Accelerometers are a valuable tool for studying animal behaviour and physiology where direct observation is unfeasible. However, giving biological meaning to multivariate acceleration data is challenging. Here, we describe a method that reliably classifies a large number of behaviours using tri-axial accelerometer data collected at the low sampling frequency of 1 Hz, using the dingo (Canis dingo) as an example. We used out-of-sample validation to compare the predictive performance of four commonly used classification models (random forest, k-nearest neighbour, support vector machine, and naïve Bayes). We tested the importance of predictor variable selection and moving window size for the classification of each behaviour and overall model performance. Random forests produced the highest out-of-sample classification accuracy, with our best-performing model predicting 14 behaviours with a mean accuracy of 87%. We also investigated the relationship between overall dynamic body acceleration (ODBA) and the activity level of each behaviour, given the increasing use of ODBA in ecophysiology as a proxy for energy expenditure. ODBA values for our four 'high activity' behaviours were significantly greater than all other behaviours, with an overall positive trend between ODBA and intensity of movement. We show that a random forest model of relatively low complexity can mitigate some major challenges associated with establishing meaningful ecological conclusions from acceleration data. Our approach has broad applicability to free-ranging terrestrial quadrupeds of comparable size. Our use of a low sampling frequency shows potential for deploying accelerometers over extended time periods, enabling the capture of invaluable behavioural and physiological data across different ontogenies.

No abstract

Context Optimal management of invasive species should determine the interval between lethal-control operations that will sustain a desired population suppression at minimum cost. This requires an understanding of the species’ rate of recruitment following control. These data are difficult to acquire for vertebrate carnivores such as the red fox (Vulpes vulpes), which are not readily trapped or observed. Aims To provide a long-term evaluation of the effects of 1080 poison baiting on the abundance and extent of movement of red foxes in a semiarid environment. Methods We used non‐invasive DNA sampling of fox hairs in semi-arid Western Australia where the population was subject to two episodes of aerially delivered sodium fluoroacetate (1080) poison baits within 12 months. Sampling took place at ~45-day intervals and individual foxes were identified by genotyping eight microsatellite DNA markers and a gender-specific marker. Open-population and spatially explicit mark–recapture models were used to estimate the density, apparent survival and movements of foxes before and following baiting. Key results Following a severe reduction in density after baiting, fox density during the ensuing 12 months increased slowly (0.01 foxes km–2 month–1), such that density had only reached 22% of pre-baiting levels ~10 months after the initial baiting. Moreover, recovery was non‐linear as population growth was negligible for 6 months, then exhibited a nine-fold increase 7–9 months after control, coincident with the dispersal of juveniles in autumn. Fox movements between recaptures were on average 470% greater after baiting than before, in line with expectations for low-density populations, suggesting that the probability of encountering baits during this period would be higher than before baiting. Conclusions Baiting with 1080 poison significantly reduced the density of foxes, and the low density was sustained for more than 6 months. Foxes moved significantly further between recaptures after baiting when at low densities. Implications Control programs in this region may be carried out at low frequency to suppress fox density to a fraction of unbaited levels. The intensity of follow-up baiting may also be adjusted downwards, to take account of an increased probability of bait encounter in more mobile foxes.

In Australia, the dingo Canis lupus dingo is the largest terrestrial predator. Dingoes contribute to ecological processes and functions throughout their continental geographic range. Their generalist diet enables daily energetic requirements to be met even in the resource‐limited deserts of central Australia, where irregular rainfall drives extreme fluctuations in mammal assemblages. Given that predation is a primary driver of declining mammal populations and failed reintroductions, it is necessary to understand the dietary intake of the continent's apex predator if conservation outcomes are to be achieved. We performed a meta‐analysis of the mammalian composition of the dingo's diet throughout Australia's arid zone, an area covering c. 5.4 million km2. We used 25 data sets from 11 studies to measure the effects of resource booms and anthropogenic disturbance (pastoralism) on diet. We explored whether the probability of prey species occurring in the diet was related to the size of the prey or to prevailing conditions. We also investigated the similarities in diet between dingo populations from across the arid zone, and how dietary overlap was affected by opposing land uses. Dingoes primarily consumed medium‐sized (35–5500 g) and large (>5500 g) mammals, whereas most small mammals (<35 g) were consumed infrequently, or not at all. The composition of the diet changed dramatically during resource booms to exploit irruptive prey species such as long‐haired rats Rattus villosissimus (native; 144 g) and house mice Mus musculus (introduced; 20 g). During resource booms, the consumption of large mammals decreased significantly, whereas the occurrence in the diet of medium‐sized mammals remained high. The diversity of species consumed by dingoes was significantly lower during boom periods than during non‐boom periods. The dingo's heightened consumption of hyper‐abundant species has the potential to release many non‐target species from direct predation pressure during boom times. This could be advantageous for remnant and reintroduced populations of mammals, which are especially susceptible to predation.

Phenotypic diversity occurs in all sexually reproducing natural populations, and is one of the basic tenets of evolution by natural selection (Fisher, 1958). Intraspecific variation in phenotypic traits such as pelage and body size is a result of the interaction between genotype and environment and is therefore predicted to vary within and between populations throughout time and space. Phenotypic differences within a population are considered important for generating variation in individual attributes such as habitat preference,