Abstract
Animals involved in studies using tracking devices or data loggers are potentially subject to a range of instrument effects which have a negative impact on survival, reproduction and behaviour. The current recommendation is that device weight shouldn’t exceed 3% of body mass for aerial species and 5% for terrestrial species. However, the 5% rule of thumb has little or no empirical basis and recent studies have found that transmitters weighing less than 3% body mass may still have impacts on energy expenditure and other behavioural parameters. Studies using telemetry and data loggers on domestic cats Felis catus have used collars carrying Kittycams, VHF and GPS tags ranging in weight between 30g - 125g and followed to 5% rule. These studies assume that the device carried is not affecting behaviour, however, this has not been tested quantitatively. The purpose of this research was to examine possible impacts of wearing collared devices of different weights on domestic cat movements and behaviour. Movement data were compared for cats wearing three different GPS collar weights: (1) light, 30g, (2) a medium-heavy, 80g and (3) heavy, a collar weighing approximately 130g. The location error and fix success rate of the GPS was also compared between habitats and different orientations (facing towards the sky, or towards the ground).
Home range size and distance travelled from home were found to be significantly smaller when cats were wearing the heaviest collar treatment. Cats also spent a greater amount of time within the home property when wearing the heaviest collar. There was no difference in habitat use between collar treatments, although habitat categories were broad. The restriction in movement and activity associated with heavy devices may have influenced predation rates by reducing the likelihood of cats encountering prey when home range size were reduced. The light collar treatment (or the medium collar treatment – 2% BM), that weighed <1% of body mass, was suggested as an optimal device weight to reduce the frequency of instrumentation effects.
The other aim of this research was to contribute to the development of an accelerometer that could be used to predict behaviours from movement patterns. Presented here is an initial phase of development, in which behaviours could realistically be distinguished from accelerometry data. Domestic cats wearing an accelerometers were videoed and matched the observed behaviours to the output. It was clear that behaviours observed from movement patterns could be manually (human assisted classification) detected when using calibration videos. Patterns of movement could be discriminated from each other for up to 15 behaviours based on acceleration and angular velocity data. Instantaneous power for both acceleration and velocity were calculated from the tri-axial accelerometer and gyroscopes and these were plotted. With more calibration video footage and a larger sample, it is possible to automatically classify behaviour. In combination with direction (magnetometers), location (GPS devices) and light meters to indicate whether the individual is indoors or in vegetation, accurate information on behaviour and movement can be collected and applied to many areas of behavioural ecology and conservation.