Abstract

Objective: This study aimed to develop and validate sex-specific regression models for predicting body weight in Indigenous Sabi sheep using linear body measurements to improve accuracy and inform flock management decisions in smallholder production systems.
Methods: A total of 173 Sabi sheep (112 ewes, 22 rams, and 39 wethers) from the Matopos Research Institute, Zimbabwe, were used in this study. Linear body measurements, including heart girth (HG), body length (BL), and chest depth (CD), were recorded and subjected to multiple linear regression analyses. Model validation was conducted using correlation and residual diagnostics to assess the predictive performance and assumption compliance.
Results: Heart girth was the strongest predictor of live weight across sexes (R² = 0.68–0.71), followed by body length and chest depth. Only two measurements were sufficient for reliable weight estimation, with model accuracy varying by sex of the individual. The optimal regression equations were as follows:
Ewes: BWT = −50.246 + 0.435HG + 0.501BL + 0.446CD (R² = 0.81)
Rams: BWT = −90.054 + 2.039BL (R² = 0.90)
Wethers: BWT = −42.722 + 0.629HG + 0.577CD (R² = 0.88)
Pooled data: BWT = −35.149 + 0.857HG (R² = 0.86)
Model validation confirmed strong correlations between actual and predicted body weights and random residual distributions, indicating robustness and reliability.
Conclusion: Sex-specific regression models using simple linear body measurements provide accurate and practical tools for predicting the live weight of Indigenous Sabi sheep. Heart girth and body length are the most reliable predictors of weight. These models can enhance smallholder decision-making in flock management, nutrition, and genetic selection, thereby supporting sustainable sheep production and improving productivity.