Large, folding Sherman traps (H.B. Sherman Traps, Tallahassee, Florida) were placed at 10-m intervals (single-trap stations) within each of six 1-ha trapping grids. To reduce trap mortality, and thereby reduce biased estimates of population parameters, we placed polyester batting and bait (rolled oats and sunflower seeds) inside each trap. We used capture-recapture methods (White et al. 1982) for sampling small mammals, trapped for eight weeks beginning July 27, 1998. Traps were set and remained open for three consecutive days during every 14-day period. Traps were checked for captured animals every 12 hours. We trapped one stand during the first 7 days and the other stand during the second 7 days of the 14-day period, thereby providing four, 3-day trapping intervals per patch. Traps were prebaited (trap locked open with a tongue depressor and baited) during each 11-day non-trapping interval to reduce behavioral effects, such as trap “shyness” or “happiness” which can affect capture probabilities and measures of precision associated with estimates of population parameters (Otis et al. 1978).
Trapping data conformed to Pollock’s robust design (Pollock 1982; Kendall et al. 1995, 1997; Kendall and Nichols 1995). The robust design is a combination of the Cormack-Jolly-Seber (CJS) live recapture model and closed capture models (Cormack 1964; Jolly 1965; Seber 1965; Kendall and Nichols 1995; Kendall et al. 1995, 1997). Captured animals, except shrews (identified to genus; Sorex sp.) and chipmunks (Tamius sp.), were identified to species, marked with ear tags (Size 1 Monel tags, National Band and Tag Co., Newport, Kentucky), and released at the capture site. For each capture, the fate (new or recapture), ear tag number, species, sex, mass (g), date, and micro site were recorded.