2.4. Aim III: resting and foraging behaviour

To assess disease-induced behavioural differences in wombats, triaxial accelerometer data loggers (AX3 Axivity) were deployed on five adult, free-living wombats in NNP (one healthy, one with ambiguous signs of early mange and three mange-infected). Wombat trapping and processing followed the protocols outlined above (see §2.3). All loggers were set out within 24 h of each other, on 20–21 April 2015. Three loggers were successfully retrieved: one from a moderately mange-infected wombat (female,W006, mange severity score 2.7), one from a wombat with ambiguous signs of early mange (male, W009, mange severity score 0.57, referred to as ‘early’), and one from a healthy wombat (female, W002, mange severity score 0.5). Despite having a similar average mange score to W002, W009 had ambiguous signs of early mange at capture (with confirmed mange in subsequent visual surveys), and thus, was conservatively classified as early-stage mange. The loggers recorded at 50 Hz from noon on 22 April 2015 to varying times on 18 May 2015.

Traces of the three cardinal axes of the accelerometer were visualized in Somnologica Studio 3.0 and activities were defined based on stereotypic patterns. To calibrate real-time wombat activities with accelerometer recordings, Axivity data loggers were also deployed on two healthy, captive wombats. Based on captive wombat accelerometer recordings, six main activities were identified: digging, steady walking, scratching, running, slow walking/grazing and inactivity. In addition, there were four unidentified activities and an activity categorized as ‘restlessness’, which was defined as a period of brief, unrecognizable activity interrupting periods of inactivity. Activities were manually scored in 3 s epochs (28 800 epochs per day), and each epoch was categorized as the activity that endured for the majority of that epoch. Activities were scored for four, 24 h periods (at 3 day intervals, excluding the first 72 h post-anaesthesia: 24 April, 27 April, 30 April, 3 May) for each wombat (115 200 epochs individually scored per wombat or 345 600 total).

Inactivity and foraging behavioural data were quantified in three ways: total number of episodes per behaviour (per day), average duration of activity bouts and percentage of day spent engaged in either state (for daily activities and averages, see the electronic supplementary material, D and E). For the average duration of activity bouts, bouts were defined as either an isolated epoch of activity (3 s) or consecutive epochs of the same activity (greater than 3 s). Differences in the number of episodes, bout durations and the proportion of time spent engaged in inactivity and foraging were analysed among wombats using ANOVAs. Inter-individual differences were assessed using a multi-comparison of means (Tukey contrasts). Plots of daily wombat activity from 12.00 on 22 April to 05.00 on 8 May show differences in circadian cycles (electronic supplementary material, F).

To assess whether mange-infected wombats can cope with the metabolic pressures of mange, realized feeding rates (energy consumed (kJ d⁻¹), derived from behavioural data) were also calculated. The average proportion of the day spent foraging (for the healthy and late-stage wombats) was used, in combination with bite rates derived from Simpson et al. [19] for healthy wombats and mange-infected wombats, to determine the number of bites taken per day. The amount of dry plant matter per wombat bite was assumed to be 0.015 g, along with 7.4 kJ metabolizable energy per gram of dry plant matter [51].