Monarch Watch Blog

Monarch Annual Cycle: Migrations and the number of generations

3 June 2020 | Author: Chip Taylor

monarchs in flight

The monarch annual cycle is quite extraordinary since it involves an overwintering phase that follows a long fall migration, a remigration in the spring and a succession of generations before the start of the next fall migration. This pattern is even more complex since it involves overlapping generations and a total of 4 migrations. Yes, 4 migrations. Further, monarchs breed over a wide range of latitudes with differing temperatures and periods of production. It can be very confusing and there is a lot of misinformation about these patterns in documentaries about monarchs and on many websites. I’ve been trying to sort out the migrations and generations since 1992, and the summary below fits the data that I’ve seen to date. Please be aware, I’m describing overall patterns – average conditions – to which there are exceptions if you look all the data that’s available.

Let’s deal with the migrations first since there is a widespread perception that monarchs take three generations/migrations to reach Canada. That’s not the case. Recolonization of the breeding area requires only two migrations. That’s well-established in the literature and you can see the pattern in the first sightings data posted to Journey North each year. Here is a quick summary:

The first migration
Monarchs leave the overwintering sites starting in late February, a process that can continue into the first week of April. This migration starts to reach Texas in the second week of March. The returning female monarchs move N and NE mating, laying eggs until they die. Most are dead by 1 May with the effective northern limits generally around 37°N.

The second migration
Offspring from the returning migration reach the adult stage mostly in mid to late April and, after a few days needed to fully mature, they begin to migrate N and NE. This generation – the first generation of the season – makes the second migration. These butterflies continue migrating until sometime late in the first week of June. At that point, all directional flight ceases, ending the second migration. By the end of the first week of June, this second migration, or wave of monarchs, has reached most of the areas containing milkweeds, and recolonization is complete or as complete as it can be for a given year.

The third migration
There is no directional movement from the first week of June until sometime in the fourth week of July. In other words, there is a six or seven-week interval during which all flight and reproduction is local. Directional flight for SOME but not ALL monarchs begins in the fourth week of July. These monarchs move S and SW and have been observed to recolonize southern areas from Georgia to Texas that had contained few or no monarchs for the previous two or more months. I had called this migration the “pre-migration migration” in previous discussions. That’s an awkward and confusing term so hereafter I’m going to refer to this migration as the “midsummer migration.” Although we know this migration occurs, we don’t know exactly when it starts or ends or the geographic origin of the butterflies. Judging by observations of directional flight and the arrival of fresh monarchs in areas from which monarchs have been absent, it appears that this migration starts in the fourth week of July and ends in the beginning of the third week of August. How long this migration takes to pass through a particular location is not clear. Reproduction follows the arrival of these butterflies from the North creating some confusion as to the number of generations produced in the South. I’ll get to that later.

The fourth migration
It’s the fourth migration that is well known. It starts at 50°N (Winnipeg) at the end of the first week of August and continues through the fall with the last monarchs arriving at the overwintering sites in early December. This migration is the longest both in terms of distance and time. It also differs from the other three migrations due to the fact that butterflies in this migration do not mate and reproduce during the migration. Nor do they reproduce at the overwintering sites in Mexico.

That monarchs reproduce in Mexico during the winter is another misconception frequently repeated in the media. The overall pattern is this: three migrations involving fast-moving butterflies that mate and reproduce as they expand the range seasonally with an outcome that generally increases monarch numbers and one fall migration that contracts the geographic range during which the population declines due to morbidity and mortality. In effect, there is an eight-month period of reproduction (March–October) that overlaps with an eight-month interval (August–March) during which the population declines. Simply put, there are two migrations going south in summer/fall and two migrations going north in the spring.

When thinking about the duration of a developing generation, it is common to add up the intervals needed for monarch immatures to complete each stage from egg through pupation — and that interval can vary from about 25 to over 50 days depending on temperatures. Egg to adult intervals are generally cited as about 30 days. That’s a good average, but it doesn’t account for a period of maturation females need before mating and egg-laying begins. That interval is also temperature sensitive but probably averages 5 days for monarchs at the northern latitudes. My point is that it’s better to think in terms of egg to egg intervals rather than egg to adult intervals when counting generations, e.g. 35 vs 30 days.

As discussed, first-generation monarchs are offspring from monarchs returning from Mexico. These first-generation offspring mate and lay eggs while moving north. At a particular point, new eggs could be deposited by passing monarchs for several weeks, in effect spreading out the second generation over time and space that later leads to overlapping generations. That’s complicated, and we won’t go there since our purpose is to characterize the general pattern of the number of generations. To do so, it helps to create a simple scenario for different latitudes.

Let’s start with eggs laid by a first-generation female on the 20th of May in northern Iowa. Using the egg to egg interval of 35 days, the second-generation females would begin laying eggs on the 25th of June. Offspring from these eggs constitute the third-generation, and they would be expected to begin egg-laying in another 35 days or around the 30th of July. As the eggs from this third-generation complete development and reach the adult stage, they constitute the fourth or fall migratory generation.

However, if we start our scenario later in Iowa, or start further north, we lose a generation. For example, eggs laid in the Winnipeg area by newly-arriving first generation females, let’s say on the 1st of June, probably wouldn’t reach the adult stage and begin egg-laying until 8 July or later, due to a longer egg to egg interval resulting from lower temperatures at that latitude. Eggs laid by those second-generation females become third generation adults in August that migrate. The message here is that most of the migratory butterflies originating from north of 40°N latitude are third and fourth generation butterflies.

But what about further south? Are there 5th and 6th generation migrants? Probably, but it’s complicated and the numbers are likely few. To explain what might be happening, we have to consider the extent to which there are continuously breeding populations in the South and the impact of the midsummer migrants. There appear to be habitat islands associated with major cities where monarchs survive in low numbers through the summer. It’s conceivable that some of these lineages produce 5th and 6th generation migrants. However, the biggest number of migrants originating from the South in September and October are probably progeny of midsummer migrants that were third generation butterflies. If this is indeed the case, then most of the monarchs originating from the South are fourth generation rather than 5th generation monarchs.

The above description applies to the eastern monarch population. It’s possible, maybe even probable, that what I’ve characterized here for eastern monarchs also applies to the western monarch population. There are many seasonal parallels but there are too many unknowns about western monarchs to know if the annual cycle differs significantly from that of their eastern counterpart.


