Are herbicides responsible for the decline in Monarch butterflies?

Monarch butterfly caterpillar feeds on a milkweed plantThe monarch butterfly is in bad shape. The number of monarchs returning to their overwintering sites in Mexico has been declining steadily for at least a decade. The consensus suggests there are several reasons for this decline, including loss of their overwintering habitat and unfavorable weather patterns. But the purported cause of monarch decline that seems to get the most coverage is the loss of milkweed (Asclepias spp) in the midwestern US migratory path. The evidence seems clear that the number of milkweed plants through this region has indeed declined.

The cause for the milkweed decline, though, is a little less certain. The same highly publicized study that documented the decline in milkweed numbers suggested that herbicide use in agricultural fields was the culprit. But the study didn’t control for herbicide use, and therefore, had no way of knowing whether it was actually the herbicide or something else associated with crop production that caused the decline in milkweed. They found a correlation between adoption of glyphosate-resistant crops and estimates of milkweed density in Iowa. Surprisingly, this limitation didn’t stop the authors from concluding herbicides were solely to blame, even titling the paper “Milkweed loss in agricultural fields because of herbicide use” (emphasis mine).

I think there is little doubt that using herbicides in corn and soybean fields has reduced milkweeds in those crops. But the authors also observed a decline in non-agricultural milkweed density over the same period, but discounted the importance of that loss because they observed more eggs in agricultural fields. However, they used vastly different data sources to estimate monarch use of agricultural vs non-agricultural milkweed, so this difference seems tenuous. Even so, if herbicides are the explanation for milkweed decline within agricultural fields, are herbicides also the reason for the decline in non-agricultural areas? Or could whatever is reducing milkweed in non-agricultural areas also be contributing to the decline in agricultural fields? It seems likely that there is more to this story than herbicide use. But the headlines predictably blamed GM crops and associated herbicides for killing all the monarchs. One of the authors of the study even claimed recently that the correlation they found was the “smoking gun.”

There is always a well-known solution to every human problem–neat, plausible, and wrong. –H. L. Mencken

The “GMO crop kills butterflies” narrative is a nice fit for many anti-GMO and anti-pesticide groups. It is certainly plausible, and it makes sense that herbicides made to kill weeds would negatively impact native plants like milkweed. But a new study by USDA-ARS and Penn State scientists suggests what most scientists already knew: the real story is probably much more complicated. It is important to note this study is not directly aimed at finding the root cause of milkweed decline, but it provides information that is very relevant. And there are a lot of things to like about this new paper. The study presents a method to untangle a very complex problem, and this approach could be quite useful to other researchers. But I think what I liked most was that the authors took a very cautious approach to interpreting their results, which is truly refreshing given the topic. Even the institutional press release was very cautious. This study could have easily been over-interpreted to grab headlines, but it wasn’t.

Certainly, there’s no doubt that herbicides sprayed in a crop field will negatively impact any weeds (including milkweed) that are growing in the field, but the impact on vegetation outside the field borders is less obvious. Egan and co-authors questioned the conventional wisdom that herbicides are the primary culprit for reducing native plant diversity in these surrounding areas. Although it is easy to blame herbicides, it is difficult to test this theory directly. Herbicide use in crops always accompanies other practices that could impact native plant diversity. Perhaps it is not the herbicides, but some other factor involved in crop production that is responsible. These other factors confound our ability to quantify the direct impact that herbicides have had on declining native plants. But Egan and co-authors used a rather elegant way to tease out the impact of herbicides. In their words:

“If herbicides are a significant driver of changes in the diversity and composition of plant communities across agricultural landscapes, then we would expect rare species to show consistently lower tolerances to nonlethal exposures than common species with similar life histories.”

In short, if herbicides are to blame for the decrease in plant diversity in areas surrounding agricultural fields, then the plant species that have become rare must be more susceptible to herbicides than plants that are still commonly found. They tested this hypothesis on 5 pairs of plant species; 5 rare species paired with 5 closely related species that remain common in the same area. Two milkweed species were included in the study: Asclepias syriaca (common) and Asclepias tuberosa (rare). They sprayed each species in the study with glyphosate, the herbicide associated with Roundup Ready corn and soybeans, and the herbicide most commonly blamed for milkweed decline. They also tested 2 additional herbicides commonly used in Midwestern corn fields: atrazine and dicamba.

