Choice Of Wintering Habitats Of Eastern Inner Mongolia Steppe Wolf (canis Lupus)

The quantity of China's wolf population was very few, it mainly distributed in our country northeast, northwest and Inner Mongolia ect,which was already included in the CITES appendix II. In recent years,along with the habitat environment worsening and excessively hunting, its population still was continuing to reduce, the distribution area also unceasingly reduced, therefore,it had been extremely essential to effectively protect wolf's wild population and its habitat.From Line transects of systematic sampling were used to characterize correlation habitat variables of the whole study area before winter. The distance between any two nearest transects was 1 km and grids (1 km×1 km) were obtained every 1 km along transects. Random transects were then applied to count the number of the tracking of wolves and record the actual positions they found after winter. Random transects as a whole traversed all the grids. November 2005 to March 2006,we investigated the wolves (Canis lupus) distribution in SW Dailai Lake, Inner Mongolia, China, as well as analyzing their habitat selection characteristics in winter with Logistic regression model.1. In the study area we fonud 1 wolf group and obtained 37 wolves trails (chain).The grids with presence of wolves were defined as active grids and valued 1 , the grids with absence as inactive ones and valued O.Morans I coefficient of dependent variables residuals was the lowest when the lag was 1 km. Therefore , the grid size was 1 km×1 km and 168 grids were obtained(57 active grids and 111 inactive ones).2. Eight ecological factors potentially related to wolves occurrence,i e,nce to the water resource,vegetation height,captive sheep abundance,captive horse abundance,captive cattle abundance,distance to the resident place and hiding cover showed significant differences between active and inactive grids. Forward elimination stepwise logistic regression was conducted with the remained variables as independent variables. Finally , regression equation with the lowest AIC_C value was regarded as the optimal model. The model could be formally expressed as : logit (P) = -8.949 + 3.190×captive sheep abundance + 0.002×distance to the resident place+0.095×hiding cover ( P was the probability of the occurrence of wolves).The model suggested that habitat selection of wolves was positively related to captive sheep abundance,hiding cover and distance to the nearest resident place.3. the optimal cut-off point is 0.5.The correctly classified ratio of model is 89.5%, the correctly classified ratio of active grids is 87.5%,the correctly classified ratio of inactive grids is 90.6%.The model could accurately predict theoccurrence of the habitat selection of wolves.4. The model predictions for habitat selection are very important to make plan for Canis lupus' protection and maintain species diversity in this area further.