  • Four migrations annually – two northward and two southward.
  • Most butterflies in the fall migration represent 3rd and 4th generations.
  • Three migrations involve butterflies that are reproductive. Butterflies in the fall migration are non-reproductive.
  • Monarchs DO NOT reproduce at overwintering sites in Mexico.
Filed under Monarch Migration | Comments Off on Monarch Annual Cycle: Migrations and the number of generations

Monarch Population Status

1 May 2020 | Author: Chip Taylor

As of yesterday (30 April), the numbers are only 4 and 16 and that’s the way I like ’em. The 4 is the number of sightings of monarchs North of 40°N on the Journey North website. Four is enough, I don’t want to see more than that north of 40°N at this time of year. And the 16, that’s the number of monarchs sighted in Kansas so far this season—and more than half of those were reported in the last 10 days.

What does that mean and why am I happy about those numbers? It means that most of the returning monarchs that produce offspring that colonize the Midwest laid most of their eggs in Texas and Oklahoma, where it is warmer—conditions that allow them to develop faster than they would further north. Further, these numbers tell me there is a possibility that a large cohort of first-generation monarchs will move north of 40°N into the summer breeding grounds in May and early June. That should go well if the long-range forecasts are accurate, especially in the second half of May. The first half of June is forecast to be a bit cooler than normal, and if that comes true, recolonization of the northernmost latitudes could be reduced. All in all, this return recolonization appears to be normal and that’s encouraging.

As many of you may know, the West Coast monarch population has a steep climb to recover from the all-time low overwintering counts of the last two seasons (each less than 30,000). Getting back to the roughly 200,000 overwintering numbers that characterized most of the years prior to 2018 will take favorable conditions for several years. While it is impossible to tell how many monarchs survived the winter to reproduce in the last two months, the numbers of sightings in California reported to Journey North is encouraging. Further, the average temperatures and amounts of precipitation for California were in the near-normal range for March and April combined, and that’s a positive, since previous declines have been associated with extreme high temperatures and droughts for those months. Still, we need to keep our fingers crossed that the Western Monarchs will number more than 30,000 this coming winter.

Filed under Monarch Population Status | Comments Off on Monarch Population Status

Monarch Watch Spring Plant Fundraiser

30 April 2020 | Author: Jim Lovett

For all of you who are located close (or close enough) to Monarch Watch here in Lawrence, Kansas our Spring Plant Fundraiser is now online at

We have thousands of plants available via online ordering and local curbside pickup (or limited local delivery). To place an order, you must live in or be willing to travel to LAWRENCE, KANSAS (we cannot ship). These plants are ideal for starting butterfly gardens or adding to established gardens and can contribute to the health of monarch and pollinator populations.

If you are not able to participate locally we invite you to contribute to this fundraiser by donating to Monarch Watch via

Thank you for your continued support during these uncertain times. Stay safe!

Filed under General | Comments Off on Monarch Watch Spring Plant Fundraiser

Monarch Population Status

20 April 2020 | Author: Chip Taylor

I have a confession to make. Well, actually, several. First, at this writing (12 April), I have no idea as to whether the monarch population is likely to increase or decrease this year. The signals are mixed. March temperatures in Texas are telling me one thing, optimal egg distribution is telling me another, and first sightings yet another. At this point, it’s likely that I won’t get a good sense of the direction the population will take until early June. I’ll get to why I’m confused below, but I have more confessions.

I’m obsessed. Yes, obsessed. It happens. This time I’m obsessed with the First Sighting Data posted to Journey North. This resource has the potential to tell us a great deal about how the population develops each year and I urge everyone interested in monarchs to report their observations to this long-standing and useful resource. There have to be patterns in a database this rich with information. Indeed. But what are they?

I’ve spent a lot of time searching for such patterns. Some are easy, and very clear, such as the distributions of sightings across latitudes going north. Each bump in the data represents cities and the surrounding suburbs, and there are gaps with few reports in areas with few towns and low-density human populations and gaps where milkweeds are scarce. There are also pathways along coasts and even a pathway, at least in most years, across northern Mississippi, Alabama, and Georgia. And there are other patterns related to terrain, prevailing northeasterly movement and the influence of weather patterns in given years. In addition, there are striking differences among years in the timing and number of monarchs that arrive in Texas in March along with the mismatch in the number of sightings relative to the size of the overwintering population the previous winter season. If you look closely, you can see gaps in the sightings that can be associated with weather events such as cold temperatures or periods of rainfall.

In addition, when trying to interpret the data, one has to be aware of its many limitations. It’s easy to show for example, that for specific regions, there are only a modest number of people willing to report first sightings. For a while, it appeared for Texas that that number was perhaps between 190–230 resulting in roughly 180–200 sightings for a number of years with the occasional low year. In this situation, only the low number of sightings is useful information since it would appear that the cohort of people willing to report sightings easily becomes saturated. In other words, due to saturation of the observers, large populations are not properly reflected in the data. A solution, of course, is to increase the number of people willing to send sightings data to JN, and that has been happening as well. That helps, but it presents another problem – how do you compare a recent year with a large number of sightings with previous years with lower numbers of sightings? For those comparisons, you can’t use the numbers and have to look for other metrics such as spatial distribution and temperature.

Below I have assembled screen shots of Journey North first sighting data as of the 11th of April each year. I haven’t given the specific years only identified them by letter, but they are easy to figure out if you are so inclined. My purpose is to challenge you to see some of the patterns I’ve mentioned and to ask you if you see anything in these records that explains the increases and decreases among these 7 years. I’ve included the average temperatures in Texas for March along with the overwintering numbers (hectares) for the previous and following winter seasons.

Figure 1. Journey North first sighting data as of the 11th of April.

Let’s start by comparing A and B. They look very similar, don’t they? C is sort of similar too, right? How about D? Wow! That pattern doesn’t look like A, B or C does it. What about H? Yes, let’s group D and H. What about F and I? again, the patterns are similar with lots of late first sightings (darker spots). That leaves E and G, which also have similar patterns but low numbers of first sightings in total and very few first sightings in the southeast.

Let’s start over with E and G. The populations in both years declined from the numbers recorded the previous winter. March temperatures were +4.9°F and 0.9°F respectively so the only common factors are the low numbers and the distributions. That’s not much to go on so there must be other factors that account for the declines in these two years, and there are, but that is another story.

