Comparison of basic methods for estimating the size and stability of grain yield in winter wheat
Nikolay Tsenov, Todor Gubatov
Abstract: Background and purpose: The complex interaction of genotype with environments alters the ranking of comparable varieties of a given group with important productivity traits and qualities for wheat. The reason for this is the biological variation, the character of which is difficult to predict and evaluate. Changing a grain variety is a reason to look for ways to evaluate it despite the interaction genotype by environment. The purpose of the study is to establish an effective method for separating varieties according to the grain yield and the stability on the background of different environmental conditions.
Methods: We studied 24 varieties of winter common wheat in 5 country-specific locations for four years. Analysis of the genotype-environment interaction is performed according to several different commonly accepted statistical methods. After determining the indices, a ranking approach has been applied to evaluate each variety of the group. An attempt has been made to visualize the results of the scatter plot to group the varieties according to their grain yield and its variation to the environmental factors.
Key results: Sorting varieties using the applied statistical methods is different. This is a prerequisite for a serious analysis of the information they carry separately and when grouped together. The detailed analysis of the visualization methods shows a number of regularities. The variety points by each of the methods are distributed "evenly" across the top. The location of the varieties in the individual areas of the figures is almost analogous to the application of the various visualization methods.
Conclusions: The evaluation of a variety of any group by statistical indices is correct and fully applicable to distinguish their value. The grain yield of the particular variety in a given group is relative to the background of the group in which it is tested. The use of simple approaches such as regression coefficient (bi), deviation from the regression line (σ2) or coefficient of variation (CV) are closer to the capabilities of large statistical programs developed for these purposes.
Methods: We studied 24 varieties of winter common wheat in 5 country-specific locations for four years. Analysis of the genotype-environment interaction is performed according to several different commonly accepted statistical methods. After determining the indices, a ranking approach has been applied to evaluate each variety of the group. An attempt has been made to visualize the results of the scatter plot to group the varieties according to their grain yield and its variation to the environmental factors.
Key results: Sorting varieties using the applied statistical methods is different. This is a prerequisite for a serious analysis of the information they carry separately and when grouped together. The detailed analysis of the visualization methods shows a number of regularities. The variety points by each of the methods are distributed "evenly" across the top. The location of the varieties in the individual areas of the figures is almost analogous to the application of the various visualization methods.
Conclusions: The evaluation of a variety of any group by statistical indices is correct and fully applicable to distinguish their value. The grain yield of the particular variety in a given group is relative to the background of the group in which it is tested. The use of simple approaches such as regression coefficient (bi), deviation from the regression line (σ2) or coefficient of variation (CV) are closer to the capabilities of large statistical programs developed for these purposes.
Keywords: correlations; genotype-environment; grain yield; stability; variance; wheat
Date published: 2019-03-22
Download full text