AKSU Journal of Agriculture and Food Sciences

ISSN(Print): 2651-5679    ISSN(Online): 2695-2556


    

Fixing Collinearity Instability in Body Weight Estimation Based on Morpho-biometrical Traits of Rabbits Bucks


Author: 
Iliya Mallam, Kuje A. Ombugu

Abstract
Morpho-biometric studies were conducted on 32 growing rabbit bucks comprising Chinchilla, New Zealand white and Dutch breeds to identify the estimation of body weight collinearity. The body weight and biometric traits taken were analyzed using statistical package for the Social Sciences (SPSS) version 16. Body weight and zoometrical factors had positive and highly significant bivariate relationships. The body weight and height at the withers showed the best relationship (0.826; p<0.01), whilst the foreleg and ear length showed the lowest correlation (0.542; p<0.05). Since none of the body measurements were greater than 10.00, the variance inflation factors (VIFs) displayed non-collinearity issues in all of the measurements. Tolerance (T) values greater than 0.10 in all of the same measurements served as confirmation of this. No collinearity was also evident in the eigenvalues of the correlation matrix, condition indices, or variance proportions. Body length was kept as the most crucial attribute for prediction among the variables. This study showed that there is no collinearity among selected morphometric traits. Stepwise multi-regression models showed that the best predictor of body weight was based on a combination of wither height, chest circumference, waist circumference, waist length, hind limb length, and head circumference (R2 and adjusted R2 =0.84). This research has the practical application that morphological indices can be utilized in the field to estimate body weight for selection and to estimate market values for rabbit bucks.