Principal Component Analysis of Morphostructural Indices of Noiler Chickens

Abstract

Morphological architecture is central to description, characterization, classification and utility of a species. Six (6) morphological indices{compactness (CMP), massiveness (MAS), condition index (CID), stockiness (STK), long leggedness (LLG) and body index (BDI)} were used to describe and predict body weight of male (52) and female (62) Noiler chickens using stepwise regression analyses of morphological indices and principal component analysis (PCA) factor scores. Descriptive analysis revealed significant (P0.05) in STK, BDI and LLG. Two PCs accounting for 89.11% and 89.16% of total variation were extracted for male and female chickens, respectively while CMP (male, 0.961; female, 0.953), CID (male, 0.889; female, 0.907) and MAS (male, 0.833; female, 0.820) had highest correlations with BWT in PC 1 of both sexes. Stepwise regression analysis indicated that CMP was the most important predicting morphological index accounting for 73.70%; and 78.10% prediction accuracy (R2 ) of BWT of cocks and hens, respectively. Regression analysis of factor scores of PCs indicated that PC 1 of the two sexes yielded the best prediction models and explained 89.70% and 91.00% of total variation in BWT of male and female Noiler chickens, respectively. Sexual dimorphism is evident and application of PC factor score models is more reliable in predicting BWT of male and female Noiler chickens than interdependent morphological indices because of its orthogonality and higher prediction accuracy.

Keywords

• compactness
• factor score
• morphological architecture
• morphological indices
• sexual dimorphism