Supplementary Information SI1. Figure S1. Significant canonical correlation of habitat use and morphology for second canonical roots (CRs; r = 0.8447, P < 0.0001, Fisher's Z refined confidence interval estimation 0.8191 < r > 0.8665). Figure S2. Logistic MAXENT model of species distribution for the Anolis cybotes species complex on Hispaniola (Dominican Republic outlined in black); black points—own sampling points and sampling points obtained from Glor et al. (2003), gray points—additional data points from specimens deposited in the Museum of Comparative Zoology, Harvard University. Red shades on map denote low occurrence probability; green shades denote high occurrence probability. Figure S3. Significant canonical correlations between the ecomorph-environment canonical roots CR2–CR4. Figure S4. Example for morphological characters whose variance is best explained by the environment. Table S1. Localities, coordinates, GenBank accession numbers, museum voucher numbers, and list of recorded parameters of studied specimens of the Anolis cybotes species complex. Table S2. Eigenvalues and explained variance for principal component analysis of variables obtained by flatbed scans of anole feet (residuals to SVL). Table S3. Factor loadings for Principal Component Analysis of variables obtained by scanning anole feet (residuals to SVL). Table S4. Eigenvalues and explained variance for Principal Component Analysis of morphological variables obtained from X-ray images of the skeleton (residuals to SVL). Table S5. Factor loadings for Principal Component Analysis of morphological variables obtained by analyzing X-ray images of the skeleton (residuals to SVL). Table S6. Eigenvalues and explained variance for principal component analysis of habitat use variables. Table S7. Factor loadings for principal component analysis of habitat use variables. Table S8. Forward-stepwise selection of the genetic principal components of the neighborhood matrix using analysis of covariance in rda (R: vegan). Table S9. Eigenvalues and explained variance for principal component analysis of environmental variables (elevation, and 19 bioclimatic variables extracted from WORLDCLIM). Table S10. Factor loadings for principal component analysis of environmental variables (elevation, and 19 bioclimatic variables extracted from WORLDCLIM). Table S11. Results of occurrence-probability weighted canonical correlation analysis between ecological and morphological variable sets. Table S12. Canonical weights for habitat use variable set. Marked are weights >0.4. Table S13. Canonical weights for morphology variable set. Marked are weights > 0.4. Table S14. Results of occurrence-probability weighted multiple regression of data set pruned to cases with no missing values. Table S15. Overall model results for each dependent variable, displaying the overall fit of all parameters in the model. Table S16. Canonical correlation of “Ecomorph” data set with environment data set obtained from 2000 random points. Table S17. Canonical weights for “Ecomorph” variable set. Table S18. Canonical weights for environment variable set. Table S19. Results of variance partitioning (R: vegan, varpart) using single morphological variables as dependent. |