The authors summarize their results very succinctly:

“Contrary to this hypothesis [that herbicides are responsible for reduced plant diversity], our bioassay experiments revealed few significant differences in herbicide tolerances between the paired rare and common species.”

Their data simply didn’t support the idea that herbicides are the primary cause of the decline in the rare species they tested. They even noted that in many cases, the rare species was actually more tolerant to the herbicides than the more commonly found species. The authors rightly note there are many limitations to their study. They only evaluated 5 genera and 3 herbicides, greenhouse studies are not always very representative of field conditions, and genetic variability was limited in their collections, among others.

“Despite these limitations, our results do suggest a challenge to the viewpoint that herbicides are a primary driver of plant biodiversity decline.”

Does this mean herbicides have played no role in the decline of important species like milkweed? Certainly not. Egan’s research was not specifically aimed at milkweed decline, so we shouldn’t extrapolate these results too far. We shouldn’t take this study to be representative of all rare species and all herbicides. And they were only interested in the plants that grow near agricultural fields, not in them. But this research does suggest that there are more important factors than herbicides responsible for the decline of native plant species near crop fields, including milkweeds. Hopefully, the methods used by Egan et al. will be used in future studies that focus primarily on milkweeds and other plant species that now reside in the areas where milkweed has become less common. This would help us unravel the real reasons for the decline in milkweeds. Until then, perhaps we should focus less on trying to blame Monsanto for the decline of the monarchs, and instead start actually doing something that could help bring them back.


Updated 11-Feb-14 to clarify importance of agricultural vs non-agricultural fields.


  1. Interesting take on the subject.

    An equally valid hypothesis for Egan et al would be that the milkweed (or other plants) tends to share the same habitat as the desirable agricultural land and so is thereby selectively targeted whereas, for example, species that prefer a different habitat would be left more-or-less intact. So selective intolerance to the herbicides in this case would be interesting but not especially relevant to the decline in milkweed plant populations.

    Monarchs are especially vulnerable because they have a single genus of plant (Asclepias)for their larval stages, and for the most part a single rapidly declining area for overwintering. In the meantime, people are in fact setting up milkweed gardens to try to offset the milkweed losses in agricultural areas. A few are also attempting to encourage alternate overwintering sites in both Florida and California (where there already is a small population that overwinters).

    Your article would have been more interesting if it had been cast as an information article rather than a bald apologia for Monsanto.


    1. Thanks for the comment, Alan. You are correct that land use will certainly negatively impact milkweed density. I have updated the post to clarify that point, specifically with respect to the Pleasants study. The Egan study was not interested in plants that grow within agricultural fields, but rather, plants that grow outside but nearby agricultural fields. Offsite movement of herbicides has been blamed for the reduction of native plants in these adjacent areas, and that is specifically what they are looking at.

      Within agricultural fields, I think the GMO or herbicide issues are red herrings. If milkweeds are growing in agricultural fields, they will probably be controlled by the farmer regardless of their production methods. Tillage and crop rotation are often quite effective for managing milkweed plants within the field. I susupect it would be equally (or more) difficult to find milkweed plants growing in organically-managed fields as Roundup Ready fields. It is the fact that we are growing crops on these lands that reduces milkweed, not the specific production practices, as you point out.

      I hardly think this is an apologia for Monsanto… I’m simply pointing out that the problem is more complex that simply blaming herbicides, and blaming Monsanto for monarch decline isn’t very helpful because it distracts from the real problem (which is habitat loss, for both milkweed and butterfly). So although you don’t like my casting of this issue, I think we mostly agree on this issue. We should be focusing on trying to do something to help the monarch, which is why I linked to an effort to do just this at the end of the post.

      1. We agree on several important points for sure – most especially let’s try to improve the chances for monarch survival by doing sensible things. I also agree that habitat reduction is the most likely culprit – but that in itself is not a simple variable because it has a number of different causes, including actual use for purposes that exclude monarchs (eg. logging in protected areas in Mexico) and or reduce milkweed (eg. exclusion as a toxic pest from agricultural land). If low monarch populations also face drought as they have recently, that makes a bad situation worse.

