Monarchs: Weather and population sizes in the past

If you have been reading about monarchs, you know they have declined in number. One account after another maintains that the population has declined 90% or, more conservatively, 80%. These declines are based on the first three years (1994-1996) in which the colonies were measured*. The populations increased from one year to the next in this short interval ending with the massive population in 1996 (18.19ha). The assumption has been that these three years are the standard – the average condition that might be expected year after year. No one has asked whether the populations were lower or possibly even higher in the past. Further, there has been no comprehensive analysis of why the sizes of the populations vary from year to year. A clue to what determines population numbers occurred in 1997. The number of hectares of monarchs dropped from 18.19 in 1996 to 5.77ha in 1997. There had to be an explanation for such a sharp decline. My pursuit for an answer for that decrease led to an analysis of the increases and declines over the last 29 years.

This quest has shown that the direction of change from one year to the next can be explained by single weather events in some cases and combinations of events in others. Extreme weather events such as late freezes (1997), cold or excessively warm summer temperatures and droughts account for a number of the declines. March temperatures in Texas as well as September temperatures in the Upper Midwest that are well above the long-term average, are also associated with declines. Extreme mortality during the winter months was a factor in several years. These changes can be seen in the data even though the population declined during a period of significant habitat loss from 1998-2012 due to the adoption of herbicide tolerant crop lines and the initiation of the Renewal Fuel Standard in 2007. Since 2012 or so, the rate of habitat loss has declined with the populations varying from year to year without a clear sign that the decline is continuing. In this text, I’m going to apply how monarchs responded to extreme deviations in temperatures in the last 29 years to years in the past to determine if conditions were similar or different from those in the present.

The Premise

Monarchs have optimal temperatures at which they function and therefore temperatures either too low or too high relative to these optima have negative effects. Since these optima are not well defined, we can only estimate how monarchs respond to ranges of physical conditions. Still, since populations develop well when temperature ranges and precipitation amounts are near the long-term averages and develop less well when these values substantially exceed these means, we can identify both favorable and unfavorable conditions for population growth. The goal is to use these criteria together with a stage specific growth approach to create a narrative for each year that explains why the population increased or decreased. This exercise has been used to explain population growth for the last 29 years (in prep.). In 27 of those years, the metrics were consistent with the outcomes. In two cases, both involving the last two years (2021, 2022), the metrics predicted lower numbers of hectares than were reported in Mexico.

In this exercise, I scanned the climatic record from 1895 to 1994 for years with extreme conditions for each of 8 metrics. The extremes for each metric were used in combination with the other metrics for those years to assess the likelihood that each population would decrease or increase.

Criteria

The first step was to define “average” conditions for temperatures and precipitation i.e., to define the ranges of values that appeared to only result in positive outcomes. For temperature, the range for “average” conditions was >-3.3 to +2.5F. In other words, temperatures less than -3.3F and more than 2.5F were deemed as having a negative impact on population growth.

Since an abundance of precipitation contributes to plant growth, nectar production, and the availability of water for monarchs, while the lack of precipitation can lead to all the negative consequences of drought, deviations from long-term means were included in the metrics. While there is no way to associate above average rainfall amounts with population growth, deviations of more than minus 2 inches were deemed as having a negative impact on population growth. This is a low standard, but since large areas are involved and precipitation is scattered, which can result in significant variation in soil moisture and a variable impact on plants and monarchs, this measure seems appropriate. Further, precipitation deficits ranging from -2 to -3 inches have been associated with low outcomes.

The ranges for both temperature and precipitation, while not precise or defined by experimentation, were informed by outcomes, i.e., deviations that appeared to have an impact on population growth.

Interpretations

The interpretations for the effects of deviations that exceed the average condition are summarized in Table 1C. Most of these interpretations reflect expected responses to extreme highs or lows for insects. There is one special case that applies to monarchs. Monarchs returning from Mexico expand their distribution to the north at rates that are determined by temperature and southwesterly winds. When March and April temperatures are high relative to the long-term average, the returning monarch advance rapidly often getting ahead of the emergence of milkweeds. These advances often take them to latitudes in warm intervals that later become quite cold – sometimes freezing. Eggs and larvae distributed through these advances take longer to develop due to the colder conditions with the result that the overall mean age to first reproduction of the offspring of returning monarchs is greater. Older age to first reproduction by cohorts slows population growth. Said another way, if the returning females lay most of their eggs in Texas and southern Oklahoma, the populations will grow faster than if, in addition to Texas and Oklahoma, eggs are also laid in Kansas and Nebraska. In effect, there is an optimal egg distribution that varies from year to year with temperature. The most favorable outcome is for most eggs to be distributed in Texas and southern Oklahoma when March and April temperature are above the long-term mean. This scenario occurred in 2018. Monarchs were confined to Texas and southern Oklahoma due to a dip in the jet stream that blocked movement to the north. The result of this combination, and other factors later in the season, was an overwintering population of 6.05ha, the largest population since 2006 (6.87ha).

Discussion

The data and interpretive statements are summarized in Table 1A, 1B and 1C.

Table 1A. Deviations from long-term means for temperature and precipitation for 15 years with all-time extreme values for one or more metric. Orange = extreme value, pink = a negative value, yellow = average range and green = a positive outcome.

Table 1B. Highest and lowest deviations from long-term means in the last 29 years.

Table 1C. Interpretations of deviations that substantially exceed the ranges for “average” conditions.

