in our environment is estrogenic hormones and
estrogen mimics. One common estrogenic chemical that is known to negatively affect wildlife, such as fish, is
human contraceptives, or birth control pills. For example, the yolk protein, vitellogenin, has been found in male
fish after exposure to estrogen (Sumpter and Jobling, 1995). Females that take birth control pills excrete
ethinyl estrodiol and its metabolic byproducts which enter the environment via municipal water treatment
plants. Constant entrance of these estrogens in the environment will continuously produce adverse effects
in exposed wildlife.
Traditional STP processing focuses on removal of nitrogen and phosphorus compounds, but often
estrogen compounds are not removed. STP involves three common treatment steps: preliminary clarification,
aerator tank, and end point clarification. Preliminary clarification removes solid materials from the wastewater.
In the second step, bacteria and protozoa further degrade the biological content. The final clarification step
prepares the wastewater for release back into the environment. Common effluent release points are natural
and man-made wetlands, rivers, or even released back into the ground water. While some processes will
remove estrogens from the sewage influent (Ternes, et al 1999), the majority of processes used fail to do so, due
to the great expense. Although studies conducted in Europe have shown a detection of estrogen in STP
discharge measured in concentrations in the lower ng/L (Ternes, et al 1999; Jobling, et al 1998), this range has
been found to have significant effects on gonad morphology. Desbrow et al. (1998) detected 17p-estradiol
in domestic wastewater ranging from ing/L up to50 and 80 ng/L. Higher concentration of sewage effluent is
directly correlated with the greater degree and number of intersexed roach (Rutilus rutilus) individuals (Jobling et
In particular, estrogens and their mimics are thought to influence wildlife species' sexual and reproductive
behavior. This is alarming because impairment of sexual behavior function coupled with possible
morphological defects could deleteriously affect the fitness of the population. The natural estrogen, 17 p-estradiol,
is known to affect hormonal pathways that regulate the development of gametes, sexual phenotype and
sexual behavior (Arcand-Hoy and Benson, 1998). One example of impairment of sexual behavior is differences
in mate choice, which could result from incorrect signaling, production of incorrect cues, or timing of
particular behavior, etc. Experimental evidence has shown that often male sexual behavior, courtship displays
and mating are decreased, impaired or even inhibited in many wildlife species exposed to EDCs (Zala and
Reproductive success depends on many factors, including the individual's ability to produce the correct
sexual behavior at the proper time. Many laboratory exposure studies have found that reproductive
success decreases with exposure to one or a concoction of chemicals, but fewer studies have linked the effect
of occurrence of EDCs in the environment to decreased reproductive success (Mills and Chichester, 2005). This
study examines the effect of sewage effluent on reproductive behavior that are explicitly linked with
reproductive success in the sailfin molly.
The sailfin molly (Poecilia latipinna) was chosen as an experimental organism because they exhibit a high rate
and variety of reproductive behavior. The natural reproductive behavior of sailfin mollies has been well
documented. The females exert particular mate choice behavior and the males display well characterized
courtship behavior such as gonoporal nibbling, gonopodial thrusting and sigmoidal curving of the body for the
male. In addition, male fish display a large colorful dorsal and caudal fin that plays a role in female choice.
They also have a highly modified structure formed from anal fin elements called a gonopodium which is used
for fertilization and has been shown to be sensitive to estrogenic chemicals in other species. These fish
are commonly found throughout the southeastern United States, occurring in fresh, brackish and even coastal
salt water in marsh edges, lowland streams, ponds, swamps and roadside ditches. Their ability to survive in
oxygen depleted and less than ideal waters makes sailfin mollies an ideal fish to study in waters receiving
The object of this experiment is to determine whether sewage effluent has an effect on the sexual and
reproductive behavior of sailfin mollies. Specifically, this experiment investigates whether male sailfin
mollies, naturally exposed to estrogen containing sewage effluent, display altered mating behavior and
whether females find males from waters containing sewage effluent less attractive. I hypothesize that male
mollies exposed the sewage effluent will display mating behavior less frequently than sailfin mollies not exposed
to effluent, and that female sailfin mollies will choose unexposed males over exposed males.
To test these hypotheses, a female was presented two males, one from a "reference site" and the other from
an "influenced site." A 20-gallon aquarium was used in the trials by dividing it in half lengthwise with a clear
plastic divider, then dividing one of these halves into three equal parts using opaque dividers. The undivided
half contained the female while the divided half, in each of the left and right compartments, contained either a
fish from the "influenced" or "reference" site. The middle compartment contained no fish, serving as a neutral
zone for females. Male fish were separated using opaque plastic dividers to avoid visual contact and male to
male competition. The dividers did not have holes so chemical cues which may diffuse through the water
where limited, but this permitted better viewing with the video camera.
