INCREASING THE EFFICIENCY OF BIOMETRICAL GENETIC-ANALYSIS IN PLANTS

The joint Cavalli test for mutual scaling was used in an experimental study attempting to reveal nonallelic interactions for nine quantitati ve traits of winter wheat. The experiment involved parental forms P-1 (cultivar Progress) and P-2 (cultivar Bogarnaya 56); the first three h ybrid generations from crossing P-1/P-2 (F-1, F-2, and F-3); the first backcrosses B-1 (F-1/P-1) and B-2 (F-1/P-2); and the offspring from t riple test crossing of F-2:L(1) (F-2/P-1), L(2) (F-2/P-2), and L(3) (F -2/F-1). Analysis of the observed generation means found no appropriat e genetic model for any trait. Adjustment of the data for average fert ility on the basis of the optimized moving mean method increased the e fficiency of the joint Cavalli test and allowed selection of an approp riate genetic model for eight out of nine traits studied. A six-parame ter additive dominant epistatic model sufficiently explained variation of three traits: plant height and main ear fertility and length. Epis tasis was not found for five traits: fertility, numbers of grains and ears per plant, number of grains in the main ear, and weight of 1000 g rains. These traits were sufficiently described by a three-parameter a dditive dominant model. A conclusion was made on the efficiency of con trol over soil heterogeneity in biometrical genetic analysis of plants .