Examining the bias of pitfall traps with enclosure experiments and removal sampling

Abstract

Abstract Pitfall trapping is among the most widely used sampling methods for ground‐dwelling beetles. However, the sampled ground‐dwelling beetle abundance in pitfall traps may be biased, which can lead to difficulties when comparing the sampled abundance between different habitats. To better understand the comparability of the sampled abundance, we focused on two sampling processes: a temporary migration process (referring to the individual movements into and out of the effective trap area) and a removal process (determining the number of individuals collected given the number of individuals within the trap area). To examine the effects of the temporary migration process, we compared the sampled abundance of 10 ground‐dwelling beetle species inside and outside enclosures in 2 habitats (clear‐cut and forested sites) and found that installing the enclosure decreased the sampled abundance in both habitats, but the effects did not differ among habitats. To examine the effects of the removal process, we examined the effects of daily soil temperature on the daily sampled abundance (removal probability) of Pterostichus thunbergi using removal sampling within enclosures. We found that soil temperature increased the removal probability, but this effect was limited and not consistent across survey periods. The results suggest that the effects of the temporary migration process may not be habitat dependent and that the effects of the removal process may be small. Therefore, the traditional use of pitfall traps is, to some extent, a robust and comparable measure of the sampled abundance of ground‐dwelling beetles among different habitats.