The effects of landscape structure and body size on the species richness and occurrence of breeding birds on Prince Edward Island

The primary goal of this thesis was to examine the effects of landscape structure, body size, latitude and longitude on the species richness and occurrence of avian species breeding on Prince Edward Island. Variation of species richness and occurrence within functional groups (landbird, seabird, shorebird, waterfowl) and species occurrence within trophic groups (carnivore, omnivore, herbivore, insectivore) is examined. Spatial data on species counts were obtained from Erskine (1992) and avian body size data were obtained from Dunning (1993). Information on land-cover types (agriculture, forest, wetland, roads, human-used areas) were obtained from the provincial government. ArcView GIS and Fragstats were used to calculate several metrics that characterized the structure and complexity of the different land-cover types (area, shape, and aggregation).

Habitat metrics, latitude and longitude, and body size information were related to avian species richness in seventy-six 10 km x 10 km sampling cells. Explanatory variables were related to avian species richness using multiple regression analysis, and models were chosen using Akaike's Information Criterion. In total, 156 species were found to breed in Prince Edward Island. Species assemblages varied between 9 and 98 species. Land-cover types, longitude, latitude and body size explained up to 49% of the variation in avian species richness on Prince Edward Island, though each functional group responded differently to particular variables. These findings suggest that avian species richness on Prince Edward Island is strongly influenced by human activities such as alteration of forest habitat and expansion of agriculture.

At the species level, habitat metrics were related to the occurrence of 102 avian species (species where occurrence >10% or <90% coverage of the study location) in seventy-six 10 km x 10 km sampling cells. Logistic regression analysis was used to examine the associations and to develop multivariate models to explain avian species occurrence. Avian species were then examined in terms of functional group (i.e. landbird, seabird, shorebird, waterfowl) and trophic level (i.e. herbivore, carnivore, insectivore, omnivore). Twenty-nine species were significantly associated with the land-cover variables. These species were primarily from the landbird functional group and the omnivore and insectivore trophic groups. There were few discernable patterns in specific metrics used by species, though forest-related variables were associated with species more often than other land-cover types.

This is the first landscape-scale study on avian community ecology on Prince Edward Island. A blended approach to the study of avian communities is useful to wildlife managers, conservation biologists, and ecologists alike. This study has examined several factors that determine species richness and occurrence of birds breeding on Prince Edward Island at a regional spatial scale. What is required now is a process-orientated investigation of some of these factors, especially those associated with anthropogenic activities. Such an approach would be best served by exploring the biology of individual species and how they interact with the physical geography of their environments.