Systemization of sort and selection lines of wheat lines by quantitative signs in conditions of controlled environment with using cluster analysis

Abstract

The article presents the results of four-year studies (2013–2016) on the determination of the balance of quantitative characteristics of varieties and selection lines of winter wheat by the use of cluster analysis. The experiment used 2 sowing lines: early (1 September, SP-1) and late (October 1, SP-3). The main goal, when analyzing the best groups and clusters, was the search for winter wheat varieties and breeding lines (SL), which in the statistical analysis had high indicators of two components of the harvest – the number of grains in the ear (KZ) and the mass of grain from the ear (M1). Also we investigate how in the best groups and on dendrograms varieties and genotypes were distinguished, which were better both in the first and in the third term sowing, not only in terms of the number of grains in the ear and the mass of grain from the ear, but also on other economic and valuable features and by yield.

It should be noted that the study of varieties, or SL by the terms of sowing, gives unique opportunities to evaluate the material under study in terms of the formation of quantitative characteristics. We can form groups of varieties or SLs, where we will have information about the level of formation of signs at early, optimal, or late sowing. We can also distinguish genotypes as carriers of a high level of quantitative attributes, especially signs of yield potential in the early or late sowing, as the usual statistical analysis does not provide information on the balance of signs in one or another variety of winter wheat. We believe that cluster analysis makes it possible to identify varieties using the Euclidean metric in terms of the balance of quantitative attributes.

Based on the research carried out on the use of cluster analysis in the technology of breeding process for 2013–2016, selected varieties and constant selection lines are balanced by economically useful features that can be used as a promising breeding material and as parent components for hybridization.

The advantage of cluster analysis with respect to statistical analysis is that cluster analysis enables identification of varieties and breeding lines on a balance basis for economically useful features, and in the statistical analysis we can select and allocate only certain quantitative attributes according to their level of formation and variability and by yield.