Study: GE crops not driving herbicide-resistant weeds, but still cause for concern

Among the many controversies spurred by genetically engineered crops, one of the biggest is whether herbicide-resistant plants prompt herbicide-resistant weeds. In a new study, published in the December 2017 issue of the journal Weed Science, University of Wyoming weed scientist Andrew Kniss finds that GE corn does not produce increased herbicide resistance in weeds relative to non-GE crops, but that soybean and cotton plantings do — but only to a limited extent.

Overall, Kniss says, herbicide-resistant weed evolution appears generally not to have accelerated with the spread of GE plants. Nonetheless, Kniss maintains that the current trajectory for herbicide usage and its connection to resistant weeds is unsustainable, and that non-herbicide interventions should become a greater part of weed management. “If we continue to see three new species per year resistant to whatever herbicide mode of action occurs in the U.S., we are going to run out of herbicide tools to combat those weeds,” Kniss says.

Kniss says he decided to study the issue based on the degree of hype around the question of “super weeds” — which he says is a misnomer for herbicide-resistant weeds — and the dearth of in-depth statistical analyses. He wanted to come up with a more rigorous understanding of the connection, and his paper is the culmination of four years of work.

Kniss’ relied on two hearty data sets: the USDA National Agricultural Statistics Service’s (USDA-NASS) comprehensive herbicide application tracking from 1990 to 2015; and the Herbicide Resistance Action Committee’s (HRAC) International Survey of Herbicide Resistant Weeds’ record of herbicide-resistance cases.

Glyphosate, used in Monsanto’s Roundup, is often at the center of the debate over resistant weeds, and it figures prominently in Kniss’ study. For corn growers, who use glyphosate but as just one herbicide among others, the diversity of weedkillers helps prevent resistance in weeds by restricting the opportunities for biological selection of resistant weeds. But in GE soybean and cotton fields, in which glyphosate is the dominant herbicide, resistant weeds were better able to take hold.

“With corn we just added one more tool to the [weed-control] toolbox, but with cotton and soy we actually replaced the tools with glyphosate,” Kniss says. In the later years of his study, a more diverse herbicide treatment in some of the soy and cotton fields seemed to reduce the instance of resistant weeds.

UC Davis weed scientist Brad Hanson called Kniss’ finding that herbicide-resistant weeds had generally not accelerated with the use of genetically modified crops important, but he echoed Kniss that the overall trend in this area is not good.

Hanson says that given the tradeoffs with different weed-eradication strategies – the effects of tilling versus herbicide use, for instance – it’s imperative that the decision about which approach to use should be tailored to local situations. And Kniss and Hanson both emphasize that, while herbicides will certainly be a part of agriculture’s future, new non-herbicide management techniques for weeds should be encouraged. They both mention robotics and artificial intelligence as particularly exciting possibilities for lessening dependence on herbicides. For robotic weed elimination, high-value crops like strawberries and lettuce will probably be where the technology matures.