Can hair traps sample wild boar (Sus scrofa) randomly for the purpose of non-invasive population estimation?

Reliable estimation of population size remains a major challenge in wildlife ecology and management. Lately, genotyping of non-invasively obtained tissue samples integrated in a modified capture–recapture approach provides new perspectives. Faeces, moulted feathers, or hairs can be easily sampled in the field. However, an important assumption is homogeneity of sampling across the population. In this pilot study, we tested the suitability of baited barbed wire hair sampling stations (‘hair traps’) for homogeneous genetic sampling for population estimation. A video system based on a new network internet protocol was used to observe the behaviour of wild boar visiting baited hair traps for gaining information about potential heterogeneities in the individual sampling probability. Within 92 monitoring nights at two sampling stations, 216 wild boar visits were recorded and 142 hair samples containing 2,124 single hairs were collected. Video analysis revealed distinct differences in the behaviour of wild boar with respect to the sampling station which are most likely to result in heterogeneous individual sampling probabilities. Adult and subadult animals differed in their behaviour dependent on their group status. This result indicates that hair sampling with baited hair traps is not suitable for representative non-invasive sampling of free ranging wild boar populations.