Male-typical reproductive behaviors vary greatly between different species of fishes with androgens playing a variety of roles that appear especially important in the gonochorist cypriniform fishes. The goldfish is an important model for the cypriniformes and while it is clear that male goldfish are fully feminized by prostaglandin F2α(PGF2α), it is not clear whether females will exhibit normal levels of male-typical reproductive behaviors as well as olfactory function when treated with androgens. To answer this question, we exposed sexually-regressed adult female goldfish to several types of androgen and monitored their tendencies to court (inspect females) and mate (spawn, or attempt to release gametes) while monitoring their olfactory sensitivity until changes in these attributes were maximized. Untreated adult males (intact) were included to determine the extent of masculinization. Treatments included the natural androgens, 11-ketotestosterone and testosterone (KT and T), administered via capsules (KT + T-implanted fish); the artificial androgen, methyltestosterone (MT), administered via capsules (MT-C); and MT administered in the fishes’ water (MT-B). Male-typical olfactory sensitivity to a pheromone (15keto-PGF2α) increased in all androgen-treated groups and by week 6 was fully equivalent to that of males. Male-typical courtship behavior increased in all androgen-treated groups although slowly, and only MT-B females came to exhibit levels equivalent to those of males after 18 weeks. In contrast, male-typical mating activity increased only slightly, with MT-B females reaching levels one-third that of males after 30 weeks. We conclude that while androgens fully masculinize olfactory sensitivity and courtship behavior in goldfish, mating behavior is controlled by a different neuroendocrine mechanism(s) that has yet to be fully elucidated.
Bibliographical noteFunding Information:
We thank the Environment and Natural Resources Trust Fund (ENRTF) and Minnesota Aquatic Invasive Species Research Center (MAISRC) for supporting this study. We gratefully acknowledge Peter Xiong for his contribution collecting behavioral data and overall assistance with most aspects of the experiment. The statistical consulting service at the University of Minnesota and Martin Bezener assisted with data analysis.