STRAIN CLASSIFICATION OF DONKEYS IN NORTHWEST NIGERIA USING CANONICAL DISCRIMINANT ANALYSIS ON QUANTITATIVE TRAITS

Abstract

Metric traits were used to determine the relationship among Red (Auraki), Black (Duni), White (Fari), Brown (Idabari) and Brown-white (Idabari-fari) donkeys. A total of 700 donkeys were used for the study. Metric measures taken were head length, head width, ear length, neck length, neck circumference, shoulder width, height at withers, heart girth, body length and tail length. Data obtained were subjected to General Linear Model Procedure of SAS to determine the relationship
among donkey strains. Canonical discriminant analysis (CANDISC procedure), was used to perform uni-and multivariate analysis to derive canonical variables (CAN), which was used to match the donkey strain groups until reached the satisfactory number of clusters (genetic groups) and to show the clustering groups among these four donkey strains. The canonical coefficients and a scatter diagram for visual interpretation of the different groups were also generated during the canonical discriminant analysis. Among the all the donkey strains, Black (Duni) strain had the highest trait loading in CAN 1 for HL (0.61), HW (0.71), NC (0.91), SW (0.71), HG (0.84) and BL (0.82). The four strains that is, red, black, white and brown donkeys were clustered or separated into a separate group while the brown-white were separated into another distinct group. The results of the study showed that Black donkeys had more canonical weight on the generalized canonical component followed by red and white donkeys. There is need to screen and conserve the four basic donkey strains to arrest further genetic erosion and dilution in these strains.