Abstract
This study evaluated precision livestock farming (PLF) tools to quantify resilience in rangeland sheep by assessing behavioral and physiological traits in relation to live body weight. We examined trait correlations, differences between activity states, and trait repeatability to determine their potential for management or selective breeding purposes. During a two‐week trial (August 19–September 1, 2023) coinciding with Hurricane Hilary, 199 ewes from the Rafter 7 flock at the Great Basin Research and Extension Center in Eureka, Nevada, were monitored using GPS collars and vaginal temperature sensors. GPS fixes were recorded every 5 minutes, filtered to remove points with speeds >72 m/min, and transformed to UTM coordinates. Hidden Markov models with multiple imputation (nsim = 1000) were used to derive land use traits—including daily horizontal and vertical distance traveled, slope usage, elevation usage, and distances from water and shade. Vaginal temperature data, recorded every 10 minutes, were summarized as daily mean, maximum, range, and coefficient of variation. Regression models examined relationships between these traits and live body weight (a proxy for resilience), while correlation analyses and comparisons between active and resting states provided further insight into energy management. Ewes with greater vertical movement and higher scaled elevation usage were significantly associated with increased live body weight (P = 0.002), suggesting enhanced access to nutrient-rich forage. In contrast, higher horizontal travel, increased slope usage, and longer distances from shade were linked to weight loss (P = 0.01), indicating elevated energetic costs. Correlations revealed that NDVI usage was positively correlated with elevation usage, and elevation usage was positively related to distance from water but negatively related to distance from shade, underscoring the influence of terrain on grazing behavior. Temperature traits (maximum, range, and standard deviation) were strongly inter-correlated. Comparisons between active and resting states revealed distinct behavioral profiles that offer insight into energy management; for instance, animals that traveled longer during active periods tended to reduce movement during rest, and ewes selecting higher resting elevations experienced lower solar noon temperatures during activity. Traits varied in repeatability, with highly repeatable traits emerging as promising candidates for selective breeding dependent on future heritability estimate analyses. Notably, multiple imputation improved trait repeatability over single imputation (mean difference = 0.18 ± 0.05, P = 0.002). PLF tools effectively capture nuanced, repeatable indicators of resilience in rangeland sheep. Advanced data processing with hidden Markov modeling and multiple imputation clarifies significant associations with live body weight and distinguishes key behavioral differences between active and resting states. These findings support the use of PLF-derived traits for real-time herd management and their integration into selective breeding programs aimed at enhancing climatic resilience in extensive production systems.