What should I as a Farmer do about Milkweeds, Monarchs and GE crops? What does Science say?

Milkweed plants are a dilemma for crop farmers like me.

We know the harm that milkweeds with their deep roots and tall tops can do to crops. Indeed, until recently, Ontario farmers were legally obligated to kill them under the Ontario Weeds Act. But milkweeds are vital for monarch butterflies.

What’s a farmer to do?


Milkweed plants in fence row beside soybean field

You can’t control milkweed by pulling them. They simply regrow. Our family members once walked through crop fields in summertime with small hand-held sprayers applying herbicides on individual milkweed plants.

When glyphosate-tolerant crops came along, the labour-intensive, hand-spraying chore was largely eliminated. Milkweeds were controlled with the routine spraying of all crop weeds. As a result, milkweeds are less common in my fields these days, though still very prevalent in field edges, stream banks and other un-cropped areas of our farm.

History suggests milkweed has simply returned to its previous status. The plant does poorly in hay fields which once dominated Ontario farmland, and Bhowmik and Bandeen, in a 1976 review, said the early use of farm herbicides allowed milkweed to become much more prevalent in grain fields by reducing competition from other weeds.

Mature forest, the indigenous ground cover for most of Eastern Canada, “is not milkweed habitat,” to quote Pleasants and Oberhauser. Crewe and McCracken, in a 2015 paper on monarch butterfly migration in Ontario, stated that “the regeneration of trees and shrubs in abandoned fields” reduces the prevalence of “monarch host and nectaring plants.”

Maybe agriculture once contributed substantially to milkweed’s prevalence as well as its recent disappearance from many fields.

But regardless of history, media have been full of claims about how modern farm crop technology – and notably genetically-engineered (“GE”; aka “genetically modified”) crop usage – is responsible for declining numbers of monarch butterflies.

Since this involves me as a farmer growing GE crops, I did some investigation of the underlying science.

A first strike occurred in 1999 when Losey et al of Cornell University claimed, based on indoor feeding tests, that the pollen from insect-resistant GE corn plants was killing monarch larvae. Further research showed that the risk outdoors was minuscule. But once stated, the claim resonated. Thereafter, the literature of anti-GE groups routinely contained statements about monarchs poisoned by GE corn pollen.

Two research studies from Iowa showed usage of glyphosate-tolerant crops may be a larger concern. Hartzler did cross-state surveys in 1999 and 2009 and concluded that while milkweed numbers increased over 10 years in road sides, the prevalence was down by as much as 90% in farm fields. He attributed the latter, in part, to wide-spread usage of glyphosate-tolerant crops.

Moarch larva Monarch Watch 4feed3

Monarch butterfly larva on milkweed plant. Source: http://www.monarchwatch.org

Pleasants and Oberhauser measured milkweed populations in seven Iowa farm fields from 2000 to 2008 and recorded a substantial decline with time. They attributed this to use of glyphosate and glyphosate-tolerant corn and soybeans. However, they stated that some of these fields were sprayed with glyphosate and others weren’t and the paper contains no information on how the pattern differed between the two. The authors also found, based on reports submitted by volunteers across the US Midwest, that there was about a 40% decline over 10 years in per-acre milkweed plant numbers in farmland in the US Conservation Reserve Program (CRP) and pasture fields – lands which likely did not receive glyphosate treatment. They found that monarch egg numbers per milkweed plant were higher in agricultural than non-agricultural fields at six locations in central Iowa, and used this to conclude that milkweed loss from cropped fields is more important than from CRP lands, pasture and roadsides across the US Midwest.

Common sense says that better weed control with glyphosate-tolerant crops should mean less milkweed in farming areas/states where lots of GE crops are grown, and that could well mean reduced monarch butterfly production. However, neither of these papers provides strong proof of a cause-and-effect relationship. Interestingly, Hartzler in his paper expresses doubt that a decline in milkweed plant numbers in Iowa is closely linked to over-wintering monarch numbers in Mexico. A more specific critique is provided by Kniss.

In 2014, Flockhart et al at the University of Guelph concluded that the widespread use of glyphosate-tolerant crops, especially in the central US, is largely responsible for the recent declines in overwintering monarch butterfly numbers in Mexico. The authors went further with a broad condemnation of “industrial agriculture” though the term was not defined and the authors considered no aspect beyond GE crop usage.

As I went through the paper, I found some serious weaknesses. First, the study was a computer simulation and contains relatively few experimental data on butterfly and milkweed numbers in agricultural fields beyond those of Haetzler and Pleasants-Oberhauser. Flockhart et al recognized the limits of the Iowa data, stating, “The functional relationship between milkweed abundance and genetically modified crops use has not been identified.” However, in their paper they appear to attribute all the decreases in milkweed numbers in farm fields to the use of GE crops, rather than better weed control in general, or any other factor.