The search through all records yielded 15 years (Table 1A) with one or more metric that exceeding the most extreme value seen in the last 29 years (Table 1B). The likely effects of these extremes are summarized in Table 1C. As part of this assessment, I looked at the record for each year that preceded the years in this summation. In each case, the prospects for growth were better than in the years in this record and it seems likely that the population decreased in all years with the possible exceptions of 1956, 1958 and 1969 which could have experienced modest growth. It is likely that substantial declines occurred in 1907, 1910, 1915, 1931,1936, 1988 and 1992 with more modest declines in all other years. Because the extreme values for 1969 and 1976 have no counterpart in the record for the last 29 years, assessing the effects of these measures is difficult. While it seems probable that the outcomes were negative, modest growth also is a possibility.

A comparison of the highest and lowest deviations from long-term means for the last 29 years (Table 1B) with those in the historical record (Table 1A) shows that most of the deviations in the latter far exceeded those of the last 29 years. It’s difficult to assess the effects of some of these extremes. There are 5 years in these records in which the mean March temperatures were -4.6F to -9.8F. These lows exceed the lowest March temperatures (-3.2) recorded in the last 29 years by a large margin. In fact, the mean lows ranged from 46.5F to 51.7F. These temperatures were too low for monarch activity of any kind and likely meant that the recolonization from Mexico was delayed an entire month or even more in 1915, 1931, 1958, 1969, and 1970. There is simply no way a population experiencing these conditions could increase relative to a more robust population the previous year. The decline in 1915 was likely to have been so severe as to taken many years to recover. That outcome was likely as well for the back-to-back declines in 1969 and 1970. Significant declines may have occurred in several other years such as those with significant droughts in the Upper Midwest in 1910, 1936, 1976, and 1988. The low summer temperatures in the summer of 1992 (the Mt. Pinatubo summer) surely led to a decline as well.

While declines were likely in all of these years, the sizes of these populations are unknown. Still, it seems probable that the number of hectares in many cases were as low or lower than lows during the worst period in the last 29 years: 2011 (2.89ha), 2012 (1.19ha), 2013 (0.67ha), and 2014 (1.13ha).

There are two general points to make about these results. First, that the physical conditions were much more variable and more extreme in the past than during the last 29 years (Figure 1). As a result, it is likely that there were many periods during which the monarch populations were quite low. Second, whether populations increase or decrease in a given year is determined by weather irrespective of the abundance of monarch habitat. In other words, even though habitat losses from 1998-2012 led to an overall decline in monarch numbers, the impact of those losses are not apparent in the data from one year to the next.

Lastly, this record only includes years with extreme deviations from the long-term means. There were certainly a number of other years from 1895-1994 in which the population declined due to an extreme event or combinations of conditions.

*It is generally not mentioned, but the 7.84ha cited as the total overwintering area was an undercount. The record shows that a number of smaller colonies were not measured that year (Garcia-Serrano, E and X. Mora-Alvarez. 1999).

Figure 1. Average March temperatures for Texas from 1895-2023. There are three trends in these records: the high variation in these records that ended in 1974, the damped variation from 1975 to perhaps 1994 and the progressive increase in temperatures from 1994 to the present. The average temperatures have increased 0.8F per decade since 1975 to 60.14 F vs the long-term average of 56.3F.

Reference

Garcia-Serrano, E and X. Mora-Alvarez. 1999. Monitoreo de las colonias de mariposa en sus sitios de invernacio en Mexico. 1997 North American Conference on the Monarch Butterfly, Morelia, Mexico.

Monarch populations during the Dust Bowl years

Introduction

The concern about the potential listing of monarchs as threatened or endangered by the Fish and Wildlife Service sometime in 2024 has led me to look into the conditions prior to 1994-1996. The populations in those years were quite large, and it has been widely assumed that the numbers of hectares measured in Mexico represented what could be expected from year to year. I’ve been skeptical, and in a previous post, I summarized data for 15 years with 1 to 3 temperature or precipitation metrics that exceeded anything that occurred in the last 29 years. The data for these 15 years suggested that significant declines were likely in all but two cases. Two of those 15 years were in the 1930s – the “Dust Bowl” or what is sometimes referred to as the “Dirty Thirties”. This led me to ask how monarchs might have fared throughout this period.

If you haven’t read about the Dust bowl or viewed the Ken Burns dust bowl documentary, I highly recommend them. The stories and lessons are compelling and sobering (Egan, 2006). Sobering because we may be headed for more of the same in the future. Farming recovered rapidly in this region after WWII as farmers tapped into the Ogallala Aquifer. This aquifer consists of Pleistocene (fossil) water stored beneath the surface of an enormous area that extends from South Dakota well into the former dust bowl areas further south. Unfortunately, water is being withdrawn from the aquifer faster than it is being replenished with the result that wells are being abandoned in many areas, taking the region back to dryland farming and the conditions that led to the dust bowl. Further, increasing temperatures and severe drought are predicted for Texas and parts of the Midwest in the years to come (Nielsen-Gammon 2021). I will leave it for you, the reader, to educate yourself about the origins and broad consequences of this disastrous yet formative time in our history. My task is to bring together three things: 1) the geography of the dust bowl, 2) the areas from which most monarchs originate that reach Mexico and 3) the broad sweep of the weather that would have affected monarchs during this interval.

The dust bowl area, which is described as the High Plains or the southwestern Great Plains, encompassed portions of 6 states (Fig.1). Much of the dust bowl was bounded by 35N-40N latitude and 100W-105W longitude. The dust storms that characterized the region in the 1930s followed the conversion of the short grass prairies to wheat production and grazing in the early decades of the 20th century. The bare soils associated with farming, droughts and high winds led to dust storms that darkened the skies and spread dust as far as the east coast.

Figure 1. Dust bowl region showing areas and years of major dust storms. See “Dust Bowl” Wikipedia entry for more detail.