F and I had similar patterns and appear to have been late recolonizations yet both increased substantially from one year to the next. In both years, the temperatures were close to the long term means for March, -0.1°F and -1.5°F respectively. That’s interesting.

D and H both have broad distributions with monarchs reaching into Nebraska and Iowa as well as Kansas and Missouri in large numbers in year D. The pattern is similar but not as dramatic for H. The mean temperatures were extremely high for both years, +7.3°F and +6.9°F above the long-term averages. However, the populations declined in both years. That’s also interesting and contrasts strongly with F and I when more restrictive distributions and temperatures at or just below the long term yielded significant increases in the population from one year to the next.

The dilemma is how to deal with A, B and C, especially A and B. The patterns for both years are similar. The numbers of first sightings up to the 15th of March were virtually identical (99 vs 104) with a pattern of the overall timing of arrivals and movement northward again highly similar. These factors seem to suggest that similar numbers of first-generation butterflies will be moving north to colonize the northern breeding areas north of 40°N in May and early June. However, the mean temperatures don’t fit the contrasting scenario presented by F and I vs D and H since the March mean temperatures were +6.9°F for A and -0.6°F for B.

We know what happened in B. The population declined when the F and I scenario would have predicted an increase or at least a decline that was not as large as observed. Indeed, all projections seemed to indicate a fairly robust population. As outlined in an earlier Blog article, two things probably account for the 2.83 hectares this past winter, the slowest moving migration seen to date and a drought in Texas and Northeast Mexico. There have been 6 droughts and two late migrations since 1994 and the population has declined in each of those years*. B, actually 2019, was the only year with both a late migration and a drought. So, what about A (2020)? Will the population decline again this year? I can’t tell you at this time. The next step is the recolonization of the northern breeding areas in May and early June and that is followed by the weather conditions from June through August. What happens in May is fairly predictive of what will happen later in the year unless of course there is a drought or a late migration.

And what about C? The first sightings pattern was similar to that of A and B but not as far north by 11 April. The real difference was that the overall weather conditions from March through August were the best seen since 2006 and 2001. The development of the population that year has been dealt with in earlier Blog articles.

To finish, let’s revisit those F & I and D & H scenarios. First let me tell you that these pairs are only examples used to illustrate a point. The entire record supports both points, with exceptions; that is, there are cool March means with declines and warm March temperatures with increases. March temperatures appear to be the strongest driver that determines population growth, but there are numerous factors that can blunt a good population buildup related to a cool March as could be seen in 2019. There are also 2 years in the record during which high March temperatures (N=11) were not followed by declines, but again, there are good explanations for those exceptions. Years during which a population gets off to a bad start, with low numbers of first sightings, or March temperatures well above the long-term mean, or both, usually decline. There is much more to say about first sightings, but that will have to wait.

Ok, I have one more confession. I started writing updates on the status of the population and various additions such as “Why monarchs are an enzyme” to the Blog with the idea that they were needed to educate everyone with an interest in monarch population biology. Yes, I do want to reach out, but the reality is that, while I write these features for readers, I use them to keep myself connected with what monarchs are doing and to work out ideas about how the population functions. So, if you have read this far, thanks for reading, and pardon my self-indulgence.

Oh, there is one more thing for those interested in the Western monarch population. The mean March temperature for California was -0.9°F less than the long-term average with precipitation just slightly above the long-term mean. Those are both good signs, since high temperatures and/or low precipitation in March and April are associated with past declines in the western population. Yet, as with the eastern population, we will have to wait until June to get a sense as to whether the western population will increase this year.

By the way, we should all be appreciative of Elizabeth Howard and her staff for creating and maintaining the first sightings and other monarch data bases on the Journey North website over the years. In case you didn’t know, these tasks have now been transferred to the University of Wisconsin under the supervision of Dr. Karen Oberhauser.

*Droughts and late migrations do not account for the monarch decline observed over the last 14–16 years. A quick review of the history of droughts in Texas and Oklahoma shows that droughts have been more frequent in the past than over the last 25 years. In fact, precipitation data for Texas shows that rainfall has increased over the last 30 years with the current mean 0.6 inches greater than the long-term average. A review of September and October temperatures suggests that late migrations occurred in the past as well. Below are three figures from Climate at a Glance that illustrate these points.

Figure 2. This figure shows the average precipitation per year for the 4-month period before migratory monarchs reach Texas in October. Note the striking difference in the precipitation amounts before and after 1956. The amplitude of the variation appears to be damped with fewer years with low rainfall later in the record. Climate at a Glance.

Figure 3. This figure summarizes average temperatures for October for Texas from 1895-2019. Note the substantial year to year variation in the period before 1976 and the distinct damping of both the low and high ranges from that year onward. Patterns of damping of weather features are common in most temperature records for both March and October indicating that monarchs are experiencing less variability over the last 30 years than in the past. Climate at a Glance.

Figure 4. This figure represents the average temperatures experienced by monarchs as they migrate through the Midwest in September (1895-2019). Again, note the range of these averages in the early years of this record and the appearance of a reduction of this variation in recent decades. The starting point for these shifts is generally associated with the mid 70s coincident with observations by many biologists of seasonal changes seen in animals and plants (phenology). Climate at a Glance.

Filed under Monarch Population Status | Comments Off on Monarch Population Status

Monarch conservation in the age of Covid-19

2 April 2020 | Author: Chip Taylor

milkweed plugs

I consistently overestimate the market for milkweeds and I’ve done it again this year. As an optimist who wants to get milkweeds in the ground to compensate for habitat losses, I ask nurseries to produce more milkweeds than we have been able to cover with sales through the Milkweed Market or funds provided by corporations for the restoration program. Last year my enthusiastic approach cost us over $30,000. This year it looks like we will need an additional $25,000 to cover all the milkweeds we have asked nurseries to produce for Texas*.

Ann and Dena remind me frequently that we don’t have the funds again this year. That’s not good. We need to figure out how to increase sales through the Milkweed Market and to get more support for the Free Milkweeds for Restoration Projects program. I’m not good at marketing. I’m an insect guy and there is nothing in my background that’s helpful in this regard. But we need to step it up, don’t we? We need to get more milkweeds in the ground.

