Get reliable rangeland science

The use of on-animal sensors to manage and monitor free-ranging livestock has advanced rapidly over the past decade, particularly with the emergence of virtual fencing technology to manage livestock distribution. Low-cost GPS tracking for purposes of virtual fencing creates new opportunities to monitor animal health and behaviour when combined with an accelerometer to quantify the animal's behavioural state. However, our understanding of how behaviors can be quantified via GPS plus accelerometer measurements, and how foraging behavior metrics relate to diet quality or animal growth rate remains in its infancy. Here, we provide an overview of multiple studies that use on-animal sensors to quantify daily foraging behaviour of both yearling steers and mature cows in semi-arid rangelands of central North America (Colorado and Wyoming). We examine analyses of behaviour at varying time steps (seconds to minutes) summarized over the daily cycle using both commercially available and custom-built GPS plus accelerometer combinations. Foraging behaviour could be most accurately predicted by analyzing both sensors at a time step of 90 seconds, but an accelerometer algorithm calculated at a 30-second time step could be linking to longer GPS fix intervals with nearly equivalent prediction accuracy. We then focus on the utility of three key behavioral metrics: (1) mean daily grazing bout duration (GBD), (2) mean velocity while grazing (VG), and (3) the tortuosity of grazing pathways quantified as the mean turn angle while grazing (TAG). Our analyses identify GBD and VG as key indicators of declining forage availability, which could be used to guide the timing of pasture rotations or provision of supplemental feed. Furthermore, VG and TAG are significantly affected by s tock density (herd size relative to pasture size), and hence could potentially be used to identify a threshold density that inhibits selective foraging and reduces weight gain.