Cluster and factor analysis of morphophysiological characteristics of soybean varieties

The article presents the results of soybean breeding work focused on a detailed study of the interrelationships of morphometric characteristics in order to optimize the period of creation of competitive, highly adaptive varieties, new in genetic structure. 43 varieties of soybeans selected by Russian scientific institutions were included in the study. According to the average indicators of the varietal characteristics, a data matrix has been compiled for correlation analysis. The studied material is grouped into 13 clusters based on cluster analysis, which is confirmed by statistical processing. The obtained correlation coefficients are grouped into six hypothetical factors, the contribution of which to the accumulated variance was more than 5 %. A general assessment of the variability of the characteristics of the studied model population is presented, a slight variation in the content of protein (3.43 %), fat (6.89 %), ash (7.01 %), fiber (4.26 %) and nitrogen-free extractive substances (5.26 %) in seeds is noted. The average variation was characterized by plant height (16.91 %) and weight of 1000 seeds (14.55 %). The height of attachment of the lower bean (30.80 %), the number of productive nodes (50.10 %), the number of beans from the plant (43.59 %), the number of seeds (39.18 %) and yield (39.10 %) are highly variable. At a critical correlation coefficient value of 5 % 0.304, the correlation of the number of productive nodes with the number of beans (r=0.85), the number of seeds (r=0.52), yield (r=0.52); the number of beans per plant with the number of seeds (r=0.47), yield (r=0.47) and the content of BEV in seeds (r=0.35). The factor loadings of the soybean model population (Z1-Z4) were 74.8 %, with the first factor accounting for 3.56 %.