Sweet cassava is a food culture of great importance because it is a source of nutrition and energy for millions of people in tropical and sub-tropical regions. For that reason, genetic diversity and population structure studies are necessary in order to obtain more information regarding the evaluated genotypes, reassuring their use in future breeding programs. The objective of this study was to evaluate the genetic diversity and population structure of 51 traditional sweet cassava accessions collected in the Southern of Minas Gerais State, Brazil, using 20 microsatellite markers. All markers used to genotype the 51 sweet cassava accessions were polymorphic (PIC = 0.4080). Four sub-populations were identified using different methods (Bayesian analysis and multivariate analysis). The PhiPT (analogous Wright Fst) index of 0.073 indicated a moderated genetic variability among the studied traditional sweet cassava accessions. The dissimilarity index ranged from 0.097 to 0.560. Among the most divergent accessions stands out BGM 690, BGM 655 and BGM 660, which are the most recommended for obtaining a heterosis in order to increase yield production.

Soybean is a peculiar crop due to its photoperiod sensitivity. The introduction of this crop to a new production region requires a detailed study of its adaptability to select the best genotypes with high production stability for the region. In Mozambique, the demand for soybean grain has been growing. However, national production is low due to little knowledge of the performance of genotypes. The objective of this study is to evaluate the grain yield of 5 soybean genotypes and evaluate genotype-byenvironment (GxE) interactions by the GGE biplot method. The trials were performed at the stations of Namapa (District of Eráti, province of Nampula), Ribáuè (District of Ribáuè, Province of Nampula), and Montepuez (District of Montepuez, Province of Cabo Delgado) in 3 seasons from 2017 to 2020. A randomized complete block design was adopted with 4 replications and 5 treatments: Wima, Wámini, 10E, Safari, and Zamboane cultivars. The results of the joint analysis of variance (p<0.05) showed a complex GxE. According to the GGE biplot method, the 10E genotype was the ideotype. Wámini was the worst genotype. PC1 was 89.64 and PC2 was 8.26. Thus the GGE biplot methodology proved to be efficient since the sum of the first two principal components (PCs) was 98.26%.