Larry C. Purcell1, C. Andy King1, and Thomas E. Carter Jr.2. (1) University of Arkansas, 1366 W. Altheimer Drive, Fayetteville, AR 72703, (2) USDA/North Carolina State University, 3127 Ligon St., Raleigh, NC 27695-7631
Under drought conditions in the field, a few soybean genotypes have been noted to have delayed wilting compared with most other genotypes of similar maturity. The reasons for the delayed wilting and the importance of this trait for conferring agronomic drought tolerance are not known. We hypothesized that delayed wilting was due to decreased radiation use efficiency (RUE) under well-watered conditions relative to genotypes with fast wilting. Field experiments were initiatied in Fayetteville, Arkansas and Clayton, NC in 2004 and 2005 comparing RUE among four slow-wilting genotypes and five fast-wilting genotypes based upon a priori classifications from previous experiments. At both locations, soybean was sown in 19 cm rows to achieve final population densities of approximately 50 to 60 plants m-2. Overhead irrigation was used throughout the season to supplement rainfall and prevent water-deficit stress. Once the crop was intercepting >90% of the incident radiation, a bordered 2-m2 section of plot was harvested at ground level, dried, and weighed. A second harvest was made 3 weeks later, and RUE was calculated as the increase in crop mass per unit of intercepted solar radiation (g MJ-1). At Fayetteville in 2004, slow wilting genotypes all had significantly lower RUE values than fast wilting genotypes. At Fayetteville in 2005, slow-wilting genotypes also generally had lower RUE values than fast-wilting genotypes. At North Carolina, RUE values were not different between fast and slow wilting genotypes. The results from Fayetteville are consistent with our hypothesis that decreased RUE for slow-wilting genotypes under well-watered conditions would conserve soil moisture that could be utilized later in the season. Environmental differences between sites in North Carolina and Arkansas may have resulted in inconsistent RUE responses.