Live-trapping in the stalk zone of tall grasses as an effective way of monitoring harvest mice (Micromys minutus)

Abstract

The harvest mouse (Micromys minutus) is understudied compared to other small mammals as a result of its small size and scansorial habits. This study in wetlands dominated by common reed (Phragmites australis) compared nest census, a commonly used technique to confirm presence/ absence and monitor populations, with live-trapping using Longworths and a home-made alternative trap (the Jordan trap) set both on the ground and in the stalk zone (∼1 m from ground or water level). Nests were found at only two of the four study sites, which may have suggested an absence of the species. However, harvest mice were caught in traps at all sites. All 108 captures of 39 individuals were made in aerial traps in the stalk zone with none caught in ground traps. Gen-eralised linear mixed model (GLMM) also showed that significantly more captures were made in Longworth traps compared to the Jordan trap, although the efficiency of the latter increased after modification. There were also significant differences in capture rate between sites and season and a preference for higher reed quality as described by PCA in the wetter areas of the reed-beds studied. We conclude that live-trapping is preferable to nest census as a population monitoring technique and that future studies of harvest mice, especially in tall wetland vegetation, should use Longworth traps and/or cheaper home-made alternatives set in the stalk zone.