Photoperiod modulation of aggressive behavior is independent of androgens in a tropical cichlid fish

Abstract

Photoperiod is a major environmental cue that signals breeding conditions in animals living in temperate climates. Therefore, the activity of the reproductive (i.e. hypothalamic pituitary gonadal, HPG) axis and of the expression of reproductive behaviors, including territoriality, is responsive to changes in day length. However, at low latitudes the seasonal variation in day length decreases dramatically and photoperiod becomes less reliable as a breeding entraining cue in tropical species. In spite of this, some tropical mammals and birds have been found to still respond to small amplitude changes in photoperiod (e.g. 17 min). Here we tested the effect of 2 photoperiod regimes, referred to as long-day (LD: 16L:08D) and short-day (SD: 08L:16D), on the activity of the HPG axis, on aggressive behavior and in the androgen response to social challenges in males of the tropical cichlid fish Tilapia rendalli. For each treatment, fish were transferred from a pre-treatment photoperiod of 12L:12D to their treatment photoperiod (either LD or SD) in which they were kept for 20 days on stock tanks. Afterwards, males were isolated for 4 days in glass aquaria in order to establish territories and initial androgen levels (testosterone, T; 11-ketotestosterone, KT) were assessed. On the 4th day, territorial intrusions were promoted such that 113 of the isolated males acted as residents and another 1/3 as intruders. Territorial intrusions lasted for 1 h to test the effects of a social challenge under different photoperiod regimes. Photoperiod treatment (either SD or LD) failed to induce significant changes in the HPG activity, as measured by androgen levels and gonadosomatic index. However, SD increased the intensity of aggressive behaviors and shortened the time to settle a dominance hierarchy in an androgen-independent manner. The androgen responsiveness to the simulated territorial intrusion was only present in KT but not for T. The percent change in KT levels in response to the social challenge was different between treatments (SD > LD) and between male types (resident > intruder). The higher androgen response to a social challenge in residents under SD may be explained by the time course of the androgen response that due to the long time it takes to fight resolution under LD, might have been delayed. This result illustrates the importance of incorporating time response data in social endocrinology studies. (C) 2014 Elsevier Inc. All rights reserved.