Short grass prairies are not major monarch production areas. Milkweeds are few, and nectar sources can be scattered and scarce for butterflies for parts of the breeding season. Further, the area is south and west of the major regions, Upper Midwest and Ohio Valley, that produce most of the monarchs that reach the overwintering sites in Mexico (Fig 2.) (Taylor, et al., in prep). So, if the region of the dust bowl is not a major production area, why is it of concern? Because what happened in the dust bowl from 1930-1939 occurred throughout the area east of the Rocky Mountains. These were the hottest consecutive 10 years during the growing season seen in the entire record from 1895-2022. This period was characterized by droughts, extreme low temperatures in the spring and higher than average summer temperatures in every year, except 1935, in that span. These conditions resulted in lower crop production in many areas to the east of the dust bowl region. And the dust itself settled widely with unknown consequences for foliage feeding insects and other herbivores.

In the following text, I’m going summarize how the growth of monarch populations was probably affected during this period. We now have a base of 29 years for which the conditions have been defined during the growing season. How these conditions affected the growth of monarch populations is now fairly clear. Average temperatures favor population growth while extreme temperatures, either above or below the long-term averages negatively affect population development. The negative factors have greater effects in March and April at the start of the growing season and lesser impacts in the summer and during the migration. Droughts that occur during the growing season or the migration can also limit population growth or the proportions of the population that survive the migration.

Methods

A series of metrics were chosen that represented critical times and locations associated with different stages in the development of the monarch populations. The long-term means and the deviations from these means are summarized in Table 1. These deviations were judged to be either favorable, neutral or unfavorable based on the assumption that there is an average range that enables population growth outside of which high or low temperatures were unfavorable. There are also a few conditions when high temperatures that occur when means are typically low are favorable. For temperature, the range for “average” conditions is >-3.3 – +2.5F. In other words, temperatures less than -3.3F and more than 2.5F were deemed as having a negative impact on population growth.

Table 1. Highest and lowest deviations from long-term means in the last 29 years.

The effects of the deviations from long-term means for precipitation are more difficult to assess. Obviously, there can be too much precipitation, but there is no scale for too much rain. Furthermore, excessive rain tends to be local. In this case, I let the outcomes define the unfavorable conditions since there were years in which deficits of rainfall of as low as -2inches were associated with declines.

These criteria were applied to the deviations for 1927-1942 in Table 2. I added years on either side of the dust bowl years to show both the conditions prior and after that period. The inclusion of 1927-1929 captured the fact that droughts actually started at the northern latitudes earlier than they did in the dust bowl region. The droughts in the Upper Midwest in 1927 and 1929 were linked to the start of dust bowl-like scenario that developed in the Canadian prairies.

Table 2. Deviations from long-term means (1901-2000) during the growing season for 1927-1941. Likely outcomes are based on population responses to variation that occurred in the 29 years from 1994-2022. The deviations are color coded as follows: blue = colder than recent, orange = hotter than recent, light orange = drier than recent, green = favorable, pink = unfavorable, yellow = neutral.

The interpretations for the effects of deviations that exceed the average condition are summarized in Table 3. Most of these interpretations reflect expected responses to extreme highs or lows for insects.

Table 3. Interpretations of deviations that substantially exceed the ranges for “average” conditions.

Results and Discussion

Two periods of contrasting outcomes are evident in Table 2. The period from 1927-1936 is characterized by a large number of unfavorable events most of which resulted in low end of season numbers. In contrast, the populations likely grew in the five years from 1937 though 1941. There were only three months in these years – all in March in Texas – with strong unfavorable measures that otherwise feature average conditions.

This record contrasts strongly with the outcomes from 1994-2022 in two ways. In that record, there was only one year in which precipitation was unfavorable in the Upper Midwest (2012, -2.27inches) while there were 7 such years in the 1927-1936 interval. Surprisingly, there were 5 years during which the March temperatures in Texas were well below the long-term averages – in contrast to none in the more recent period.

Cold temperatures in these years surely delayed the recolonization of Texas by monarchs returning from Mexico and slowed the growth of the first generation. These conditions probably limited the size of the first generation as well. In 1931, low March and April temperatures may have delayed the development of the population for an entire month or the equivalent of an entire generation. Low May temperatures in Minnesota are associated with declines in both the early and most recent records. These low temperatures delay the colonization of the summer breeding area north of 40N and the development of the second generation. During the summer months in the Upper Midwest, both low (1927) and high (1933,1936) extremes contributed to declines. High temperatures during September in the Upper Midwest may also have contributed to the declines in 1931, 1933 and 1936.

Although it is impossible to infer the number of monarchs that survived to overwinter from 1927-1941, it is likely that the numbers were quite low during the first 10 years of this period, perhaps even lower than the 0.67hectares recorded in 2013. There were 7 years from 1927-1936 with one or more deviation that exceeded any deviation seen in the last 29years. The lowest of the low points could have been at the end of 1931, the end of three consecutive years with strongly unfavorable conditions or perhaps at the end of 1936, a year with several unfavorable deviations from the long-term means.

So, while the region of the dust bowl didn’t contribute to declines in monarch numbers from 1930-1939, the conditions in those years from Texas to the central Dakotas and east to the Ohio Valley surely did. This was a remarkable interval. There is no other period in the record from 1895 to the present in which the growing season conditions were so similar from year and so damaging. And, no other period that has had such a profound impact on the people in affected areas. The lessons learned changed agriculture, land management and more. We need to remember those lessons as we draw down the Ogallala aquifer.

References

Climate at a Glance: Regional and State Time Series, published June 2023, retrieved on July 10, 2023.

Egan, Timothy (2006). The Worst Hard Time: The Untold Story of Those Who Survived the Great American Dust Bowl. Houghton Mifflin Harcourt. ISBN 978-0-618-77347-3.

Nielsen-Gammon JS, Holman A, Buley S, Jorgensen J, Escobedo C, Ott J, Dedrick J, Van Fleet A (2021) Assessment of Historic and Future Trends of Extreme Weather in Texas, 1900- 2036: 2021 Update. Document OSC-202101, Office of the State Climatologist, Texas A&M University, College Station, 44 pp.

