Selective signatures and high genome-wide diversity in traditional Brazilian manioc (Manihot esculenta Crantz) varieties

Abstract

Knowledge about genetic diversity is essential to promote effective use and conservation of crops, because it enables farmers to adapt their crops to specific needs and is the raw material for breeding. Manioc (Manihot esculenta ssp. esculenta) is one of the world's major food crops and has the potential to help achieve food security in the context of on-going climate changes. We evaluated single nucleotide polymorphisms in traditional Brazilian manioc varieties conserved in the gene bank of the Luiz de Queiroz College of Agriculture, University of São Paulo. We assessed genome-wide diversity and identified selective signatures contrasting varieties from different biomes with samples of manioc's wild ancestor M. esculenta ssp. flabellifolia. We identified signatures of selection putatively associated with resistance genes, plant development and response to abiotic stresses that might have been important for the crop's domestication and diversification resulting from cultivation in different environments. Additionally, high neutral genetic diversity within groups of varieties from different biomes and low genetic divergence among biomes reflect the complexity of manioc's evolutionary dynamics under traditional cultivation. Our results exemplify how smallholder practices contribute to conserve manioc's genetic resources, maintaining variation of potential adaptive significance and high levels of neutral genetic diversity.

Figure 1

Map of Brazil showing the geographical locations of the municipalities in which manioc (Manihot esculenta) samples were originally collected (some points were slightly moved for easier visualization). The black square indicates the location of the gene bank at the Luiz de Queiroz College of Agriculture, Piracicaba, São Paulo, Brazil. ND = no data on toxicity. The map was drawn with maptools 0.8–36 (https://CRAN.R-project.org/package=maptools).

Figure 2

Summary of genome scans for signatures of selection considering different groups of manioc (Manihot esculenta) samples. Venn diagrams showing the number of outlier SNPs detected for each test (within parenthesis) and the overlap among them (numbers inside ellipses) for (a) wild and cultivated manioc, and (b) the groups of varieties per biome. The genomic context of outlier SNPs is illustrated in circular plots for (c) the groups of wild and cultivated manioc, and d) the groups of varieties per biome. Each manioc chromosome is represented by a different box (10 Mb tick sizes), and their names are coded according to the manioc genome Manihot esculenta v6 (NCBI PRJNA234389). The outlier SNPs are represented by dots for each test, which are shown in different layers. The outlier SNPs detected by at least two tests are highlighted in red.

Figure 3

Genetic structure of 92 manioc (Manihot esculenta) varieties based on 10,917 neutral SNPs. (a) Plot of cross-entropy estimates for different numbers of ancestral populations (K) in sparse non-negative matrix factorization (sNMF) showing no evident flatting point in the curve corresponding to the most-likely number of ancestral populations. (b) Bar plots of sNMF ancestry coefficients for K = 2, 3, and 5. Discriminant analyses of principal components (DAPC) considering: (c) the groups of wild manioc and the different biomes, and (d) only the cultivated varieties grouped by biomes. The respective membership coefficients of each DAPC are shown as bar plots below scatter plots. Cultivated manioc is ordered in the sNMF and DAPC bar plots according to the biomes and their reputed toxicity (B = bitter, S = sweet, ND = non-designated).