Study On The Habitat Selection Of Plateau Zokor (Myospalax Baileyi) In Northern Qilian Mountain

Plateau zokor is special species in the Qinghai-Tibet plateau and plays the keyrole in the alpine meadow ecosystem. The study about the habitat of Plateau zokor isquite important for controlling rodent and biodiversity conservation on the grasslandecosystem. From2012to2013, the structure of plants community (herbage coverage,height, density, frequentness and biomass), terrain (aspect and slope) and soilcondition (compaction, moisture and pH) had been studied in habitat of Plateau zokorin different spatial scale in northern Qilian mountain. In the study, the selectionfactors of rodent habitat had been cleared with unary linear regression equation,resource selection functionals and independent sample T test according to differentstudying scale. The results are as follows:1. Under landscape scale（3types grassland in northern Qilian mountain）, theselection factors of zokor habitat are different.On alpine grassland, there was a significant correlation between the soilcondition, plants and the zokor population density. Except that the soil temperature in0~10cm depth has negative correlation with zokor population density（P<0.05）, thesoil compaction, temperature, moisture and pH in the three depths (0~10cm,10~20cm,20~30cm) had significant correlation with zokor population density（P<0.01）.The coverage of grasses has significant negative correlation with zokorpopulation density and the coverage of forb plants has significant positive correlationwith zokor population density（P<0.05）, the species richness has significant positivecorrelation with zokor population density（P<0.05）.On alpine shrub meadow, the soil compaction had significant negativecorrelation with zokor population density in three depths (0~10cm,10~20cm,20~30cm)（P<0.01）. The soil temperature in two depths (0~10cm,20~30cm) hadsignificant negative correlation with zokor population density（P<0.05）. The coverageof grasses had significant positive correlation with zokor population density and thecoverage of forbs had significant negative correlation with zokor population density （P<0.01）.The soil temperature, moisture and plant species richness had nosignificant correlation with the zokor population density.On alpine meadow, the soil density had significant negative correlation withzokor population in different soil layer（P<0.01）.The soil temperature in the depths(0~10cm) had significant negative correlation with zokor population density（P<0.01）.Other soil properties had no significant correlation with the zokor populationdensity.The vegetation functional groups had no significant correlation with the zokorpopulation density,but the plant species richness had significant correlation with it（P<0.05）.In conclusion, the soil compaction in three depths (l0~10cm,10~20cm,20~30cm)and the soil temperature in0~10cm were the main selection factors of zokor habitat.2.Under home-range scale(a hill about40hectare in Zhuanxixiulong vliiage ofTianzhu county), the six factors including biomass, height, coverage of sedges, pH ofsoil, biomass of sadges and forbs had significant influence on the selection of zokorhabitat. The model with resource selection functionals for the zokor habitatselection is Logit(p)=35.765－3.176×soil pH－1.65×Simpson index－1.33×slope+0.869×Shannon-wiener index－0.368×soil moisture+0.079×forbs coverage－0.025×elevation－0.004×soil compaction. Which slope and forbs coverage withstatistically significant (P <0.05).The fitting model is P=elogit(p)/（1+elogit(p)）.Thepredict rate of model was85.4%,Multiple correlation coefficient was0.419(Nagelkerke R2=0.542). The fitting model works well.3. Under microhabitat (a special habitat region of Plateau zokor about2.2hectareon alpine meadow of Tianzhu). The coverage of vegetation functional groups,plantdiversity didn’t differ between home range and non-home range of Plateauzokor(P>0.05).The soil compaction（1105.5MPa）in the zokor home range wassignificantly lower than non-home range（3194.7MPa）（t=1.575，P<0.01）.Meanwhile,the soil pH（7.596）in home range was also significantly lower than non-home range（7.713）（t=-4.658，P<0.01）. The area of home range increased fromAug. to Oct.The rank of the home range size as follows: subadult <adult female <adult male.