In the analysis of the workers alone, all the 18 partial warps contributed significantly (α = 0.05) to the separation of the five groups. MANOVA demonstrated that the groups were significantly different (Wilk’s λ = 0.000003; P < 0.00001; Figure 4). The cross validation test of the workers was able to correctly identify 100% of the individuals in each respective group and all individuals were classified with a probability of 100%. The graphical distribution of the groups was very similar to the one for the male analysis; again, workers of Plebeia remota and Nannotrigona testaceicornis were the closest two groups. The Mahalanobis square distances between the groups were greater for workers than for males (Table 2).