Comparison of Selection Systems for the Identification of Exotic Soybean Lines for Use in Germplasm Development

Abstract

Hybridization programs utilizing exotic soybean [Glycine max (L.) Merr.] germplasm might be more successful if selection systems were used to identify plant introductions (PIs) that combine well with adapted genotypes to produce high‐yielding breeding populations. The objectives of this study were to compare PIs selected for visual appearance vs. yield using a topcross procedure (TCP), compare the performance of PI parents with their progeny utilizing a TCP, to compare the performance of testers used in the topcross, and to identify PIs that combine well with adapted genotypes to produce high‐yielding progeny. Nine PIs selected for desirable visual appearance and 14 yield‐selected PIs were each crossed to four adapted lines in a TCP. The 92 F2 bulks were evaluated along with the 23 PIs and four adapted testers in three environments with two replications per environment. The F2 bulks produced from the yield‐selected PIs had a mean yield of 2669 kg/ha, which was significantly (P < 0.01) greater than the yield (2566 kg/ha) of the F2 bulks produced from the visually selected PIs. A positive correlation (r = 0.65; P < 0.01) between PI parent yield and PI group yield (mean of four F2 bulks with a common PI parent) demonstrated good association for yield between parent and progeny. Selecting PIs for yield, per se, appeared to be the more efficient method to identify desirable PIs for germplasm development programs. Our results indicate that the choice of a tester would not be critical if a TCP was utilized. The plant introductions PI 85469, PI 88306, and PI 88486 produced the three highest‐yielding PI groups and, therefore, are recommended for germplasm development programs in the mid‐Atlantic states. We suggest that these three PIs be evaluated in other geographic regions. The F2 bulks generally had acceptable lodging resistance, seed qualities, and seed weights.