A comparison of statistical methods for assessing winter wheat grain yield stability

• Main Authors: A. F. Cheshkova, P. I. Stepochkin, A. F. Aleynikov, I. G. Grebennikova, V. I. Ponomarenko
• Format: Article
• Language: English
• Published: Institute of Cytology and Genetics of Siberian Branch of the Russian Academy of Sciences 2020-05-01
• Series: Vavilovskij Žurnal Genetiki i Selekcii
• Subjects: phenotypic stability; genotype x environment interaction; plant breeding; statistical methods; winter wheat
• Online Access: https://vavilov.elpub.ru/jour/article/view/2594

The multitude of existing methods for assessing the phenotypic stability of plants makes breeders be faced with the problem of choosing an appropriate variant. The purpose of this study was to compare different methods of analyzing the genotype x environment interaction and, on their basis, assess the stability of the yield of 7 varieties of winter wheat. The article compares 17 stability statistics by applying them to data obtained from agrotechnical experiments carried in 2009-2011 for evaluating the grain yield of 7 varieties of winter common wheat of Siberian selection (Novosibirskaya 32, Novosibirskaya 40, Novosibirskaya 51, Novosibirskaya 3, Novosibirskaya 2, Obskaya winter, Omskaya 6). Analysis of variance revealed a significant (p < 0.001) genotype x environment interaction in the experiments, which indicates a different reaction of genotypes to changes in environmental conditions. Genotypes were ranked according to the level of stability. Based on the analysis of the rank correlation matrix, the stability statistics were categorized in five groups. Recommendations were made on which group of methods to use depending on the objectives of the study. In the case when the goal of breeding research is the selection of the most biologically stable varieties with the minimum variance across a range of environments, one should use the methods of the static concept. If it is necessary to choose a genotype with a predictable reaction to changes of environmental conditions, corresponding to the calculated level or forecast, the regression approach is the most appropriate. The stability statistics generally identified Novosibirskaya 32 as the most stable variety from a biological point of view. The regression approach showed that Novosibirskaya 3 was the genotype with the smallest deviation from mean yield in all environments, while methods accessing the contribution of each genotype to the genotype x environment interaction defined Novosibirskaya 51 as the most stable variety.