Remote characterization of locomotion, grazing and drinking behavior in beef cattle using GPS and ruminant temperature dynamics

A low-cost GPS herd activity and well-being kit (GPS HAWK) was developed as an alternative to commercial GPS tracking collars. The operational goal of the GPS HAWK was to collect GPS location data at a high frequency and store the data in a secure format. The use of one animal to represent herd behavior and the effect of GPS sampling rate on the characterization of the animal locomotion behavior were assessed. Though the percentage of time spent at each activity varied over days, cows spent 33.2%, 25.5%, 40.4% and 0.9% of four 8-hr visual observation periods on lying (L), standing (S), grazing (G) and traveling (T), respectively. Cows within the herd in this study spent 56.7% of the time performing the same activity simultaneously; however, mean daily duration for each activity was similar for cows. Use of a cow location vector (CLV) provided a simplistic method to illustrate the dynamic locomotive behavior of cows over time. The CLV provided a visual means to discern grouping of cows based on locomotion. Though differences in location occurred due to sub-grouping of cows throughout each day, the cumulative distances traveled were similar across cows. Therefore, monitoring a single cow will suffice in the quantification of average time spent at activities of L, S, G and T and CTD. However, a single animal would not be sufficient to illustrate dynamics in herd location. In addition to tracking locomotion, the utility of using rumen temperature to estimate water consumption was investigated on six Angus steers. The rumen temperature and temperature responses to drinking activity were characterized using stainless steel temperature boluses placed in the following locations: (1) the reticulum floor (RET), (2) the rumen floor (RUM), and (3) tethered to the rumen-cannula by an 18-inch nylon string (TET). Mean temperature response (DeltaT) to water volume (VOL) were -1.7°C (+/-0.125), -1.0°C (+/-0.125) and -0.8°C (+/-0.125) for the reticulum floor (RET), rumen floor (RUM) and TET, respectively. RET was different (p \u3c 0.001) than RUM and TET. Multiple regression of temperature response against water volume and water temperature for each steer explained 84% of total variation. A drinking event algorithm was created using the large drops in rumen temperature associated with drinking. The algorithm detected 64 (82%) of the drinking events recorded and 3 drinking events that were not recorded. The total predicted volume consumed by six steers over 2 days, using the drinking event algorithm, was 10.5% greater than actual volume. The use of reticulum temperature response to a drinking event may provide a means to remotely quantify patterns in drinking activity for individual animals in remote or group settings over a period of time