top of page
  • Andy Davis

Summary and comments on the 2016 study on the suitability of roadside habitat for monarchs

Greetings all,

In light of my last blog post on road mortality of monarchs, I thought I’d continue on the road theme in the blog post today. So, here I’ll tell you about a research study that came out late last year, and which I didn’t have time to cover before now. It is a project that examined the conservation value of roadside habitat for monarchs, which is a very important, but dicey conservation issue, as I’ll explain in this post.

The project was led by a gentleman in the University of Minnesota monarch lab, Kyle Kasten and he was the first author on the paper. I think he is a staff researcher there - and the paper was published in the Journal of Insect Conservation - not bad! It is titled, “Can roadside habitat lead monarchs on a route to recovery?” Here’s a link, but the whole paper is not downloadable. It is interesting that the title of the paper is an open-ended question, because as I’ll explain, I’m not sure this paper answers it. Perhaps the authors sensed this too.

BTW, to get you thinking about this, here is a picture of a monarch larva next to a road - this is a screengrab of a short youtube video I found -

So let’s start with the gist of the study. From my read, it looks like Kyle and his colleagues set out to determine how important roadside habitat is for breeding monarchs (the strips of land right next to highways), and he did so using a series of surveys of roadsides, where he looked for both milkweed and monarchs. It looks like he (or his coauthors) made a series of road trips throughout the Midwest, stopping at specific places along the way. He covered a lot of ground (literally) in this project, since there were a total of 212 surveys conducted over 4 states in the Midwest (Minnesota, Wisconsin, South Dakota, and Iowa). Below is a map copied from the paper, which shows the extent of the surveys. This is very impressive!

One thing to take note of with this map is that the sites were picked at random from a map (prior to setting out each day), which is very important from a statistical standpoint – this gives some strength to the results, and it minimizes human bias. It means they chose the roadside sites to survey before they knew what they would find – very smart.

It looks like they visited each site only once, probably for logistical reasons. This comes into play later, because it’s important. They did all surveys within a 3 month span from mid-July to mid-Oct in 2015.

At each site they visited, they documented all milkweed plants they saw within a transect of specific length. From this information, they could calculate the density of milkweeds per unit distance of road. They also recorded the presence or absence of all monarch eggs and larvae on plants, and they calculated their density too, on a per-plant basis.

Next they recorded some information about the sites themselves – the type of landcover around the site, like if the site was forested, next to a cornfield, urban, etc.

Now here is the very cool part - the authors also wanted to know how the roadside milkweed and monarch densities compare with other sites, and to do this they used citizen-science data from the Monarch Larva Monitoring Project. As most readers here know, this long-running program has volunteers watch milkweed patches during the summer and record the presence or absence of larvae on a weekly basis. According to this paper, the comparison sites in the MLMP include backyard gardens, natural areas (such as state parks and other nature preserves), pastures, old fields, and other areas that contain milkweed. And most importantly, Kasten et al. used the MLMP data from the SAME time period as their roadside surveys (July-Oct), and from the same survey region in the Midwest. So essentially, they used MLMP data as their ‘control’ or ‘natural habitat’ data, and asked, how does the abundance of monarch larvae on roadside habitats compare to what you would find in the ‘natural’ habitat of monarchs?

This about covers the gist and the methodology of the paper. Now on to the results.

I’ll skip the statistical jargon, and basically just point out that from my read, their analyses appeared to be sound. So for the field surveys, the authors found that there was at least one milkweed plant on 622 of the 1045 sites (put another way, 60% of the roadside sites had milkweed). Of these, the vast majority of plants were common milkweed, Asclepias syriaca. They also determined that the density of milkweed did not differ across the different landscape types they covered. For the monarch results, Kyle found monarch eggs or larvae on 23% of the surveys.

