Objective To understand the relation between environmental variables and suitable region of T.ciliata var. pubescens, which can provides reference to resource conservation, introduction and plantation for this plant.

Method An MaxEnt model, because of the advantages of using presence-only data and performing well with small sample sizes, incomplete data and gaps, was employed to simulate the habitat suitability distribution. And the area under the receive operating characteristic curve(AUC)were used to examine the model's accuracy, the AUC is larger and predictions is more accuracy. So, based on the location in Yunnan province of T.ciliata var. pubescens, the MaxEnt was used to set up it's distributional model of potential habitat, one topographic variable and the six main bioclimatic variables influencing species distribution.

Result The results show that the Mean training AUC and Mean test AUC were 0.891, 0.885, respectively, it illustrated that the prediction of T.ciliata var. pubescens's suitable habitats are reliable. CV of precipitation and Precipitation of the coldest quarter are significant factors, Range of annual temperature, Mean temperature of the coldest quarter, Precipitation of the wettest quarter, Precipitation of the driest quarter are secondary factors to determining T.ciliata var. pubescens's suitable habitat. Habitat suitability for current and future climate warming(2050S, 2070S) under scenarios RCP2.6 in Yunnan province and China were calculated. The result predictions the T.ciliata var. pubescens' suitable habitats change under different climate in intuitive and quantitative.

Conclusion The habitat suitability of T.ciliata var. pubescensin Yunnan Province and China are predicted to deteriorate in slight with globalwarming.