Monarchs are down once again with only 2.83 hectares being reported at the overwintering sites this year. The large year-to-year population increases seen in the past (1996, 2001, see Why monarchs are an enzyme part 3) are no longer possible due to the losses of milkweeds in the Upper Midwest. An analysis of the year-to-year variation in monarch numbers along with the need to have a population that will be able to rebound from extreme winter storm related mortality** will require an overwintering monarch population that AVERAGES 6 hectares. Since 2004, the population has only averaged 3.24 hectares. To achieve an overwintering average of 6 hectares will require the restoration of 1.3 billion milkweed stems. That’s a lot of milkweeds.

We are trying to do our part and through our work with nurseries we have distributed 1 million milkweed plugs since 2010. That’s a start, but given the scale of the problem, we need to do more and to do so we need your help. You can help us market milkweeds by reaching out to others to buy milkweeds through the Milkweed Market or by connecting us with individuals, groups, businesses or corporations that can contribute funds in support of the Free Milkweeds for Restoration Projects program. We know that we are asking a lot in this time of the Covid-19 virus, but monarch conservation doesn’t stop. Let’s plant some milkweeds!

* We are offering Texas customers milkweeds at cost. Flats of 50 milkweed plugs (A. viridis, A. Asperula), appropriate for their region, are available in mid April for $98 each through the Milkweed Market.

** The need to have large overwintering numbers to survive winter storms is clear. Due to increasing temperatures in the mid Pacific, there have been four instances in recent years during which warm air masses moved into central Mexico in the winter months. In each instance, as the air masses encountered cooler air in the mountains, the moisture condensed into rain and sometime sleet or snow, frequently with freezing conditions, that killed the majority of the overwintering populations in 2002, 2004, 2010 and 2016. Temperatures are continuing to increase in the Pacific and such winter events are likely to occur again in the near future.

Filed under Monarch Conservation | Comments Off on Monarch conservation in the age of Covid-19

Gardening for monarchs in the age of COVID-19

19 March 2020 | Author: Chip Taylor

We appear to have entered a new era of uncertain duration, one possibly characterized by waves of reinfection by Covid-19. Yet, we must carry on with monarch conservation – somehow.

We are in new territory. Our lives will change in many ways. Social distancing and quarantines will ground many of us, confining us to our properties yet giving us time to garden for monarchs. By adding milkweeds and nectar plants to our gardens, while waiting for the virus to subside, we can contribute to the health of monarch and pollinator populations. Why not? Gardening gets you out of the house, it’s good therapy, you’ll be in a virus-free environment, the exercise will be good for you and your actions will have a positive impact.

Through our partner nurseries, we have milkweed plugs of many species available for much of the country. For those of you in Texas and Oklahoma, we will be able to ship in April. Why not order a flat today for yourself or your house-bound friends?

Visit our Milkweed Market:

For those who already have milkweed, or have no space to plant your own, you can donate to Monarch Watch in support of our Free Milkweeds for Restoration program to help get milkweeds into the ground.

Donate to Monarch Watch:

Chip Taylor
Monarch Watch

Filed under Monarch Conservation | Comments Off on Gardening for monarchs in the age of COVID-19

Monarch Population Status

13 March 2020 | Author: Jim Lovett

World Wildlife Fund Mexico in collaboration with CONANP and the Monarch Butterfly Biosphere Reserve (MBBR) announced the total forest area occupied by overwintering monarch colonies today. Eleven (11) colonies were located this winter season with a total area of 2.83 hectares, a 53.22% decrease from the previous season (6.05 ha):

Figure 1. Total Area Occupied by Monarch Colonies at Overwintering Sites in Mexico.

WWF release (in spanish): La mariposa Monarca redujo en un 53% su ocupación en los bosques mexicanos de hibernación

Why overwintering numbers were lower this year
Chip Taylor, Monarch Watch Director

In my November 2019 “Monarch Population Status” blog post (Why overwintering monarch numbers will be lower this year) I outlined my reasons for expecting lower monarch numbers.

I finished the article with a summary of bullet points which made sense to me given the data at hand at the time:

Population growth
• Less than optimal egg distribution in March and April
• Later recolonization of the Upper Midwest
• Low monarch production in Iowa and maybe western portions of the upper Midwest
• Lower summer temperatures than in 2018

• Late migrations are associated with lower numbers reaching Mexico
• Droughts are associated with lower numbers reaching Mexico
• High numbers in the northeast do not translate to high overwintering numbers
• Northeast butterflies take too long to migrate southwest

In the text that follows, I will elaborate on these points.

Conditions were less favorable for population growth in 2019 than in 2018. Although the temperatures in March 2019 were near the long-term average, monarchs still moved too far north too soon with many laying eggs at latitudes with cooler weather. With cooler temperatures, the immatures develop more slowly, and the overall effect is to produce a first-generation cohort with an older average age to first reproduction. Growth rates of populations decrease as age to first reproduction increases. Because most first-generation monarchs migrate north in May and early June, the conditions during that time period are critical. In 2019, those conditions – mostly lower temperatures — delayed the recolonization of the summer breeding area north of 40°N. Delayed arrivals can ripple through the rest of the breeding season, resulting in a late migration.

Reports we receive during the summer alert us to booming populations, but silence from areas that are normally productive can be informative as well. Last August we heard about large numbers of monarchs in several areas of the northeast, especially coastal Maine and parts of Wisconsin and southeast Minnesota, but there was silence from, or reports of low numbers from, the western portion of the Upper Midwest. Summer temperatures can be important with both extreme high or low temperatures leading to population decreases (see Monarchs are an enzyme – Part 1). Populations grow well with summer temperatures close to the long-term mean, and that was the case in 2019. However, they grow even better when temperatures are a couple of degrees above normal as they were in 2018.

While conditions for growth of the population weren’t as favorable in 2019 as they were in 2018, the two biggest factors that appear to account for the lower numbers this winter are the lateness of the migration and the drought in Texas. Both late migrations (Taylor, et al. 2019) and droughts (Saunders, et al. 2019) have been associated with lower overwintering numbers. In the earlier post to the blog, I commented on the extreme lateness of the migration in September which I attributed to long periods of high temperatures north of Kansas that inhibited migratory flight. Although there have been a few late migrations through Kansas, since the late 1980s, the two-week-late passage of monarchs in 2019 was absolutely the latest to date. Although we have no way to be certain, hot dry weather probably takes a toll on the monarchs. Longer migrations in terms of the number of days in flight likely add to this attrition. As the migration moved into Texas, nectar was in short supply due to the drought (Figure 2).

