| Summary: | Thesis (M.Sc.)--Memorial University of Newfoundland, 1995. Biology Bibliography: leaves [128]-150. Theoretical models of habitat selection generally assume that organisms behave optimally, that population density correlates with resource abundance, and that increasing density reduces habitat quality. My purpose was to determine whether current models could explain the distribution of Atlantic Puffins (Fratercula arctica) breeding on Great Island, Newfoundland and, if not, to propose a model that could. Theoretical models have rarely been applied to explain the distribution of a colonially-nesting species that may gain fitness benefits from increasing density. -- I used breeding success as a representative measure of fitness that I compared among three habitats, maritime slope, maritime level, and inland slope, sampled at three locations, north, east and south on Great Island in 1992 and 1993. Based on results of previous studies, I predicted that distance from the shore edge of the colony, slope, and aspect would be the most important habitat variables that discriminate habitat quality for puffins on Great Island. -- Nest density was highest in maritime slope and lowest in inland slope habitat and was best predicted by distance from the shore edge of the colony. Breeding success, as measured by the proportions of burrows that fledged chicks, was highest in maritime and inland slope habitats and lowest in maritime level habitat and was related to distance from the edge and slope. Aspect was an important predictor of timing of breeding but was not significantly related to breeding success. Thus, it appeared that preferred areas were close to shore, while optimal areas were on slopes. -- High breeding success and nest density in maritime slope habitat was predicted by current habitat selection models, but high success at low density in inland slope habitat was not. Breeding success increased with density within habitats, also contrary to model predictions. Current models proved unable to explain the distribution of puffins if breeding success was used as the sole measure of fitness. A cost-benefit model is proposed that acknowledges habitat related fitness costs that are not accounted for by typical measures of breeding success. Unique cost-benefit ratios for different habitats can explain observed patterns of dispersion and breeding success.
|
|---|