Abstract: It has been argued that intraspecific mate-choice copying can be adaptive under certain conditions. Dugatkin's (1992,Am. Nat.139, 1384-1389) work with guppies,Poecilia reticulataremains the most influential experimental demonstration of this phenomenon. We replicated Dugatkin's work using several choice criteria to ensure that our results were not dependent upon any single method of judging mate choice. We also tested our findings against two null hypotheses of differing stringency. Irrespective of the choice criteria or null hypothesis used, we did not observe any relationship between female mate choice and copying. We conclude that further experimental evidence of female mate-choice copying is required before the existence of this behaviour can be affirmed.

Abstract: At present, the most general evolutionary theory of honest communication is Grafen's model of Zahavi's 'handicap' signalling system, in which honesty of signals about the signaller's quality (e.g. mate suitability or fighting ability) is maintained by the differentially high cost of signals to signallers having lower quality. The latter model is here further generalized to include any communication between signallers and receivers that are genetically related (e.g. parents and begging offspring, cooperative or competing siblings). Signalling systems involving relatives are shown to be evolutionarily stable, despite a potential pay-off for false signalling, if the Zahavian assumption of differential signal costs holds and there are diminishing reproductive returns to the signaller as the receiver's assessed value of its attribute increases, or if, regardless of whether the Zahavian assumption holds, signallers with high values of the attribute benefit more from a given receiver assessment than signallers with low values (e.g. begging chicks that are hungrier benefit more from being fed). In stable systems of signalling among kin, it is also shown to be generally true that (i) levels of signalling and thus observed signal costs will decline as relatedness increases or as the receiver's reproductive penalty for erroneous assessment increases, and (ii) receivers will consistently, altruistically overestimate the true value of the signalled attribute.

Abstract: We investigated whether capuchin monkeys cooperate to solve a task and to what extent they take into account the behaviour of another individual when cooperating. Two groups of capuchin monkeys (N=5 and 6) were tested in a task whose solution required simultaneous pulling of two handles which were too far from one another to be pulled by one monkey. Before carrying out the cooperation study, individual monkeys were trained to pull one handle (training phase 1) and to pull two handles simultaneously (training phase 2) for a food reward. Nine subjects were successful in training phase 1, and five in training phase 2. In the cooperation study seven subjects were successful, that is, pulled one handle while a companion pulled the other. Further analyses revealed that capuchins did not increase their pulling actions when a partner was close to or at the other handle, that is, when cooperation might occur. These data suggest that capuchin monkeys acted together at the task and got the reward without understanding the role of the partner and without taking its behaviour into consideration. Social tolerance, as well as their tendency to explore and their manual dexterity, were the major factors accounting for the capuchins' success.

Abstract: Lactating female baboons, Papio cynocephalusoften maintain close associations with particular males. There are at least three proposed benefits of 'friendships' to females: (1) male protection against potentially infanticidal males; (2) male protection against harassment by dominant females; (3) male attachment to an infant that develops into future care of juveniles. These hypotheses were examined in a population of chacma baboons, P. c. ursinusin which male infanticide accounted for at least 38% of infant mortality. Almost all mothers of young infants formed strong bonds with one or two males with whom they had copulated during the cycle in which they conceived their infants. Females were primarily responsible for maintaining friendships during lactation, but they terminated these relationships if their infants died. In playbacks of females' screams, male friends responded more strongly than control males. They also responded more strongly to the screams of female friends than to the screams of control females. Following an infant's death, however, male friends responded less strongly than control males to the same females' screams. Finally, male friends responded more strongly than control males to playback sequences in which female screams were combined with the threat vocalizations of a potentially infanticidal alpha male, but not when female screams were combined with the threat calls of a non-infanticidal male or the alpha female. Both observations and experiments suggest that the benefits of friendships to females derive from the protection of their infants against infanticide.1997The Association for the Study of Animal Behaviour

Abstract: In the literature on dominance hierarchies, “winner” and “loser” effects usually are denned as an increased probability of winning at time T, bated on victories at time T-l, T-2, etc, and an increased probability of losing at time T, based on losing at T-1, T-2, etc., respectively. Despite some early theoretical work on winner and loser effects, these factors and how they affect the structure of dominance hierarchies have not been examined in detail. I developed a computer simulation to examine winner and loser effects when such effects are independent of one another (as well as when they interact) and when combatants assess each other's resource-holding power. When winner effects alone were important, a hierarchy in which all individuals held an unambiguous rank was found. When only loser effects were important, a dear alpha individual always emerged, but the rank of others in the group was often unclear because of the scarcity of aggressive interactions. Increasing winner effects for a given value of the loser effect increase the number of individuals with unambiguous positions in a hierarchy and the converse is true for increasing the value of the loser effect for a given winner effect Although winner and loser effects have been documented in a number of species, no study has documented both winner and loser effects (using some controlled, pairwise testing system) and the detailed nature of behavioral interactions when individuals are in groups. I hope the results of this model will spur such studies in the future.