U.S. Drought Monitor, Texas, 2019-10-15
Figure 2. U.S. Drought Monitor, Texas 15 October 2019.

Because Texas and northeast Mexico share weather patterns in addition to a border, it’s likely that monarchs continued to experience drought conditions as they entered northeastern Mexico. If so, the drought could have taken a considerable toll of the monarchs with low lipid reserves. However, how much drought monarchs experienced in Mexico in October isn’t clear. The end of the migration monarchs we sampled here in Lawrence, Kansas were both smaller than average and substantially below average in mass and thus ill-prepared to reach Mexico.

In the earlier text, I predicted that the overwintering number would be 4.7 hectares. Having made that claim I pointed out that:

I will be both right and wrong. I will be right about the size of the overwintering population relative to that of last year (6.05 hectares). The number this year will be lower. That’s been clear since late March and early April. I’ll explain why it will be lower below. The question I’ve been wrestling with is how much lower will the number be this winter. That’s where I’ll be wrong. I’ll never hit that number precisely. There are too many variables.

Given the numbers tallied by the World Wildlife Fund Mexico for the total hectares for 2019–2020 (2.83), where I was right and where I was wrong is clear to all. I’ll keep working at it.

Saunders, S. P., Ries, L., Neupane, N., Ramirez, M. I., Garcia-Serrano, E., Rendon-Salinas, E., and Zipkin, E. F. (2019). Multi-scale seasonal factors drive the size of winter monarch colonies. PNAS 116: 8609-8614.

Taylor, O.R., Lovett, J.P, Gibo, D.L., Weiser, E.L., Thogmartin, W.E., Semmens, D.J., Diffendorfer, J.E., Pleasants, J.M., Pecoraro, S.D., and Grundel, R. (2019). Is the timing, pace and success of the monarch migration associated with sun angle? Frontiers Ecology Evolution. published: 10 December 2019

Filed under Monarch Population Status | Comments Off on Monarch Population Status

Why monarchs are an enzyme – Part 3

6 March 2020 | Author: Chip Taylor

Why monarchs are an enzyme – Part 1
Why monarchs are an enzyme – Part 2

In Part 2 of this tutorial on monarch demography, I dealt with realized fecundity and age to first reproduction with the promise that the next topic would be reproductive success. Realized fecundity, as you may recall, is the total number of offspring produced relative to the potential reproductive output. For monarchs and most insects, that means the total number of eggs laid per female lifetime. Reproductive success represents the proportion of those offspring (eggs) that hatch and mature to reach the adult stage. Some might argue that reaching the adult stage isn’t the stopping point. Rather, the end point for defining reproductive success should be the number of offspring that survive to reproduce themselves. Indeed, when it comes to monarchs, it is not enough to talk about the number of monarchs produced in the last generation, the end point for success is the number of offspring that survive the migration, the overwintering stage and the return migration to reproduce in the spring in the southern United States. So, while maximizing the number of eggs laid is an important aspect of population growth, we need to consider all the mortality that follows the egg laying process together with the condition of newly emerging adults.

Once returning monarchs reach Texas, we need to consider the conditions for monarch reproduction in the spring in terms of realized fecundity. Those conditions can be favorable or limited. Limited egg production can be sufficient to produce a large cohort of larvae and eventually adults – IF – the fire ant population is low but insufficient when fire ants are abundant. Even when realized egg laying is high, fire ants can limit reproductive success, thus curbing population growth. To understand this dynamic, we need to know what influences the size and number of fire ant colonies. It’s rainfall. Rainfall governs plant growth, the development of insects that feed on the plants and more. In other words, the abundance and diversity of prey allows fire ant colonies to grow and to produce queens that produce more colonies. Droughts, especially long droughts, reduce fire ant numbers, and cool temperatures reduce their foraging rates. Recovery from droughts requires rainfall and warm temperatures.

So, what do you suppose might happen if there was a long drought before monarchs returned to Texas in March with temperatures that were substantially below the long-term mean? My guess is that milkweeds, being perennial and deep-rooted, would emerge while other vegetation would lag with the result that milkweeds would be “apparent” to returning females allowing them to maximize the number of eggs laid. At the same time, due to lower numbers because of the drought, predation and foraging* by fire ants would be reduced, resulting in what ecologists have referred to as “ecological release.” Has this scenario ever happened? Yes, indeed. Those may well have been the conditions in 1996**, the year of the largest overwintering population in the record (18.19 hectares). There had been a lengthy drought that extended into the spring of 1996, and the mean temperatures for March of that year were -3.2°F below average. It’s easy to imagine that these conditions produced a large cohort of first-generation monarchs that moved into the Upper Midwest. That was before the widespread adoption of herbicide-tolerant crops, and milkweeds were still abundant in corn and soybean fields. That year, the fall migration was spectacular, and overwintering monarchs were reported at colony sites that hadn’t been used by monarchs in years. The net increase in the population from 1995–1996 was 5.58 hectares (12.61 to 18.19). That number prompted me to see if there were other increases as large in the record, and there is one.

From 2000 to 2001 the population increased from 2.83 to 9.35 hectares – an amazing 6.52 hectares! Hmm, what were the conditions in the spring of 2001? Get ready. There had been a severe drought in 2000 (Figure 1), which evidently had an impact on the number of monarchs reaching the overwintering sites that fall (a decline from 9.05 to 2.83).

Figure 1. U.S. Drought Monitor, TEXAS 11 October 2000.

The effects of that drought on fire ants probably carried over into the spring of 2001 (Figure 2) due to cooler than average temperatures (-1.2°F, October – February) that limited colony growth and reproduction.

Figure 2. U.S. Drought Monitor, TEXAS 20 March 2001.

Though rainfall had been well above average during that interval (15.37 inches vs 4.95 for 1996) effectively eliminating the drought, the mean March temperatures were -3.0°F below average once again. Since lower than average temperatures tend to limit most egg laying by returning monarchs to Texas, it’s probable that this egg laying also produced a large cohort of first-generation monarchs that moved north in May due to low fire ant numbers. The greater increase in 2001 vs 1996 may have been due to higher mean temperatures in May (+2.4°F vs -2.8°F) and the summer (+1.5°F vs -0.5°F) that more strongly favored population growth than in 1996. Again, in 2001, there was still an abundance of milkweeds in corn and soybean fields in the Upper Midwest. Those are the only two spring drought and low temperature scenarios in the record. This similarity is a discovery. I had never compared these years before. But there is one other drought to consider, that of 2011 (Figures 3 and 4).