Taylor, O.R., et. al., (in prep) Geographic and temporal variation in monarch butterfly migration success.

Figure 2. Google Earth image of dust bowl region. There are no detailed weather records for this region in the 1930s – not even for cities like Pueblo and Amarillo. As far as state records are concerned, those of Kansas are most appropriate even though about half the state is to the east of the dust bowl area.

Figure 3. Part of Haskell County, KS showing center pivot irrigation.

Figure 4. Region from which most monarchs originate that reach the overwintering sites in Mexico.

Monarchs: Reaching 50N and beyond

This text is all about the journey north by monarchs to reach 50N and beyond in Canada. And it’s based on first sightings data posted to Journey North. In this short piece, I will discuss factors associated with the number of monarch first sightings reported in Manitoba and Saskatchewan from 2014-2023. In theory, the numbers arriving in these provinces are a function of the temperature, timing and the size of the first-generation population of butterflies moving north in May and early June.

First generation monarchs originating from the South Region in the United States are on the move to the north from the end of April until about the 12th of June, the approximate date at which directional flight ceases at 50N (Taylor, 2022). Thus, arrival at 50N needs to occur before this date. Since monarchs, even under the most favorable conditions, are unlikely to reach 50N before the 15th of May, and most often later, the window of opportunity to reach the prairie provinces is close to three weeks*. Because the window is narrow, the weather during this interval is critical and the temperatures and wind conditions have to favor directional (migratory) flight. Interestingly, the window can narrow if cold weather in late May and early June inhibits directional flight and the arrival of monarchs, but it can’t be lengthened beyond 12 June. Narrowing the window of arrival is something that occurs at all northern latitudes from time to time – with earlier end dates at lower latitudes (Taylor, 2022).

Support for this interpretation comes from an answer to a simple question. Why did Fred Urquhart say in his book “The Monarch Butterfly”: “In the summer of 1958 the numbers had been so reduced in some areas that the Monarch was considered rare.” The answer involves low temperatures throughout most of the growing season. The mean March and April temperatures in Texas and the April temperature Oklahoma were too low for normal colonization and population development (49.3F, 62.6F and 56.8F respectively). These conditions delayed the development of the population for weeks, and probably resulted in a small first generation that moved north in May and early June. The first generation advanced into areas where May and early June temperatures were below long-term averages. For example, the mean temperatures for May and June in Buffalo, NY were 53.9F (-4.7F) and 61.9F (-5.7F). These temperatures may have prevented recolonization in some areas and delayed it in others. The temperatures that followed the rest of the breeding season were also below average (-2.4F in the Upper Midwest). This sequence of conditions, none of which favored normal population growth, surely resulted in a low fall population that was deserving of Fred Urquhart’s assessment. The sequence of weather events in 1958 appear to be a case in which a delayed start to the population led to an outcome that carried over into the following year.

Because directional flight seems to decline when temperatures are less than 66F, I use that temperature to list the number of days less than 66F as a measure of the days with unfavorable conditions. This is a surrogate measure, an approximation, but it seems to be associated with negative and positive outcomes.

Our only proxy for the number of first-generation monarchs moving north from late April to mid-June is the number of first sightings recorded by Journey North. We have summarized these numbers by longitude and latitude into 10-day intervals from 1May-9June. This tabulation is also a proxy for the number of butterflies moving north during this period. These numbers also reflect the conditions that enable migratory flight and the opportunities for monarchs to be observed.

Table 1. First sightings from 1May to 21June reported to Journey North for Manitoba and Saskatchewan from 2014-2023 in relation to mean temperatures for May and June in Winnipeg, days <66F, and first sightings from 90-100W longitude north of 35N. The later represents the first sightings from 1May to 9June. Green=good, yellow=neutral and pink=bad conditions.

There are several patterns in the data summarized in Table 1. First, the temperatures in Winnipeg appear to be getting warmer from 2014 to 2023, especially in June. Second, along with these increases, the numbers of first sightings in Manitoba and Saskatchewan have been increasing, particularly in years with higher May and June temperatures. Third, there appears to be an association between the number of first sightings in 90-100W (north of 35N) and the total sightings in the provinces. All years with less than 169 sightings had low numbers in the provinces. The number of days with temperatures greater than 66F may also be a factor.

However, there are inconsistencies in these patterns which is to say that the expectations set by some conditions were not met in some years. For example, based on the outcomes from 2019-2023, the number of monarchs sighted in the provinces in 2018 should have rivaled those of 2021 and 2023, particularly since the temperatures for both 2018 and 2023 were nearly identical. This result tells me that it’s likely that something occurred during or after 2018 that the metrics didn’t account for – such as a sharp increase in reporting that occurred after 2018. Statistics have not been applied here since there is no way to account for the increase in the number of citizens interested in monarchs since 2014.

While there appears to be an association between temperatures and the number of monarchs in the first generation that influences the colonization of the prairie provinces, more data is needed. Earlier records are not helpful. Although the Journey North first sightings record began in 2000, the numbers were not robust enough for this type of analysis until 2006. Further, the temperature records for Winnipeg only date back to 2014. So, to be sure of these relationships, we will have to track these numbers into the future.

Still, this is an interesting dynamic, and going forward, as the temperatures increase, it is reasonable to expect that the numbers of monarchs arriving in the provinces in late May and June will increase. Actually, that applies to most of Canada. Whether that will lead to good or bad outcomes is an open question. Much will depend on the abundance of milkweeds**, if increasing temperatures delay the departure or progress of migrating monarchs and whether the increasing distance itself takes a toll on the migrants.

