Estimation of Liveweight from Body Measurements through Best Fitted Regression Model in Dezhou Donkey Breed.
Authors: Zhang Zhenwei, Zhan Yandong, Han Ying, Liu Ziwen, Wang Yonghui, Wang Changfa
Journal: Journal of equine veterinary science
Summary
# Estimation of Liveweight from Body Measurements in Dezhou Donkeys Accurate liveweight assessment is essential for breeding programmes, nutritional management and health monitoring, yet scales are not always accessible in field settings. Researchers analysed 162 Dezhou donkeys (aged 2–10 years) using multiple regression modelling to develop prediction equations correlating bodyweight with nine key body measurements including thoracic girth, body length, thoracic width, cannon bone circumference and rump dimensions. Thoracic girth alone proved the strongest single predictor of bodyweight (R² = 0.72), but combining all five measurements—thoracic girth (1.88), body length (1.27), thoracic width (2.55), cannon bone circumference (4.61) and rump width (1.27)—into a stepwise regression model achieved substantially greater accuracy (R² = 0.906, P < 0.01), with validation predictions showing R² > 0.94. For equine professionals managing donkeys in resource-limited settings, these breed-specific equations offer a practical alternative to scales, though the authors emphasise that existing prediction models developed for British and Moroccan donkey populations significantly over- or underestimate Dezhou bodyweights, underscoring the importance of breed-specific calibration for accurate weight estimation across different populations.
Read the full abstract on PubMed
Practical Takeaways
- •Thoracic girth alone provides a quick, reasonably accurate estimate of donkey bodyweight (R² = 0.72) in field conditions where scales are unavailable
- •For improved accuracy in weight estimation of Dezhou donkeys, use the six-variable regression equation rather than relying on breed equations from other countries
- •Breed-specific prediction equations are necessary as morphological differences between donkey breeds significantly affect the accuracy of bodyweight estimation
Key Findings
- •Thoracic girth showed the strongest single correlation with bodyweight (R² = 0.72, P < 0.01) in Dezhou donkeys
- •A six-variable regression model combining thoracic girth, body length, thoracic width, cannon bone circumference, rump width and rump length achieved R² = 0.906 for bodyweight prediction
- •Existing prediction equations for British and Moroccan donkey breeds were not suitable for Dezhou donkeys due to breed-specific morphological differences
- •Predicted bodyweight from the derived equations showed very close agreement with actual bodyweight (R² > 0.94, P < 0.01)