The frog factor and the ingredients of alarm

Study on the effect of a glyphosate-based herbicide formulation on amphibian tadpoles is an example of how misinterpretation or misleading research can be quickly taken up and disseminated through the internet to swing popular opinion on pesticides.

Posted: August 11, 2009

Does typical use of glyphosate in agriculture or forestry really lead to widespread mortality of amphibians? The weight of available scientific evidence suggests that it does not, says a leading researcher. However, faulty research combined with over-extrapolation and extremist headlines can quickly mislead the public and policy-makers, putting the continued use of a herbicide which has been repeatedly considered as environmentally acceptable by many regulator agencies and scientific reviews at risk.

Dr. Dean Thompson of Natural Resources Canada discussed his perspective at the Canadian Weed Science Society meeting in Banff.

In a presentation to the Canadian Weed Science Society, Dr. Dean Thompson of Natural Resources Canada in Sault Ste. Marie, Ontario focused on a widely publicized 2005 study by Dr. Rick Relyea of the University of Pittsburgh that raised concern about the impact of glyphosate-based herbicides on frogs. The study drove headlines such as "Popular herbicide kills tadpoles" and lead to implications, intended or otherwise, that glyphosate-based herbicide formulations widely used in agriculture and forestry were linked to global amphibian declines.

Thompson and his colleagues publicly refuted the findings of the study, which claimed that that current application rates for the glyphosate formulation Roundup are highly lethal to many amphibian species. Thompson argued that the Relyea study did not represent typical use scenarios in agriculture or forestry, nor did it employ environmentally realistic dose and exposure levels that might occur in natural wetlands associated with typical use patterns in these major use sectors. In fact, they argued, all other major studies, including their own, indicate wetlands organisms including amphibians are not at significant risk from typical use of this herbicide.

"This situation was a classic example of what can happen when you put those ingredients together – environmentally unrealistic dose and exposure regimes, misinterpretation and over-extrapolation of singular study results and extremist headlines. The lay public reads the headline and accepts the related statements at face value and conclude that there must be a major problem here when in fact, if you look at the scientific details of that particular study, the weight of scientific evidence related to the specific issue or the general consensus of major scientific and regulatory reviews, that is clearly not the case."

Risk assessments should not be based on any single study, but on the weight of scientific evidence principle, says Thompson. "That means applying the cumulative knowledge base. It's critical that we understand that knowledge base and incorporate it into an overall assessment with a focus on environmental realism, instead of just going with what the latest experimental study has purported to show."

Factors behind alarm

There are several factors that caused Relyea's study to gain credibility in the eyes of the public, says Thompson. A key element was that simplistic derivative statements started to become fact through repetition in a manner similar to the way in which old wives tales become generally accepted as truth. However, in the modern era, the process is vastly accelerated through transmission via the internet. In contrast scientific verification of any study results is a much slower process involving independent repetition of the study by independent researchers or generation of other corroborating evidence leading to the same conclusion. Another element is that alarmist headlines and misinformation on the internet play to general public fear of pesticides and big industry.

Flaws in study

There is some truth to Relyea's study, says Thompson. Research conducted by Thompson's team and by others prior to the publication of Relyea's study confirm that amphibians are, in fact, one of the most sensitive species to formulated products of glyphosate, particularly those containing a particular type of surfactant generally referred to as POEA. The problem, he says, lies with Relyea's methodology and the inferences drawn from a single study being over-extrapolated.

The fundamental flaw in Relyea's study was the unrealistic aqueous exposure concentration employed which exceeds general concentration levels found in surface waters in Canada, the USA and internationally by more than 100-fold. Even the upper maximum of concentrations found in directly oversprayed forest wetlands are less than the level tested by Relyea by at least five-fold, says Thompson. Moreover, the data from the Relyea study were used to implicate glyphosate use in the global amphibian decline phenomenon without any other evidence to support that contention. "Most of us were taught in school that such extrapolation, particularly from a single study, is scientifically unacceptable," he says.

Although Relyea's results revealed that an aqueous glyphosate concentration of 3,100 parts per billion (ppb) was lethal to amphibian tadpoles and conform to other previously published concentration-response relationships, real-world data show actual glyphosate concentrations to be far less than that, says Thompson. "Average concentrations published from monitoring surface waters associated with roadside applications, U.S. agriculture, Canadian agriculture, and European Union orchard situations are less than 25 ppb, far below the test concentration employed by Relyea and also well below water quality guidelines established by the Canadian government to protect aquatic organisms. As such, we would not expect to see any substantial levels of mortality because both the magnitude and duration of real-world exposures are far less than those examined by Dr. Relyea."

Thompson's prior and subsequent research supports this. In a multi-tier study conducted prior to the Releya work, his team conducted laboratory studies showing that amphibian larvae were quite sensitive to formulated glyphosate products. However, the studies found no significant mortality, avoidance response or growth impairment among amphibians exposed to concentrations below 1,430 ppb when studied under more natural in-situ wetland mesocosm experiments or in biomonitoring studies involving caged amphibians tadpoles of two species exposed under worst-case scenarios of direct overspray under operational forestry conditions.

Not a done deal

At the same time, Thompson warns not to close the book on the effects of glyphosate on amphibians. He cites research from a colleague who recently published a study showing differences in wood frog populations under different forest management regimes including those that that involve herbicide applications versus those that do not. While the mechanism by which such differences might occur is not understood, it is plausible that indirect effects through changes to habitat quality may have an influence.

"Following from our previous study results we are less concerned about direct acute toxicity and more interested in potential subtle indirect effects on habitat quality that might have an effect on amphibian populations. In fact we are currently conducting studies to examine this aspect as well as other studies investigating potential multiple stess interactions. It seems intuitively reasonable to focus on potential subtle sublethal effects since in years of conducting studies in forestry sites recently treated with glyphosate-based herbicides we have never seen massive mortality of tadpoles, juvenile or adult frogs nor are we are unaware of any such instances in agricultural or other use sectors".