Abstract:Habitat assessing and prediction is important when attempting to protect endangered species. In the Ebinur Lake National Nature Reserve, Xinjiang, A total of 92 recorded GPS coordinates showed that Cervuselaphus was present in the reserve in September 2013 and October 2014. We grouped 23 habitat predictor variables into three classes-terrain, vegetation structure, and climatic factors. We used the MAXENT model to predict the potentially suitable autumn habitat distribution for C. elaphus and determined the contribution each habitat factor made to the distribution characteristics. The accuracy of our prediction models was accessed by the area under the curve (AUC) values for a receiver operating characteristic (ROC) curve, and the validation showed that the results had high average AUC of 0.976. The results of a Jackknife test indicated that the maximum temperature in the warmest month had the strongest influence on autumn C. elaphus habitat suitability, followed by altitude, annual precipitation, mean diurnal range, and mean temperature during the warmest quarter. Vegetation type and slope had little effect on habitat distribution. The simulated habitat was divided into four classes-the most suitable habitat, moderate suitability habitat, low suitability habitat, and unsuitable habitat. The most suitable habitats for C. elaphus were mainly distributed in the northern parts of the study area, and habitats with moderate and low suitability were in the marginal areas of the most suitable habitats. In contrast, the western and eastern parts of the study area were classified as unsuitable habitats for C. elaphus. This study provides information on the actual distribution of C. elaphus in the Ebinur Lake National Nature Reserve and identifies important indicators that could be investigated by future studieson this species and its habitat.