Research Article

Choosing Organic Pesticides over Synthetic Pesticides May Not Effectively Mitigate Environmental Risk in Soybeans

  • Christine A. Bahlai,

    Affiliation: School of Environmental Sciences, University of Guelph, Guelph, Ontario, Canada

  • Yingen Xue,

    Affiliation: School of Environmental Sciences, University of Guelph, Guelph, Ontario, Canada

  • Cara M. McCreary,

    Affiliation: School of Environmental Sciences, University of Guelph, Guelph, Ontario, Canada

  • Arthur W. Schaafsma,

    Affiliation: Department of Plant Agriculture, University of Guelph, Ridgetown, Ontario, Canada

  • Rebecca H. Hallett mail

    Affiliation: School of Environmental Sciences, University of Guelph, Guelph, Ontario, Canada

  • Published: June 22, 2010
  • DOI: 10.1371/journal.pone.0011250

Reader Comments (2)

Post a new comment on this article

Implications of this study??

Posted by PWeston on 26 Jun 2010 at 16:59 GMT

In my opinion, the original commenter brought up some valid concerns which have yet to be addressed by the authors of the paper. If the goal of the study was to evaluate the efficacy and environmental impact from the use of a limited number of pesticides provided by AAFC to control soybean aphids, why was the principle finding: "organic approved insecticides had a similar or even greater negative impact on several natural enemy species in lab studies, were more detrimental to biological control organisms in field experiments, and had higher Environmental Impact Quotients at field use rates."? That's a terribly general conclusion. This is especially true considering neither "natural" control has been recommended for the control of soybean aphids in organic systems that I know of in Canada.

The study also contains many statements of conjecture at best, without backing evidence or references. One example is the critique of Reganold et al. (2001) where the authors state the study results in this Nature paper are likely influenced by differences in practices and assessment methodology without even identifying such differences. Scientific publications must be held to at least the same standards as public Wikipedia articles.

This "study" sounds more like an open letter to the Canadian government instructing them not to make policy decisions for political reasons that favor "organic" controls over synthetic. I agree, but there are other communication forums for lobbyists that may be more effective. I also think that government should not make policy decisions based upon the powerful corporate lobby either. For some odd reason, this influence has been totally disregarded by this paper, and instead makes the assumption that the organic growers lobby is the only danger and influence because of unsubstantiated conjecture that the majority of Canadians think that all natural substances are less harmful than synthetic ones. I can assure the authors that this likely isn't the case, given that the organic produce market is still significantly smaller than the conventional market. If the corporate lobby was less powerful in Canada, we researchers would have succeeded in convincing policy makers to ban organophosphates for indoor domestic use two decades ago instead of just a few years ago. I could bring up numerous other examples, however I'm sure the authors of this study already know them.

Being a worker-bee who is relied upon to provide advice to policy makers from time-to-time, I welcome changes such as Ontario's restrictions on the residential use of pesticides for cosmetic use. I can assure you I'm not alone. Considering safety, diversity, and lowering environmental impact are supposedly goals for policy makers, we would prefer if they went much further and enacted legislation that would discourage the creation of cosmetic artifacts (lawns) entirely!

Like the first commenter, I also welcome an evaluation of various options for the control of soybean aphids that may prove to be a serious threat to production in Canada. However, the unscientific and superfluous side-discussion on the evils of organic production methods that form the bulk of this paper are both unnecessary and unsubstantiated.

No competing interests declared.

RE: Implications of this study??

cbahlai replied to PWeston on 28 Jun 2010 at 14:50 GMT

The purpose of this study was to evaluate novel, reduced risk products, to assess the suitability of their inclusion in a soybean aphid management program. Because previous work by us and others (see references #22-26, for example) indicate that natural enemy populations are key regulators of soybean aphid, a major component of this assessment was to evaluate the effect the use of these pesticides would have on natural enemy communities. It is useful to evaluate the effect a pesticide has on a natural enemy community, because preserving this community is essential to prevent pest population resurgence- essentially, an effective, but non-selective pesticide is more likely to require re-application than effective, selective product, as pest populations can grow much more quickly in the absence of natural enemies. Additionally, as we mention in the paper, natural enemies are useful indicators of environmental impact because their exposure levels would be high relative to most organisms, because they tend to be present when the pest is present, and, like the pest, they are primarily arthropods, and thus are more likely to have similar physiology, and similar physiological responses to a pesticide than less-related organisms.

At the onset of this study, only dimethoate and cyhalothrin-lambda were registered for use as foliar treatments for soybean aphid. All the other pesticides tested - both the two novel synthetic chemistries and the two organic pesticides- were evaluated in this study to examine their potential for use in this crop in the future.

