Changing land cover and small mammal habitats: Implications for landscape ecological integrity

Abstract

This paper adds to the landscape ecological integrity (LEI) literature by emphasising the need to move beyond the assumption of a linear relationship between fragmentation metrics and ecological integrity. This study argues that such a relationship is weak or even non-existent when not associated with specific species or ecological process. Thus, using the Greater Accra Region of Ghana as a case, this study assessed the quality and spatio-temporal distribution of terrestrial small mammal habitats, as influenced by changing land cover, highlighting the implications for LEI. First, the study used fragmentation statistics to model changing land cover from 1986 to 2020. This was followed by the use of the Leo Breiman's random forest algorithm to predict small mammal distribution in the same period. The results showed that normalized difference vegetation index (NDVI) has a high predictive importance on small mammal distribution. Decreasing the vegetated areas does not necessarily lead to a decrease in species habitats. However, the decreasing forest areas and shrub-lands leads to a statistically significant increase in small mammal population in built-up areas while their population reduced within farmlands. The paper concludes that an increase in small mammal habitats in built-up areas does mean improved landscape ecological health, considering that these built-up areas are not natural abodes.