*The window for first generation monarchs to reach the western provinces in Canada appears to be just over three weeks. Of the 24 first-of-the-year sightings for Manitoba, 21 (87%) occurred between 20 May and 13 June (25days). Fifty percent of these first sightings occurred from 26May to 4June. The earliest first sighting occurred on the 12th of May 2012. The arrival window is as long as five or six weeks at the most southerly latitudes.

**The prairie provinces are intensely farmed as seen in the screen shot of an area east of Winnipeg below. The distribution and abundance of milkweeds in these landscapes will set the upper limit for the size of the fall population.

Reference
Chip Taylor, 2022, Monarch Puzzle Wrap Up, Monarch Watch Blog.

First sightings in the greater Winnipeg area in 2023.

First sightings in Manitoba and Saskatchewan in 2023.

Screen shot of land use east of Winnipeg

The decline and recovery in western monarch numbers from 2020-2022

In a previous post to this Blog (The Western monarch puzzle: the decline and increase in monarch numbers), I presented a long explanation for why the number of monarchs counted at the overwintering sites in California numbered a mere 1849 in 2000 yet gave rise to an overwintering population of 247K in 2021. I offered two related explanations for this outcome. First, that the extremely hot conditions in September – November 2000 along the California coast prevented monarchs from reaching the typical overwintering locations or staying non-reproductive if they did so. Rather, I suggested that many of the monarchs in that fall migration overwintered in small, scattered clusters at elevations of 1500-3000’ in the low ranging hills inland from the coast. Second, that a sufficient number of the overwintering monarchs, and perhaps some breeding butterflies from the coast, produced a large first generation in the spring of 2021 that dispersed to colonize the inner-mountain west to the east of California and to a lesser extent the states to the Northwest. Thereafter, the 2-3 generations that followed culminated in a fall migratory population of 300k (or more) that resulted in a Thanksgiving count of over 247K monarchs.

My calculations suggested that such an increase would only be possible if the breeding population that gave rise to the first generation consisted of at least 10,000 females. While this interpretation seemed reasonable, it lacked support until now. Recently, when going through iNaturalist records to answer a particular question, I decided to see what those records said about what happened during the last few breeding seasons in California – BINGO! The iNaturalist records are consistent with the expectation that the population in the fall of 2020 was much larger than the 1849 tallied during the Thanksgiving counts that year. Further, they are consistent with the hypothesis that there was a robust breeding population in the spring of 2021. In fact, the numbers in 2021 (Tables 1,2) are similar to those of 2022 that resulted in an even larger migratory and overwintering population. These records also suggest that the number of monarchs overwintering along the coast in the fall of 2023 are likely to be lower (Table 2).

In the records cited in Table 1, the March-15June records represent the beginning of the spring breeding season which effectively stops with the end of directional flight (migration) by first generation monarchs in mid-June. The total for all records ends in mid-September. This interval covers all breeding and the first 3 weeks of the migration but stops before migratory monarchs usually reach the coastal areas.

The number of iNaturalist records for the Central Valley in 2022 contrasts strongly with other years as can be seen in the maps below. The early movement into the Central Valley in that year was evidently due to high temperatures in March. The deviations from the long-term means in March were as follows: 2020 -1.2F; 2021 -1.6F; 2022 +2.3F; 2023 -4.5F. These early movements into the Central Valley and the foothills to the east appear to account, in part, for the large Thanksgiving count in 2022 – the largest count since 2000.

Table 1. iNaturalist records for the spring and summer breeding seasons for monarchs in California (2017-2023). The low percentage for March-June for 2019 may have been due to the low temperatures in February. The mean temperatures that month along the coast (45-46F) were the coldest since 1966. Higher percentages, as in 2018 and 2020, may have been due to a combination of more favorable temperatures. The records for early season and full season intervals from 2020-2022 are quite similar given that community science records can be influenced by weather and other factors. The counts for 2020 indicate that the migratory population that year was quite substantial.

Table 2. iNaturalist records for March-June 2020-2022. These records suggest that the populations developed in a similar manner from month to month in 2020-2022. The population build up started more slowly in 2023 and the lower number of records for 2023 thus far indicate that overwintering numbers are likely to be lower this coming winter season.

Reference
Taylor, Chip, 2023. The Western monarch puzzle: the decline and increase in monarch numbers. Monarch Watch Blog, monarchwatch.org/blog, May 2023.

iNaturalist maps for monarch records for the years and intervals indicated.

The pending decision: Will monarchs be designated as threatened or endangered?

Introduction

As many of you know the Fish and Wildlife Service (FWS), a division of the Department of the Interior, has been mandated to make a determination as to whether the monarch butterfly should be warranted for protection under provisions of the Endangered Species Act (ESA). The choices seem to be either not warranted or warranted and either threatened or endangered. In a previous ruling in 2020, protection for the monarch was determined to be warranted but precluded on the basis that other species were more deserving of protection at that time. Included in that ruling was a provision to monitor the monarch numbers yearly and to reevaluate the status of the monarch in three years (Taylor, 2020). That time has passed, and the FWS is expected to issue an updated Species Status Assessment (SSA) for monarchs this month.

Because the previous decision was “warranted but precluded” it is probable that “not warranted” is off the table and that the decision will favor either a threatened or endangered status. Realistically, rather than endangered, monarchs are likely to receive a “threatened 4d” designation meaning that monarchs will receive some of the protections and support authorized under the ESA but will not attain the level of protections mandated by endangered status. The threatened 4d status allows for exclusions or exceptions and the overall impact of a 4d determination will be defined by those exclusions (fws.gov/sites/default/files/documents/section-4d-rules_0.pdf). The issuance of the SSA will be followed by a public comment period. Following that, a final determination will be made in November, although there is some possibility the announcement could be pushed into 2024.

