Management of agroecosystems has become increasingly difficult due to the rapid evolution and dispersal of herbicide resistant weeds. To better understand the evolutionary outcomes of continued, strong selection from exposure to glyphosate, the main active ingredient in RoundUp, we screened for variation in GR in Conyza canadensis, a highly self-pollinating weed common to no-till soybean fields. We hypothesized greater prevalence of GR in Ohio (GR first reported in 2002) compared to Iowa (GR first reported in 2011), greater GR in plants from agricultural habitats compared to non-agricultural habitats, and no early growth costs associated with resistance in the absence of glyphosate. We collected 74 biotypes (maternal seed families) per state with approximately half coming from soybean fields and half from non-agricultural habitats. GR was characterized using concentrations of 0x (control), 1x (840 g ae/ha; recommended application rate), 8x, 20x, and 40x. Biotypes with at least 80% survival at 1x were designated as “weakly resistant” and at 40x as “extremely resistant.” During summer 2016 and 2017, we used 10 susceptible, 10 weakly resistant, and 10 extremely resistant biotypes from Iowa in a common garden experiment to test for fitness related costs of GR in the absence of glyphosate.
Results/Conclusions
Within sampled areas of Ohio, 93% of sampled biotypes from agricultural sites were extremely resistant compared to 63% from non-agricultural sites. In contrast, 39% of sampled biotypes from agricultural sites in Iowa were extremely resistant compared to only 9% from non-agricultural sites. Although our sampling was targeted, these results suggest that extremely resistant biotypes are common and that non-agricultural sites serve as a habitat for extremely resistant plants. No significant patterns of growth differences have emerged between susceptible and resistant plants based upon early growth measures of rosettes at ~8 weeks. If early growth correlates well with lifetime plant fitness, and there is no fitness cost associated with GR, then the extremely resistant biotypes will persist within both agricultural and non-agricultural habitats. If there are no fitness costs associated with GR, we can expect those populations from agricultural sites to continue to disperse to non-agricultural habitats where they should persist. As GR continues to evolve and disperse across both agricultural and non-agricultural landscapes, farmers who rely heavily on herbicides will continue to face mounting challenges for effective weed management.