The experimental trials were performed within 4 days of the female releasing a brood. The experimental
aquarium was prepared no later than a day before the test to roughly match the holding tank conditions with
6ppt salinity. The water was dechlorinated and aerated for at least 24 hours. Males were isolated in their own
tanks 24 hours prior to female exposure. This isolation has been determined to induce natural mating rates
typically found in natural environments for fish of all sizes (Ptacek and Travis 1997 as in Travis 1994). Females
were left alone until the beginning of the trial. Behavioral observations were conducted between 1100 and
1600 hours. The fish were placed in the experimental tank and given 15 minutes to acclimate to their
new environment, after which video recording began and lasted 25-30 minutes. After the experiment
was completed, the female fish was placed back into her aquarium and the males were euthanized in 1%
neutral buffered MS222 and fixed in neutral buffered formalin for later body length and gonopodium measurements.
The males selected for testing were +/- 10mm of the female's total length to increase interactions, as studies
have shown that males are not attracted to significantly larger females and vice versa (Ptacek and Travis 1997 as
in Travis 1994). The color of each male was controlled for to avoid coloration preference exhibited by the
female. Previous observations have shown that females prefer males that are more brightly colored over males
that were not brightly colored. Males were not used from Otter Creek to avoid natal population preference by
Sailfin Molly Mating Behavior
Male sailfin mollies display three prominent mating behavior towards females. In a courtship display, the male
will display an enlarged dorsal fin complemented by a sigmoidal curving of the body and tilting the body towards
the female. A second behavior, gonopodial thrusting, occurs when the male attempts to forcefully inseminate
the female. The third behavior, gonoporal nibbling, occurs when the male tries to make facial contact with
the female's gonopore. In this experiment, because the male and female fish are divided into
separate compartments, only sigmoidal curving can be considered in the behavioral observations of males.
Female choice was be measured by the amount of time she spent with each male.
Measurements and Statistics
Standard length (SL), gonopodium length, and weight were measured for each male fish used in the
experiment. Standard length was measured after the fish was euthanized and gonopodium length was
measured after the fish were preserved in formalin.
The amount of time each female spent with either male or with neither (middle compartment), and the amount
of time each male spent courting the female were recorded from video recording, using the JWatcher program.
Because there is often a relationship between SL and gonopodium length, a Univariate Analysis of
Covariance (ANCOVA) was performed on these measures. However, no relationship was found for either Lake
Alice or Tumblin Creek sites (F=1.345, df=2, p=0.286) possibly due to the fact that body size was controlled for
and thus did not vary. Therefore, independent t-tests were computed to compare male standard length
and gonopodium length separately.
The relationship between total time a female spent with a male from each site and gonopodium (or body) length
was analyzed using a univariate ANCOVA. Specifically we analyzed the main effects of site and gonopodium
(or body) length and their interaction (the interaction provides a test for the assumption of equal slopes).
Slopes were equivalent in both of these comparisons. Similarly, the relationship between total time males from
each site spent courting and gonopodium (or body) length was analyzed using a univariate ANCOVA. Specifically
we analyzed the main effects of site and gonopodium (body) length and their interaction.
The total time the male spends courting the female, the mean duration of male courting per bout, the total time
the female spends with each male, and the mean duration per bout that the female spends with the male
were analyzed using a paired samples, one-tailed t-test. The confidence level was set at 95% and the alpha level
= 0.05 for all statistical tests.
Collection and Care
Fish were collected between July and September 2006 in North Central Florida. Female sailfin mollies were
collected at Otter Creak in Dixie County, Florida. Male sailfin mollies were collected in Tumblin Creek in the city
of Gainesville and also in Lake Alice on the University of Florida campus. Sites classified as "reference" received
no sewage effluent and were not used as dumping grounds for other chemical waste products. Sites classified
as "influenced" received direct sewage effluent.
Fish from Otter Creek and Tumblin creek were collected by seining in several areas along the creek. Fish from
Lake Alice were collected using fish traps randomly placed throughout the stream leading into the main lake
area. Only mature male fish were collected so they would be receptive to mating. Male fish were determined
as mature if the total length was about 25mm. Female sailfin mollies signal their receptivity to mating upon
maturity and also during a two to three day period after releasing a brood (Ptacek and Travis 1997; Farr and
Travis 1986; Travis 1989). Male mollies also display a higher number and duration of courtship displays when
the female is receptive (Farr and Travis 1986). Studies have shown that nonreceptive females do not exercise
any preference given a choice of males and have high rates of apathy (Ptacek and Travis 1997). Therefore,
only gravid females were collected and maintained until releasing her brood and only receptive females were used
in the experiment to increase rates of female preference.
