author(s)
Usman H. Dukku

abstract
Samples of workers of honeybee were collected from 49 colonies belonging to 12 localities in Nigeria and analysed using classical morphometry. In addition, 10 colonies from three localities in neighbouring Niger and Cameroon were included in the analysis. Measurements of 35 morphological characters of body size, colour and pilosity were taken from 10 workers per colony and the data subjected to various statistical analyses. Although ANOVA revealed a considerable variation of morphological characters between the sampled localities, principal component analysis indicated that this variation was not sufficient to group the colonies into geographically separable groups. However, stepwise discriminant analysis confirmed the three a priori clusters into which the colonies were classified by hierarchical structural analysis. One of the clusters consisted of colonies from the plateaus while the remaining two consisted of colonies from the plains. Pearson correlation analysis revealed a significant positive correlation between altitude and the discriminant functions.