This pending designation brings forth a number of questions. First, is a threatened or endangered status merited based on the size of the monarch population, its geographic range and the habitat data on hand? In other words, do we have the evidence required to justify such a determination? We can also ask how either designation would benefit monarch conservation. Again, how could our actions, with respect to sustaining and restoring habitat, serve to maintain the monarch migration? That brings us to a matter of scale and investment. How many plants and how many acres would we need to restore each year? It is also fair to ask whether limiting “take” or contact of any kind by citizens with monarchs would be beneficial**.

Endangered vs Threatened

Endangered status is conferred to species that are in danger of extinction in all or in a significant portion of their historical range. Most species receiving this designation are uncommon to rare and have lost a portion of their range or have lost a significant amount of the habitat or resource base that is needed to sustain the population. Rarity can also be due to disease or forms of displacement by introduced species.

The threatened designation is used for species that appear to be likely to become endangered in the future due to declining numbers, a continuous degradation of critical resources in its preferred range or the growing threat of diseases or the introduction of superior competitors. Climate change is another consideration in some cases.

Interestingly, some species designated as endangered have recovered sufficiently to no longer be consider endangered, e.g., giant panda, gray wolf and others. Species once considered to be threatened could also be removed from that status if the original assessment proves to have been overly cautious.

Extinction

As applied in this case, extinction refers to the loss of the monarch migration and not the species per se. Given the link between the increase in greenhouse gas emissions and increasing temperatures and the world’s slow response to these changes, yes, the monarch migration will eventually be lost. The question is when. Is extinction imminent, or likely in the near future? Do we know enough about how monarchs respond to weather variability, and can we predict the course of the changes in climate accurately enough to forecast the demise of the monarch migration? No, we don’t and can’t. Rather, in this instance, due to our lack of certainty, we are applying the “precautionary principal” writ large. Fair enough. But what are the consequences and down-to-earth realities of designating the monarch threatened or endangered?

The Decline

Monarch numbers have declined since the early 1990s. The high population numbers from 1994-1996 are taken as a baseline when the numbers were probably much lower many times in the past. There are no data supporting a supposition that these 1994-1996 populations were “average”. They may well have been the exception. Aside from declines due to specific weather events, e.g., the late spring freeze in 1997 and the drought of 2000, etc., there is ample evidence that the decline was due to the adoption of herbicide tolerant (HT) crop lines (1998 to 2006) (Pleasants, 2017) and the renewable fuel standard (RFS) from 2007-2011 (Lark, et al 2015). In both cases, millions of acres with milkweed that supported monarchs were eliminated from the landscape (Pleasants and Oberhauser, 2012, Pleasants, 2017). There seems to be an assumption that the decline has continued following the end of the surge in corn growing that was spurred by the adoption of the RFS. Perhaps it has, but if so, such effects are too small to be detected given the variability in the annual cycle and the measurements of the colonies in Mexico. Rather, as shown by Meehan and Crossley (2023), there is reason to believe that the monarch population is relatively stable. Further, there is no indication that the population is continuing to decline. One can argue that the population can be expected to continue to fluctuate with overwintering counts varying from .7 to 6hectares within the present climate and amount of available habitat. Under these conditions, running 5-year averages can be expected to vary from 2.5-3.5hectares. Obviously, we need to restore more habitat to improve those numbers.

Status

So, what is the status of the monarch population? How many were there in the Eastern and Western overwintering populations last winter? We also need to know how the populations respond to negative conditions and whether are they capable of rebounding from low numbers?

Abundance

Although the winter count in Mexico of 2.21hectares was 22% lower than the previous year due to weather conditions in the summer breeding range and high temperatures during the migration, the number of butterflies was still substantial. Since there are an estimated 21.1million monarchs per hectare, it follows that roughly 46.6million monarchs overwintered in Mexico this past winter. There have been 6 years in the last 29 during which the overwintering numbers were lower.

In California, the Thanksgiving counts, coordinated by the Xerces Society, indicated that at least 335 thousand monarchs overwintered at roost sites along the coast from San Diego in the south to Sonoma County in the north. This number was the highest recorded since 2000 (Taylor, 2023C).

Resilience

Have both populations experience lower numbers in the past? The answer is yes.

In the East, the lowest measured number occurred in 2013. During that year, the population measured 0.67hectares. The population increased to 1.13hectares in 2014 and to 4.01hectares in 2015 (Taylor, 2021). The number in 2015 was virtually identical to the number in 2010 (4.02), the year prior to the three-year crash that started with the 7month drought in Texas in 2011. In 2018, five years after the crash, the population measured 6.05hectares, the highest number since 2006.

The overwintering numbers in the West reached an all-time low along the California coast of 1849 monarchs in 2020. Yet, somehow, the population rebounded to 246,253 in 2021, the 8th highest in the last 25 years. The counts increased again in 2022 to 335,479 (Taylor, 2023C).

The recoveries from low numbers in both cases demonstrate the remarkable resilience of this species. Have monarchs recovered from low numbers such as these in the past? Yes, very likely.
The recent collapses are related to weather events, sometimes a series of negative conditions, that reduce reproductive, migratory or overwintering success either separately or in combination (Taylor, et. al. 2020, Taylor, 2023A, B). Using the association of negative conditions to the declines in the last 29 years to the probable success of monarch populations in the past reveals that there have been many episodes in the record during which the populations declined to low numbers. The conditions during the droughts and high temperatures of the “dirty thirties” were much more extreme than any encountered by monarchs in the last 29 years*.

The Choices

Should monarchs be classified as not warranted or warranted but threatened or endangered? The answer depends on whether we are talking near term, that is, the next few decades or look to the future such as 2040 and beyond. In the near term, the population is robust, though not as abundant as in the early 1990s, has maintained the same geographic range and shown itself to be resilient under the current habitat limitations and variation in weather. It is also likely that while monarchs thrived in the early 1990s, and did so from 1975 to 1996, the least variable climate interval from 1885 to the present, their numbers were as low or lower than in recent years many times in the past*. With that perspective, the designation should be “not warranted.”

