Environmental factors influencing the distribution of bats (Chiroptera) in South Africa

Doctoral Thesis


Permanent link to this Item
Journal Title
Link to Journal
Journal ISSN
Volume Title

University of Cape Town

Environmental factors limiting the distribution of 37 of the 56 bat species in a warm temperate region (South Africa) were determined using GIS software and the Maximum Entropy modelling technique (MaxEnt). Undertaking such a study in a warm temperate region like South Africa is essential as the outcomes of this study could inform our general understanding of distributions of other animals in other parts of the world. Hypotheses related to the ecological niche-based characteristics of species were tested to identify the most important variables influencing the distribution of South African bats and to predict the probability of occurrence for bat species in South Africa. A database that included locality records for bat species from different museums in South Africa was compiled and then combined with the researcher's own data for the Northern Cape Province as there was insufficient knowledge of bat distributions in this province. A total of 23 environmental variables were considered, of which 20 were downloaded from the WorldClim database as potential environmental variables influencing the contemporary distribution of bats in South Africa based on previous studies that use environmental variables from WorldClim to predict the distribution of species. The environmental variables were grouped into broad categories, temperature, precipitation, and biophysical (i.e., vegetation biomes, land use/land cover, geology) variables. As predicted, taxonomic affiliations appear to have no bearing on which factors influenced the geographic distribution of South African bat species. The distributional limits of even closely related species within the same genus appear to be influenced by disparate environmental factors. Geology appeared to be the most important limiting factor for 15 of the 37 species, all of which are known to use roosts associated with geological features (i.e., caves, mines and rock crevices). Land use/land cover influenced the distribution of six bat species most of which are known to use human structures or domesticated crops as roosting sites. Roost availability thus appeared to be an important factor limiting the distribution of bats. The distribution of only one South African bat species, the endemic Rhinolophus capensis, was associated with a biome as being the most important predictor variable. Temperature variables appeared to be the most important factors influencing the distribution of 12 of the 37 species of bats in South Africa. This might be linked directly to the roosting ecology and thermoregulation capability of each species and their need for hibernation and/or torpor. Precipitation parameters were the most influential in the distribution of 9 of the 37 South African bat species whose distribution is centred towards the wet east of the country. This could probably be linked to its effects on the availability of food in the form of fruit or insects. However, the results of this study should be interpreted cautiously. The majority of the environmental variables employed in this study to model the distribution of bats, were correlated to some degree, which could affect the contribution of an individual environmental variable to the performance of a model. Furthermore, certain bat species included in this study have their centres of distribution ranges further north in Africa and have only marginal intrusion into South Africa's political boundary, which means that only a portion of the distribution of these species is modelled; this could yield erroneous results that might not be transferable to other parts of the ranges of these species. Finally, field verification of the occurrence of species in areas where they are predicted to have a high probability of occurrence is crucial in order to verify the reliability of the models.