A common Iowa sight: a rig spreading herbicide on glyphosate-tolerant soybeans. Photo credit: clisenberg John Deere 4730 via photopin (license).

A year or so ago I was listening to a candidate (who will remain unnamed) for a federal office (that will remain unnamed). This candidate sounded reasonable and I agreed with most of what I heard – except for a call to label products made with genetically modified organisms (GMOs).

I challenged this assertion. There is no evidence, despite extensive testing, that consuming GMOs poses a health hazard. Labeling would feed unfounded fears.

Another person turned to me and said that may be so, but planting GMO crops that resist specific weed-killers and insects encouraged farmers to overuse pesticides, putting the environment at risk.

It only occurred to me later to ask: How would labeling fix that?

And in any case, recent research out of Iowa, Kansas, and Michigan state universities and the University of Virginia suggests the GMO-pesticide connection may be more complicated than we think. In some cases, planting genetically engineered (GE) crops may lead to more pesticide use, but in other cases it leads to less. And when a pesticide’s environmental impact is taken into consideration, the picture gets even cloudier.

The paper, published August 30 in the journal Science Advances, is unusual because it relies on actual pesticide use data from farmers growing the two most common crops: corn (maize) and soybeans. A market research company surveyed around 5,000 corn and 5,000 soybean growers each year from 1998 to 2011. That allowed the agricultural economists (Kansas State’s Edward Perry, Virginia’s Federico Ciliberto, Michigan State’s David Hennessy and ISU’s GianCarlo Moschini) to track the amount of pesticides applied to individual plots and estimate their impacts over several years.

“We essentially have ground-level data,” Moschini said. That means the researchers “do the comparison between farmers that have a plot where they plant genetically engineered seeds and one where they don’t. And then we look at the actual use of pesticides.”

Most previous studies looked at how much pesticides farmers who adopt GE crops use, then estimate what farmers would have used without GE technology. But in making the estimates, Moschini said, they calculate it based on the application rates recommended on the pesticide container. “These recommended doses usually are quite excessive. Farmers use far less,” he added, so estimates relying on manufacturer’s instructions tend to be inflated.

In making their comparisons, the team looked at the data in two ways: First, the raw amount of pesticide active ingredients applied, calculated by kilograms per hectare; second, by adjusting those numbers according to the pesticide’s calculated environmental impact. They examined three cases: herbicide use for soybeans and for corn engineered to tolerate glyphosate, popularly known as Roundup® after the most commonly known glyphosate brand; and Bt corn, engineered to resist European corn borer caterpillars.

There is little dispute that Bt corn has cut insecticide use, and the study does nothing to contradict that. On average, farmers planting Bt corn from 1998 to 2011 used about 0.13 kilograms per hectare (kg/ha) less than insecticide than farmers who didn’t adopt the technology, a difference of 11.2 percent. Weighting the results for environmental impact did little to change that outcome.

Farmers who planted glyphosate-tolerant (GT) corn also used slightly less herbicide – on average, about 0.03 kg/ha – 1.2 percent less than farmers who didn’t plant GT corn. There’s more to both of these figures, which we’ll get to in a minute.

Soybeans, with a side of glyphosate

The story was different for GT soybeans: Those farmers planting GT soybeans used, on average, 28 percent more herbicide (0.30 kg/ha) by weight than farmers who avoided GT soybeans. To me, less than a third of a kilogram seems like a small difference – but remember, we’re talking about averages, which means some farmers used more and some used less.

The studies measure total herbicide use, not just glyphosate, but the trend is clear that with adoption of GT varieties, glyphosate use has steadily risen, taking the place of other chemicals.

That’s where the economists’ other yardstick comes in. Essentially, not all pesticides are created equal. Some are more toxic to humans and animals than others and linger longer in the environment.

So to even the differences between farm chemicals, the economists considered their environmental impact quotient, an index that rates each herbicide or pesticide on its toxicity, depending on its impact on farm workers, consumers and the ecology. “Whether it does a good job” of summarizing a chemical’s actual risk or not “is an open question,” Moschini said, “but it’s an objective measure and it’s something we have numbers for for all the herbicides and pesticides that are out there.”

