Problems of Live-Trapping and Population Estimation for the Wood Mouse, Apodemus sylvaticus (L.)

Publisher Website
Google Scholar
Abstract
Live trapping of Apodemus sylvaticus to obtain capture/recapture data proved unsatisfactory. Captures were low each summer but increased considerably during the autumn and it was clear that many animals present during the summer had gone undetected. Reliable estimates of population size/ could not therefore be made for the summer. The use of population indices based on the number of animals caught is shown to be of doubtful value. The reliability of population estimates made by Darroch''s method was affected by several sampling factors. The chance of capture was not equal throughout the population, attributable partly to the presence of "trap-prone" and "trap-shy" animals and partly to the pronounced "edge effect" on the 5.2 ha study area. The chance of capture remained fairly constant for any animal during the period in which it was caught, so there was no evidence of learning to enter or avoid traps. Up to 15% of the animals on the area were known to have migrated between trap periods, and the equivalent influx of new animals tended to inflate estimates. The satisfactory estimation of a truncated negative binomial for the frequency of capture data suggests that the hypothesis of a gamma distribution of catchability is acceptable and leads to a possible new method for estimating population size over a period, allowing for unequal chance of capture in the population.