Environmental predictors for spatial variation in temperature in a South African savanna and their predictive value for distribution of Southern White Rhinoceros (Ceratotherium simum simum)
MetadataShow full item record
High temperature extremes are projected to increase in frequency and severity in southern Africa. This could negatively impact large animals more than smaller animals. The southern white rhinoceros is experiencing population declines across southern Africa, and efficiency of protection efforts are needed. Adverse effects from a warming climate could further decrease the potential for rhino population sizes to grow. If rhino distribution is limited by high peak temperatures, spatial variations of temperature in the landscape could be a predictor for rhino distribution. Certain landscape features are assumed to be potentially influential with regard to local air temperature. This research investigated the effect of canopy cover, vegetation density and dominant slope aspect on local air temperature measured at 160 cm height in the Kempiana reserve in South Africa. Subsequently, rhino distribution based on these and additional landscape features (elevation, waterhole availability and dominant vegetation type) was modelled. Patches with high canopy cover, low vegetation density and south-facing slopes were hypothesized to be cooler than patches with no canopy cover, high vegetation density and north-facing slopes, respectively. During relatively hot days, rhinos were hypothesized to predominantly be in areas with landscape features associated with lower temperatures. 24 iButton thermometers were used to measure temperature in 2 groups of 4 landscape features in separate experiments: dense versus sparse vegetation and closed versus open canopy in the first experiment, south-facing slopes versus north-facing slopes in the second experiment, and east-facing slopes versus west-facing slopes in the last. Distribution patterns of the white rhino in Kempiana were modelled on a scale of 500*500 meters against canopy cover, vegetation density, elevation, dominant aspect, dominant vegetation and waterhole availability using Generalized Linear Mixed Models. 2 GLMMs were used, one with presence-absence data, and another with presence-only data. This was done for hot-season data, comparing rhino location data of cooler days with that of hotter days. The rhino location data was collected by spotter plane in irregular intervals during the years of 2014-2019. Patches under tree canopy were on average 0.5°C cooler than intercanopy patches. Rhino distribution did not show different correlations with any of the landscape features between hot and cold days. In the study area as a whole, rhino density in the cold period was twice as high as in the hot period, suggesting larger scale limitations to rhino distribution as an effect of temperature. This could be a finding to investigate in future research.