The objectives of this research were: to compare productivity and adaptation of conventional and glyphosate-resistant genetically engineered soybean genotypes, from different maturity groups, developed by the breeding program at Embrapa Soja for Paraná State, Brazil; to study the relative importance of the effects of location, year, cultivar and their respective interactions; and to verify the possibility of stratifying the State in more homogeneous regions, in order to reduce the number of locations for line competition experiments. Paraná State yield data from regional experiments carried out in randomized complete block design from 2001 to 2005 were used. The possibility of stratifying the State in homogeneous regions and discarding similar locations was checked by the significance of genotype x environment interaction among locations. Yield was not significantly different between genetically engineered and conventional soybean, regardless of the maturity group. The effect of location was more important than the effect of year in the environment composition. The stratification of Paraná State in regions for yield testing soybean lines brought neither significant nor consistent advantages for the evaluated years. Only locations within the South region showed some degree of similarity.
Glycine max; adaptation in soybean; environmental stratification; plant breeding; soybean breeding
