New technologies for understanding livestock behaviour in conservation management: A case study at RSPB Lake Vyrnwy

The use of grazers for vegetation control and restoration is common practice in agri- environment schemes and in wider habitat management exercises. However, there is a fine balance in achieving the optimum grazing intensity on a given site, with many habitats currently under- or over-grazed. In order to achieve best practice, it is important to understand the animal-environment interactions occurring on a given site. This refers both to the effect the animals are having on their surrounding environment, and also how the animals are being influenced by their environment. This latter point is important because knowledge of which environmental variables are driving grazer distribution can facilitate manipulation to allow more targeted grazing regimes. This study explored new methodologies and technologies that could be used to measure these animal-environment interactions on large-scale sites. Using a 72 ha upland site at RSPB Lake Vyrnwy, a combination of custom developed unmanned aerial vehicles (UAVs) and GPS tracking collars were used to collect data relating to animal movement and environmental factors. A series of analytical approaches were developed and employed, including simple practitioner methods which could be used by land managers, as well as more complex statistical techniques which can address academic hypotheses relating to the site. These included a bespoke livestock unit (LSU) estimator based on GPS recordings, individual and combined home range analyses, GIS visual interpretation, customised software for estimating livestock behaviour, and species distribution modelling. Results from practitioner methods revealed a number of interesting aspects as regards the behaviour of the sheep, as well as the extent of the unevenness of the grazing pressure being applied across the site. Statistical outputs revealed water, standing shelter and vegetation greenness to be key factors affecting animal distribution, and thus these would be prime candidates for manipulation to achieve more targeted grazing.