Scale-dependent habitat selection by reintroduced Eld’s deer (Cervus eldi) in a human-dominated landscape

Context Knowledge of the habitat selection of reintroduced species is crucial to successful re-establishment of viable populations and effective conservation decision-making. Aims The aim of our research was to examine habitat selection by reintroduced Eld’s deer (Cervus eldi) in a human-dominated landscape. Methods The study was conducted during the period from July 2005 to November 2007 in the Chihao region, a human-dominated area located in western Hainan Island, China. Radio-telemetry was used to monitor 15 collared deer to gain their location information. Resource selection functions were used to quantify habitat selection of the study population at the landscape and home-range scales in both wet and dry seasons. Key results At the landscape scale, Eld’s deer showed selection for habitats with scrubland, high elevation, gentle slope, close to water sources and roads. At the home-range scale, Eld’s deer showed selection for habitats with dense forest, scrubland, grassland, low elevation and far away from roads, but they randomly used habitats without special consideration to the distance to water sources. At both landscape and home-range scales, Eld’s deer showed strong avoidance of villages. In addition, Eld’s deer showed increased selection of sparse forests and decreased use of grasslands in the dry season, as compared with the wet season at both spatial scales. Sexual differences in habitat selection existed in reintroduced Eld’s deer. Males showed stronger avoidance to human disturbance, whereas females selected vegetation with higher forage availability but poor hiding cover, especially during the antler-growing period (i.e. wet season). Conclusions The habitat selection of reintroduced Eld’s deer was scale-dependent. As a non-fatal anthropogenic factor, human disturbance had a strong influence on habitat selection of Eld’s deer. They more strongly selected slope habitats at relatively high elevations. However, our results also indicated that the reintroduced Eld’s deer had certain adaptive ability and tolerance to the disturbed environment. Implications This work provides insight into the habitat selection of reintroduced Eld’s deer in a human-dominated landscape. If the essential food resources are available, the regions at a relatively high elevation with low human disturbance can be considered as potential sites of future Eld’s deer reintroduction.