The influence of patch and landscape attributes on species richness and occurrence of terrestrial mammals in Hong Kong

Abstract

Studies on forest loss and fragmentation have found that proportion of forest cover, forest patch size, shape and isolation, and matrix quality affect mammal species richness and occurrence in tropical landscapes. However, how do these characteristics influence mammal species in a landscape like Hong Kong where the deforestation process has been lengthy and interrupted by waves of forest contraction and expansion, and that has resulted in a mammal community of habitat generalists? We have a good understanding of the mammal distributions in Hong Kong; however, we know very little about the factors affecting this.

This study investigated if forest at the patch- and landscape-scale influences the species richness and relative abundance of mammal species in Hong Kong. Camera trapping data from a separate, Hong Kong-wide study, undertaken from 2000-2003 (which recorded 20 non-flying mammals species), was used to evaluate the influence of patch metrics (patch size/area and shape) and landscape attributes (forest cover, number of forest patches, mean inter-patch isolation distance, matrix quality and urban developed area cover) at two spatial scales (100 hectare buffer and 450 hectare buffer) on mammal species richness and relative abundance (using occurrence index). Patch metrics were measured from 32 focal patches and landscape attributes were measured from 21 landscape buffers, across Hong Kong.

At the patch-scale, patch shape and size/area were not found to have an influence on extant and common mammal species richness or relative abundance. This is probably because the most sensitive species have already disappeared from Hong Kong, and the wide range of habitats in the matrix may be sufficient for the extant mammal species to survive and reproduce outside of forest patches.

At the landscape-scale, Melogale moschata, Felis catus and Prionailurus bengalensis were associated with landscape buffers comprising a lower patch number and reduced patch isolation, indicating that whilst not forest obligates, these species may have some reliance on forest. In addition, the study found greater records of presence in forest habitat compared to matrix habitat for Hystrix brachyura, Macaca mulatta, Paguma larvata, Viverricula indica, Sus scrofa and Prionailurus bengalensis. Therefore, while the mammal species are likely to be able to use the matrix for survival, forest habitat is still likely to be important in Hong Kong for its terrestrial mammals.

Matrix quality was not found to influence mammal species richness or relative abundance. The use of the matrix, which predominantly comprises natural habitats (e.g. shrubland and grassland) in Hong Kong, by the mammal generalists may help to mitigate the negative effects of forest fragmentation.

The results of the study appeared to be strongly influenced by the geographic locations of the forest patches and landscape buffers. As the obvious patterns associated with patch metrics and landscape attributes were not strongly apparent, further examination of the mammal communities, particularly within different regions may be necessary. In addition given the rarity of some of the ‘outlier’ species omitted from the analyses, association or possible reliance on forest habitat/patches may make them more vulnerable. Therefore, further assessment of rare species is recommended.