The long view is different. In time, it will be clear that monarchs are threatened and then endangered. As we all know, the climate is changing, increasing temperatures, along with droughts, both of which are projected to increase in Texas in the coming decades. These changes will progressively limit monarch reproductive and migratory success and, in time, will reduce the ability to overwinter as well. But monarchs will not be alone in suffering the consequences of these changing conditions. Virtually all other species in the United States will be impacted by these changes and many, many other species will be threatened and then endangered. It’s fair to wonder when we will be overwhelmed with trying to protect species from becoming extinct. There is a reality ahead of us that we can’t ignore. We need to be mindful of the rates of change of a multitude of drivers, and we need plans, strategies and resources to cope with these changes. Adaptations will be costly and resources will be limited which will lead to cost/benefit assessments, priorities and triage. At lot of species will be left behind. That already seems to be happening.

So, what is the time frame for the envisioned changes? Do we have a decade, two, maybe three before the monarch migration is lost? We don’t know, but it will happen given the increases in greenhouse gas emissions. So, what should we do? Clearly, we need to maintain and increase monarch habitat, especially in the Upper Midwest, the source area of >80% of the monarchs that overwinter in Mexico (Taylor, et. al., in prep). As to maintaining habitat, we need to know how much habitat is being lost each year in the areas that produce the greatest number of monarchs. And we have to offset those losses with restoration. There are many restoration efforts underway now, but it is not at all clear that restoration is keeping pace with annual rate of habitat loss. Beyond annual losses, the goal is to replace the losses incurred as the result of the adoption of herbicide tolerant crop lines and implementation of the renewable fuel standard. The calculation by experts has been that we need to establish 1.8 billion milkweed stems to attain a monarch population that could sustain a mean of 6hectares of occupied trees at the overwintering sites in Mexico (Thogmartin, et al 2017). There are several difficulties with this projection. First, there is relatively little federal land in the Upper Midwest meaning that the restorations would have to involve marginal lands of tens of thousands of private landowners. That would be a major undertaking. Secondly, that would involve funding – massive funding – since it takes about $2 to establish each new stem. Third, there is neither the seed capacity for wide spread establishment via seeds nor do we have the nursery capacity to produce a sufficient number of plugs (small plants) to meet the restoration targets. And to repeat, those targets are extreme. To reach the 1.8 billion stem goal, we would need to establish at least 100 million stems a year for a decade or more. A hundred million is beyond our capacity to establish stems with seeds or plugs by at least a factor of 10.

Given our limitations, what should our goals be? First, we have to keep pace with the current rate of habitat loss – which is unknown at present – but which could be as much as two million acres per year. Secondly, we will need to do better than to keep pace with habitat losses since there will be losses due to the progressive increases in March temperatures in Texas and the September temperatures in the Upper Midwest and Northeast. Both of these conditions will result in lower numbers of monarchs reaching the overwintering sites (Taylor, et. al., 2020, Taylor, 2023A,B, Culbertson et. al. 2021). That means, that in addition to knowing the mean rates of annual habitat loss, we need to develop a deep understanding of the role of increasing temperatures and droughts on monarch numbers.

What will follow the decision?

Will declaring monarchs threatened or endangered enable us to reach these goals? Will either declaration, along with its associated provisions, lead to more pesticide regulations? If so, would that involve both EPA and the Department of Agriculture? Would such regulations be opposed by organizations that represent farming and ranching interests? Could this become a political issue? These are important questions. The decision by FWS is supposed to be based on science alone, but regulations could trigger a number of issues and conflicts that are independent of monarch biology.

Will the people of the United States be forbidden from all contact with monarchs as is now the case in California? If so, will separating monarchs and the people who care about them increase or decrease the incentives to sustain monarch numbers? Negatives seldom motivate people to do the right thing. Positive incentives that encourage the public to become part of a solution are more effective, but will such incentives be possible if monarchs are declared threatened?

Will a declaration of threatened 4d constitute a threat in itself to land owners who currently have milkweeds on their lands? This possibility seems real since many land owners fear regulations, and along the lines of the “shoot, shovel and shut up” practice that is spoken of in connection with the endangered species act, they might simply eliminate milkweeds from their lands.

More questions involve the possible benefits of a threatened or endangered status. How would either benefit monarchs? Would more money be available for restoration? If so, would the amount be a token relative to the need, and would such funds conflict with the need to protect other species? An initial proposal to sustain and restore monarch habitats could garner support, but the dollar amount would have to be in the 100s of million to mitigate the annual and past habitat losses, and that mitigation would have to be continuous.

While it seems wise to adopt the precautionary principal in the face of threats, we also need to be mindful that all actions along these lines have consequences and it is best “to do no harm”. It’s a delicate balance.

There are many more questions. This is just a start. But here is one more. If the monarch migration will be lost eventually, why make great efforts to sustain it? Faith. We have to have faith that the world will come to its senses and work collaboratively toward the reduction of greenhouse gases to save the natural systems that sustain us. There is hope. The rate of increase in CO2ppm has declined in recent years.

*The data supporting this statement will be posted along with a text describing the known responses of monarchs to temperature extremes.

**The term “take” refers to the impact of human actions on a candidate species and can be intentional or unintentional. Intentional “take” involves any action that would bring harm such as harassing, hunting, collecting, etc. Unintentional “take” would involve incidental harm that occurs as the result of normal activities. For example, you are not allowed to harm a threatened or endangered species, but you would not be at fault if you hit and killed one with your car. Of interest in this case is whether rearing of any sort would be considered “take” and therefore prohibited.

