Figure 5

Results of the kdr index value distribution spatial analysis. (a) and (b) indicate the spatial patterns and the corresponding regions identified from the kdr index value spatial distributions in Brazil for Aedes aegypti. These patterns correspond to the most explanatory eigenvectors of a principal coordinates analysis of a given neighbor matrix (see Supplementary Text S1). White and black squares consist of negative and positive eigenvector component values, respectively, and square size is proportional to the absolute value of the vector components. These patterns were selected because they present a significant Moran’s I spatial autocorrelation and explain 51.6% (29.6% and 22.0% for A and B, respectively) of the total variance of the observed kdr index values, by their inclusion as explanatory variables in a multiple linear regression model. (c) and (d) display the kdr index value distributions as a function of the identified regions. The background map is the “OSM TF Landscape” product (Maps Thunderforest, Data OpenStreetMap contributors, under license CC-BY-SA 2.0). Map was generated with the free and open source software (GNU General Public License) QGIS version 3.12.3 (developd by the Open Source Geospatial Foundation Project, http://qgis.org).