Some aspects of the breeding biology of the Dominican gull Larus dominicanus (Lichtenstein 1823) in Nelson province, New Zealand.

The thesis addresses some aspects of the breeding biology of Dominican Gulls (Larus dominicanus) in Tasman Bay, Nelson Province, New Zealand. A synopsis of a 10 year (1980-91) nest count and banding programme of chicks is given, which illustrates an apparent reduction in pairs nesting over that period. There were significant changes in food resources available to the gulls in 1987 with the closure of several rubbish tips and fish offal dumping sites. These closures did not appear to make any difference to mean clutch size and yet the numbers of gulls breeding continued to drop. In the thesis, I review Lack's hypothesis (1954, 1968) of population regulation and conclude that the hypothesis is unlikely to explain a number of anomalies in the breeding behaviour of local populations of Dominican Gulls. Field work in three relatively small breeding colonies on Rabbit and Bells Islands in the Waimea Inlet near Nelson, is described over two breeding seasons. Sexual dimorphism of adults, pair bonding, and nest site fidelity were studied, together with the recording of egg size and weight, date of laying, clutch size, hatching dates and chick weights. Egg temperature studies were maintained throughout the incubation period. Results indicate that strong pair bonding occurs, and nest site fidelity is developed equally strongly after advantageous sites within the colony have been gained. There is a hierarchy within the nesting colonies with a gradient of increasing breeding success from the outer perimeter of the site toward the epi-centre of the colony. Nest density plays an important part in breeding success and densities in excess of 350 nest/ha initiate a tension factor within nesting colonies, which leads to parasitism and other behaviour inimical to breeding success. It is high nest densities which may eventually lead to colony abandonment. Nesting colonies develop and wane in a cyclic manner with nest densities appearing to play an important role in the dynamic pattern of the colonies. Incubation appears to start as soon as the first egg is laid, and mean egg temperatures increase as incubation progresses. Embryonic heat generation can control the rate of cooling of eggs and also the rate by which eggs are restored to full incubation temperature after being uncovered by the incubating parent. Eggs can survive temperatures in excess of 40 degrees C and low temperatures of 20 degrees C during the incubation period and still produce healthy chicks. Parental investment in the offspring, is not in terms of clutch size, nor is brood reduction attained primarily by asynchronous hatching. Rather, female parents manipulate egg quality within the clutch, and apportion investment between A, Band C eggs depending on body weight of the female and the circumstance of the nesting colony in terms of its cyclical development. It is concluded that the local populations of Dominican Gulls react in an opportunistic way to the immediate circumstances of the environment, and that factors of experience and learning are likely to influence breeding patterns of behaviour more than the adaptations by natural selection, seen by Lack as those factors which ensure the raising of the greatest number of offspring possible. I conclude that the local population of Dominican Gulls is behaving in a similar way to that hypothesised by Wynne-Edwards (1962), whereby animals attain a homeostatic state and regulate their own population numbers in order not to over-exploit the local food resource. High nest densities, infertility in eggs, and high levels of predation of both eggs and fledglings are the probable factors which have most effect on population density and breeding success, whereas food resources do not appear to have an immediate limiting effect