Field testing of LoRa-WAN sensors for real-time tracking and biosensing of Brangus and Raramuri Criollo cattle foraging on a small pasture

TitleField testing of LoRa-WAN sensors for real-time tracking and biosensing of Brangus and Raramuri Criollo cattle foraging on a small pasture
Publication TypeConference Proceedings
Year of Publication2022
AuthorsNyamuryekung'e S., Duff G., Cibils AF, Estell RE, Utsumi S.A, Funk M., Cox A, Gong Q, Spiegal S, Gouvêa VN, Brandani C
Conference NameJournal of Animal Science
Date Published3/21/2022
ARIS Log Number392874
Abstract

The study’s objective was to investigate the use of Internet of Things (IoT) biosensors using Long Range Wide Area Network (LoRa-WAN) communication to compare the foraging behavior of two desert adapted cattle breeds. Two groups of mature Brangus (BA) and Raramuri Criollo (RC) cows were introduced into a small pasture (38ha) of native grassland and irrigated winter wheat at the New Mexico State University Clayton Livestock Research Center during Dec-Jan, 2021 (41-days). The real-time position and activity count data (15- and 1-minute interval, respectively) were retrieved continuously from 4 BA and 5 RC cows collared with commercially available trackers. Response variables derived from GPS records included daily distance traveled (Dist) and daily area explored (Area). Response variables derived from accelerometers included daily percentage motion (%Act) and cumulative activity data (CumAct). All variables were grouped into six consecutive weekly periods and analyzed by a repeated measures mixed model that included the fixed effects of BREED, WEEK, and the BREED*WEEK interaction in a completely randomized design. A significant BREED*WEEK interaction was detected on all GPS-derived variables, with RC exhibiting a trend to explore larger areas than BA (P=0.06). For the accelerometer-derived variables, the BREED*WEEK interaction was only significant for CumAct. Overall, BA spent a larger percentage of the day active (%Act), while investing a significantly  greater overall activity (CumAct) compared to RC cows. The lack of sensitivity of GPS-derived variables is hypothesized to be the result of aggregating errors associated with less frequent GPS location measurements coupled with position errors (~5m) and a low success rate of data packet transmission (~40%). In contrast, the more frequent inherent data packaging of successive accelerometer data represented as cumulative guaranteed low data loss rates. The spatial scale of the study is hypothesized to have influenced the differences in foraging behavior observed between breeds and the better performance of the accelerometer vs. GPS-derived variables.