VARIABILITY OF YIELD AND CHEMICAL COMPOSITION IN SOYBEAN GENOTYPES GROWN UNDER DIFFERENT AGROECOLOGICAL CONDITIONS OF SERBIA

Study of the interaction between genotype (G) and year (Y) provides good estimates of genotypes breeding values. In order to investigate the main effects of G, Y and G x Y interactions on yield and quality components of NS soybean genotypes, an experiment with genotypes of different maturity groups was carried out during three-year period. The average yield for all genotypes was 4,716 kg ha-1. Genotype, year and interaction G x Y had statistically significant effect on the yield, p<0.05, p<0.01. The highest grain yields per unit area had, on an average, was recorded in genotype Venera (4,962 kg ha-1), significantly higher than genotype Vojvodjanka (4,522 kg ha-1), p<0.05. The highest yield stability was recorded by genotypes of MG 0. The average protein content of all examined genotypes was 37.60%. Year, genotype and G x Y interaction had statistically significant effects on protein content. Genotype Galina had on an average the highest protein content (38.11%), significantly higher than genotypes Trijumf, Valjevka and Venera, p<0.05. Significantly higher protein content was achieved during 2008 and 2009 compared with 2010, p<0.05, p<0.01. Average protein yield for all genotypes was 1,711 kg ha-1. The highest protein yields had genotype Venera, significantly higher than genotype Vojvodjanka, p<0.05. The average oil content for all examined genotypes was 21.51%. The most favorable year for oil synthesis in the test period was 2008 (22.41%). Statistically significantly higher oil content was recorded in 2008 (22.41%) compared to 2010 (20.22%) and significantly higher than in 2009 (21.89%), p<0.05. Genotype Valjevka (21.78%) had on average significantly higher oil content than genotype Trijumf, p<0.05. The average oil yield for all genotypes was 1,014 kg ha-1 . On average the highest oil yield was recorded in genotype Venera. The yield was negatively statistically significant correlated with protein content, highly negatively significant correlated with air temperature and positively significant correlated by protein yield, oil yield and precipitation. The goal of the breeder was to create highly productive soybean varieties, followed by quality grain. This research can constitute the basis for further soybean breeding.