Escalating fights for rank-order position between male swordtails (Xiphophorus helleri): Effects of prior rank-order experience and information transfer

Summary A previous hypothesis which we originally based on standardized mirror tests was examined. It stated that subordinate males from socially stabilized groups, when confronted with a strange male of similar fighting ability, tend to take the offensive; in contrast, dominant males tend to wait and see and to increase their aggressiveness only as far as necessary to counter the aggressive actions of the opponent. 30 highly escalating fights between well matched opponents, either with different or with the same prior rank-order experience, were recorded on video tape and then analysed in detail. Rank-order experience affected neither the escalation level nor the outcome of the fights. However, in the first nonescalated stage of the fights all measures of unilateral aggressive actions were higher in the omega than in the alpha males. This is contrary to expectations based on the theory of social conditioning; nor can it yet be explained by game-theory models. We then discuss why it should pay subordinates to be more offensive and to show higher levels of unilateral aggression than dominants. Some current aspects of game theory are supported by our data: (1) Threat display is a poor predictor of attack and fleeing, and simply informs the opponent that the threatening individual will probably just stay and avoid further escalations. (2) Relative fighting ability is demonstrated during highly-escalated fighting by delivering Fin Grips, the most effective and costly attack pattern. The high energetic costs of Fin Grips can be compensated by reducing the rate of Biting. (3) Probably, the combatants did not transfer information about their intentions, i.e. they concealed their intentions as to what they were going to do next. Contrary to the predictions of game theory, winners and losers showed some differences in Pecking early in the nonescalated stage of the fights, the prospective losers exhibiting more Pecking than the winners.This study was supported by grants from the Deutsche Forschungsgemeinschaft