By Mark Loux, Ohio State University Extension herbicide specialist
We have been screening waterhemp populations over the past few years for their response to various herbicide sites of action. Our overall goal is to get a sense of the types of resistance that have developed or may be developing, so that growers and agronomists know what they should be looking for. We summarized some of this in a fairly lengthy C.O.R.N. article last year. We recommend that readers also view the Take action video, “Why care about metabolic herbicide resistance” by Pat Tranel, University of Illinois. One of the discussion points in this video is that metabolic resistance is likely to be extremely variable, resulting in waterhemp with resistance to anywhere from one to 5 sites of action, and any combination of these. This is in addition to the target-site based resistance to glyphosate, group 2, and group 14 that already exists in most populations. Reminder that the Take Action Herbicide Classification chart is a great tool for determining herbicide site of action group number.
In the previous C.O.R.N. article, we covered results with populations collected through September 2019. Results with 2020 populations were overall similar, and we are still in the process of screening populations collected in 2021. We initially screened for resistance to glyphosate, imazethapyr (group 2), and fomesafen (group 14). Based on the consistent trends in that screening, and the metabolic herbicide resistance developing in states to the west of us, we dropped the glyphosate and imazethapyr and broadened the screen to include herbicide groups for which metabolic resistance is important. Screens of 2019 and 2020 populations included atrazine (group 5), mesotrione (group 27), 2,4-D (group 2), fomesafen, and s-metolachlor (group 15). We used foliar applications ofnemerged plants for the first four and measured mortality. For the s-metolachlor, we assessed control of plants emerging from seed in treated soil. In the 2022 screen we have dropped the fomesafen and added glufosinate. We have not included dicamba because using this in the greenhouse did not seem especially wise, given this it’s potential to volatilize and follow any type of air movement and ruin a lot of other research. We have also for several years submitted populations to a regional project where waterhemp has been screened for resistance to glufosinate and dicamba. Our research has been funded by the Ohio Soybean Council, and the regional project is funded by USB. A summary of findings and implications follows.
• Most waterhemp populations in Ohio are resistant to glyphosate and group 2 herbicides, and growers should not count on these to provide any control.
• Depending upon the year of screening, up to 75% of the populations screened had some level of resistance to group 14 herbicides. This includes not only the POST herbicides — fomesafen, lactofen, acifluorfen — but also the residual herbicides — flumioxazin, sulfentrazone. Resistance to the residual herbicides is not manifested as a complete lack of control, but rather reduced longevity of control so that there is a loss of overall effectiveness. Growers trying to control waterhemp in non-GMO or RoundupReady soybeans will have to use group 14 herbicides for POST control, and should take steps to determine whether group 14 resistance is present before counting on these to work. The abundance of Enlist, XtendiFlex, and LibertyLink soybeans has reduced our use of POST group 14 herbicides overall, which may have slowed the development of additional resistance.
• We observed some level of variable response to atrazine, mesotrione, s-metolachlor, and 2,4-D among the 38 total populations we screened from 2019 and 2020. About half the populations had a reduced response (less than 80% control) to foliar application of atrazine at 1 lb ai/A, although most of them were still controlled at 4 lbs/A. About half of the 2019 populations had a reduced response to mesotrione at 3 oz product/A, but all were controlled at 4 times this, and all of the 2020 populations were controlled by both rates. The response to 2,4-D was more difficult to assess, because it appeared that the 1X rate of 0.75 lb ae/A was overall less effective in the greenhouse than is usually is in the field. Regardless, it appeared that a couple of populations had reduced response to 2,4-D. The same issue with rate occurred for the screening with s-metolachlor. We observed a similar issue with rate for the s-metolachlor screen, where the 1X rate of 1.5 lb ai/A appeared overall less effective than it should have been. But – we still observed variable response among populations to both rates, and 20 to 33% of populations had a reduced response to the 4X rate.
• A summary of the preceding paragraph would be that we have evidence of resistance developing to those 4 herbicides, with the caveat that we did not do the necessary follow up research to confirm this (does response and heritability studies). These are mostly populations collected during our annual preharvest survey, from fields where there appeared to be indicators of resistance. Some populations were poorly controlled at the standard herbicide use rate, and for some of these, using four times that rate still did not result in more than 80% control. If we have metabolic based resistance developing, it is following the predictions of Tranel from Illinois. The presence of this type of resistance for one site of action does mean that resistance to the others occurs. However, there were several populations with reduced response to all four sites of action we used, for which metabolic resistance is an issue.
• The regional project screening did not find reduced response to dicamba or glufosinate in the populations we submitted. Our feeling is that it’s probably a matter of time until we have populations in Ohio with resistance to these herbicides, and also 2,4-D, given the intensive use of these in POST herbicide programs for soybeans.
Putting this in perspective, waterhemp continues a trend of developing resistance to any herbicide used against it repeatedly. Metabolic resistance introduces the additional problem of having a broader assortment of herbicides and herbicide mixtures select for a single mechanism that confers resistance to multiple sites of action, reducing the effectiveness of a number of control options at once. It won’t be enough to alternate herbicide sites of action (which doesn’t happen much anyway) and mix multiple sites of action. Herbicide management does not go far enough toward preventing resistance problems in waterhemp. Ultimately, we need to scout fields often enough to discover control problems, which could be due to resistance, and prevent plants from going to seed.