Figure 3. U.S. Drought Monitor, TEXAS 18 October 2011.

Figure 4. U.S. Drought Monitor, TEXAS 20 March 2012 .

The dynamics were different in the spring of 2012. Instead of cool temperatures, the mean March temperature was +6.8°F. The vegetation had rebounded somewhat from the drought of the previous season. In 2011 there had been a seven-month drought (rainfall of 4.06 inches from February through August) – a drought of historic proportions that changed the landscape due to the death of hundreds of millions of trees. The drought ended in September and was followed by seven months during which rainfall exceeded the long-term average (20.44 inches). There were several results of this shift from drought conditions to warm and wet that pertain to monarchs and the points made earlier and in Part 2. First, the fire ants, which had declined during the drought had not yet rebounded by March of 2012. Second, the high March temperatures and favorable winds allowed the returning monarchs to expand their distribution rapidly with the result that large numbers of overwintering monarchs were recorded north of 40N in late April and early May (see Journey North first sightings for 2012). They arrived too far north too soon. For monarchs, this meant that reproduction in Texas was minimal in contrast to years with cooler temperatures, and, by distributing eggs further north into areas with cooler temperatures, it meant that age to first reproduction for the entire returning female cohort was much older than normal – a negative for population growth. Mean age to first reproduction was likely one of the factors that accounted for the decline from 2.89 hectares in 2011 to 1.19 hectares in 2012. This outcome, in contrast to that of 1996 and 2001, suggests that cooler March temperatures result in high rates of egg laying in Texas which favors population growth, especially when the fire ant population is low. So, if fire ants were down in the spring of 2012, was there evidence of “ecological release” that spring as well as in 1996 and 2001? The answer is yes.

There was a massive migration northward of other butterfly species that spring. Conditions for the development of butterfly populations were favorable yielding large numbers of dispersing butterflies which I attributed to “ecological release.” Starting with an abundance of red admirals in the first week of April, I recorded the arrival of 16 species of butterflies in eastern Kansas that originated from Texas. The numbers were like nothing I had ever seen. All species were extremely abundant relative to their usual numbers. The red admiral migration was so spectacular that it created headlines as far north as Toronto***. I prepared a slide show and gave talks about the uniqueness of this event which was likely due to low fire ant numbers, as well as those of other predators and parasites. Part of this dynamic is due to the fact that predators and parasites have lower reproductive rates than their prey or hosts.

The point of this discussion is that predation can have a major impact on reproductive success. We can add other predators, parasites (Tachinids) and pathogens (O.e., etc.) to that list, but there are other factors to consider such as density dependence (too many larvae per plant), milkweed condition, latex production that traps first instar larvae and extreme physical conditions that affect survival of immatures and even adults during the migration. Understanding the impact of each of these factors on reproductive success and how they vary regionally and from year to year, along with variation in realized fecundity and age to first reproduction, will help us understand inter-annual variation in monarch numbers.

*Fire ants appear to require surface temperatures of at least 80°F (27°C) to forage.

**Although there is no Drought Monitor map for March of 1996, that drought is well represented in the literature (see references).

***Sample headlines included the following:

“Canada’s butterfly migration is largest on record”
300 million red admiral butterflies estimated from Windsor to New Brunswick
CBC News Posted: May 16, 2012 5:12 AM ET

“Southern Ontario sees ‘irruption’ of red admiral butterflies”
Published on Saturday April 21, 2012

“Red Admiral Butterfly Invasion Update: Some Will Settle In NYC!”
The great butterfly migration: Unnoticed by most, Red Admiral silently invades N.J.
Published: Friday, May 11, 2012, 8:02 AM

“Syracuse hit by swarm of butterflies”
Published: Thursday, May 03, 2012, 9:20 PM

“Red Admiral, ‘Butterfly of Doom,’ population explodes in NY Special”
By Victoria N. Alexander May 5, 2012 in Science
(Abundant the year the Russian Tsar Alexander II was assassinated.)

“The Migration Event Of The Week Wasn’t Birds, It Was Butterflies!”
Salmon Creek Tree Swallow Project, May 5, 2012


Allen, C. R.; Epperson, D. M.; and Garmestani, A.S., “Red Imported Fire Ant Impacts on Wildlife: A Decade of Research” (2004). Nebraska Cooperative Fish & Wildlife Research Unit — Staff Publications. Paper 31.

Brown, W. O. and R. R. Heim, Jr. Drought in the United States: 1996 Summary and Historical Perspective. 1997.

Lu, Y., Wang, L., Zeng, L. and Y. Xu. 2012. The Effects of Temperature on the Foraging Activity of Red Imported Fire Ant Workers (Hymenoptera: Formicidae) in South China. Sociobiology 59(2):573-584.

NOAA. Update on Drought Conditions in the Southern Plains and the Southwest. 1996.

Suitable Habitat for Imported Fire Ant Colonization Under Natural Rainfall and Irrigated Conditions. USDA APHIS.

Filed under Monarch Biology | Comments Off on Why monarchs are an enzyme – Part 3

Monarchs and climate in the West

25 February 2020 | Author: Chip Taylor

Many months ago, I received a request to be a keynote speaker at a Monarch Summit in California. I accepted the invitation with some reluctance realizing that most of the audience wouldn’t be interested in what I had to say about the eastern monarch population. In addition, I knew almost nothing about what was happening in the West. I had purposefully ignored the West for several reasons. First, I had plenty to do trying to manage Monarch Watch and understand what was happening in the East. Second, I was generally ignorant about the dynamics of the Western population, and third, there always seemed to be a lot of conflict in the West that wasn’t exactly inviting. Further, a number of colleagues had begun publishing papers on monarchs and habitats with milkweeds in the West. Nevertheless, no longer being as smart as I was when younger, I decided to determine if I could learn anything new about the Western monarch population by applying what I know about how the eastern monarchs respond to seasonal conditions. My first approach was to examine the inter-annual variation in monarch numbers in the West relative to monthly and seasonal averages in temperature and precipitation. In other words, I wanted to determine if linked seasonal conditions, or weather, explained any of the year to year variation. They do, sort of, but that’s another story that will require more study. What I want to deal with in this piece is how weather led me to climate, and it’s the climate story I want to tell since that story is critical to understanding what is happening to the Western monarch population and what might happen to monarchs in the near future.

