Figure 4. Population structure of Atta cephalotes as obtained from the STRUCTURE and DAPC analyses. A.Discriminant analysis of principal components (DAPC) scatterplot between samples. The key describes the colors attributed to each region and inertia ellipses describe the general distribution of points. Eigenvalues for each PC axis are shown (PC1, vertical; PC2, horizontal). The number of PCA axes retained in each DAPC analysis is shown in the bottom-right inset (gray bars). B. Multilocus genotype clustering of A. cephalotes populations, using STRUCTURE and following Evanno et al. (2005). Each region is characterized by color and each location is represented by a vertical bar. Locations are organized by region, starting with Pacific (BVT – GUA) followed by Andean 1 (DAG and CAL), and Andean 2 (PPY – MNZ). A different simulation was run for the southern cluster. C. Map for sampling localities according to the STRUCTURE clustering: Pacific (BVT, QBD, and GUA in blue), Andean south (DAG, CAL, and PPY in violet), and Andean north (DOV, QBY, YOB, and MNZ in orange).