top of page
  • Andy Davis

New study shows monarch breeding range expansion in Canada - why that's good and bad

Hi everyone,

There have been a bunch of new studies published on monarchs this summer, which strangely, have largely gone unnoticed by the monarch-y. It may be that their findings were not as earth-shattering as the recent PNAS paper on captive-reared monarchs. In fact, by my count there have been over a dozen new studies in the last month (did you know?). This is great though. Any new research on monarchs is good, even the if the findings aren't earth-shattering.

One of the studies that did not make any waves was a new one by Tyler Flockhart and colleagues, which was just published in a Canadian journal called Facets. I'm going to explore this new study today, and also tell you how these new findings affect what we know about monarchs.

First, here is a link to the study, which is completely online -, The authors were, in order, Tyler Flockhart, Maxim Larrivée, Kathleen L. Prudic and D. Ryan Norris. I encourage you to read it, or at least have a look. There's a lot of complex stats, some of which are even over my head. But one can easily get the gist of the study nevertheless.

This project appeared to be aimed at trying to determine the distribution of monarchs in Canada. That means, they wanted to figure out where the monarchs breed in Canada - like where are the prime spots with lots of breeding monarchs, where is the northern limit of their range, etc. In other words, what does the breeding range in Canada look like for both the western and eastern monarch population? In addition, they also set out to determine if this distribution could be statistically predicted by basic variables like temperature and precipitation, or even human population in a given region.

They accomplished this by examining data from two large-scale citizen science projects, Journey North, and e-butterfly. Both of these programs are large-scale, and long term in nature, and so are perfect for this use. I believe the authors obtained permission to use the Journey North data, and Katie Prudic (one of the authors) is one of the people in charge of e-butterfly. Both programs have people report when and where they see monarchs, although each has slightly different approaches. For Journey North, anyone who sees a monarch, anytime, can simply send in the sighting, which gets logged into the long-term database, and then the sighting appears on the map. For e-butterfly, I think people can upload more of a checklist of butterflies in a particular area. I'm a little less familiar with this program, but I have played around with their nifty mapping feature. For example, if you ask it to pull up all records of monarchs, you get the following map:

Keep in mind that this map has an undefined timescale, so I think this is all records of monarchs, from all years of the program. But, this map shows that there are a lot of records of monarchs in this program!

Next, readers of this blog are probably more familiar with the journey north maps. Here is the journey north map for monarchs, as of July 26 - these are the "first" sightings of adult monarchs. Look at all of the monarchs this year!

Now, go back and look at both maps again and notice carefully where the Canadian border is, and where the Canadian monarch sightings are. If you look close you can see that compared to the US sightings, the Canadian portion of the breeding range is fairly small - I would estimate that the Canadian portion is less than 20% of the entire breeding range of either the western or eastern population. But interestingly, it looks like this percentage is growing, based on the results of this new paper, but I'm getting ahead of myself now.

The researchers looked at 16 years of sightings from these two programs to answer their questions, from 2000 to 2015. It looks like they combined both datasets for their analyses, which seemed logical. But, for reasons I can't understand, they statistically examined each year separately. That means they took all monarch sightings for the year 2000, and then asked where are the monarchs located, and what variables predict these locations. Then they did the same thing for the year 2001, and then 2002, and so on. This seems odd to me.

In their statistical analyses of the monarch sightings, the researchers asked if the temperature or precipitation of a given region in Canada predicts the number of sightings in that region. They also asked if the number of people is a predictor (more people = more sightings). And then they asked if the latitude or longitude was a predictor.

They presented the outcome of their analyses using a series of maps, which I'll paste below. Each map has red shading to indicate where the monarchs were predicted to occur. That means, this is where the statistical program estimated monarchs would be, based on the variables that it knew were important to monarchs. Another way to think about this is that the statistical program took the the points with known monarch sightings, and extrapolated this to areas where there weren't sightings (or people). I know, it's complicated - just keep in mind that the darker the red, the more sure they were that monarchs were in that area.

One thing that jumps out from looking at these maps is that there are some years where the Canadian breeding range was larger, and some when it was small. And, if you squint you can see that the later years tended to have larger breeding distributions. This idea comes up again next.