Did you know that there is a difference between how the terms weather and climate should be used? Lots of folks don’t – including politicians and some scientists. Weather is defined as the state of the atmosphere at a place and time as regards heat, dryness, sunshine, wind, rain, humidity, barometric pressure and cloudiness. In other words, it’s what is happening now or forecast to occur in the near future. On the other hand, climate represents the long-term average of weather over a period of 30 years or more. Another way to think about the difference is that weather is a snapshot in time while climate represents past events whose trends might inform us of both future weather events and long-term changes.

The transition from thinking about how weather might explain year to year variation in monarch numbers to the potential impact of long-term climatic changes in California produced a number of surprises. Let’s start with the main breeding season. The mean temperatures for the 5-month period from May through September for California are summarized for the last 30 years in Figure 1.

Figure 1. California average temperature May–September. Modified from

The trend line shows temperatures increasing through this 30-year period. Although the long-term trend from 1895 to 2019 shows an increase of 0.2 degrees per decade, the mean for the last 30 years shows an increase of 1.63 degrees or 0.54 degrees per decade. Added to that figure are the 30 year means and rates of increase per decade for AZ, the NW and UMW (Upper Midwest), with the latter showing the lowest rate of change. It’s clear from these data that the West is heating up faster than the Upper Midwest, in fact, faster than the rest of the lower US. That result is sobering, but the real shocker is what has happened along the California coast over the last 20 years (Figure 2).

Figure 2. Average temperature January–February in four counties in California. County locations shown in Figure 4. Modified from

Figure 2 displays both the long-term mean and the mean for the last 20 years for January–February for 4 counties from Marin in the north to San Diego in the south. The differences range from +1.7°F for Marin County to +2.4 degrees (.85–1.2 degrees/decade) for San Diego County. Wow! That’s an incredible rate of increase over such a short period. When I saw those numbers, I asked what might be happening at the overwintering sites. It had been rumored that monarchs were moving progressively northward to overwinter. I looked for references where this interpretation had been published but found nothing. So, I decided to dig into the Thanksgiving Day Count Database maintained by the Xerces Society. The result of my partial analysis is shown in Figure 3.

Figure 3. Changing distribution of overwintering monarchs in California.

Since many new overwintering sites had been added to the Thanksgiving counts after 2011, sites added after 2011 were excluded from my tabulation to avoid biasing the interpretation. As you can see in Figure 3, the number of sites and the number of butterflies in the southernmost counties (San Diego, Orange, and Los Angeles) has declined to almost nothing while the counts in Marin, the most northerly county, indicate that there has been a significant increase in the proportion of the overwintering population in that county from 1998 to that of 2015–2018. Although these results indicate that sites are being lost in southern counties, and that monarchs are progressively moving northward to overwinter, these trends deserve a more in-depth analysis by someone broadly familiar with the monarch overwintering in California.

map of California counties
Figure 4. Map of counties in California. Starred counties referenced in Figure 2.

So, what might be happening in the southern most counties? The temperature data for San Diego County in January and February (Figure 2), and the preceding months*, may signal that it is simply too warm for monarchs to form clusters, or stay in them, resulting in progressive movement northward to cooler locations. If this interpretation is correct, what can we expect in the future if mean January-February temperatures continue to increase at 0.80 degrees, or greater, per decade? We don’t want to go there, do we? We don’t want to think about it, but the reality is that, by the end of the next decade, mean temperatures could be 50.8–51.0°F for all counties along the California coast. Such temperatures could further limit the ability of monarchs to successfully overwinter in traditional sites. Would monarchs move inland to cooler overwintering sites or continue moving northward along the coast in search of the temperatures and cluster sites that favor overwintering? That’s a good question. Folks familiar with coastal and inland conditions in January and February might be able to answer that question. I can’t.

The loss of overwintering sites and lower numbers of monarchs through the winters in the southern California counties raises another question. What about spring recolonization? With fewer overwintering butterflies, recolonization of the southern areas of California and Nevada will be limited further contributing to the decline of monarchs in those areas, a decline that could move northward as it continues to warm along the coast. At the same time, monarchs overwintering further north would be closer to the summer breeding grounds in the Northwest. That might be beneficial since the limited isotope data reported by Yang, et al (2015) suggests that, at present, roughly 50% of the overwintering monarchs originate from that region.

As to why the mean temperatures along the California coast are rising so rapidly, you can blame it on the increase in temperatures in the Pacific, and if you ask why those temperatures are increasing, point your finger at yourself and at the system that has failed to curb the emission of greenhouse gases (CO2, methane, nitrous oxide) that trap heat energy – 90% of which is absorbed by the oceans.

*November–December mean temperatures for San Diego County increased from 53.5°F (the long-term mean) to 55.0°F from 1990–2019 or 0.5°F per decade.

Office of Environmental Health Hazard Assessment
Yang, L.H., Ostrovsky, D., Rogers, M. C., and J.M. Welker. 2015. Intra‐population variation in the natal origins and wing morphology of overwintering western monarch butterflies Danaus plexippus. Ecography 39:998-1007.

Filed under Western Monarchs | Comments Off on Monarchs and climate in the West

Why monarchs are an enzyme – Part 2

25 February 2020 | Author: Chip Taylor

See Why monarchs are an enzyme – Part 1 posted earlier this month.

What the heck is realized fecundity/fertility and why is it important?

A term I mention from time to time in my talks is realized fecundity. Add to that, I might mention fertility, reproductive success and age to first reproduction. I first encountered these terms when taking an ecology course in graduate school in the late 1960s. There was a unit in the course that dealt with demography – birth and death rates, stable age distributions and factors leading to declines or growth of populations. Among these lessons were a few exercises designed to inform the students of factors that most strongly influenced population growth. The two that stood out were realized fecundity and age to first reproduction. Both are key to understanding population growth in monarchs and inter-annual variations in the sizes of the population at the end of the growing season.

When assessing population growth, the focus is on females. Each monarch female has a potential to produce offspring – a maximum number of eggs that could be laid given the size of the female, the fat body carried over from the larval stage, it’s inherent fitness as defined by its genes, etc. Total lifespan is a factor as well. There are only so many wing beats and degree days per life time. These are all intrinsic factors, that is, properties of each female. Realized fecundity deals with extrinsic factors that have the effect of limiting the number of eggs laid by a female in her lifetime. The list of extrinsic factors is long and involves both physical factors such temperature, precipitation, and wind speed as well as biological factors that include plant quality, nectar availability, number of and fertility of mates, predators, etc. In addition, we need to consider that there may be a cost in terms of eggs not laid due to search time required to find suitable host plants and nectar sources. Search time involves habitat fragmentation which I may get to in other posts.

