Response to inoculation in Illinois bundleflower

Illinois bundleflower [Desmanthus illinoensis (Michx.) MacMillan] has potential as a pasture and grain legume in the Midwestern USA, but has proven to be somewhat problematic in inoculation response. We identified inoculant-quality Rhizobium strains from among 120 isolates trapped from 12 sites within the native range of Illinois bundleflower, then determined the effect of inoculation with four selected strains on plant dry matter and plant total N in growth chamber, greenhouse, and field experiments. We also studied nodule occupancy in the field. In field experiments at Salina (Kansas) and Becker (Minnesota) inoculated plants yielded 170% greater dry matter and total N on average than did uninoculated and N fertilized controls at each site in the seeding year. There were no differences in plant response among the Rhizobium strains used, uninoculated plants were devoid of nodules, and the response to inoculation was similar for the three plant accessions considered. In the second growing season, plant regrowth was different at Salina and Becker. At Salina, all plant accessions regrew and no differences were found between inoculated and control treatments in the second year. At Becker, toward the northern extreme of the range for Illinois bundleflower, only a MN ecotype regenerated in the second year, and inoculation was a requirement for plant persistence. Regrowth was similar with each of the four inoculant strains used. While 90% of the rhizobia recovered from nodules in the seeding year at Becker had genetic fingerprints similar to the inoculant strains, in the second year this percentage had declined to 70%. Our results highlight the importance of Rhizobium strain selection if the use of native legumes is to be increased.