They presented these same yearly data in a graphical format in the paper, which I've pasted below. These graphs show the overall size of the Canadian breeding range each year, broken down into the western population, and the eastern one. The bottom graph shows the overall range for both populations. The different colors of each bar refer to the probability of occurrence, or basically, how sure they were about these estimates.

I'm sure you see what I see with these graphs - there is some variation from year to year in the size of the Canadian breeding range, but overall, it looks like the range has been expanding over the 16 years. And, keep in mind that these data end at 2015. Since that time, the monarch population, at least in the east, has continued to grow. In 2017, monarchs had an above-average breeding season, 2018 was absolutely stellar, and this year they also seem to be doing very well.

In fact, I sat down at my computer, pulled up some Journey North maps for the last 4 years, and I eyeballed them to see if I could estimate what the Canadian breeding range would be. I did this by comparing maps of these years to those from the study. Then, I came up with my own graph for the eastern population, which combines the data from the figure above (graph B), and my estimated numbers for the last 4 years. Below is what I got.

As I suspected, once you factor in the (estimated) numbers from these latest years, they seem to show pretty clearly that the (eastern) monarch breeding range in Canada is expanding. You know it's weird, but nowhere in the paper did the authors mention this, when this trend is clear even from their 16 year dataset.

As for the other findings in this paper, recall they had tried to find a link with breeding density and temperature and precipitation - it turns out there is none! Areas that tend to be warmer did not have more monarchs, and the same was true with precipitation. Human population density, on the other hand, definitely was correlated with monarch density. This is something the authors expected to find, because more people equals more monarch sightings. Also, the human population centers in Canada tend to be huddled near the most prime real estate for monarchs anyway.

One final interesting finding from the paper was from a small statistical test at the end of the results. Almost as an afterthought, the authors tested if the annual size of the Canadian breeding range was related to the size of the overwintering colonies - either in the west (California) or the east (Mexico). They first tested if winter colony size in any given year predicted the size of the subsequent Canadian breeding range later that same year - it didn't. They also tested if the size of the Canadian breeding range predicted how big the winter colonies would be in any given year - they didn't. So in other words, the winter colony size in any given year does not matter to the number of breeding monarchs in Canada. Another way to think about this is that a year with a really good breeding season in Canada will not necessarily lead to large overwintering colonies.

OK, so that's the gist of this paper. Now here is where I tell you why the findings are both good and bad.

On the one hand, an expanding breeding range (of any animal) is a sign of a healthy, robust population. The converse is also true - a shrinking breeding range is a really bad sign. As a case in point, you might have heard about the rusty-patched bumble bee in the last couple of years. It was recently placed on the endangered species list in the US largely because its breeding range has diminished. It was once found in over 26 US states, and now it is only found in 13 or so.

So at least in Canada, we know now the monarch breeding range is growing - this is great, right? More signs that monarchs (in the east, anyway) are doing well, despite what we keep hearing in the news. drawback here (here's the bad) is that this could hinder the annual migration to Mexico (or to California). The northward expansion of the breeding range is essentially adding more miles (or kilometers, for Canadians) to the fall journey, and given how treacherous the whole trip is, this is in no way helpful. It would be like moving the starting line back each year in the Boston Marathon, making it longer each year. For monarchs, longer travel means more time exposed to all of the dangers encountered along the way, which will lead to fewer monarchs successfully reaching the winter colonies. Oh, the irony.

Finally, any guesses as to why the northern breeding range is expanding? I'll give you a hint, it's two words - yup, climate change (or global warming, if you prefer those). Warming temperatures are most likely pushing the northern limit of milkweeds higher, so that they can grow now in places that were once too cold for them. In fact, there have been a couple of studies on this. I covered one of these that came out in 2015, written by Nathan Lemoine (see that blog here). Another was done in 2014 by Reba Batalten, Karen Oberhauser and Townsend Petersen (link here to their abstract). - basically, the gist of each study was that global warming is predicted to push the northern limit of milkweeds further north in the next few decades, and, the monarchs should follow this northward shift during the breeding season. Well folks, that's exactly what's happening now.

Well, I think this about covers this new study, plus my own thoughts on it. Hope everyone learned something today!


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