Predicted and realized gains of different selection strategies in soybean populations: a comparative study

Abstract

Genetic gains from selection on yield were predicted in three soybean populations using original data and data corrected by the variation in agronomic traits of check cultivars. Predicted gains were compared to realized gains after correction for agricultural season effects -caused by environmental differences during the prediction and checking growing seasons -to indicate the best selection strategy and check the influence on yield of selection on the other traits assessed in the field. Individual selection, selection among and within families and combined selection were used for grain yield alone and for simultaneous selection on grain yield, plant height and number of days to maturity traits. Selection was made on F5 generation plants grown in an experimental field in Viçosa, Minas Gerais State during the 1997/98 agricultural season. The experiment used intercalated check cultivars. The results indicated lower within family heritabilities compared to the other types. Little variation was detected in gain prediction due to data correction. At field level, simultaneous selection performed better for grain yield improvement. Simultaneous gains for number of pods per plant, number of seeds per pod and mean weight of one seed, which are primary yield components, were obtained only when selection for the three components was balanced.

Analysis of direct and indirect selection and indices in soybean segregating populations

“…Similar amplitudes of gains were obtained by Reis et al (2002), in the selection among and within families as much as in the mass selection for NDM, PHM and NP. For the latter two, the authors obtained some superior gains.…”

“…For the latter two, the authors obtained some superior gains. Lower gains were observed by Reis et al (2002) regarding trait GY.…”