Female fish were kept in 10-gallon aquaria divided into three equal parts by plastic dividers. One female was kept
in each section so I could detect when she gave birth. Male fish were kept separately in 10-gallon aquaria
according to their place of origin. Fish were kept at 6ppt salinity on a 12-121ight dark cycle. They were fed
TetraMin tropical flake food once or twice a day.
Gonopodium length (t=1.388. df=18, p=0.091) and standard length (t=0.338, df=18, p=0.369) did not
differ significantly between LA males and TB males. The total time the female spent with the male was not related
to gonopodium length (F=2.143, df=1, p=0.165 i.e. slope was not different from zero), and did not differ
between the sites (and (F=2.149, df=l, p=0.165, respectively). Total time each male spent courting the female
was also not related to gonopodium length (F=2.090, df=l, p=0.170), and did not differ between the two sites
(and F=0.713, df=l, p=0.413, respectively). Female mollies were not affected by the size of the males in her
choice of mate (F=0.491, df=1, p=0.494). Male molly courting time did not vary with the males' SL (F=1.001,
The female mollies spent on average 36% of the total observation time (t=l, df=9, p=0.172, Fig 1) in the
proximity of the Lake Alice (LA) males and 45% with the Tumblin Creek (TB) males. The LA male mollies spent
an average of 24% and TB male mollies spent 35% of the observation time courting the females (t=1.242,
df=9, p=0.123, Fig. 2). Neither of these differences was significant.
TT female with LA male TT female withTB male
Figure 1. The total time the female spends with each male and standard deviation.
TT LA male courting TT TB male courting
Figure 2. The total time the male spends courting the female and standard deviation.
While the results do not show a significant difference in directionality of female choice, there was a tendency for
the female to choose TB males. The male mollies from Tumblin creek, the reference site, also tend to exhibit
more courting behavior and for longer periods of time. These "tendencies" support the hypothesis that females
will choose fish from the non-contaminated site and contaminated males will show decreased courting behavior.
The lack of significance could in fact be due to the small sample size; thus the interpretation of these results are
open for further experimentation and no definite conclusions can be made.
While this experiment was inconclusive, results are consistent with other studies that show the decrease of
sexual behavior when exposed to estrogens. Adult male guppy (Poecilia reticulata) mating behavior, quantified
by duration and number of sigmoidal displays, decreased after the fish were exposed to a natural estrogen or to
a xenoestrogen (Bayley et al, 1999). Baatrup and Junge (2001) exposed the adult male guppy to the
antiandrogens, p,p'-DDE, flutamide and vinclozolin, and found that male sexual behavior, specifically
sigmoidal display and posturing behavior had decreased in duration. On the contrary, similar exposure
experiments during development of the male guppy saw an increase in sexual behavior (Toft and Baatrup,
2003) although the explanation for this converse effect is unclear. Toft and Baatrup (2003) suggested that
as androgens must be converted to estrogens in the brain to regulate sexual behavior, the increase in
sexual behavior could be due to altered neural development or other unknown estrogenic effects. There are no
other studies in the literature resulting in an increase in sexual behavior as known. Variation in the duration
and total time courting is higher in the LA males than in the TB males, suggesting that there is another
mechanism affecting these behavior, possibly timing of exposure. Further experiments are needed to
determine whether this variation is due to insufficient sample size or if there are other biologically
Careful selection of the fish by size proved to be successful. The lack of difference between sampled sites in
standard length (SL) and gonopodium length suggests that we were successful in matching the males for body
size. Additionally, the females' choice was not influenced by SL, suggesting that the male pair was
appropriately matched to the females' SL. The lack of a difference in gonopodium length between the two sites
could possibly be a result of only selecting males that appeared to be mature. In mosquitofish (Gambusia
affinis), gonopodium length is set at maturity and does not experience growth even as the body may continue
to grow larger (Turner, 1941; Angus et al, 2005), and it is likely the gonopodium is also set at maturity in
sailfin mollies. The fact that there was no difference in gonopodium length between the influenced site and
the reference site was actually contrary to expectations. Exposure studies of 17p- estradiol and 17a-
ethynlyestradiol during the development of poeciliids, show elongation of the gonopodium (Doyle and Lim,
2002; Angus, et al, 2005; Toft and Baatrup, 2003).
Although sailfin mollies are common within North Florida, the accessible areas for collection were difficult to
find, resulting in lack of site-type replication. Future studies should find more contaminated and
uncontaminated sites to produce replication within the experiment and increase sample size.
