Breeding wheat for weed-competitive ability: II–measuring gains from selection and local adaptation

Although the effect of local adaptation is well documented in evolutionary biology, few studies have quantified the impact of local adaptation in plant breeding. Decentralized plant breeding programs have the potential to harness local adaptation for crop improvement, but the effectiveness of such models is understudied. We quantified the ability of a decentralized participatory plant breeding program to improve Weed-competitive ability (WCA) in organic spring wheat. After four farmers in the northeast United States selected wheat populations for WCA and its correlated trait of early vigor, we tracked gains in selection and local adaptation. On average, farmers enhanced competitive ability of selected genotypes by 11.46%. Measured gains from selection for early vigor and early canopy cover, however, varied among testing environments. Gains in selection were highly related to the genetic correlation coefficient between selection and testing environment (r = 0.77 and r = 0.80 for early vigor and canopy cover, respectively). To accurately measure gains from selection for decentralized breeding programs, testing environments should be chosen that are similar to where selection took place. Inconsistent weed competition among site-years limited conclusions from the analysis of local adaptation for weed competitive ability. Detecting local adaptation in plant breeding, which typically uses a small number of selection cycles compared to evolutionary biology, likely requires many genotypes, environments, and years for adequate statistical power. The ecological complexity of weed competitive ability further complicates experimental design and challenges the ability to measure local adaption.