Amphibians currently represent the most at-risk as well as the least-conserved vertebrate class on earth and factors contributing to these declines are a subject of active research. However, amphibians present a unique challenge to scientists and conservation managers mainly because of their bi-phasic life cycles. Furthermore, it is now generally accepted that each amphibian life history trait is affected differently by anthropogenic threats. In exploring this aspect further, focusing on the amphibians of the southern African sub-continent, frog species were grouped according to their breeding habitats (stream-, permanent aquatic-, temporary aquatic- and terrestrial-breeding groups). First, using a random draw technique focusing specifically on South Africa, at both the national and the biogeographical scales (the latter being defined as sub-regions within the study area identified based on similarities in frog species distributions; see Chapter 2); I evaluate whether areas where different frog breeding-groups occur are characterised by higher levels of anthropogenic threats (human population density, percentage land transformation, percentage protected area, and invasive alien plant richness) than expected by chance. Terrestrial-breeders were often spatially congruent with areas of high threat than expected by chance at both national and biogeographical scales with land transformation and invasive alien plant richness being most significant. Areas where stream-breeders occur were spatially congruent with anthropogenic threats (with alien plants being most consistent) in five of the seven regions examined while protected areas were well-represented in four of the seven regions. Non-significant results were found for permanent and temporary aquatic-breeders at both national and biogeographic scales. Second, I then evaluated the relationship between different frog breeding-group richness and the proportion of protected areas per grid cell (i.e., the proportion of the grid cell that constitutes protected land) and how this relationship is affected by controlling for the wider unprotected matrix. I found that the current southern African conservation portfolio (that is, all proclaimed statutory conservation areas of the sub-continent) only moderately captures anuran distributions. I found a negative relationship between terrestrial-breeding species and the proportion of protected areas. Furthermore, I found that the strength of the model (irrespective of the anuran breeding group) decreased after controlling for the wider unprotected matrix suggesting, at least for now, that amphibian species, in general, are maintaining viable populations within the wider unprotected matrix of the sub-continent. This study indicates that in addition to species with an aquatic tadpole stage, terrestrial-breeding anurans of the southern African sub-continent are in dire need of conservation action.

Submitted to fulfil the requirements for an MSc degree.