Please login first
Sensor-Derived Heat-Tolerance Traits in Sheep: Heritability, Prediction, and On-Farm Decision Triggers
* 1, 2 , * 2, 3 , * 2, 3 , * 2, 3
1  College of Animal Science and Technology, Tarim University, Alar 843300, China
2  Key Laboratory of Tarim Animal Husbandry Science and Technology, Xinjang Production and Construction Group, School of Animal Science and Technology, Tarim University, Alaer 843300, China.
3  College of Animal Science and Tarim University, Alar, Xinjiang 843300, China
Academic Editor: Andrea Pezzuolo

Abstract:

Sheep reproduction, survival, and welfare are increasingly challenged by heat, humidity, and unstable forage. Modeling in Nature Food estimates heat stress already causes ~2.1 million lamb losses annually in Australia, rising to ~3.3 million with +3 °C warming, with mating on hot days (≥32 °C) especially harmful. The Temperature–Humidity Index (THI) for small ruminants provides actionable thresholds—moderate 82–<84, severe 84–<86, extreme ≥86 useful for both management and analysis. Precision Livestock Farming (PLF) offers a data driven response. Sensor streams accelerometers, location data, and computer vision can generate selection-worthy phenotypes and enable early stress detection. In a Merino cohort (n=160), accelerometer-derived grazing time showed heritability h² = 0.44 ± 0.23 and repeatability 0.70 ± 0.03, indicating sensor traits are improvable through genetic selection. Genomic work in dairy sheep reports heat-resilience heritability around 0.26, suggesting thermotolerance can be selected without sacrificing production. Modern CV pipelines (e.g., YOLO/DeepSORT families) reliably classify eating, lying, and rumination, enabling automated welfare/intake monitoring at scale. Management strategies triggered by PLF signals—timely shade/ventilation and ration adjustments improve comfort and performance, reducing heat-load behaviors. We propose an integrated five-stage system for climate-resilient sheep breeding and operations: (1) quantify heat load with on-farm loggers and THI; (2) extract PLF features (activity bouts, rumination, shade-seeking, grazing time); (3) link features to fertility and survival via mixed-effects models and gradient boosting; (4) estimate genetic parameters and breeding values for sensor-derived heat-tolerance indicators; and (5) deploy real-time triggers (shade/soakers) when locally validated THI thresholds are exceeded. This pipeline converts environmental pressures into measurable, heritable, and decision-useful phenotypes. By coupling continuous sensing with predictive analytics and genomic selection, producers can improve reproductive success and lamb survival while safeguarding welfare—offering a practical path toward climate-resilient sheep systems.

Keywords: Sheep; heat stress; temperature–humidity index (THI); accelerometers; computer vision; genomic selection; climate resilience.
Top