Estudos genéticos em germoplasma de mandioca no estado do Espírito Santo, Brasil

Abstract

Cassava (Manihot esculenta Crantz), a tropical root rich in carbohydrates, is one of the most traditional agricultural crops, of socioeconomic relevance, cultivated in all regions of Brazil. Thus, the objective of this study was: (1) to evaluate the diversity, estimate the genetic parameters and quantify the gains with the selection of cassava genotypes, using the methodology of mixed models (REML/BLUP) and (2) to evaluate the structure population and genetic diversity in cassava genotypes using SNP markers via DArTseq technology and morphoagronomic descriptors. 106 and 87 cassava genotypes were evaluated in studies 1 and 2, respectively, in experiments carried out in the 2018/2019 and 2019/2020 agricultural years in the cities of Mimoso do Sul - ES, Venda Nova do Imigrante - ES and Linhares - ES. The parameters evaluated were: shoot height (HPA), total number of tuberous roots (NR); commercial root weight (PESORC); total root weight (PESORT), commercial root weight/total root weight ratio (PESORR); root cortex color (RCC), root pulp color (CP) and cooking time (TC). Estimates of genetic parameters showed the existence of genetic diversity and selection potential of the studied cassava genotypes. The highest heritabilities (h 2 = 0.90; 0.75 and 0.75) were recorded for the traits total root number (NRT), commercial root weight (PESORC) and total root weight (PESORT), respectively. The gains with selection were positive for all characteristics evaluated. Greater selection gains were observed (GS = 21.77% and 20.15%) for traits PESORC and PESORT, considering animal and human consumption and (GS = 10.89% and 17.45%) for traits NRT and PESORC, when intended for exclusively human consumption. The population showed diversity, allowing the selection of 20 superior genotypes. Genotypes 82, 76, 46, T3 and 2 stand out for selection purposes for animal and human consumption. Genotypes 81, 69, 49, 12 and 2 stand out for selection purposes for exclusively human food. In chapter 2, a wide genetic variability was observed among the cassava genotypes studied. The Bayesian grouping did not separate the analyzed collection sites, but it shows the formation of three genetic groups (K=3). The Principal Coordinates analysis (PCoA) showed PC1 explaining 27.05% of the variation between genotypes, and PC2 explaining 23.05% of the total variation, confirming the low geographic differentiation between the sample groups, with the formation of a mix of genotypes between the microregions, mainly in the southern and mountain regions of the state. Principal Component Analysis (PCA) shows most positive eigenvectors, with the exception of the amylose characteristic. The first two axes of PCA explained 62.79% of the data variation. The first axis with 42.69% of the data variation, being positively correlated with the number of total roots (NRT), total root weight (PESORT), commercial root weight (PESORC) and shoot height (HPA), contributing for the first component, with emphasis on the variables PESORC (0.5658) and PESORT (0.5870), being the most representative variables. The SNP markers were effective for characterizing the genetic diversity and population structure of the genotypes studied, with the greatest diversity distributed within each geographic region, proving the importance of the role of farmers in the flow of genetic material, since they promote both the introduction of cassava varieties, such as the dissemination of local materials. The SR_G76, S_G2, T3 and N_G82 genotypes are possible superior genotypes, susceptible to selection when we consider the root productivity traits (PESORC and PESORT). The SR_G69 and SR_G64 genotypes presented high levels of starch, being possible superior genotypes when we consider this characteristic.