Abstract:
Burrowing mammals, i.e., mammals that dig in the ground for shelter creation, food, and nesting, play an important role in the ecosystem, since their burrows support plant and animal assemblages that differ from that of the surrounding environment. Previous research suggests that environmental parameters, such as soil texture and vegetation, influence the distribution of burrows in the landscape. However, burrowing mammals are not only influenced by soil texture but also influence the soil texture. The current study investigated the two-way interactions between aardvark (Orycteropus afer) burrowing and the characteristics of the surrounding landscape at Rietvlei Nature Reserve, South Africa. First, soil type, geology, rockiness, vegetation, distance to water lines, and distance to roads were explored as potential determinants, both of burrow presence (using Fisher exact tests), as well as burrow density (using Generalized Linear Models (GLZs)) in quadrats. Second, spatial clustering of aardvark burrows was explored using Average Nearest Neighbour Analysis, Multi-distance Spatial Cluster Analysis, and Optimized Hot-Spot analysis in ArcGIS Pro. Finally, mean, sorting, skewness, and kurtosis were calculated to compare soil texture between burrow mounds and control samples. Controls were taken 2 meters away from burrow mounds and represent areas not recently burrowed by aardvark. The study found no association between burrow presence in a quadrat and any of the environmental determinants investigated. In addition, none of the potential determinants was a significant predictor of burrow density within quadrats. Cluster analysis results from three quadrats with highest number of burrows (Q2, 3 and 28) show the dispersion of burrows in two (Q3 and 28) of the three quadrats used, whereas the third (Q2) quadrat shows clustering at distances less than 28 m and dispersion at distances greater than 43 m. The study further found that aardvark burrowing affects the size and distribution of soil particles but not sorting. The mean, kurtosis, and skewness of soil samples from burrow and controls differ significantly (P>0.05). While sorting of soil samples from burrows and controls did not differ significantly (P<0.05). From the results of GLZs, it is possible that other factors, instead of the environmental parameters explored here, are driving aardvark distributions. These may include more nuanced differences in plant cover and grass height between quadrats. Quadrats are, for example, rotated when it comes to managed veld burning. This affects grass cover, which in turn may affect aardvark burrowing. It is thus recommended that this be investigated
Keywords: aardvark, burrow distribution, burrowing mammals, cluster analysis, diggings, Orycteropus afer