Especially puzzling to me was a conclusion that a 20% reduction in milkweed numbers in central North American had caused up to a 90% decline in monarch populations. If milkweed plant numbers were the critical factor, one would expect a loss relationship closer to one-to-one or maybe a lesser reduction in monarch numbers than milkweed plants – the rationale being that more larvae per plant would be expected if plant number is the critical limitation.


Monarch butterfly annual migration map. Source: http://www.monarchwatch.org

Yet, my own casual observations on farm milkweed plants in late summer in recent years around Guelph have shown the same phenomenon – still a large number of milkweed plants present in fence rows and non-cropped areas, but few monarch larvae to be seen. If 100 plants have few larvae, would it be different with 200?

One explanation is it’s an absence of milkweed plants en route during migration that’s the problem. But there are serious flaws with that logic too: 1) monarch adults do not need milkweed plants for nourishment during migration but only for egg laying and food supply for larvae, 2) “monarchs are excellent long-distance fliers,” to quote Pleasants et al, and fully capable of travelling from the southern US to Ontario without need for a generational change (i.e., egg laying and larval feeding) in between, and 3) monarch numbers have been also been said to be down in the mid-Atlantic region of the United States where corn and soybean production is far less intensive and where the migration pattern does not involve travel across the US Midwest.

USDA stats show that while about two-thirds of total land area in Iowa is corn and soybeans, with about 90% of this being GE, the acreage of GE corn, soybeans and cotton crops is far lower in the southern and eastern US (typically 10% or less; references  here and here, and the link to “Tableau Public” here). It’s about 10% for southern Ontario, too.

In stark contrast to the work of Flockhart et al is a paper published in 2016 by Inamine et al at Cornell University. Using survey data provided by volunteers of the North American Butterfly Association on monarch butterfly numbers at various locations across the USA over 22 years, they assessed over-time relationships for two migration routes. One route was up the eastern US, and one through the Midwest. What they found were statistically significant time relationships for the migrations northward:  Numbers of monarchs in the southern US in spring were related to previous over-winter numbers in Mexico. Summertime numbers in both northeast and Midwest were related to springtime numbers the same year in the south. However, there was no virtually no relationship between late summer numbers in the Midwest and northeast and numbers the next winter in Mexico.

They concluded the biggest and most critical losses occurred during the autumn migration south. And since adult monarch butterflies do not need milkweed plants for nutrition, including during migration, “lack of milkweed, the only host for monarch butterfly larvae, is unlikely to be driving the monarch’s population decline.” The Cornell researchers offer two likely reasons for the decline during southward migration – poor weather (they refer specifically to “the severe ‘100-year’ drought in Texas, 2010-2015”), and habitat loss, the loss of flowering plants and nectar supply over the migration route.

Notwithstanding the conclusions of Inamine et al, it does seem reasonable to expect a major decline in milkweed plant numbers in a place like Iowa, where GE corn and soybeans treated with glyphosate are the predominant land cover, to have some effect on monarch numbers. But for a locale like southern Ontario with far lower concentration of glyphosate-tolerant crops and lots of marginal land where milkweeds flourish, it’s far less obvious.

Fortunately, monarch numbers measured as hectares of overwintering adults in Mexico have recovered somewhat since a low of 0.67 hectares in 2013/14, up to about 4 ha in 2015/16 (see here). That’s despite no reduction in use of glyphosate-tolerant crops or apparent growth in milkweed numbers. However, the 2015/16 number is well below the peak of 44 ha measured years earlier. (It’s fascinating to speculate what the number would have been in pre-settlement days, when forest and tall-grass prairie covered nearly all of eastern and central North America.)


Milkweed growing on marginal farmland near Orangeville, Ontario

So what does all of this “science” mean for me as an Ontario farmer wanting to support monarchs while controlling weeds?

I see no scientifically credible reason to be less diligent in removing milkweeds and other unwanted plants from my crop fields and no reason not to continue the use of genetically engineered crops and milkweed-controlling herbicides. GE crops cover only a small portion of the land area of southern Ontario, and food production is of critical importance to everyone. Sustainable agriculture includes consideration of food supply, farm financial stability and many environmental features beyond milkweed plants (eg., reduced tillage associated with the use of glyphosate and glyphosate-tolerant crops).

But for fence rows, wet lands and other natural areas on our farm (which collectively represent about 1/6 of our total acreage) milkweeds are welcome – as are most other nectar-producing wildflowers, many of which may be equally important for monarch well-being.