부산대학교 도서관

Reliable assessment of wildlife population parameters is essential to ensure conservation and management. Data on spatio-temporal variations in population density are needed to understand habitat requirements and anthropogenic impacts. We investigated the impact of forest loss and fragmentation on the population density of a forest specialist mammal like the pine marten. We assessed it in two contrasted landscapes, a continuous forest and an agricultural land where forests are fragmented into small woodlots and hedgerows. A non-invasive genetic sampling was carried out over two consecutive years to assess population density using capture-mark-recapture models implemented in CAPWIRE and CAPTURE. Across the different sampling sessions, 86 distinct pine martens could be identified, 59 in the hedged farmland and 27 in the forest site. Contrary to our expectations, the hedged farmland carried a similar level of pine marten density (around 0.85 (95%CI 0.44–1.30) individual/km2) to that estimated in forest (0.64 (95%CI 0.41–1.09) individual/km2) but experienced a high rate of turnover. Consistently with other studies in Europe, we showed that at between 20 and 35% of forest cover, landscapes can carry high-density levels for pine marten populations. This study underlines the need to keep forest patches in agricultural landscapes safe from trapping and hunting to maintain sustainable pine marten populations.