The adaptive nature of agonistic behaviour in winter flocks of silvereyes is assessed by examining the advantages and disadvantages incurred by individuals that behave in different ways.
Elements of agonistic behaviour are described and defined from film of interactions. Interaction sequences involve static displays (made up of one or more elements in combination) and overt acts such as attacks, chases and aerial combat.
One element, wing fluttering, occurs also in the reproductive and food begging contexts and a comparison of the form of this element within and between contexts is carried out. Intensity differences within contexts occur and differences between contexts are only apparent at high intensity. The ambiguity inherent in such a signal is discussed and its role as a 'qualifier' suggested.
Interactions between silvereyes at a feeding perch after prior food deprivation lead to relationships which can be expressed in peck-dominance hierarchies. Relationships between birds remain stable although reverse packs are not uncommon and male birds tend to be dominant over female birds.
Access to food correlates significantly with rank under certain conditions of prior food deprivation. Three behavioural strategies are used to gain access to food: direct competition (threat display or attack); submissive waiting and scrambling competition. The use of these strategies varies with rank and also with length of prior food deprivation.
The type of behaviour used in initiating direct competition varies between individuals. Relationship to the opponent, rank proximity and familiarity with the opponent significantly affect initiator behaviour. Birds of alpha rank initiate interactions with low intensity elements of attack but escalate in retaliation. Sub-dominant birds tend to initiate frequently with high intensity attack behaviours. The frequency of interactions decreases with decreasing rank but initiations by low ranking individuals involve threat display. Subordinate birds rarely initiate interactions and behave submissively while feeding.
Relationships between elements of agonistic behaviour are sought by considering the following: the pattern of their use by individuals and rank combinations of participating individuals; the co-occurrence of elements in initiator behaviour; covariation with changes in the situation in which the behaviour occurs; and interactional sequential associations.
The results of these analyses indicate that two sub-groups of elements of initiation exist, viz., elements of attack and elements of threat display, each of which can be used to remove a potential competitor from the feeding perch. Escalation to aerial com.bat and chase follows behaviours from each group. The probability that the opponent will retaliate is higher for elements of threat display than elements of attack.
The cost (in terms of energy expenditure and probability that the opponent will retaliate) associated with each element is assessed. The elements of initiator behaviour adopted by a bird may thus encode the 'cost' the individual is willing (at an unconscious motivational threshold) to incur in order to win a contest.
A variety of behavioural strategies in terms of the nature and frequency of interactions are adopted by individuals competing for access to food. The costs and benefits associated with each strategy vary from individual to individual. Each bird adopts strategies which maximize net energy gain in relation to its competitive abilities: functions of age, sex, physiological state and inherited ability.
Agonistic behaviour in the winter flock may have been selected for as means by which individuals maximize energy input per unit energy expenditure, i.e. as optimal behavioural strategies.