CLASSIFICATION OF COAT COLOUR PATTERN AMONG SOME CATTLE BREEDS IN NIGERIA USING STEPWISE DISCRIMINANT ANALYSIS

Abstract

Study was carried out to determine the distributions of coat colour patterns among Adamawa Gudali, Sokoto Gudali, Bunaji and Rahaji cattle breeds using stepwise discriminant analysis. A multistage approach was used to select the cattle from 4 states with large population of agro-pastoralists who rear these breeds. The four States selected were Adamawa, Taraba, Gombe and Sokoto with 4, 4, 3 and 2 Local Government Areas respectively. Snowball method was used to sample 1008 cattle consisting of 162 Adamawa Gudali, 306 Sokoto Gudali, 234 Bunaji and 306 Rahaji cattle from agropastoralists in many communities of the LGAs. Stepwise discriminant (STEPDISC) analysis was employed to identify variables for their relative importance in discriminating between groups. The distribution of the measured qualitative traits of cattle was assessed for significance using Chi-square analysis. Results revealed that 72.92% of the coat colour patterns were classified as patched, 55.69% as solid) and 83.33% as spotted. Significant (p<0.001) canonical functions was observed for patched and spotted coat colour patterns. Patched coat colour pattern (0.7725) in the first function and spotted coat colour pattern (0.7676) in the second function had the best and high discriminating power compared to solid coat colour pattern. Patched and spotted coat colour revealed the best canonical discriminant functions, which had high discriminating power in classifying the coat colour patterns into genetic groups.