Competitive distorting of sunflower plants traits and background traits for correction of productivity estimations

Low efficiency of breeding potential prognosis for individual plants from segregating populations is due mainly to display their different competitive ability in field cenosis. In conducted in Krasnodar and Armavir 2004-2005 years experiments it is determined that due to competition selection traits variation is accompanied by regular background traits variation. It ensures opportunities for correction of selection traits estimations. In particular, seed yields and oil yields covariations with seed oil content are positive if caused by genotypic differences. But that traits covariations are negative if are determined by different conditions of plant nutrition. Use of these different directions of covariations permitted academician V.S. Pustovoit to identify desirable genotypes with high precision. In conducting last ten years abroad full-sibs sunflower recurrent breeding much less number of population individuals are testing in S0 and S1 generations in creating lines trials. Accordingly probability of unique genotypes discovering is much less too. Efficiency of sunflower hybrids creation may be raised two recurrent breeding version combination: at first stage to increase desirable alleles frequencies in population by V.S. Pustovoit semi-sibs method and at second stage - to select full-sibs families from improved populations on the base of early yield testing.