EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Principal component-based regression models for predicting body weight from morphometric traits in Hararghe Highland sheep, Ethiopia

Mohammed Yousuf, Kefelegn Kebede, Yesihak Yusuf, Gemeda Duguma, Solomon Abegaz

Veterinary and Animal Science · 2026-12 · 34 · 100918

DOI: 10.1016/j.vas.2026.100918 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Veterinary and Animal Science, 2026-12. Verfasst von Mohammed Yousuf und 4 weiteren Autor:innen. Beteiligte Einrichtung: Dambi Dollo University, Department of Animal and Range Sciences, Dambi Dollo, Ethiopia. Die Arbeit nennt 48 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

In smallholder farming systems across the Ethiopian highlands, weighing scales are rarely available, and farmers rely on visual assessment to estimate sheep body weight (BW) for management. This study evaluated multivariate approaches for predicting BW from morphometric traits in Hararghe Highland ewes, Ethiopia. A cross-sectional survey was conducted on 392 ewes from three districts, recording BW and 16 measurements. Data were analysed using Pearson correlations, multicollinearity diagnostics, principal component analysis (PCA), and multiple linear regression (MLR), with LASSO and Elastic Net evaluated using 10-fold cross-validation in SAS 9.4. BW showed the strongest correlation with chest girth (r = 0.94), followed by rump height (r = 0.92), body length (r = 0.90), tail length (r = 0.90), and height at withers (r = 0.87; p 10), whereas PC regression removed it (VIF =1.0) at the cost of lower cross-validated accuracy (R² = 0.905 vs 0.940 for MLR). LASSO and Elastic Net (R² = 0.939) did not differ from MLR while controlling multicollinearity, and all three models outperformed PC regression and chest girth alone (P < 0.05). CG alone (R² = 0.881) offers a practical field option, whereas PC regression provides an interpretable description of body conformation but lower predictive accuracy.

Schlagwörter

Stichwörter der Autor:innen

Body weight predictionEthiopiaHararghe Highland ewesMorphometric traitsPrincipal component analysis

Autor:innen und Institutionen

  • Dambi Dollo University, Department of Animal and Range Sciences, Dambi Dollo, Ethiopia.
  • Haramaya University, School of Animal and Range Sciences, Haramaya, Ethiopia.
  • Wollega University, Department of Animal Sciences, Nekemte, Ethiopia.
  • International Livestock Research Institutes , Addis Ababa, Ethiopia.

Zitieren

APA

Yousuf, M., Kebede, K., Yusuf, Y., Duguma, G., & Abegaz, S. (2026). Principal component-based regression models for predicting body weight from morphometric traits in Hararghe Highland sheep, Ethiopia. Veterinary and Animal Science, 34, 100918. https://doi.org/10.1016/j.vas.2026.100918

Vancouver

Yousuf M, Kebede K, Yusuf Y, Duguma G, Abegaz S. Principal component-based regression models for predicting body weight from morphometric traits in Hararghe Highland sheep, Ethiopia. Veterinary and Animal Science. 2026;34:100918. doi:10.1016/j.vas.2026.100918

BibTeX
@article{Yousuf2026,
  author = {Yousuf, Mohammed and Kebede, Kefelegn and Yusuf, Yesihak and Duguma, Gemeda and Abegaz, Solomon},
  title = {Principal component-based regression models for predicting body weight from morphometric traits in Hararghe Highland sheep, Ethiopia},
  journal = {Veterinary and Animal Science},
  year = {2026},
  volume = {34},
  pages = {100918},
  doi = {10.1016/j.vas.2026.100918}
}
Kostenlos starten →