A basic tenant of population growth is that populations with a short age to first reproduction grow faster than populations in which reproduction is delayed. When talking about monarchs, age to first reproduction can be defined as the interval from when an egg is laid until a female that has developed from that egg lays her first egg. That interval can be as short as 30 days (4 egg, 12 larva, 10 pupa, 4 mating to first egg) and at least as long as 50 days in the spring. This range is due to differences in the temperatures experienced throughout the developmental period. In other words, the temperatures experienced throughout development determine age to first reproduction. As we will see, the distribution of eggs across the latitudes in the spring has a big role in determining age to first reproduction and therefore population growth.

While I used the potential egg laying capacity of individual females to introduce this topic, we need to consider females as a group or cohort to determine the impact of extrinsic factors at the population level. So, what are the conditions that would enable a cohort of females to achieve a high-level egg laying or, conversely to reduce egg laying? I’ve examined the weather conditions during the annual cycle for the eastern monarch population for every year since 1994. That analysis has yielded lists of factors that favor and do not favor monarch population growth (Tables 1 and 2). As you go down the list in each of these figures, you will note that a number of the factors listed, such as winter mortality or survival as monarchs migrate north from the overwintering sites into Texas in March, have to do with losses that ultimately define cohort size. Some unknown portion of that mortality may be due to the condition of the butterflies that arrived at the overwintering sites in the fall. For example, butterflies that experienced drought, either during development as a larva or during the migration itself, are more likely to die during the winter and migration north than butterflies that developed and migrated under more favorable conditions. Further, it is probably the case that the potential reproductive capacity of female cohorts arriving in Texas varies from year to year. It should be noted that both nectar and water availability during both the winter and the spring exit from Mexico play a role in survival during these periods.

Table 1. Optimal conditions for population growth.

Table 2. Negative conditions.

Once the cohort of returning monarchs reaches the milkweed rich areas of Texas (12-15 March), realized fecundity and age to first reproduction becomes important. Conditions favoring egg laying include temperatures that are less than -1.5°F below the long-term average, abundant emerging milkweeds, adequate nectar and water, moderate winds and relatively low precipitation from 15 March to 15 April. The lower than average temperatures extend the life of the females (lower number of degree days) but also have the effect of limiting northward movement. The net effect of lower temperatures is that most eggs are laid in Texas and southern Oklahoma where it is relatively warm allowing the immatures to grow rapidly – thus reaching reproductive age in a minimum number of days. In contrast, with warmer temperatures females continue moving north laying their eggs at latitudes with cooler temperatures. The effect is to produce offspring with longer ages to first reproduction at these latitudes and to increase the average age to first reproduction for all offspring produced by the cohort arriving from Mexico. The importance of where eggs are laid by returning females and age to first reproduction is supported by the observation that in all four years with mean temperatures of less than -1.5°F for March in Texas, the populations grew from one year to the next. However, the populations declined in 9/11 years during which temperatures were greater than 1.9°F above the mean. One of the two years with high mean March temperatures in which the population increased was 2018 (+5.4°F). In other years with similar high temperatures, the returning monarchs moved northward into Oklahoma, Kansas and sometimes Nebraska, but not in 2018. This unusual dynamic was due to a low that settled over north Texas and southern Oklahoma in late March. Temperatures were low enough during this period to keep monarchs confined to central Texas well into April*. Thus, egg laying was largely confined to Texas where warmer than average temperatures accelerated the development of the immatures. The result was a large cohort of first-generation monarchs that migrated northward in May, a cohort with a low and therefore favorable age to first reproduction. This combination of warm conditions favoring rapid development of immatures yet cold that largely restricted egg laying to Texas in March and into early April has only occurred once since 1994. Yet, it was one of the major factors that contributed to the increase in monarch numbers from 2017 (2.48 hectares) to 2018 (6.05 hectares).

Up to this point I’ve just hinted at how realized fecundity can be modified by extrinsic factors. Let’s consider drought, high temperatures and extended periods of rainfall. There are other factors, but these cases will provide examples of how deviations from optimal conditions can influence population growth.

Summer droughts affect monarchs, nectar sources and host plants. Monarchs need water which is usually obtained from nectar or dew during the summer, and both are scarce in droughts. In addition, monarchs need the carbohydrates (and amino acids) found in nectars to fuel flight, egg development and egg laying, etc. Lack of water and nectar can result in fewer eggs laid and even shorten life span for a reproductive cohort. Higher than average temperatures (>+2°F) can have similar effects. Plants develop faster resulting in shorter flowering intervals, often with lower nectar production and more rapid senescence, the latter making the milkweeds less attractive to females for oviposition. Again, adult life span is reduced at higher temperatures. Rainfall, if prolonged over several days, or any weather that restricts flight and egg laying for a number of days, also reduces realized fecundity. There are only so many degree days in the life of an adult monarch (530)**, time is ticking, and, as pointed out in Zalucki and Rochester (2004), there is no recovery from lost opportunities to lay eggs.

If you have been able to follow this tutorial, it should be apparent how deviations from optimal conditions in terms of realized fecundity are the basis for the stage-specific model I’ve mentioned in previous posts to this Blog. But, it’s not the only factor. We need to consider reproductive success as well. Beyond that, we need to discuss how populations recover from a series of negative conditions that have significantly reduced the size of reproductive cohorts.

*Although the optimal temperatures that allow returning monarchs to move northward are not known, it’s clear that advances are limited when temperatures are less than 70°F.

**The estimated number of degree days (530) represents a life span of roughly 3-4 weeks for reproductive monarchs under average summer conditions (Zalucki and Rochester, 2004). Longer life is possible during periods with daytime temperatures in the 60s. Shorter life spans are expected when temperatures exceed 90°F.

Zalucki, M.P. and W.A. Rochester. 2004. Spatial and temporal population dynamics of monarchs down under: Lessons for North America. In The Monarch Butterfly: Biology and Conservation, eds., Oberhauser, K. S. and M.J. Solensky. pp. 219-228.

Filed under Monarch Biology | Comments Off on Why monarchs are an enzyme – Part 2