Assessment of risks with herbicide tolerant wheat varieties

December 9, 2005, Via Agnet/ISAA CropBiotech Net

Wheat varieties grown all over the world may be classified as conventional, produced by genetic engineering, or generated by mutagenic techniques. With three such methods in place, Robert K.D. Peterson and Leslie M. Shama of Montana State University carry out A comparative risk assessment of genetically engineered, mutagenic, and conventional wheat production systems using the risk assessment paradigm. Their paper appears in the current issue of Transgenic Research.
Among others, researchers found that the herbicides glyphosate and imazamox, which are used to protect wheat, actually present lower human health and ecological risks than many other herbicides associated with conventional wheat production systems. The researchers acknowledge that their assessment is not comprehensive, but state that the approach they presented demonstrates the potential risk trade-offs (especially for herbicides) when implementing the newer biotechnologies.

A comparative risk assessment of genetically engineered, mutagenic, and conventional wheat production systems
Robert K.D. Peterson* & Leslie M. Shama
Agricultural and Biological Risk Assessment, Montana State University, 334 Leon Johnson Hall, Bozeman, MT,
59717-3120, USA
Received 25 March 2005; accepted 27 July 2005
Some comments made about herbicide risks:
Human dietary risk
Imazamox was the only herbicide evaluated that is exempt from having a USEPA tolerance (USFR, 2003a); therefore, there is no established chronic RfD and no dietary risk estimate. Dietary risk as a percentage of the chronic RfD for the other herbicides ranged from 0.005 to 565.5 for the U.S. population (Table 1). Children from 3 to 5 years of age represented the most sensitive U.S. subpopulation for each herbicide. Dietary risk for that subpopulation ranged from 0.01 to 1336.8% of the respective chronic RfD. Clodinafop was the only herbicide active ingredient to exceed 100% of the chronic RfD. Ten of the 16 herbicides had greater dietary risks than glyphosate. All of the herbicides had greater dietary risks than imazamox.

Applicator risk
Applicator risk ranged from 0.004 to 109.9% of the chronic toxicity NOAEL (Table 2). Thirteen of the 15 herbicides had greater applicator risks than glyphosate. As with dietary risk, all herbicides had greater applicator risks than imazamox.

Ecological risk
Non-target terrestrial plants. All 13 active ingredients for which data were available had seedling emergence risk quotients (RQ’s) greater than glyphosate, which was not unexpected given that glyphosate was the only herbicide assessed which is practically non-toxic with respect to seedling emergence (Table 3). Seven of the 14 herbicides for which data were available had vegetative vigor RQ’s greater than glyphosate (Table 3).

Aquatic organisms. Fourteen of the 15 herbicides for which toxicity data were available had vascular aquatic plant RQ’s greater than glyphosate. Four of the 16 active ingredients had aquatic invertebrate RQ’s greater than glyphosate. Five of the 16 active ingredients had aquatic vertebrate RQ’s greater than glyphosate.

Herbicide concentrations in groundwater. Eleven of 16 herbicides had higher predicted groundwater concentrations than glyphosate. Seven herbicides with lower maximum single use rates than glyphosate had higher predicted groundwater concentrations.


Homology of the CP4 EPSPS protein [conveying glyphosate herbicide tolerance] with native EPSPS protein. Lee et al. (2001) sequenced the modified EPSPS protein in glyphosate-tolerant maize and observed that it was 99.3% identical in its amino acid sequence to the native maize EPSPS protein (Lee et al., 2001).

Toxicity
Acute and subchronic effects. Harrison et al.(1996) examined the acute toxicity of CP4 EPSPS protein in mice by feeding high doses of soybean meal or seed containing the CP4 EPSPS protein at 572 mg kg BW administered once by gavage, which exceeds by 1000-fold the estimated consumption level for food products containing CP4 EPSPS protein. No adverse effects occurred in the mice dosed with CP4 EPSPS protein. Their body weight, cumulative body weight, and food consumption did not show significant differences between the control groups and CP4 EPSPS protein treated groups. Hammond et al. (2001) compared the responses of rats fed grain for 13 weeks from glyphosate-tolerant maize to the parental variety of non-transgenic maize and commercial varieties of non-transgenic maize. There was no observed toxicity and no differences in organ weight among all maize varieties. Lee et al. (2001) performed an in vitro test on mice to determine the toxicity of CP4 EPSPS protein in maize. A high dose of 45.6 mg of transgenic EPSPS protein per kg body weight was administered orally each day for 90 days. There was no observed toxicity from the protein (Lee et al., 2001). Chang et al. (2002) subjected Sprague Dawley rats to a toxicity test by orally administering 0.5 or 2.0 mg kg BW of CP4 EPSPS protein in saline solution three times per week for three weeks.

The dosages of CP4 EPSPS protein in this study were considered to be the approximate amounts of EPSPS protein in soybean consumed annually by humans. No toxicity was observed when compared to the saline control (Chang et al., 2002).

Discussion
Human risk
Both glyphosate and imazamox present lower risks than many other herbicides associated with conventional wheat production systems evaluated in this study. Because no acute or chronic toxicity for imazamox has been identified, dietary risk for it is essentially zero. The dietary risk for glyphosate also is low; 7 of the 17 herbicides had lower tier 1 dietary risk.

Ecological risk
Despite the variation in RQ’s, ranking each herbicide with respect to RQ for each ecological risk category allows us to draw several conclusions. For risks to duckweed, groundwater, and non-target plant seedling emergence, glyphosate had less risk than most other active ingredients. For risks to fish and waterflea, imazamox had less risk than most other active ingredients. Imazamox had greater risk than glyphosate for non-target plant seedling emergence, duckweed, and groundwater.
As with human risks, the differences in ecological risks were most pronounced when comparing glyphosate and imazamox to active ingredients with substantial market share, such as 2,4-D, MCPA, triallate, dicamba, and bromoxynil (USDA, 2003). Currently, the broadleaf herbicides 2,4-D, MCPA, dicamba, tribenuron, and bromoxynilare used on approximately 58, 26, 23, 20, and 13%, respectively, of total spring wheat acres in the US (USDA, 2003). The herbicides 2,4-D, MCPA, and bromoxynil had greater RQ’s than glyphosate for five, five, and four of the six ecological receptors evaluated. Dicamba and tribenuron had greater RQ’s than glyphosate for four and two of the six receptors, respectively. The herbicides 2,4-D, MCPA, bromoxynil, dicamba, and tribenuron had greater RQ’s than imazamox for four, six, three, five, and zero of the six ecological receptors evaluated (Table 3).