This study proposes a strategy for improving the bulk method within progenies in self‐pollinated plants. The proposal focuses on the use of a multigeneration index, which considers the whole information across generations and the covariance among successive generations with a view toward identification of the best progenies. The results of the simulation and analysis of field data from common bean (Phaseolus vulgaris L.) were concordant and showed that the estimates of the accuracies were near 85% using the multigeneration index and 70% for the traditional methodology based on comparisons among progeny means of the generation F3:6 or of the last generation evaluated. Another advantage of the index is mitigation of the effect of the genotype × environment interaction, because it considers all the generations of assessment, thus involving interaction of the progenies with the sowing times. When selection is performed using only the information from the last generation assessed, performance of the progenies is inflated by this interaction. The efficiency of the strategy proposed leads to a 12% gain in relation to that which is normally used. This gain is considerable and equivalent in most cases to that which would be obtained through one additional selection cycle. It should be highlighted that this increase is obtained without any additional cost on experimentation, but simply through changing the procedure of identification of the best progenies after successive generations.

Caracterização biométrica de frutos e sementes de mirindiba (Buchenavia tomentosaEichler R E S U M OAs espécies arbóreas tem potencial para aproveitamento em diversos setores e cadeias produtivas. Neste contexto, a biometria constitui-se em uma importante ferramenta para identificar as principais características intrínsecas ao fruto e a semente. Portanto, objetivou-se com este estudo caracterizar biometricamente os frutos e sementes de mirindiba e inajá e, correlacionar com os parâmetros dos frutos e sementes de cada espécie, a fim de obter informações sobre a variação das características biométricas. Analisaram-se 100 frutos frescos e 100 sementes, provenientes de 10 árvores de cada espécie, da região do sul do Piauí, quanto ao diâmetro equatorial, comprimento longitudinal, massa fresca, massa fresca da polpa do fruto e da semente, volume do fruto, índice de volume das sementes, umidade da polpa e umidade da semente. As características biométricas foram analisadas mediante distribuição de frequência, calculando-se o coeficiente de correlação de Spearman (rS) e o nível de significância através do teste t. Os frutos e sementes de mirindiba e inajá apresentam variabilidade para a maioria das características biométricas avaliadas, bem como correlação positiva predominante entre elas. Os frutos apresentam grande potencial econômico pelo alto rendimento de polpa observado, sendo possível a seleção de frutos com maior teor de polpa por características como o peso e tamanho do fruto.Palavras-chave: biometria, Cerrado, correlação Spearman, espécies arbóreas. A B S T R A C TTree species have potential for use in various industries and supply chains. In this context, biometrics constitutes an important tool to identify key features intrinsic to the fruit and seed. Therefore, the aim of this study was to characterize biometrically the fruits and seeds of mirindiba and inajá and to correlate the fruit and seed parameters of each species in order to obtain information on biometric variation. Equatorial diameter, longitudinal length, fresh pulp, fresh weight of the fruit pulp and seed, fruit volume, volume index of seeds, humidity of the pulp and seed moisture collected in south of Piaui, were analyzed, considering 100 fresh fruits and 100 seeds from ten trees of each species. Biometric features were analyzed by frequency distribution, calculating the Spearman correlation coefficient (rs) and the level of significance using the t test. The fruits and seeds of mirindiba and inajá exhibit variability for most biometric characteristics evaluated, as well as predominantly positive correlation between them. The fruits have great economic potential for the high yield of pulp observed, with the possible selection of fruits with higher pulp content by characteristics such as weight and fruit size.

No abstract

Genotypes x environment (G x E) interaction consists of different behavior of genotypes cultivated in different environments. This interaction occurs due to the performance variation of each genotype in different environments. To reduce the effect of the interaction in soybean crops, some studies have been reported in the literature to study their adaptability and stability. However, these studies are still scarce in Minas Gerais State. Thus, the aim of this study was to verify the adaptability and stability of soybean cultivars and identify the cultivars that contribute least to the G x E interaction in Minas Gerais. Six soybean cultivars were evaluated in 9 different environments. The plots were composed of 4 rows of 5 m with a spacing of 0.5 m between rows, and only the two central rows were harvested. The inoculation with Bradyrhizobium japonicum was performed through sowing furrow. The fertilization followed the recommendations of the Soil Fertility Commission of Minas Gerais. Grain yield was evaluated in kg/ha after conversion to 13% moisture. After individual analysis, the joint analysis was performed by grouping the phenotypic means by the Scott and Knott (1974) test. Wricke's ecovalence methodologies and the Annicchiarico confidence index were applied for the adaptability and stability analysis. The interaction was decomposed into a simple and a complex part. The cultivars BRSMG 820RR and BRSMG 760SRR have wide adaptability and stability. The first one presents a better index of confidence and a small contribution to the interaction.

ABSTRACT.Estimating genotype x environment (GxE) parameters for quality and yield in soybean seed grown in different environments in Minas Gerais State was the goal of this study, as well as to evaluate interaction effects of GxE for soybean seeds yield and quality. Seeds were produced in three locations in Minas Gerais State (Lavras, Inconfidentes, and Patos de Minas) in 2013/14 and 2014/15 seasons. Field experiments were conducted in randomized blocks in a factorial 17 x 6 (GxE), and three replications. Seed yield and quality were evaluated for germination in substrates paper and sand, seedling emergence, speed emergency index, mechanical damage by sodium hypochlorite, electrical conductivity, speed aging, vigor and viability of seeds by tetrazolium test in laboratory using completely randomized design. Quadratic component genotypic, GXE variance component, genotype determination coefficient, genetic variation coefficient and environmental variation coefficient were estimated using the Genes software. Percentage analysis of genotypes contribution, environments and genotype x environment interaction were conducted by sites combination two by two and three sites combination, using the R software. Considering genotypes selection of broad adaptation, TMG 1179 RR, CD 2737 RR, and CD 237 RR associated better yield performance at high physical and physiological potential of seed. Environmental effect was more expressive for most of the characters related to soybean seed quality. GxE interaction effects were expressive though genotypes did not present coincidental behavior in different environments.