Allometry and Geographic Variation in Forewing Shape and Size Among Honey Bee (Apis mellifera L.) Populations from Türkiye

Wing shape provides an important source of morphological information for evaluating population differentiation in honey bees, but the extent of geographic variation relative to allometry and within-population variation remains unclear. This study examined forewing shape and size variation in 1.040 honey bee (Apis mellifera) specimens collected from seven locations in Türkiye. Nineteen homologous landmarks were digitized and analysed using Generalized Procrustes Analysis, Procrustes ANOVA, allometric models, and clustering, and reduced landmark-displacement features were used to assess individual-level geographic assignment. Overall forewing size, measured as centroid size, did not differ significantly among locations (ANOVA: F₆,₁₀₃₃=0.89, P=0.502; Kruskal–Wallis: P=0.483). In contrast, forewing shape differed significantly among locations (Procrustes ANOVA: F₆,₁₀₃₃=3.06, Z=6.84, P=0.001), although the geographic effect was small, explaining only 1.75% of total shape variation. Allometry explained a larger proportion of shape variation (R²=4.83%, P=0.001), while the size × location interaction was not significant (P=0.072). Ward hierarchical clustering suggested two broad shape groups, but these were not meaningfully associated with geographic origin (Cramér's V=0.058, P=0.739). Individual-level geographic assignment based on reduced landmark-displacement features was limited, with the highest test accuracy reaching 27.6%, only marginally above the majority-class baseline of approximately 26.6%. Together, these results indicate that the sampled honey bee populations from seven locations in Türkiye show statistically detectable but biologically modest location-associated variation in forewing shape, which is outweighed by allometric and within-population variation and is insufficient for reliable individual-level geographic assignment.