Impact of Genetically Engineered Crops on Pesticide Use: US Organic Center Report Evaluation
News Nov 20, 2009
PG Economics welcomes the Organic Center (OC) latest release Impacts of genetically engineered crops on pesticide use: the first thirteen years by Charles Benbrook, which confirms the positive impact biotech crops have had on reducing insecticide use and associated environmental impacts. However, the OC’s assessment of the impact of biotech herbicide tolerant traits (HT) is disappointingly inaccurate, misleading and fails to acknowledge several of the benefits US farmers and citizens have derived from use of the technology.
For those reviewing the issues examined in the OC report, the following should be noted: • Confirmation: of biotech insect resistant (IR) impact on insecticide use: the OC paper confirms the findings of other work that the use of IR technology has resulted in important reductions in insecticide use on these crops that would otherwise have been used with conventional technology;
• Failure to acknowledge the environmental benefits arising from use of HT technology. These include facilitation of no/reduced tillage production systems2 which has resulted in important reductions in greenhouse gas emissions. For example, US HT biotech crops contributed, in 2007, to the equivalent of removing 9.48 billion pounds (4.3 billion kg) of carbon dioxide from the atmosphere or equal to removing nearly 1.9 million cars from the road for one year. In addition, whilst usage of broad spectrum herbicides, notably glyphosate (and to a lesser extent glufosinate) has increased significantly, usage of less environmentally benign products such as pendimethalin, metribuzin, fluazifop and metalochlor has fallen substantially, leading to net benefits to the environment 3;
• Inaccuracies: It uses assumptions relating to herbicide use on biotech crops in the US that do not concur with actual practice. As a result, it overstates herbicide use on US biotech crops significantly. For example, it overstates herbicide use on the HT crops of corn, cotton and soybeans for the period between 1998 and 2008 by 63.4 million pounds (28.75 million kg) of active ingredient;
• Misleading use of official data: The OC report states many times that the pesticide impact data is based on official, government (USDA NASS) pesticide usage data. Whilst this dataset is used, its limitations (namely not covering pesticide use on some of the most recent years and not providing disaggregated breakdowns of use between conventional and biotech crops) mean that the author’s analysis relied on own-estimates of usage and cannot reasonably claim to be based on official sources. As a result, the herbicide usage assumptions on conventional crops, if they replaced biotech HT traited crops, are significantly understated and unreliable.
Combined with the overstated use assumptions on HT biotech crops, it is therefore not surprising that the document concluded that biotech crops lead to an increase in US herbicide use. This contrasts sharply with the findings of PG Economics’ peer reviewed analysis4 that estimated that biotech crop adoption in the US has reduced pesticide spraying in the US, eg, by 357 million lbs (162 million kg: -7.1% 1996-2007) relative to what might reasonably be expected if the crops were all planted to conventional varieties;
• Weak approach: the approach of the OC report author is based on personal assumptions of herbicide use for biotech versus conventional crops and extrapolation of average trends in total crop active ingredient use (from an incomplete dataset). It also does not present any information about typical weed control regimes that might be expected in conventional systems. Not surprisingly, this resulted in significant over estimation of herbicide use on biotech HT crops (see above) and under estimation of usage on conventional alternatives. As such, the approach delivers unreliable and unrepresentative outcomes. It is noted that the OC author is critical of the approach used by other analysts5 to estimate the herbicide usage regimes that might reasonably be expected on conventional crops if biotech HT traits were not used in the US corn, cotton and soybean crops over the last thirteen years. The NCFAP/PG Economics approach, criticized by the OC report, is to present and estimate the conventional alternatives based on a survey of opinion from over 50 extension advisors in almost all states growing these three crops. Observers should note the key differences between the two approaches with the NCFAP & PG Economics approach being much more reliable and representative.
Given the complexities of agricultural production systems and the nature of weed and pest control systems, more detailed comment and critique of the OCS report is detailed below.
More detailed points relating to the OC report and the real impacts of biotech crops in the US
1. The claims made in the OC report about changes in pesticide use on corn, cotton and soybean
crops in the US during the first 13 years of biotech crops (1996-2008) are cited as being based on
data from the USDA NASS, which annually produces reports on ‘Agro-chemical usage on crops’.
The most recent of these reports was published in 2008 and covers usage in 2007. The NASS
surveys do not collect pesticide usage data on all field crops, in every year. The only one of the
three crops of soybeans, corn and cotton covered in the latest report covering 2007 was cotton.
The last time pesticide usage data on soybeans was collected related to 2006 and the last time data
on corn was collected covered 2005. Whilst the OC report draws on, and uses this data, its
analysis ‘fills in the missing years’ based on trends and extrapolates forward to 2008.
Furthermore USDA NASS data does not differentiate pesticide usage between biotech and non
biotech crops. In order to make such comparisons the OC report author made assumptions on
use of pesticides on each type of production for all years. Therefore the frequent reference in the
report to NASS-based data (notably for the last few years for total usage on each crop and all
usage differentiated into biotech versus conventional) is misleading and disingenuous to USDA
NASS – many readers might gain the impression that the report is using the government data
source when, in fact, crucial parts of the data used, on which conclusions and arguments are
drawn, do not draw from this source but are founded on the author’s use assumptions (see below
for additional comments).
2. The only comprehensive source of pesticide usage data on field crops in the US is DMR Kynetec,
an independent, private sector source of data on agricultural input usage in the US6. This dataset
goes back to 1998 and covers the period up to, and including 2008. It also provides data
disaggregated into usage on biotech versus conventional crops. A comparison of the actual
average usage volumes for herbicide active ingredient use per acre on biotech HT crops from this
dataset compared to the assumed usage rates in the OC report shows that the OC report
overstates herbicide active ingredient on biotech crops. Over the period 1998-2008, the OC paper
overstates the amount of herbicide active ingredient used on the biotech HT corn, cotton and
soybean crops by 63.4 million lbs (28.75 million kg) compared to actual usage recorded in the
DMR Kynetec dataset (equal to 6% of the total herbicide active ingredient used on these crops in
this eleven year period).
3. Assessment of the amount of pesticide usage that would be used on the three crops of corn,
cotton and soybeans in the US, if the entire crops were conventional requires the use of
assumptions, about what herbicides and insecticides might reasonably be expected to be used in
the absence of biotechnology. Applying usage rates for existing conventional crops is one
approach (eg, using the average values identified from the disaggregated data in the DMR
Kynetec dataset). However, this is likely to provide significant under estimates of what usage
would be in the absence of biotechnology, when the conventional cropping dataset used to
identify pesticide use relates to a relatively small share of total crop area. This has been the case
in respect of the US corn, cotton and soybean crops for several years. The reasons why this
conventional cropping dataset is unrepresentative of the levels of pesticide use that might
reasonably be expected to be used in the absence of biotechnology include--- (cut)
Detailed report at http://www.pgeconomics.co.uk/
2nd World Congress on Epilepsy and Brain Disorders
Jun 13 - Jun 14, 2019