|dc.description.abstract||According to the mesopredator release hypothesis, when apex predators disappear from an ecosystem, the number of mesopredators increase and thus predation of prey species increases. Since the extermination of large predators in KwaZulu-Natal (KZN), South Africa, the caracal (Caracal caracal), a feline mesopredator, is presumed to have spread throughout the province. At the same time, populations of the rare blue duiker (Philantomba monticola), a small forest antelope, have markedly declined; probably due to habitat loss and degradation, yet possibly also by increasing numbers of mesopredators. As forest patches become smaller and more isolated, the blue duiker’s ability to disperse and find refuges becomes limited. Any additional predation threats due to increased probability of encounters between blue duiker and caracal could drive local populations to extinction. However, quantitative data on current broad-scale distribution ranges and habitat requirements of both species are scant. The aim of this study was to use an environmental niche modelling approach to test the hypotheses that the caracal has expanded its distribution range; that its suitable habitat space (i.e. Grinnellian niche) overlaps considerably with that of the blue duiker; and that human impact affects suitable habitat space.
I compiled a comprehensive occurrence database for caracal and blue duiker using historical data, recent sightings and survey responses from conservancies and protected areas throughout KZN. I then used niche-based environmental models (Maxent) to map the potential suitable habitat space of caracal and blue duiker, based on the occurrence data and different combinations of environmental and human impact variables. I calculated two niche overlap indices to quantify habitat niche overlap and compared these to 1000 expected overlap indices in a null model to test for niche equivalence.
After spatial filtering, my database comprised 94 caracal and 90 blue duiker occurrence points for 1994 to 2013. For both species, vegetation biome, specifically forests, was the most important variable, with rainfall second most important. Approximately two-thirds of KZN was moderate to highly suitable habitat for caracal and one third was moderate to highly suitable habitat for blue duiker. The potential suitable habitat maps matched mostly with published distribution ranges for both species. Nonetheless, occurrence data within comparably lower habitat suitability, suggest caracal may have expanded their range further towards the north east of the province. The niche overlap scores were considerably high (D = 0.683 and I = 0.924), yet the caracal and blue duiker niche models were not identical, which can be expected of a generalist carnivore with a broad niche compared to the narrow niche of a specialist herbivore. At the highest threshold of habitat suitability the overlap between caracal and blue duiker covered 7% of KZN with 20% of the overlap occurring in 70% of the forest biome. The overlap consisted primarily of a large region in the midlands, smaller regions in the south and forest patches throughout KZN. Human impact variables contributed much less to the models than environmental variables and although the negative impact was mostly < 10%, the effect was widespread over ~60% of KZN for both species.
My results confirm that the caracal distribution range has expanded historically from the western highlands to the south coast of KZN and thus overlaps considerably with the blue duiker distribution range in the midlands and southern coastal forests. Although the proportion of blue duiker in caracal diet reported in studies elsewhere is low (5% or less), any new threat to a blue duiker population already under pressure from a variety of other threats, could lead to an irrecoverable decline. Thus further studies investigating the effects of the expanding caracal population on the diminishing blue duiker population and other vulnerable prey species are warranted.||en_US