Journal Article

Environmental stratification in trials of unbalanced multiyear soybean (Glycine max (l.) Merril) via the integration of GGE Biplot graphs and networks of environmental similarity

Rodrigues, Fernanda Cupertino, Bhering, Leonardo Lopes, da Silva, Felipe Lopes, Carneiro, Pedro Crescêncio Souza, Peternelli, Luiz Alexandre, Silva, Francisco Charles Santos

2022

Overview

Genotype × environment (GE) interaction can difficult soybean breeding programs to achieve the aim of more productive cultivars. Environment stratification is a way to circumvent this problem. However, multiyear data studies are difficult to work with, because they are, usually, unbalanced. GGE Biplot is an efficient method to find mega-environments, however, it allows for, at most, 30% of the unbalanced data. Thinking about how we could resolve this problem we came up with the idea of test a method that englobes GGE Biplot graphs, environment coincidence matrices and networks of environment. Wherefore, this work aimed to gather GGE Biplot graphs of a network of trials unbalance multiyear soybean via matrices of coincidence and networks of environment to optimize environmental stratification. Data from an experimental network of 43 trials was used, these experiments were implanted in 23 municipalities during the crop seasons of 2011/2012, 2012/2013, 2013/2014 and 2015/2016 in Brazil. The trials were implanted under an experimental block design with randomized treatments and approximately 30 genotypes were evaluated, of which most of these genotypes were not repeated between the trials evaluated. The GE interaction were statistically significant for all 43 trials. The step by step of our analyses was: GGE Biplots graphs were obtained; the environment coincidence matrices were calculated; the values of matrices were used to obtain the networks of environmental similarity. The study demonstrated that by the method was possible to identify, using unbalanced multiyear data, four mega-environments. The region under study can be represented by the municipalities of Palotina, Maracaju, Bela Vista do Paraíso and Rolândia. Therefore, integrating GGE Biplot graphs and networks of environmental similarity is an efficient method to optimize a soybean program by environment stratification.

• Publisher: Springer Nature B.V
• Subject: Cultivars / Genotypes / Glycine max / Graphs / Municipalities / Networks / Plant breeding / Similarity / Soybeans / Statistical analysis / Stratification