Social interactions tune aggression and stress responsiveness in a territorial cichlid fish (Archocentrus nigrofasciatus)

We examined the relative influences of pre-fight housing condition, contest intensity, and contest outcome in modulating post-fight stress hormone concentrations in territorial male convict cichlids (Archocentrus nigrofasciatus). Individuals were housed either in isolation or in semi-natural communal tanks. Pairs of male cichlids that differed considerably in body mass were selected from the same housing regime. Pre-fight water-borne cortisol levels were obtained before allowing the dyad to interact until contest resolution, after which time post-fight cortisol levels were obtained from the winner and loser. There were no outcome-related differences in post-fight cortisol concentrations following escalated or non-escalated contests, a result that held true for both housing regimes. Pre-fight cortisol levels were significantly higher than post-fight cortisol levels, suggesting that initial confinement in a beaker for the water-borne hormone samples was a stressor, but that the animals acclimated quickly to confinement. Fights involving previously isolated participants were significantly more intense than those involving group-housed animals, which we explain as being a function of established relationships between social isolation, heightened acute cortisol responsiveness, and the expression of excessive aggressive behavior. Only group-housed losers demonstrated the ability to modulate aggression or hypothalamic-pituitary-interrenal (HPI) activity in a graded fashion to acute increases in cortisol or changes in contest intensity, respectively. We discuss a variety of factors that could disrupt the ability of isolates to appropriately modulate interactions between social behavior and the HPI axis, and we examine a number of functional hypotheses underlying the sensitivity of group-housed losers to changes in contest dynamics. Crown Copyright (c) 2006 Published by Elsevier Inc. All rights reserved.