Ochotona ladacensis
Univariate models - At the Ladakh spatial scale, many variables related to rocks (lower percentage cover of rocks, lower percentage of large-sized rocks, lower number of large-sized rocks), heights of plants (low herb and forb height, tall cushion plants) were important in determining the presence of the species (Table C3). At the South-East Ladakh scale, only topographic variables such as elevation and slope significantly predicted presence (Table C3). All univariate models run are presented in Supplementary information (Table S9, Table S10).
Multivariate models - At both spatial scales, the purposeful modeling approach best predicted the occurrence of the species (Table 3). Variables that strongly predicted the absence of the species and those that predicted their presence were important in explaining where the species was most likely to be found. At both scales, the presence of OL was associated with a lower cover of rocks of different sizes, a lower number of large-sized rocks, a higher relative percentage of cushion plants, lower heights of shortgrass-sedges and shrubs, high elevation, and a more prominent aspect.