Industrial landscapes promote small carnivore diversity and modulate the predation experienced by small mammals

Abstract

The adverse ecological effects of landscape modification by humans are well known and continue to be studied. Habitat disturbances arising from such modifications result in various levels of biodiversity loss. Amongst mammals, large carnivores are usually the first to disappear, leaving a gap in the trophic pyramid. However, ecologically adaptive small carnivores, free from the effects of intraguild predation can readily fill this gap (i.e. mesopredator release). In highly productive disturbed habitats, small mammals may thrive, reaching densities that can support an abundant and diverse suite of small carnivores. After frequent sightings of serval (Leptailurus serval) on the highly disturbed Sasol Synfuels Operations property, three camera trap surveys were conducted between 2014 and 2015 to study its ecology. The resulting data was subsequently used to study the site’s small carnivores. Multispecies occupancy modelling was used to determine detection (p) and occupancy (ψ) probabilities and estimate species richness across the study area. Site-specific covariates were then modelled against abundance values to identify any correlations. Out of 23 small carnivore species predicted to occur on the site, 11 were detected. Spatially, estimated species richness was highest in disturbed habitats, while it was slightly lower in Grassland possibly due to undersampling. Detection and occupancy probabilities were low (except for serval), with interspecies variations. It is suspected that this is the result of survey bias towards serval. The only covariate showing any significant effect was livestock presence, negatively affecting serval occupancy. Estimated species richness was used as a proxy to identify high- or low-predation areas. Within these areas, three vegetation treatments were selected (low, medium and high cover). Small mammal foraging behaviour under varying predation pressures within these treatments was then studied using the giving-up density (GUD) framework. Differences in GUDs were examined using generalised linear mixed models (GLMM). Small mammal trapping in the study area showed that four-striped grass mouse (Rhabdomys sp.; diurnal) and multimammate mouse (Mastomys sp.; nocturnal) were dominant. Nocturnal GUDs were lower than diurnal, which is interesting as four-striped grass mice are more abundant than multimammate mice. This indicates that density alone does not influence GUDs. Additionally, no significant difference in GUDs was observed between predation and vegetation treatments. I suggest that observed GUDs are linked to individual or combined impacts of interspecific differences in foraging behaviour, metabolic requirements or temporal variations in perceived predation pressure. GUDs also varied between surveys, being lower in mid- compared to early winter. This might be attributable to decreased availability of food and greater metabolic requirements in mid-winter. The results show that modified landscapes (such as the study site) can contribute to biodiversity conservation, especially of small carnivores, the adaptability of which allows them to flourish in disturbed habitats. Under favourable conditions, ecologically flexible small mammal species seem to be unaffected by the risks associated with an abundance of small carnivores. While the processes governing the dynamics of predator and prey communities in disturbed systems are not entirely clear, the conservation potential of such areas cannot be ignored and deserves more attention from researchers.