This study may be broadened in scope to include assessment of more male reproductive behavior. Of the
three common sexual behavior, only sigmoidal displays were measured, so decreased (or increased) display of
the other two behavior, gonopodial thrusting and gonoporal nibbling, has yet to be assessed in
estrogen contaminated mollies for sailfin mollies. Also quantification of estrogens found in both reference
and influenced sites would shed light on how the fish are physiologically affected.
Given the considerable amount of literature on the adverse effect of EDCs on development and reproduction
of wildlife, it is important to continue to develop biomarkers so that we may quickly evaluate the potency of
EDC concoctions within our water systems.
1. Arcand-Hoy, L. D., and Benson, W. H. 1998. Fish reproduction: An ecologically relevant indicator of
endocrine disruption. Environ. Toxicol.Chem. 17, 49-57.
2. Angus, R.A., Stanko, J., Jenkins, R.L., Watson, R.D. 2005. Effects of 17a-ethynylestradiol on sexual development
of male western mosquitofish (Gambusia affinis). Comparative Biochemistry and Physiology, Part C 140 330 - 339.
3. Baatrup E, and Junge M. 2001. Antiandrogenic pesticides disrupt sexual characteristics in the adult male
guppy (Poecilia reticulata) Environmental Health Perspectives. Oct. 109 (10): 1063-1070.
4. Bayley, M., Nielsen, J.R., and Baatrup, E. 1999. Guppy Sexual Behavior as an Effect Biomarker of Estrogen
Mimics. Ecotoxicology and Environmental Safety 43, 68-73.
5. Desbrow C, Routledge EJ, Brighty GC, Sumpter JP, Waldock M. 1998. Identification of estrogenic chemicals is
STW effluent. 1. chemical fractionation and in vitro biological screening. Environ. Sci. Technol. 32:1549-1558.
6. Doyle, C.J., Lim, R.P., 2002. The effect of 1713-estradiol on the gonopodial development and sexual activity
of Gambusia holbrooki. Environ. Toxicol. Chem. 21, 2719-2724.
7. Farr, J.A. and Travis, J. 1986. Fertility Advertisement by Female sailfin mollies, Poecilia latipinna. Copeia, May.
No. 2, pp. 467-472.
8. Jobling, S., Nolan, M., Tyler, C.R., Brightly, G., and Sumpter, J.P. 1998. Widespread Sexual Disruption in Wild
Fish. Environ. Sci. Technol., 32, 2498-2506.
9. Mills, L.J. and Chichester, C. 2005. Review of evidence: Are endocrine-disrupting chemicals in the
aquatic environment impacting fish populations? Science of the Total Environment 343: 1- 34.
10. Ptacek, M. and Travis, J. 1997. Mate Choice in the Sailfin Molly, Poecilia latipinna. Evolution, Vol. 51, No. 4.
11. Sumpter, J.P., Jobling, S., 1995. Vitellogenesis as a biomarker for estrogenic contamination of the
aquatic environment. Environ. HealthPerspect. 103 (Suppl. 7), 173-178.
12. Ternes, T.A., Stumpf, M., Mueller, J., Haberer, K., Wilken, R.D. and Servo, M. 1999. Behavior and occurrence
of estrogens in municipal sewage treatment plants I. Investigations in Germany, Canada and Brazil. The Science
of the Total Environment 225: 81-90.
13. Toft, G. and Baatrup, E. 2003. Altered sexual characteristics in guppies (Poecilia reticulata) exposed to 17b-
estradiol and 4-tert-octylphenol during sexual development. Ecotoxicology and Environmental Safety 56: 228-237.
14. Travis, J. 1994. Size-dependent behavioral variation and its genetic control within and among populations. Pp.
165-187 in C.R.B. Boake, ed Quantitative genetic approaches to animal behavior. Univ. of Chicago Press, Chicago.
15. Turner, C.L., 1941. Regeneration of the gonopodium of Gambusia during morphogenesis. J. Exp. Zool. 87, 181-209.
16. Tyler CR, Jobling S, Sumpter JP. 1998. Endocrine disruption in wildlife: a critical review of the evidence. Crit
Rev Toxicol. 28: 319- 61.
17. Van Der Kraak G, Hewitt M, Lister A, McMaster ME, Munkittrick KR. 2001. Endocrine toxicants and
reproductive success in fish. Hum Ecol Risk Assess; 7:1017- 25.
18. Zala, S.M. and Penn, D.J. 2004. Abnormal behavior induced by chemical pollution: a review of the evidence and
new challenges. Animal Behavior, 68, 649-664.
Back to the Journal of Undergraduate Research
College of Liberal Arts and Sciences I University Scholars Program I University of Florida |
ï¿½ University of Florida, Gainesville, FL 32611; (352) 846-2032.
*UP UNIVERSITY of
The;fi,'irri8id'ano fw T Gor I- - - . ',