Principal component analysis explains the associations between biometric characters of Indian Sheep Breeds and non-descript sheep

This study identifies the principal components (PCs) that best characterise the biometric characters of 21 Indian sheep breeds and non-descript sheep by explaining the largest percentage of total variance. Data on seven biometric characters viz., body length (BL), height at withers (HW), chest girth (CG), paunch girth (PG), ear length (EL), tail length (TL) and body weight (BW) of 5960 adult ewes were used. Varimax rotated principal component analysis (PCA) with Kaiser's criterion (eigen value >1) was carried out to know relationships among morphometric variables. Significant positive correlations (P<0.05) between the morphometric characters demonstrated the degree of harmony among them. The phenotypic correlations among the characters were ranged from -0.57 to 0.87. BW was strongly correlated with CG and PG, weakly correlated with EL and TL, and moderately correlated with BL and HW. PCArevealed two principal components in all the sheep breeds and non-descript, with the exception of Patanwadi and Sonadi breeds, explaining between 54.2 and 74.9% of the total variance in breed and non-descript morphometry. For 10 sheep breeds and non-descript sheep, prominent characters (CG, PG, BL, HW, and BW) were loaded on the first PC, and EL and TL were loaded on the second PC. The remaining sheep breeds and non-descript sheep had one or two characters on the second PC loaded with either EL or TL or EL and TL. The study explained the body size and shape of sheep breeds and non-descript sheep and can be used for selection of animals based on a group of morphometric characters that are related to one another. The first PC can be utilized in phenotypic selection programme to explain body shape of adult sheep.