2.2.1 Warthog numbers derived from annual transect counts
NGR has eight permanent line transects used for annual game counts, ranging from 1.5 km to 8 km with a combined length of 50.915 km. The transects cut across the four vegetation types. ArcGIS 10.6 (ESRI, Redmont, California, USA) was used to measure lengths of transects based on their position within different vegetation types (Table 1, supplementary material). These transects were used to conduct annual warthog counts and bushpig track counts. To maximise the accuracy of estimates, the eight transects in NGR were walked 16 times annually, during the dry seasons (April-September) of 2017 and 2018 and using distance sampling methodology. The researcher, together with NGR game scouts and students from Tshwane University of Technology (TUT), conducted the counts in the morning (5-8 am) to maximise chances of detecting warthogs because they are difficult to detect later in the day when they rest in the shade. Two observers, each focusing on one side of the transect, counted warthogs observed and, for each encounter, recorded the coordinates of the observer, distance from observer to animal (r), group size (n), and angle of the animal from the transect (Ө). In addition, transect count data from 2013 to 2016 were made available for analysis by Ezemvelo KwaZulu Natal Wildlife (EKZNW). The number of warthogs in the reserve was estimated using the Distance sampling software (Distance V8) using the negative exponential cosine model as the detection function. This function computes the likelihood contributions for off-transect sightings distances, scaled appropriately, for use as a distance likelihood. Only those years with consistent and reliable count data that fitted statistical models were used. Warthog transect sightings were mapped according to vegetation types using ArcGIS V10 and recorded in Excel (Microsoft Corporation, 2018) as follows: a) year of transect counts b) transect number c) vegetation type in which that observation was made d) length of the transect within that specific habitat e) number of warthogs seen and f) a the number of warthogs per km of transect. Descriptive analyses and linear mixed effect regression (Pinheiro and Bates, 2006) were performed. Summary statistics, including mean and standard error of the mean (SEM), were computed.