Exploring spatio-temporal variations of habitat loss and its causal factors in the Shenzhen River cross-border watershed

Understanding the mechanisms that underlie habitat loss is fundamental to nature conservation. Habitat loss and its related factors are location dependent, but their spatial variations across geographical space are poorly understood. This research attempts to explore the spatio-temporal variations of human and biophysical factors of habitat loss in the Shenzhen River watershed, a border region shared by the Hong Kong Special Administrative Region (SAR) and Shenzhen in China. Habitat loss was quantified through the comparison of historical land cover maps in 1988/1998/2008, derived from the Landsat imagery. The different degree of habitat loss is further considered by referring to loss of ecosystem services during habitat alteration or destruction processes. In order to examine how influencing factors change with space, the model of geographically weighted regression (GWR) was employed. The GWR model was proved to be effective in revealing a spatially varying relationship, and has better performance compared with the global regression model. The results also provided the detailed site information of the different roles of the related factors across the different locations. Various factors, especially the human-related ones, played a significant role in determining both the “rate” and “location” of habitat loss, and the influences also varied during the entire study period. These results could provide comprehensive information for future habitat conservation and cross-border watershed management.

► We explored the spatio-temporal variations of habitat loss and its related factors.
► A cross-border watershed was used to illustrate the distinct variations.
► The GWR model could effectively reveal a spatially varying relationship.
► Human factors determined both the “rate” and “location” of habitat loss.
► Results can provide comprehensive information for future habitat conservation.