Cassava (Manihot esculenta Crantz) is a species whose natural diversity has great food, industrial and energy potential. This study aimed to evaluate the genetic diversity, and population structure of 1,580 accessions belonging to the international germplasm conserved in Brazil using 20,601 single‐nucleotide polymorphism markers. The average values of polymorphic information content (PIC), inbreeding (f), observed heterozygosity (Ho) and expected heterozygosity (He) were 0.24, 0.21, 0.23 and 0.30, respectively, presenting levels of genetic diversity compatible with the nature of the marker (predominantly biallelic) and with the species reproductive system. The mean values of He, Ho, PIC and f were similar for the 18 chromosomes of the species. At the individual level, f values ranged from 0.49 to 0.97, with an average of 0.69, whereas some cassava accessions showed f > 0.90. The values of linkage disequilibrium ranged from 15 to 20 kb (r2 = 0.20). In the discriminant analysis of principal components (DAPC), 22 groups were formed with probabilities of assignment of individuals above 0.99. No association was observed between DAPC grouping and clusters based on phenotypic data (cyanogenic compound content and colour of the root pulp) and on genetic and geographic origin. However, the DAPC grouping resulted in greater variation between groups (14.85%). This information suggests extensive exchange of cassava germplasm in Brazil and wide molecular and phenotypic diversity. These results will provide a better understanding of the preserved genetic variability and the population organisation of the germplasm.

No abstract

Genetic redundancy in cassava (Manihot esculenta Crantz) presents a challenge to efficient management of genetic resources. This study aimed to identify and define the genetic structure of duplicates in cassava germplasm from various Embrapa research units, using single-nucleotide polymorphism (SNP) markers. We evaluated 2,371 accessions with 20,712 SNPs. The identification of duplicates was performed based on multilocus genotypes (MLG), adopting a maximum genetic distance threshold of 0.05. The population structure was defined based on discriminant analysis of principal components (DAPC). A total of 1,757 unique and 614 duplicate accessions were identified. The redundancy of the collections ranged from 17 % (Belém, PA-Brazil) to 39 % (Petrolina, PE-Brazil), with an average of 21 %. This redundancy between different research units is probably due to the historical sharing of accessions, as well as collections carried out in the same region, or even to the intense germplasm exchange between farmers with different genotype names. In terms of genetic structure, the 250 principal components explained 88 % of the genetic variation of the SNP markers and defined the hierarchical structure of the duplicate cassava germplasm in 12 groups. Since heterotic groups have not yet been identified for cassava, crosses between accessions of the 12 DAPC groups may be promising. All MLGs were allocated within the same DAPC group, corroborating duplicate analyses yet still revealing high variability between groups that were quite distinct based on the first two discriminant functions. Our results contribute to optimizing the conservation of genetic resources, together with understanding diversity and its use in crop improvement.

No abstract

Demand for the development of cassava varieties with different native starches has guided the search for these characteristics in the germplasm of Manihot esculenta Crantz. Therefore, the objective of this study was to estimate the genetic diversity of cassava accessions for root and starch granule characteristics to guide the future industrial application of this species. Starches from 56 accessions were evaluated for the number of granules in 1 g of starch (NTG), area (AG, μm), length (LG, μm), width (WG, μm), starch granule roundness (Round), dry matter content in the roots (DMC, %), pulp color (PulCo), and cyanogenic compounds (HCN). Images captured by light microscopy were used to determine the average phenotypic values, and these were further analyzed by principal component analysis (PCA) considering mixed data (quantitative and qualitative). Significant differences between the cassava accessions for all traits measured revealed wide variability in starch granule characteristics. Four diversity groups with better fitness for the classification of cassava accessions based on PulCo were identified, in comparison with HCN. Accessions with differential starch characteristics were identified, and crossings for the generation of segregating populations in order to obtain table and industry varieties have been proposed.