Tropical Animal Health and Production Животноводство и зоология

Regional morphometric variation and weak phenotypic clustering of Moorish zebu cows (Bos indicus) in Mauritania

El Hadj, Youssouf, Ahmed, Mohamed Ould, Hamrouni, Abir, Hamady, Sidi, Djemali, M’Naouer, Boukhary, Ali Ould Mohamed Salem
Springer Nature · 2026-11-20 · DOI 10.1007/s11250-026-05493-w

Аннотация

The Moorish zebu supports pastoral livelihoods in Mauritania, but regional morphometric variation has not been quantified using a common field protocol. To describe phenotypic variation among adult lactating cows from three Mauritanian regions and explore whether the sample contained distinct morphometric groups. A purposive cross-sectional sample included 352 cows from 158 herds in Hodh Chargui, Brakna, and Trarza. Sixteen direct measurements, estimated body weight, 16 zoometric indices, and six qualitative traits were recorded. Regional comparisons used one-way analysis of variance with herd-aware confirmation; LDA, PCA, and K-means were exploratory. Associations with recorded milk yield were examined in 349 cows. LDA classification was validated by leave-one-herd-out cross-validation across the 158 sampled herds. Nominal regional differences were most evident for thoracic dimensions, estimated weight, and udder traits. Herd-grouped LDA classified 59.8% of cows to their region of origin, 2.6 points lower than non-grouped cow-level validation (62.4%). Three principal components explained 50.3% of variance. A three-group partition had weak separation (silhouette = 0.281, versus 0.298 for two groups). Teat length showed the largest unadjusted correlation with daily milk yield ( r = 0.381). The exploratory three-trait model had cross-validated R^2 = 0.162 (non-nested) and 0.137 (nested). Sampled cows displayed overlapping regional morphometric variation rather than discrete regional types. This study provides, to our knowledge, the first multi-region, herd-aware morphometric description of Moorish zebu cows in Mauritania, establishing provisional descriptive references while showing why evidence does not yet support a breed standard, regional genetic subdivision, or phenotypic selection criterion; the proposed clusters and milk model require independent, herd-aware validation.

Читать у издателя →