top of page
  • Andy Davis

Migratory mortality - the most important, yet understudied and underappreciated component of monarch

Greetings everyone,

This post is on a somewhat morbid topic, but as you will see, this is perhaps one of the most important things we need to know about the monarch migration today. And, I'll also tell you about how even most scientists are in the dark about this, very critical, piece of information.

So this blog post was instigated by a couple of recent events. First, there was a new scientific study published this month in the journal, Ecological Entomology, by a big team of scientists, including Karen Oberhauser. This study (link here to view the abstract) was a very technical and dense modelling (statistical) project that was aimed at trying to figure out where we should be focusing our efforts at conserving monarchs - that is, which life stage and/or region of the country is most important for growing the numbers of monarchs that reach Mexico. Essentially, the authors collated a whole bunch of data and information on the monarch, in all of its life stages, and including a lot of citizen science data, and then did some very fancy math to come up with an answer to this question. I'm going to paste a figure from the paper that shows which life stages, and regions of the continent they examined. this is sort of like a migration and range map that we're used to seeing for the eastern population.

The way to think about this map is that the researchers examined how many monarchs there were each year in each of the colored sections, and then they asked how many of them move from one color to the next during the year, during spring or fall migration, and then finally how each color section, or the transitions to and from, impacts the numbers of monarchs that are seen in Mexico. So just follow the arrows.

Now here's the kicker that I had with this analyses - the researchers attempted to test whether the fall migration is important in terms of predicting the winter numbers - that is, how many many monarchs start the journey each year but fail to reach the finish line. It was great that they thought of this. If you've read this blog before you know that this is something I feel is crucial to the ongoing conversation about monarch conservation. However, from what I can tell, this part of the project was problematic from the start. The researchers needed to estimate how many of the monarchs fail to make it from the yellow to the orange, or from the yellow to the purple each year, etc., but to do this, they had no actual data to go on (it does not even exist). So they used a commonly-used technique by modellers whereby they asked for expert opinion on this. They asked about 6 monarch scientists to give them 'gut-feeling' estimates for the proportion of monarchs that start in the yellow in the summer, and what proportion successfully make it to the overwintering site. They then averaged these estimates and used this average in their statistical model. The estimates they got ranged from 60% to 80%. This is an absurdly high estimate in my book. Essentially, the scientists they asked (few of them actually study the monarch migration, anyway) all thought that most of the monarchs who begin the migration actually do reach Mexico. This is crazy - have they not read my blog on road mortality? Just do the math people, or better yet, read my blogs on road mortality. From my extrapolations of the data we have on road mortality, at least 25 million monarchs die from car strikes each fall in the United States, and at least 100 million die on roads in Mexico during fall migration (the purple region above). Compare this to the 50-100 million monarchs that maybe arrive in Mexico each year, and you can easily see that at least 50% of the migratory cohort probably dies during migration, and that's just from car strikes alone. There is also weather-related mortality, plus those that die from lack of nectar, and a whole slew of natural causes, etc. So from my estimation the proportion of monarchs that successfully reach Mexico during the fall is probably closer to 30%. In other words, the cohort of monarchs we see in Mexico is probably only about one third (or less) the size of the breeding population, because of all of the mortality along the way.

So because the statistical model in this paper did not think the amount of migratory mortality was high, when the numbers were crunched it failed to recognize this life stage as being 'important' for conserving monarchs in Mexico - so the researchers erroneously concluded that promoting habitat in other life stages (breeding) were more important than migratory habitat. All of this, because the researchers themselves provided misleading estimations for this life stage. To be fair, there was mention in the paper about how conserving habitat during both stages would be beneficial, but the take-home message of the paper was that breeding habitat was most important. Don't get me wrong, I am fully on board with restoring native prairies, and fields, and meadows. But I and others believe this won't solve the monarch problems in Mexico.

There was another recent event that triggered this blog post, which was a new, competing, blog post at MonarchWatch that examined the tag recovery data. From my reading of it, the blog article (link here) was a pseudo-sciency analysis of the number of monarchs recovered in Mexico vs the number that are initially tagged, or the tag recovery rate. The goal of this investigation was to find out if the rate of tag recoveries has gone down over the last 15+ years, which would be evidence that the migratory mortality rate has increased over time. If you've read my blog, you know that this is something I and other monarch folks have been asking about for a while now, because of the conflicting evidence we have from long-term datasets on monarchs. Basically, since the number of monarchs in the breeding range has not declined over the last 20 years, yet the numbers reaching Mexico have clearly declined in that time frame, I and others believe there must be a growing problem they face in the migration - that is, there must be greater mortality during the fall migration now than there used to be. Examining tagging data is one way to look into this idea, and I applaud the MonarchWatch people for attempting this. I wish they had done this in a peer-reviewed, scientific journal though, because then it would be actual 'science'. As it is now, these are just thoughts on a website, and website analyses don't really count as scientific evidence.

So the MonarchWatch blog appears to show that there has been no significant change in tag recovery rates in the last 15 or so years, which calls into question the migratory mortality idea. I do have many questions about this though, which would need to be addressed before this evidence is to be believed. One of the biggest questions has to do with the effort expended by the locals to find the tags. It seems that the way we have the tag recoveries in the first place is the local people go into the colonies and search for tagged monarchs (dead ones on the ground, I guess), and then the people from MonarchWatch buy the tags from them. But does anyone really know how many locals actually do this each year? And has the number of searchers increased in the past 15 years? And/or have they increased the amount of time they spend looking for tags? Most importantly (and perhaps a reader can answer this), has the price of tags gone up over these years, which would likely cause an increase in searching effort? The bottom line is that if the search effort for tags has gone up over time, it would mask any declines in tag recovery rates. This is a crucial piece of information that we need to properly interpret these data.

By the way, the other important thing to note from the tag recovery analyses is that the rate of tag recovery is really, really small, based on my reading of the graph. You could say this is further evidence that the size of the overwintering cohort in Mexico is a only tiny fraction of the numbers that actually began the journey.

When you really stop and think about this - about just how many monarchs die along the way to Mexico - you start to wonder why it is that we use the size of the wintering cohort as the measure of "population size". A better way to think about the Mexico numbers is that these represent the number of monarchs that successfully complete the journey, and these are the lucky few who are destined to reseed the next generation.

That's all for now folks. I hope this got you thinking.


Direct link to this blog article:


The science of monarch butterflies

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

bottom of page