Abstract
Multi-scale approaches have been used to determine scales at which mammal species are responding to habitat destruction due to invasion, but the impacts of weeds on mammals have not been extensively studied, especially in Africa. Inside the Groenkloof Nature Reserve (GNR), we assessed how mammals are affected by an invasive weed Lantana camara. A series of models were applied to determine the differences in species abundance as well as richness, separated for large and small mammals. When diversity indices were used, an Analysis of Variance (ANOVA) revealed no statistically significant difference between treatments (F(5) = 0.233, p = 0.945) for large mammals. The results of a Generalised Linear Mixed Model (GLMM) showed that vegetation type (Wald χ(2)(2) = 120.156; p < 0.01) and foraging guilds (Wald χ(2)(3) = 76.771; p < 0.01) were significant predictors of large mammal species richness. However, for small mammals, the results of a GLMM showed that only treatment type (Wald χ(2)(5) = 10.62; p = 0.050) was a significant predictor of the number of small mammals trapped. In addition, the ANOVA revealed statistically significant differences in species diversity between treatments (F(5) = 0.934; p < 0.001) and by season (F(1) = 9.122 p = 0.003) for small mammals. The presence of L. camara coupled with other predictors was associated with differences in large mammal abundances and diversity, and differences in how these large mammals were distributed across the landscape. Furthermore, the highest species diversity was found in the spring for small mammals. Therefore, for all the mammals studied, the presence of L. camara negatively affected species abundance, richness, and diversity, as well as how these species were distributed across the invaded and cleared areas.