Abstract:
The current study assessed the ambient temperatures, and those selected, by captive adult Nile crocodiles on a commercial farm in South Africa. Non-invasive data capture techniques were developed to ensure the crocodiles natural behaviours were not disrupted or altered. Thermal and climate data, over summer and winter seasons, were collected from local weather stations, an on-site Internet of Things system, and a Mavic 2 Enterprise Dual drone. The method developed in this paper transformed relative thermal maps (produced by a DJI Mavic 2 Enterprise Dual drone) into a predictive model in which temperatures were derived to within 2.6°C per pixel of a processed orthophoto. Crocodile thermal and behavioural data were extracted from the drone imagery and juxtaposed with climate and thermal data from the pen. The greatest number of crocodiles were counted during early morning winter flights and the lowest number during late afternoon summer flights. Material (concrete, water, nest, grass/sand) selection by crocodiles varied with season, time of day and daily climatic conditions. Crocodile back temperature (10.2–49.6°C, µ = 30.4°C) ranges fell within those of their positional/environmental (10.6–66.6°C, µ = 28.7°C) temperature range selections. Strong, positive, significant correlations were found between crocodile back temperatures and positional temperatures for both winter and summer seasons, highlighting ectothermy. Application of this methodology on a commercial crocodile farm facilitated the inspection of potential shortfalls of the pens design from a thermal perspective, as well as suggestions for improvements that would ameliorate crocodile thermal discomfort (relating to hyperthermia).
