Thesis (M.S.)--Wisconsin. / Includes bibliographical references (leaves 42-46).

Endocrine disrupting compounds (EDC) are environmental contaminants that disrupt reproduction, development and behaviour in aquatic organisms. A thorough evaluation of the impacts of EDCs on aquatic organisms is currently limited by a lack of robust biomarkers in small model fish, particularly for assessing EDCs with (anti-)androgenic activity. Male sticklebacks build nests using spiggin, an androgen-responsive glycoprotein, which can be used to assess (anti-)androgenic exposure. EDC assessment in the field using threespine stickleback and the spiggin biomarker is limited to coastal and estuarine environments. However, their freshwater relative, brook stickleback (Culaea inconstans), also possess spiggin and their widespread distribution suggests that they may have applications as a bioindicator of EDCs in freshwater systems. Therefore, the overall objective of this thesis was to determine if brook stickleback are a suitable bioindicator species for EDCs by evaluating their response and sensitivity to estrogenic and (anti-)androgenic chemicals. Basal transcript levels of spiggin in kidney and vitellogenin in liver were first measured in wild-caught brook stickleback using qPCR and found to be differentially expressed in males and females. Brook stickleback were then exposed to two model compounds, 17α-ethinylestradiol (EE2) and 17α-methyltestosterone (MT), at 1, 10 and 100 ng/L for 21 days (sampled at 7 and 21 days) via static-renewal to determine the responsiveness of these transcripts to exogenous hormones. The effect of hormone exposure on condition factor, organosomatic indices and histopathology of kidneys was also measured. Exposure to MT and EE2 significantly induced spiggin and vitellogenin transcripts in female kidneys and male livers, respectively. Exposure to EE2 also significantly increased the hepatosomatic index in females after 7 days and in both sexes after 21 days whereas the gonadosomatic index was reduced in females after 21 days. An increase in kidney epithelium cell height was also observed in MT-exposed females and males after 7 days. These results mirror those of threespine stickleback and suggest that brook stickleback are responsive to androgenic and estrogenic chemical exposure and more specifically, possess quantifiable and sensitive biomarkers for exposure to compounds with androgenic activity. In a third experiment, female fish were co-exposed to MT at 500 ng/L and an anti-androgen (flutamide; FL) at 25, 150 and 250 µg/L for 14 days (sampled at 4 and 14 days) to validate this bioassay for the evaluation of anti-androgens using the same endpoints as in the previous two experiments. In females, exposure to MT increased spiggin transcript levels and nephrosomatic index (NSI) but co-exposure to FL did not result in a significant suppression of these endpoints because of high inter-individual variability. In males, exposure to MT increased NSI and co-exposure to FL resulted in a reduction in this endpoint, illustrating anti-androgenic effects. Although the response of brook stickleback to hormone exposure was endpoint-specific and was at times lower than other small model fish species, the ability to simultaneously assess estrogenic and (anti-)androgenic chemical exposure in a single fish using quantitative endpoints is an advantage exclusively held by members of the stickleback family. The results of this thesis suggest that brook stickleback hold promise as an additional small fish model for the evaluation of EDCs, with potential application in EDC biomonitoring in the freshwaters of North America.

Throughout their lives, prey organisms must balance the tradeoff between fitness-related activities and the risk of predation. To successfully mediate such tradeoffs, prey must have an accurate method to gauge current predation risk. For many aquatic organisms, the use of chemosensory information has been shown to be a ubiquitous and useful tool in mediating predation risk. The chemical cues to which aquatic organisms respond include the odour of known predators and the odour of a damaged conspecific or known or closely related heterospecific. In fishes, the response to damage-released cues from conspecifics or closely related heterospecifics has been shown to be innate, while the response to distantly related unknown heterospecific cues are likely learned. In a series of laboratory and field studies I examined the role of learning in the ability of fathead minnows to respond to damage-released cues of brook stickleback as an indication of predation risk. My results indicate that minnows from a population without stickleback do not recognize stickleback cues as dangerous. However, following the introduction of stickleback, minnows learn to recognize stickleback cues as dangerous. Further study indicated a low ratio of stickleback to minnows in a given population will decrease the likelihood of learning when compared with a similar sized population containing a higher ratio of stickleback to minnows. I also demonstrated that an increase in habitat complexity decreases the ability of minnows to learn to recognize stickleback cues. Studies have further demonstrated that in the face of predation (as indicated by chemical cues from minnows and stickleback) minnows will decrease their antipredator response when in the presence of a fish shoal, especially a shoal of conspecifics. Finally, an examination of the effects of a minnows length, body condition and breeding status indicate that morphological parameters can play a significant role in the intensity of response to heterospecific and conspecific damage-released cues. Previous works have demonstrated that behavioural responses to heterospecific damage-released cues increase the probability of the receiver surviving a predation event. This has important implications for predator/prey interactions and consequently factors affecting the learning and use of such cues deserve increasing attention from ecologists.

Throughout their lives, prey organisms must balance the tradeoff between fitness-related activities and the risk of predation. To successfully mediate such tradeoffs, prey must have an accurate method to gauge current predation risk. For many aquatic organisms, the use of chemosensory information has been shown to be a ubiquitous and useful tool in mediating predation risk. The chemical cues to which aquatic organisms respond include the odour of known predators and the odour of a damaged conspecific or known or closely related heterospecific. In fishes, the response to damage-released cues from conspecifics or closely related heterospecifics has been shown to be innate, while the response to distantly related unknown heterospecific cues are likely learned. In a series of laboratory and field studies I examined the role of learning in the ability of fathead minnows to respond to damage-released cues of brook stickleback as an indication of predation risk. My results indicate that minnows from a population without stickleback do not recognize stickleback cues as dangerous. However, following the introduction of stickleback, minnows learn to recognize stickleback cues as dangerous. Further study indicated a low ratio of stickleback to minnows in a given population will decrease the likelihood of learning when compared with a similar sized population containing a higher ratio of stickleback to minnows. I also demonstrated that an increase in habitat complexity decreases the ability of minnows to learn to recognize stickleback cues. Studies have further demonstrated that in the face of predation (as indicated by chemical cues from minnows and stickleback) minnows will decrease their antipredator response when in the presence of a fish shoal, especially a shoal of conspecifics. Finally, an examination of the effects of a minnows length, body condition and breeding status indicate that morphological parameters can play a significant role in the intensity of response to heterospecific and conspecific damage-released cues. Previous works have demonstrated that behavioural responses to heterospecific damage-released cues increase the probability of the receiver surviving a predation event. This has important implications for predator/prey interactions and consequently factors affecting the learning and use of such cues deserve increasing attention from ecologists.