The
training AUC and test AUC of the distribution model of red panda were
0.96 and 0.94 respectively. Similarly, the training AUC and test AUC of
the potential bamboo species. were 0.93 and 0.92 respectively (Fig. 2).
Both training and test AUC were higher than the random that indicated
the model performed better than random. In western Nepal, potential
habitat of red panda and bamboo distribution area were estimated to be
3,222 km2 and 6,294 km2respectively. At total of km2 area was estimated as
the potential habitat when combing overlapping area among distribution
of Himalayan red panda, bamboo species and forest cover. (Table
2). The distribution model
predicted more than two third of the total habitat available in only
three districts, namely Jajarkot, Jumla and Kalikot, while the remaining
five districts constituted less than one third of the habitat (Table 2).
Table
2 Potential suitable habitat for red panda based on predicted red panda
habitat, bamboo distribution, and forest cover