Assessment of Environmental Variables in the Ensemble Model and
Their Impact on Species Distribution
In the construct of this study, nine environmental variables were
integrated into the predictive modeling framework. Of these, the
precipitation during the coldest quarter (bio19), the Normalized
Difference Vegetation Index of March (NDVI0321), and the topographic
variable of slope emerged as pivotal, each contributing in excess of
10% to the model’s predictive capacity, as elucidated in Table 2. This
denotes their substantial relevance and influence within the ecological
modeling construct. In contrast, the other variables demonstrated a
relative importance below the 10% benchmark, suggesting a more marginal
role in the model’s overall predictive accuracy. Notably, the leading
trio of environmental parameters exhibited a significant positive
correlation with the probability of habitat suitability for Manis
pentadactyla (Chinese pangolin), as illustrated in Figure 2. In
comparison, the secondary environmental variables did not exhibit marked
correlations with the species’ distribution probabilities, underscoring
the differential impact of various ecological factors on pangolin
habitat suitability.