Vetnuus | September 2025 21 Doing this could have acclimatised the various species visiting the waterholes to the presence of the cameras, as the researcher did not observe any avoidance behaviour prior to, or during the study, when time was spent at the various waterholes. Camera traps were checked every seven days for damage, to download imagery, clear memory cards, replace batteries, and ensure that they were still properly positioned. Data collected from the camera traps provided information about the number of days each waterhole was effectively photographed. Camera traps were positioned to maximise the percentage of waterhole edge covered, with additional camera traps used to cover larger waterholes. For waterholes that had more than one camera trap, dates and times on the various camera traps were synchronised so that when viewing images from various cameras, it was possible to identify the same animals i.e. images of animals taken within a few minutes of each other on different cameras could be isolated and the animals identified as being the same ones. Also, when there was more than one camera at a waterhole, the cameras were placed in such a way that there was a slight overlap of their field-of-view, resulting in animals moving from one camera’s field-of-view to another camera’s field-of-view being identified as the same animals. Invariably, there were occasions when camera traps had to be removed for repairs or replacement. To cater for days when waterholes did not have cameras, a sightings-per-day value was calculated for all cameras by dividing the number of sightings at a waterhole by the number of days the waterhole was effectively monitored. This allowed for direct comparison of visitation data between waterhole types. The number of camera traps deployed at Singwe Big Dam and Oxford Big Dam varied depending on available surface water, which was influenced by rainfall. When these dams were very full, more camera traps were placed out than when the dams were empty and had less surface water. Photographs from the camera traps were digitally dated and time-stamped for accuracy when doing comparisons. Information obtained from the camera traps included the time of day that elephant, black rhino and white rhino utilised the various waterhole types, social grouping types frequenting the waterholes, number of animals, and the duration of stay at waterholes. Elephant group types identified were bachelors, bachelor groups and breeding herds. Black rhino group types were bachelors, bachelor groups, cows, cow groups, bull and cow groups and unknown adults. White rhino groups were the same as for black rhino, except for cow groups, which were not observed during this study, as this group type likely used waterholes without camera traps. A single visitation to a waterhole was defined as a single photograph or sequence of photographs isolated by a minimum of five minutes from any other photographs according to the timestamp. The five-minute interval is based on direct observations made by the researcher. The timings of visitations (time of day) and duration of stay by elephants and rhinos were used to determine waterhole preferences and utilisation patterns. Duration of stay at a waterhole was calculated by subtracting departure times from associated arrival times at waterholes. Data analyses Binomial tests with confidence intervals were used to test for waterhole preference by the different species and their social group types [46]. The rationale for using a binomial test was that it is an exact test, providing a rigorous examination of the observed outcomes against expected probabilities for two distinct possible outcomes, selection (preference) or rejection (no preference) [46]. The precision of the binomial test is particularly beneficial when evaluating waterhole preferences, as it ensures accurate statistical inferences, even in scenarios with small sample sizes or discrete datasets. The use of binomial tests in this study extends beyond a general preference assessment. By conducting tests across different seasons (wet and dry) and daily periods (night, morning, midday, and afternoon), we capture a comprehensive view of waterhole utilisation patterns. Only significant results with a lower 95% confidence interval above the chance threshold of 25% were reported, ensuring that detected preferences were not due to chance. The 25% threshold was established based on the equal probability of selection among the four distinct waterhole types. Additionally, a series of Poisson General Linear Models (GLM’s) were run to determine the impact of five predictor variables, Season (1 Wet or 2 Dry), Daily Period (1 Morning, 2 Midday, 3 Afternoon, 4 Night), Waterhole type (1 Earth Dam, 2 Concrete 3 Pan, 4 Reservoir, 5 Trough), Waterhole Size (1 Small, 2 Medium, 3 Large), and Social Group Type (1 Bachelor, 2 Bachelor group, 3 Cow, 4 Cow group, 5 Cow and calf, 6 Unknown adult, 7 Breeding herd, 8 Bull and cow) on the frequency of visits to waterholes by the three study species [47]. Poisson GLMs were chosen because the response variable represented count data, characterised by discrete, positive integer values, and this model type accounts for non-normality and variance that is not constant [47]. All predictor variables were treated as fixed effects, with no need to control for random effects, thus eliminating the necessity for generalised linear mixed models [47]. Grouping variables used in the GLMs included Season, Daily Period and Waterhole Type. Table 1. Olifants West Nature Reserve artificial waterhole types with their local names, waterhole type, capacity, size categories and number of camera traps placed at the waterholes Artificial Waterhole Characteristics Name Type Capacity (m3) Size category Nr of camera traps Ngala Earth dam 100–1 000 Medium 1 Oxford big dam Earth dam > 1 000 Large 2/3 * Singwe big dam Earth dam > 1 000 Large 3/4 * Leopard’s view Concrete pan < 100 Small 1 Singwe bush camp Concrete pan < 100 Small 1 Toni’s dam Concrete pan < 100 Small 1 Nyala Reservoir 100–1 000 Medium 1 Nzulwini Reservoir 100–1 000 Medium 1 Van Wyk’s Reservoir 100–1 000 Medium 1 Nyala Trough < 100 Small 1 Van Wyk’s Trough < 100 Small 1 * The number of camera traps deployed at Singwe Big Dam and Oxford Big Dam varied depending on the water surface area, which was affected by rainfall. Research Article >>>22
RkJQdWJsaXNoZXIy OTc5MDU=