Here is one of the most important findings (I think), and I’m copying and pasting the exact text here – “roadside sites had significantly lower mean egg and larval per plant densities than MLMP sites monitored in the same area over the same time". Of all MLMP sites, there were 0.144 eggs/plant, compared to 0.039 eggs/plants on roadsides (3.5 times fewer eggs on roadsides). For larvae, the average density in MLMP sites is 0.0596 larvae/plant, and they found 0.0199 larvae/plant on roadside milkweeds, which is 3 times less larvae. Keep in mind that they were comparing monarch densities from essentially the same region (the Midwest), and from the same time period. Bottom line here, there were far fewer monarch larvae next to roads than would be expected based on the milkweed density there.

Recall that they only surveyed each site once – this is unfortunate (though entirely reasonable given the logistics), because that means they could not tell how the SURVIVAL of these larvae compared between roadsides and natural areas. That would have been very cool. To do this, you would need to have multiple surveys of the same milkweeds, and keep track of how many larvae are present at each successive life stage. It would have been great to know, for example, if roadside larvae had lower survival to adulthood.

OK, so that covers the main results (there are more but you can read the paper). Let me point out here that from my read, the results of the project were solid, though their interpretation is another story. Remember the overall question they wanted to answer was whether roadsides have any conservation value to monarchs. In their discussion, they seemed to indicate that they do. I’m going to paste some text here from their discussion – “Sites monitored by MLMP volunteers had significantly higher per plant monarch densities than roadside habitats throughout the summer. However, this does not mean that roadsides are not an important source of habitat for monarchs. The presence of 5th instar larvae on roadside milkweed plants indicates that monarch caterpillars are able to develop within the confines of roadsides.”

So, this would not be my interpretation of their data. The authors specifically found that there were far fewer monarch eggs and larvae on roadside milkweeds than they would have expected. Why is that? I note that this result was downplayed in the paper, but it really is pertinent. Since there were fewer eggs in these roadside habitats, that means there were fewer adult females laying eggs in these places. So the question then becomes, why are female monarchs choosing not to lay eggs on roadside milkweed? Incidentally, Kyle also kept track of any adults he saw, and they report that adults were seen in 49 of 240 sites with milkweed (20%).

The authors’ interpretation of their data goes as follows – given that they witnessed at least some monarch larvae on roadside plants, that must mean this is 'suitable' habitat. There is some logic to this. However, reaching this conclusion, given what they really found, makes me think that they were planning on reaching this conclusion before the project even started, and no matter how many larvae they found. Hear me out here – if they had found that roadside milkweed had even fewer larvae than what they found, say, 20 times fewer larvae than normal, they could have reached the same conclusion – that since at least SOME monarch larvae were present (no matter how few), that must mean it is 'suitable' habitat, or at least, capable of supporting larvae. By this logic, if I were to go to a Walmart parking lot and find a single, lowly, ugly-looking, milkweed plant growing between some pavement cracks, and if a single monarch larvae happen to be on that plant, I could conclude that Walmart parking lots must be ok for monarchs, because this one larva was capable of growing there – see what I mean? The mere presence of one, or a few, larvae does not signal habitat suitability.

My point here, is that the monarchs themselves are telling us something different. Based on their own data, the authors showed (very convincingly) that female monarchs choose not to lay eggs in roadside habitat as much as they would in a regular habitat. The question is why?

There are a number of potential dangers associated with roadsides, which the monarchs could be sensing somehow. There is pollution from car exhaust in the air (which can get into the plants), there is road salt and grime from the roads that can leach into the soil, and there is often excessive noise. Any one of these, or in combination, could be something that warns female monarchs to stay away. And let’s not forget about the ever-present risk of getting smashed by cars, which I’ve blogged about before. Given all of this, it’s hard to see how one could conclude that this is suitable habitat for monarchs, and that we should enhance roadside habitats by adding more milkweed.

So, that’s my coverage of the paper, including my two cents on it. My overall impression of this paper was that it was an excellent bit of science, but the authors’ interpretation of their data is not well thought-out.

That's all for now.


Direct link to this blog entry:



The science of monarch butterflies

A blog about monarchs, written by a monarch scientist, for people who love monarchs

bottom of page