(In a National Public Radio story about the study, one weed scientist criticized the EIQ as a toxicity gauge. Moschini agrees it’s imperfect, but EIQ data was readily available and comparable to other studies. “I have absolutely no stake in defending the EIQ,” he said, but some rough calculations with other toxicity measurements methods shows little change in overall herbicide use trends.)

When herbicides are weighted according to their EIQ, there’s little difference in use between farmers planting GT soybeans and those planting non-GT soybeans, “reflecting the relatively lower EIQ values for glyphosate,” the authors wrote. In other words, when you consider glyphosate’s more benign environmental impact, the 28-percent gap in herbicide use essentially disappears.

In corn, the small, 1.2 percent average edge the GT variety had in lower herbicide use becomes a 9.8 percent margin when weighted for EIQ.

So it would appear that farmers planting GE soybeans use significantly more herbicide than those who don’t – unless you consider that glyphosate, the main herbicide those farmers choose, replaces other, more toxic herbicides. In corn, planting GT varieties appears to lead to slightly less herbicide use, but significantly less when glyphosate is compared to other, more toxic chemicals.

The “interesting” story

But average use over time hides the more interesting story, Moschini said. “The bigger story is that the impact has changed a lot over time. In the early years, in fact, there was big saving of herbicides, but in later years it has gone the opposite.” For both GT corn and soybeans, herbicide use increased steadily, rising sharply in the late 2000s.

A: Year-specific impacts of GT soybeans on herbicide use, by weight and EIQ, showing a steady increase in the later years. B: Year-specific impacts of GT corn on herbicide use, by weight and EIQ, also showing a steady increase in later years. C: Year-specific impacts of Bt corn on insecticide use and EIQ. For all panels, vertical bars denote 95 percent confidence intervals. From “Genetically engineered crops and pesticide use in U.S. maize and soybeans,” Edward E. Perry, Federico Ciliberto, David A. Hennessy and GianCarlo Moschini (August 31, 2016) Sci Adv 2016, 2:. doi: 10.1126/sciadv.1600850. Click to enlarge.

There are several possible reasons, all of which probably contribute. First, glyphosate is a cheap herbicide that became even cheaper after Monsanto Co.’s patent expired in 2000, leading to inexpensive generic versions.

Second, commodity prices were high in that period. “For farmers, the crop was much more valuable and there was further incentive to make sure they used as much herbicide as needed,” Moschini said, boosting consumption.

Finally, there’s the emergence of herbicide-resistant weeds. When glyphosate and GT crops first came out, one pass with herbicide probably was enough to kill weeds. In later years, that may have no longer been true, Moschini said. A chart in the paper shows that in 1998, 70 percent of the hectares planted to GT soybeans were sprayed exclusively with glyphosate. That started falling around 2005 and by 2011, had reached between 40 and 50 percent. Farmers were using additional, non-glyphosate weed-killer on more plots of their GT soybeans.

For GT corn, the percentage of hectares on which farmers exclusively sprayed glyphosate fell from around half to around 20 percent. The figures are “telltale evidence that the emergence of herbicide-resistant weeds is having a bite,” Moschini said.

So what are the study’s implications for the debate over GMOs and pesticide use? Moschini demurs. “We are just trying to inform the discussion. It’s easy to make statements predicated on ignorance and a lot of people do that in this area in particular.”

The bottom line is that using Bt corn clearly reduces insecticide use, while the picture is cloudier for herbicides. But “you have to think about the benefits.” Farmers use fewer resources to produce the same amount or more crops, and that efficiency gets transmitted through the system, presumably to the benefit of consumers.

That’s the important part for me. Yes, much of the corn and soybeans farmers are harvesting this fall will go to produce fuel or to feed cattle and hogs – which most of us eat – but some will go straight to consumers. In any case, we want to pay as little as possible for biofuel and food.

In some cases, GMO crops increase the use of pesticides – but often pesticides that are more environmentally benign than the alternatives. If consumers want greater yields and cheaper fuel and food, they’ll have to ignore the scare tactics surrounding GMO consumption.

Or we can forgo this technology and risk losing crops to weed and insect infestations, cutting yields and driving up food prices while perhaps encouraging farmers to use even more dangerous chemicals.

Those are our choices when it comes to large-scale agriculture.