Divergent selection at ultra-low plant density for seed protein and oil content within soybean cultivars

Soybean [Glycine max (L.) Merr.] seed is a major source of protein for animal feed and oil for human consumption. Selection within elite soybean cultivars for the improvement of agronomic and seed traits is assumed to be ineffective due to the belief that cultivars are highly homogeneous. Previously reported data suggest that latent variation among the single plant selections within a cultivar exists and that mechanisms that generate de novo variation may also be present. The main objective of this study was to perform divergent single-plant selection at ultra-low plant density and investigate the presence of genetic variation for seed protein and oil within three elite soybean cultivars. A secondary objective was to investigate the variation for fatty acid composition. In 1995, single plants from the three cultivars were grown in a honeycomb design using a plant-to-plant spacing of 0.9 m. A total of 333 plants from 'Benning', 392 plants from 'Haskell'. and 371 plants from 'Cook' were evaluated. Divergent single-plant selection for protein and oil content was performed to select a total of 20 plants for high or low protein and 20 plants for high or low oil from each cultivar. The selected plants were further evaluated in replicated row-plot experiments for 3 years. Our results indicate that single-plant selection at low plant density was successful in discovering significant variation for seed protein and oil within each of the three soybean cultivars. For protein content. the magnitude of intra-cultivar variation between the highest- and lowest-protein lines averaged 19 g kg(-1) across the three cultivars and ranged from 13 to 24 g kg(-1). For oil composition, the magnitude of variation between the most divergent lines averaged 12 g kg(-1) across the three cultivars; and ranged from 9 to 14 g kg(-1). Significant variation among the selected progeny lines was also discovered for specific fatty, acid composition. The magnitude of intra-cultivar variation averaged from 6 to 29 g kg(-1) across the five fatty acids of soybean. The genetic variation discovered within the soybean cultivars is most likely due to latent variation and/or newly created variation. Our data provide evidence that single-plant selection at ultra-low plant density within elite cultivars can be effective in improving the seed composition of a soybean cultivar. (C) 2004 Elsevier B.V. All rights reserved.