Document Type

Honors Paper


Robert Askins

Publication Date



Bird and vegetation surveys have been conducted periodically in the Connecticut College Arboretum since the 1950s, permitting analysis of long-term relationships of bird populations with the structure and floristic composition of the vegetation and with forest cover changes in the surrounding region. Principal component analyses created vegetation factors encompassing variables such as total basal area, basal area of common tree species, and tree diversity, and created a forest fragmentation factor based on GIS analysis of the landscape within 2 kilometers of the study site. Poisson regression models revealed relationships between these factors and population changes for twenty individual species and seven groups of ecologically similar species, most of which have declined sharply since the 1980s.

Most habitat and migrant groups showed a strong negative relationship with the first vegetation factor, which positively correlates with increasing total basal area and dominance of oaks and maples but negatively correlates with tree and shrub diversity, suggesting that floristic diversity may be critical for determining habitat quality. All groups showed at least a slight positive relationship with the second vegetation factor, which is related to eastern hemlock basal area, suggesting that the recent decline in hemlocks due to hemlock woolly adelgid has had adverse impacts on the forest bird community. All groups except short-distance migrants showed a negative relationship with the landscape factor measuring forest fragmentation, suggesting that conservation efforts to protect bird communities should take the wider landscape into account. The strongest relationship in most cases was with the first vegetation factor.

Included in

Ornithology Commons



The views expressed in this paper are solely those of the author.