Sampling protocols for estimating time budgets of roosting bats

Abstract

This study was designed to evaluate sampling protocols for estimating time-activity budgets of colonial roosting bats during the day-roosting period, based on data collected from roosting activities of adult eastern pipistrelles (Pipistrellus subflavus) at a maternity colony in eastern Massachussetts. A comparison of three different scan-sampling protocols was used to quantify day-roosting activities (LONG, 72 scans at 60-min intervals; SHORT, 20 scans at 15-min intervals; and BRIEF, 10 scans at 15-min intervals) relative to a standardized protocol (HOUR, continuous replicate scans over a 60-min session). These analyses indicate that day-roosting activity is most reliably quantified using the SHORT and BRIEF sampling protocols (those with the highest sampling frequencies). The rarity of some activities (e.g., crawling) may not warrant an increased frequency of sampling to quantify daily time budgets of roosting bats. Comparisons of an intensive version of SHORT (20 scans at 10-min intervals) and focal sampling protocols used during the two most active periods of day-roosting (immediately following return to the roost before dawn and just prior to departure before dusk) suggest that these sampling methods provide comparable estimates of roosting activity. We suggest that frequent short scan-sampling protocols may be used to reliably quantify day-roosting activities of bats both during active and inactive periods, although focal sampling may be more appropriate if individual behavior is of interest during intense periods of activity (postfeeding return and pre-emergence periods).