Fig 1.
Map of Ny. darlingi collection sites in Amazonas and Acre States, Brazil, with insets for each municipality group: SAO: São Gabriel da Cachoeira, PRE: Presidente Figueiredo, CRU: Cruzeiro do Sul area, LAB: Lábrea.
Each collection site within each municipality group is indicated by an orange dot, with Levels 1–5 in the municipality insets indicating the forest cover level (Level 1 = ~0–20% forest cover in a 1km radius, Level 2 = ~20–40% forest cover, etc.). Green color on inset map indicates areas of tree canopy cover as of the year 2000, pink color indicates forest loss during the period 2000–2014 (Hansen/UMD/Google/USGS/NASA, http://data.globalforestwatch.org/); and blue color indicates bodies of water (ESRI World Water Bodies/World Linear Water layers). Basemap: Natural Earth (https://www.naturalearthdata.com/).
Table 1.
Forest cover, collection date, and sample size for the 16 collection sites.
Fig 2.
Results of STRUCTURE analysis of the 5,561 SNP dataset, comparing Ny. darlingi collected from different municipality groups (CRU: Cruzeiro do Sul area, LAB: Lábrea, PRE: Presidente Figueiredo, SAO: São Gabriel da Cachoeira) and forest cover levels (1–5), depicting K = 2–6 inferred clusters.
Fig 3.
Results of Discriminant Analysis of Principal Components (DAPC) using the 5,561 SNP dataset. Ordination plots for DAPCs of three (A) and four (B) clusters, with insets showing the distribution of eigenvalues for the principal components analysis (PCA) and discriminant analysis (DA). For both analyses, 100 principal components were retained in the PCA, accounting for 73.8% of the variance in the dataset. CRU: Cruzeiro do Sul area, LAB: Lábrea, PRE: Presidente Figueiredo, SAO: São Gabriel da Cachoeira. *SAO/LAB cluster in (A) also includes one individual from PRE.
Table 2.
Analysis of molecular variance (AMOVA), with individual Ny. darlingi nested within forest covers nested within municipality groups.
Table 3.
Pairwise FST values between municipality groups (CRU: Cruzeiro do Sul area, LAB: Lábrea, PRE: Presidente Figueiredo, SAO: São Gabriel da Cachoeira) with 95% confidence intervals based on 999 bootstrap samples (below the diagonal), and geographic distances between the centroid of collection sites in each municipality group (above the diagonal).