Hard lessons emerge from new study on the impacts of GE crops on pesticide use

I have worked for many years on pesticide use, risks, and regulation, as well as the design, implementation, and benefits of Integrated Pest Management (IPM) systems.  Given that all of the commercially significant, first-generation traits in genetically engineered (GE) crops are related to pest management, the real-world impacts of GE crops on pesticide use has always been on my radar screen.

I started tracking the development of herbicide-tolerant technology in the late 1980s during my tenure as the Executive Director of the NAS Board on Agriculture (1984-1990).  Even back then, years before the technology’s commercial launch in 1996, weed management experts were expressing concern that glyphosate-tolerant, Roundup Ready (RR) crops could lead to the emergence of resistant weeds. 

In the first years of use, RR crops triggered a shift in herbicide selection from several low-dose imidazolinone and sulfonylurea herbicides to a relatively high-dose herbicide, glyphosate. USDA pesticide use data released in the late 1990s clearly reflects such shifts.  This is why I was skeptical from the start over claims that GE crops would and were reducing the pounds of herbicide applied on the three major RR crops.

To track changes in pesticide use brought on by GE crops, I started work in 2002 on a model that quantifies GE-crop-pesticide use impacts for both herbicide-tolerant, RR crops and insect-protected, Bt corn and cotton.  The model uses publicly accessible USDA data on pesticide use and GE crop adoption.  My first report based on model results estimated the impact of GE crops on pesticide use over the first nine years of use (1996-2004), and a 2009 report extended the analysis through the first 13 years.  Each analysis documented increasingly high rates of herbicide use on GE acres, compared to acres planted to non-GE cultivars.  It is now regrettably clear that the still-common claim that GE crops are reducing herbicide and overall pesticide use is no longer true.

My model and basic findings through the first 16 years of commercial use have now been published in the peer-reviewed, open access journal Environmental Sciences Europe.  I chose this journal because it has been publishing the major analytical papers from a series of international conferences on the impacts of GE crops on large spatial scales.  This June in Bremen, Germany, I attended the third conference in this series and presented the preliminary results of my 16-year analysis.  Another advantage — this journal provides free, online access to papers, facilitating the rapid dissemination of results.

The paper is available online, as is a WSU press release announcing its publication.  Because the paper is lengthy and technical, I prepared a short document that highlights the major findings.

Limited, publicly accessible data make the job of quantifying the impact of GE crops on pesticide use much more complicated than it needs to be.  When the USDA’s National Agricultural Statistics Service (NASS) surveys pesticide use on a corn, cotton, or soybean field, it does not ask the growers two simple questions – Did you apply these pesticides on land planted to a conventional or GE crop variety? And if GE, which variety?

Budget cuts in USDA have created another hurdle and source of uncertainty.  In the 1990s and through around 2005, NASS had ample funds to annually survey pesticide use on these three major crops.  But starting around 2006, NASS could afford to cover only a subset of these crops each year.  As a result, corn was not surveyed from 2005 until 2010, cotton was surveyed only in 2007 and 2010, and new soybean pesticide use data has not been released since 2006 (it will be next summer).  To deal with gaps between NASS surveys, I assume that pesticide use changes in equal, annual increments.  Since the last survey and through 2011, the pounds of pesticides applied are extrapolated conservatively, based on recent trends.

Despite these uncertainties, the overall trends in pesticide use on GE crops are clear.  Herbicide use is much greater on GE acres compared to conventionally managed acres planted to non-GE cultivars.  The spread of resistant weeds has been driving herbicide use up for a decade, and will continue to do so for years.

In order to deal with resistant weeds, farmers are being forced to expand use of older, higher-risk herbicides.  To stop corn and cotton insects from developing resistance to Bt, farmers planting Bt crops are being asked to spray the insecticides that Bt corn and cotton were designed to displace.

Without doubt, GE crop technology has profoundly changed corn, cotton, and soybean pest management, but the unintended impact on pesticide use is a harsh reminder that farmers should not put all their eggs in one pest control basket.