The assessment methodologies used to estimate the impact of pesticide regimes in different cropping systems varied quite dramatically between the Reganold et al (2001) paper versus the Suckling et al (1999) paper. We did not detail the differences in the text of the paper, but Reganold et al’s primary means of environmental assessment of pesticides consisted of using a dose-independent pesticide risk assessment tool- the “Responsible Choice” (RC) program developed by a grower group in Washington state, described in that paper as “similar to Cornell University’s Environmental impact quotient” (but see reference #15, by Avery, for a commentary on why RC is not like EIQ). Suckling et al’s primary means of assessment was direct field measurement of pest and natural enemy populations. Differences in assessment methodologies often lead to differences in results and conclusions, as an examination of these studies demonstrates- and this is why we included multiple experiments in this study- an EIQ risk assessment, laboratory direct-contact bioassays and field trials.

I, personally, do not endorse policy decisions made on any basis other than empirical data. All pesticides, regardless of whether they are of synthetic or natural origin, must undergo rigorous testing and thorough environmental assessment, and should only be applied by qualified applicators, under appropriate conditions, at appropriate doses, as a last resort, to maximize both safety and sustainability- and, when these conditions occur and options exist, the pesticide with the most favourable environmental profile should be selected. Thus, I also welcome restrictions which limit the injudicious use of pesticides- these are bioactive molecules that warrant policy limiting their use to appropriate circumstances only.

-Christie Bahlai

Competing interests declared: Author of study under discussion

RE: RE: Implications of this study??

PWeston replied to cbahlai on 29 Jun 2010 at 15:21 GMT

Thank you for the reply Christie. It did offer some clarification concerning your personal position regarding the responsible use of pesticides. However, the concerns expressed in my original post were meant to be criticisms of this paper rather than any beliefs the authors may hold. I apologise if I gave a different impression.

The term “likely” conveys a general measure of statistical significance and therefore must be backed with empirical evidence. Of course differences in assessment methodologies may cause differing results, but backing science must be presented if this is stated as a likely cause. Your paper presents no evidence to support your conclusion that the use of RC in the Reganold et al. (2001) paper was a likely factor that caused such divergent results to that of Suckling et al. (1999). It appears instead to be an untested hypothesis that has little relevance to the stated subject of your paper, and should be identified as such.

The goal of Reganold’s was to evaluate the differing apple production systems used on 3 specific farms located in Washington State. There are a host of other factors that may have influenced the results considering that the three farms evaluated by Reganold et al. (2001) were not the same three farms evaluated by Suckling et al. (1999). In fact, the farms were evaluated for different periods, had different farmers, and were located in different hemispheres on different sides of the planet! The weather patterns (wind, temperature, humidity, and rainfall) might have even been different between the two sets of farms studied in these two papers. Yet you state that Reganold’s use of RC is likely the cause of differing results, and the differences in study methodology somehow caused the New Zealand farmer using the organic system to have difficulty controlling certain apple pests that affected economic sustainability of the operation due to marketability/export problems associated with damaged/infected produce? I’m not trying to be funny here.

The Reganold et al. (2001) study was one of the most comprehensive of its type undertaken at the time of publication, and included pest monitoring and the creation of a soil quality index based on a number of factors that provided a measure environmental health. You critically state that Reganold used an environmental rating system that is dose-independent, but it is specifically stated in this paper that “The active ingredient of each pesticide and the dose and frequency of application were used to calculate the environmental impact ratings”. Details that include the number of applications, dose, and how they affect ratings for every treatment used on each farm are provided in the supplementary information.

All that being said, I still fail to see the relevance of the Reganold versus Suckling reference (and other uncorroborated discourse) in your paper that you confirmed is supposed to be an evaluation of control options for soybean aphids. Why instead is there no mention or reference to other studies that evaluate the efficacy and effect on predators of some of the same treatments for soybean aphids you are also evaluating? Even mineral oil has been examined as a control of Aphis glycines in studies such as Kraiss & Cullen (2008). Instead of building upon these studies and uncovering questions worthy of additional study, your paper seems to ignore them all together.

Please don’t misunderstand, my criticism is directed at your paper and not to the authors responsible for it. The actual field work undertaken may prove to be very useful despite the questionable relevance of some EIQ ratings, and I wish that your paper simply stuck to the presentation of the work and your results.

