Prior residency and the stability of dominance relationships in pairs of Green Swordtail fish, Xiphophorus helleri (Pisces Poeciliidae). Jacques Beaugrand and Martin Beaugrand Département de Psychologie, Université du Québec à Montréal, CP 8888, Station «A», Montréal, Québec, H3C 3P8 Canada Abstract The stability of dominance relationships between pairs of male Green Swordtail fish was followed daily for 20 consecutive days. In one experimental sample composed of 21 pairs, dominance of one of the fish had been favoured on the first day by giving the fish prior familiarity (prior residency) with the aquarium where it was to meet an intruder. In a control sample composed of 12 pairs, two intruders met in an unfamiliar aquarium. It was expected that the advantage given to the dominant by familiarity with the aquarium on the first day would disappear as the subordinate acquired in turn familiarity with the milieu. In comparison with pairs composed of two intruders, this would show up by more frequent inversions of the initial dominance relationship in pairs composed of a prior resident and an intruder. Only two inversions occurred over the 20 days of follow up and they occurred equally in the experimental (5%) and control (8%) samples. These results confirm the great stability of dominance relationships in dyads and invalidate the hypothesis that the prior residency advantage would decay as the subordinate became familiar with the aquarium. Unexpectedly, 13 of the 66 (20%) fish died over the 20 days. Death equally occurred in both samples but 12 (92%) cases implied initial subordinates. The exception was an initial dominant which had become the subordinate pair member three days before death. Various hypotheses are suggested to account for the selective deaths of subordinates. Key-words: Prior residency, temporal stability of dominance/subordination relationships, death of subordinates, Green Swordtail fish, Xiphophorus helleri. Introduction Familiarity with the surrounding in which the meeting takes place favours dominance in the individual over one in a place which is unfamiliar to him. Thus, the advantages of a three hour familiarization period with the future meeting place were clearly demonstrated in the Green Swordtail for individuals of equivalent size which had been put in isolation for 18 hours beforehand (Zayan, 1975a, 1965b, 1976; Beaugrand and Zayan, 1985). In these studies, the opponents were equally manipulated and simultaneously introduced into the place of residence of one of the two fish. Residency effect can therefore not be attributed to differences in manipulation or even to the fact that an opponent had been present in the residence at the time the other fish was introduced. Prior residency can thus correspond to the opponent's greater familiarity with the environment in comparison to the other partner, or even to the fact that the intruder must become familiar with the additional environment of the resident. So, "prior residency " is expressed through a very revealing tendency of the resident to dominate over the intruder. Under these conditions where dominance is favoured from the outset by a familiarity with the environment, one would expect this advantage to disappear as the intruder in turn gains in familiarity. In comparison with a control group made up of pairs of intruders, this advantage should make itself known with more frequent subsequent dominance relationship inversions in the case of pairs composed of one resident and one intruder. Beaugrand Beaugand, 1991 file:///C|/Documents and Settings/Administrator/My ...ents/publications/Beaugrand-Beaugrand-1991-Mort.htm (1 of 6) [11/30/2001 23:46:25]