Genetic of the 3D morphology of jumping horses using geometric morphometrics

Archive ouverte

Ricard, Anne | Pourcelot, Philippe | Dumont Saint Priest, Bernard | Crevier-Denoix, Nathalie | Danvy, Sophie

Edité par CCSD ; Wageningen Academic Publishers -

International audience. Morphological data were recorded on 2,097 jumping horses aged 4 and 5 with 3D technology using 3 cameras. For each horse, the anatomical landmarks of the main joints, including the head, have been identified, leading to a 25 point file describing the morphology of the horse with their coordinates in 3 dimensions. The study was made by dissociating the shapes of the general size through a Procrusted type analysis. After creation of new Procrustes coordinates, a principal component analysis reveals the predominant forms in playing simultaneously on all points. The Geomorph package under R was used. The heritability of these components has been estimated after correcting the influence of the person who made the marks on the images of the horses, the place and the date of measure, the age and sex of the horse, and the angle of the anterior and posterior canons with the vertical on the chosen image. The genealogies over 6 generations were used (18,029 horses). ASREML software was used. 10 components are needed to explain 80% of the variance. The first two components are the orientation of the limbs in the frontal plane: the first distinguishes the horses whose posteriors marks are close to the midline of the hoof point. The second is similar for the forelegs with a width at shoulders following the same movement. The third component is mainly explained on the sagittal plane and distinguishes big and short horses of long and small for the same general size. The heritability of the overall size is 0.21. The two first components are not heritable (0.09 and 0.06). Only components 3, 6 (length of the neckline) and 8 (high neck tie, long shoulder, short withers and hollow back line) are really heritable (respectively 0.27, 0.26 and 0.21). Genetic correlation with overall size were not null, denoting a strong allometry. Multiple phenotypic regression with performances in jumping competition do not exceed a r2 of 2%.

• Description
• Sujet/Public
• Outil
• Outil d'anticipation
• Mémoire et thèse
• Gestion