I’m a little surprised that mineral oil instead of neem oil soap was selected as a treatment option. While the choice might have been out of your control, the many papers that evaluate the use of neem for aphid control seem to show good to fair results, while the many papers that evaluate mineral oil for this option report much less efficacy, and often poor results. Neem oil may also have an improved profile as far as environmental impact because of reduced dosing, and also differs by being currently allowed by all certification organisations in Canada for such use.

Now if you excuse me, I must assess an environmental disaster to determine how many towns downstream need to be evacuated after 500 kilograms of Beauvana bassiana were spilled into the river. Kidding of course!

No competing interests declared.

RE: RE: RE: Implications of this study??

cbahlai replied to PWeston on 29 Jun 2010 at 19:58 GMT

Hi Again,

The statement from the paper that we’re discussing is:

“Differing outcomes may be attributed partially to differing geography, climate and pest complexes at the two locations, but it is likely that differences in assessment methodology and the inconsistencies between specific practices classed as organic or conventional at each location were also influential in obtaining the observed results.”

In this statement, we acknowledge, essentially, that though these two studies are talking about the same commodity system, they are quite different for a number of reasons. The word ‘likely’ was not meant in quantifiable, statistical way, and we are not saying that differences in assessment methodologies account for all the differences in results and conclusions, but that, simply, the choice of assessment methodology can influence results. In this case, we’re comparing a field study (in New Zealand) to a theoretical measure of environmental impact (in Washington State). However, assessment methodology can, and does, have quantifiable impact on the results obtained.

Referring to the paper by Avery (2008) that I mentioned earlier, the author presents a clear example of how the differences in just choosing a difference pesticide selector can impact results. As you mentioned, the rates, number of applications, and products used are all available in supplementary information associated with the paper online , so the author goes back to Reganold’s original appendices, and using the data presented, calculated the EIQ scores for each pesticide regime. He found that, using EIQ resulted in similar scores between the organic and conventional systems (ratio of scores is 1.32), where using the RC system, a much more marked difference of scores (ratio 6.21) was found between the two systems.

The Reganold study was quite comprehensive, but I believe their choice to use the RC pesticide selector was flawed, and this is related to the mechanics of how RC is calculated as compared to EIQ. The Avery paper offers a very comprehensive discussion on this subject, but I will summarize. Though Reganold et al report the dose of each pesticide used, an examination of the RC calculation shows that, for a single application, the RC score for a given pesticide is dose-independent- it is pest specific (that is, a heavily weighted parameter in its equation is the efficacy against the target pest), and assumes the pesticide is applied at label rate for a given pest. This means that a pesticide which may be applied at the same rate for two different pests can result in two different RC scores, depending on what it’s being applied for. Dose is only accounted for in the number of applications required- ie: RC scores are summed for each application of pesticide. EIQ is certainly not without its flaws and simplifications, however, the mechanics of its calculation, in my opinion, leave less room for bias and ratings are more broadly applicable. EIQ differs from RC in that the ratings are based on a per-unit-weight assessment of properties of the active chemical, then multiplied by the application rate, and the number of applications. Because it is dose, rather than pest specific, it results in the same score for an application at a given rate, regardless of the pest for which it’s being applied.

Neem would have been an interesting product to include in our study. We did have plans to include several botanically derived pesticides in this study from the outset, however, the manufacturers we contacted were not interested in participating in the trial at this time. Soybean is a relatively low-value crop, and these manufacturers felt that it would not be financially feasible for farmers to use these more expensive products.

Good luck with your hypothetical B. bassiana spill- I’d aim to evacuate the dragonfly larvae first. :)

-Christie Bahlai

Competing interests declared: Author of the study under discussion

RE: RE: RE: RE: Implications of this study??

PWeston replied to cbahlai on 30 Jun 2010 at 01:52 GMT

Christie, I'm sure you realize that the term "likely" when used in a scientific context doesn't mean the same thing as "possibly". Scientific publications are held to higher standards than lay public discourse in neighborhood bars ;-)

When researchers use the term "likely", they really mean likely. If you really think Reganold did it so wrong that it invalidates his results, you better be prepared to back it up with real empirical evidence. This is what separates science from opinion.

So enough of that. Now I'm curious about bassiana in water. While I knew about liquid cultures, I was unaware that spore germination in natural aquatic environments actually produced mycelium that attacked aquatic insects such as dragonfly larvae. From the literature available, it seems that B. bassiana is mainly a danger when ingested by humans, causing mushrooms to grow from their ears which when ingested causes an intoxication that makes them enjoy the Jonas Brothers. We have quarantined all 11 year old girls in the towns affected by this environmental disaster as a precaution.

We will keep you updated!

No competing interests declared.