        I also agree that GMOs per se are a red herring in this issue, but they are implicated insofar as they are the main reason for the spectacular rise in use of glyphosates.

        One minor disagreement, milkweed was not especially well managed before the introduction of Round Up and RR corn and soybeans. Milkweed is effectively killed by glyphosates but not as easily with other herbicides or tillage practices. The significantly improved management of milkweed as a pest resulted in more than a 50% reduction in milkweed in fields. (Soybean glyphosate use 1994-2006 increased nearly 3,000% and in corn fields 2000-2010 increased over 1,500%). Given that half of the monarchs in Mexico overwintering populations come from the Midwest and major reductions in milkweed populations occurred in that same area, it is a reasonable place to start looking for cause and effect. So while a correlation absolutely does not indicate cause and effect, this one is not unreasonable given that monarchs are pretty much obligated to milkweed to host their larvae. As you point out correctly, some intersecting testing or correlations would help to tease out if there is a real cause and effect relationship.

      2. Thank you for all your research I would like to see land set aside for animals and insects. Good job Andrew Kniss just putting the blame on someone doen’t help anyone we all need to work together. I know I have blamed Monsanto I am sorry I would hope everyone wants the best for our planet.

  2. Alan,
    great summary! It does make a great deal more sense to intentionally cultivate milkweed along roadsides, in field borders etc rather than to farm weeds in crops (it wouldn’t be easy to encourage milkweed there without doing so for other weeds with no redeeming role).

  3. There are as you point out several factors contributing to monarch declines in addition to the widespread and intensifying use of glyphosates, and thats actually why we need integrated wewed management problem-solvers–both farmers and weed scientists- to get on board and help with this issue. It is not a them versus us issue, but one in which collaboration across disciplines,ideologies and professions will be the onlysystemic solution. But this is not just about protecting butterflies or milkweweds. Its about protecting crop pollinators, which most fruit and vegetable farmers need more of. And its about protecting farmers from the rising costs of herbicide use against the growing number of superweeds. Lets found common ground and create solutions, rather than seeing this as one more pro vs con stalemate where people draw a line in the sand. Lets inspire positive change, and get beyond finger pointing.

    1. Hi Paul,

      It is great that you found a resting site for the monarchs. Obviously no milkweeds in the area and from their behaviour, it is clear they are using the trees as their temporary refuge. If you can convince Monsanto to plant a dozen 10 acre fields of milkweeds in the surrounding few square miles, that would probably significantly increase the survival of the monarchs you are watching and vastly increase their potential for breeding in the area. There are probably many local volunteers who would be happy to monitor monarch breeding success, and Monsanto would be making a contribution to Monarch success.


  4. On corn, before Roundup Ready and Roundup use, the herbicide was atrazine. Atrazine is a broadleaf only herbicide. So it killed milkweed too. Now, if a farmer sprays a ribon of land adjacent to the cornfield, everything dies. This leaves bare land, where a pioneer plant such as common milkweed should actually be favored. With atrazine, such over spray just killed the broadleaf weeds. Clearly, since there was atrazine before, herbicide use cannot be THE cause. The chart showing the correlation in Roundup use and monarch decline claiming a cause is just correlation. A similar chart documenting a correlation between the decline of atrazine use and the decline of monarchs could be used to claim that atrazine was good for butterflies.

    1. Thankyou Kay and others.

      A perverse fundamental of the eco-attack is that they say that milkweed is common in crops, on roadsides, and in ditches – all human creations.

      So the decline in Monarch butterfly populations they use to attack herbicides and GMO plants in general would just be a decline toward a situation without widespread human activity – IOW, to what they consider more “natural”.

      As for corn, a common target, widespread planting of it only occurred in recent centuries, after development of corn by tribal people from precursors of maize.
      (Some alarmists also talk about soybean – is that native to North America? (In eastern Iowa, which is in the “Midwest”, the main crops are corn and soybean which its soil and climate foster.)

Comments are closed