Principal component analysis of morphometric and growth traits in crossbred piglets

Abstract

This investigation was undertaken to derive fewer independent common factors through principal component analysis to characterize 75% Landrace crossbreds. Records of 26 body measurements and weights at 4, 6 and 8 week from 279 crossbred (75% Landrace × 25% Bareilly local) piglets, born between September 2017 and April 2018 were used. Principal component analysis was done using PROC PRINCOMP Module of SAS 9.3 software. High correlation coefficients were observed among most of the morphometric and growth traits. Kaiser-Meyer- Olkin measure of sampling adequacy, Bartlett’s Test of Sphericity and communality at different ages were calculated. Three PCs were extracted at 4 week, two at 6 week and two at 8 week. These PCs accounted for 73.7, 73.76 and 75.45% variance, respectively of total variance. PC1 was represented by high loading of body length, heart girth and height at back at 4 week; and body length, heart girth and body weight at 6 and 8 week. PC2 had high loading for height at foreleg, height at shoulder, chest width and inner orbital width at 4 week; and height at foreleg, height at shoulder and snout circumference at 6 and 8 week. PC3 was characterized by snout circumference and ear length. PC1 seemed to explain the body of pig, i.e. general size; whereas PC2 seemed to represent the front view of the pig. Accounted PCs at different stage may be exploited in breeding and selection programmes to acquire highly coordinated animal bodies using fewer measurements.