Date of this Version
Roosting behavior is common to most avian pests of agriculture. Movements from highly aggregated distributions in roosts to highly dispersed distributions on foraging grounds determine pattern and severity of avian pest problems. This research seeks an understanding of how roosting behavior influences the dispersion of avian agricultural pests and the damage they cause. My focus is on why birds form communal roosts and how communal roosting influences the selection of foraging sites. I document patterns of roosting behavior in European starlings (Sturnus vulgaris) through population level studies, followed by analysis of individual behavior using radio telemetry. Starlings maintain long-term fidelity (up to 130 days) to the same diurnal activity center (DAC), while using a variety of roosting sites at night. DACs tend to be at the center of the distribution of roosting sites used by individual birds. These and other results contradict expectations based on the most widely held explanations for roosting behavior and have led us to a new interpretation based on an association between large roosts and high-quality feeding sites (e.g., agricultural fields). Examination of previous attempts to manage avian pest problems in light of these new findings helps explain some earlier successes and failures, and may also promote development of new more efficient approaches to avian pest problems.