Phenotypic divergence of <i>Glossina morsitans</i> (Diptera: Glossinidae) populations in Zambia: Application of landmark‐based wing geometric morphometrics to discriminate population‐level variation

An important consequence of the discontinuous distribution of insect populations within their geographic range is phenotypic divergence. Detection of this divergence can be challenging when it occurs through subtle shifts in morphological traits with complex geometries, such as insect wing venation. Here, we used landmark-based wing geometric morphometrics to investigate the population-level phenotypic variation of the two subspecies of <i>Glossina morsitans</i>, <i>G. m. centralis</i> Machado and <i>G. m. morsitans</i> Westwood that occur in Zambia. Twelve homologous landmarks digitised on the right wings of 720 specimens collected from four and five sites (80 per site with 1:1 sex ratio) within the <i>G. m. centralis</i> and <i>G. m. morsitans</i> range respectively, were subjected to generalised Procrustes analysis to obtain wing centroid size (CS) and wing shape variables. Linear permutation models and redundancy analysis were then used to compare CS and wing shape between male and female <i>G. morsitans</i>, the two subspecies <i>G. m. centralis</i> and <i>G. m. morsitans</i>, the sexes of each subspecies and between sample locations within each subspecies range, respectively. Significant differences in CS and wing shape were observed between <i>G. morsitans</i> sexes, subspecies and sample locations within each subspecies range. A neighbour-joining cladogram derived from the analysis of Procrustes distances showed that tsetse within each subspecies range were highly divergent. We conclude that <i>G. morsitans</i> populations in Zambia exhibit significant population-level variation in fly size and wing shape which suggests high levels of population structuring. The main drivers of this structuring could be random genetic drift in <i>G. m. centralis</i> demes and local adaptation to environmental conditions in <i>G. m. morsitans</i> populations. We therefore recommend molecular studies to estimate the levels of gene flow between these populations and identify possible barriers to genetic flow.