Acknowledgements

Once again, I’m thankful to Janis Lentz for correcting the many things I missed in this draft and to Jim Lovett for assisting with the formatting and posting it to the Blog.

References

Culbertson, K. A., Garland, M. S., Walton, R. K., Zemaitis, L., & Pocius, V. M. (2021). Long‐Term Monitoring Indicates Shifting Fall Migration Timing in Monarch Butterflies (Danaus plexippus). Global Change Biology. doi.org/10.1111/gcb.15957

Lark T.J., Salmon JM, Gibbs HK (2015) Cropland expansion outpaces agricultural and biofuel policies in the United States. Environ Res Lett 10:044003. doi.org/10.1088/1748-9326/10/4/044003

Meehan, T.D. & Crossley, M.S. (2023) Change in monarch winter abundance over the past decade: A Red List perspective. Insect Conservation and Diversity, 1–8. doi.org/10.1111/icad.12646

Pleasants, J.M., and Oberhauser, K.S. (2013). Milkweed loss in agricultural fields because of herbicide use: effect on the monarch butterfly population. Insect Conserv. Divers. 6, 135–144. doi.org/10.1111/j.1752-4598.2012.00196.x

Pleasants, J. (2017) Milkweed restoration in the Midwest for monarch butterfly recovery: estimates of milkweeds lost, milkweeds remaining and milkweeds that must be added to increase the monarch population. Insect Conserv Divers 10 :42–53. doi.org/10.1111/icad.12198

Taylor O.R. Jr, Pleasants J.M., Grundel R., Pecoraro S.D., Lovett J.P. and Ryan A. (2020) Evaluating the Migration Mortality Hypothesis Using Monarch Tagging Data. Front Ecol Evol 8:264. doi.org/10.3389/fevo.2020.00264

Taylor, O.R. 2020. ESA listing decision for the monarch.
monarchwatch.org/blog/2020/12/15/esa-listing-decision-for-the-monarch

Taylor, O.R. 2021. Monarch population crash in 2013
monarchwatch.org/blog/2021/06/11/monarch-population-crash-in-2013

Taylor, O.R. 2023A. Monarch numbers: dynamics of population establishment each spring. monarchwatch.org/blog/2023/03/27/monarch-numbers-dynamics-of-population-establishment-each-spring

Taylor, O.R. 2023B. Monarch numbers: trends due to weather and climate. monarchwatch.org/blog/2023/03/27/monarch-numbers-trends-due-to-weather-and-climate

Taylor, O.R. 2023C. The Western monarch puzzle: the decline and increase in monarch numbers. monarchwatch.org/blog/2023/05/29/the-western-monarch-puzzle-the-decline-and-increase-in-monarch-numbers

Taylor O.R. Jr, Pleasants J.M., Grundel R., Pecoraro S.D., Lovett J.P., Ryan A., and C. Stenoien. (In prep) Geographic and temporal variation in monarch butterfly migration success.

Thogmartin W.E., Diffendorfer J.E., López-Hoffman L., Oberhauser K., Pleasants J., Semmens B.X., Semmens D., Taylor O.R., Wiederholt R. Density estimates of monarch butterflies overwintering in central Mexico. PeerJ. 2017 Apr 26;5:e3221. doi.org/10.7717/peerj.3221 PMID: 28462031; PMCID: PMC5408724.

Supplementary Materials

The Western monarch puzzle: the decline and increase in monarch numbers

Preamble

The first draft of the text that follows was written without consulting the literature. I wanted to work out my interpretations without the weight of the points of view of others. While I drew on my memory of the literature for the portions of the narrative dealing with demography, the behavioral interpretations were based on my experience. The idea was to add references later, and I have done so, but the list of relevant texts is surely incomplete. In my search of the literature, I came across a paper published in 2018 (Fisher, A., et al. Climatic Niche Model for Overwintering Monarch Butterflies in a Topographically Complex Region of California) that I should have remembered. It’s likely that prior knowledge of that paper would have changed my approach to the subject. As it stands, some of what follows validates the niche model outlined in that paper and it could be said, that the model, at least the portion that deals with overwintering, supports my interpretations. Just as there are no quantitative data to assess the validity of the model at this time, there is no way to validate my interpretation of what happened in the fall of 2020 and the 2021 growing season. Both are based on known properties of overwintering sites and conditions and known behavior of monarchs and both appear to be reasonable based on our current state of knowledge.

Introduction

The Western monarch population declined sharply in 2017 followed by further declines in 2018 and 2019. Even more alarming, and yet perplexing, were the counts of 2020 and 2021. The Thanksgiving counts in 2020 yielded only 1849 monarchs at California overwintering sites from San Diego County in the south to Sonoma County in the north (Western Monarch Thanksgiving Count Data, 1997-2021). Of the 249 sites examined, the vast majority had zero to less than 20 monarchs, and there were only 5 sites with more than 50 (westernmonarchcount.org). These numbers were taken as a clear sign that the population was below the theoretical extinction threshold and that recovery from such a low number was problematic (Pelton, et al., 2020, Semmens, et al., 2016). Yet, the Thanksgiving count in 2021 yielded 246,253 monarchs at California sites – an over 100-fold increase – and the 8th highest count in the records since 1997 (Western Monarch Thanksgiving Count Data, 1997-2021). There have been many attempts to explain this rate of increase since it’s impossible for a cohort of 1849 overwintering monarchs to initiate a cascade of reproduction over 3-4 generations that would result in a large fall migratory population. This result is improbable since the number of surviving females would have been less than 600, and the known rates of mortality for all life stages indicate that rates of increase are strongly constrained. Clearly, the population growth in 2021 was initiated by thousands of females, perhaps 10s of thousands. The need to understand the increase leads to questions about where those females came from.

In the text below, I will make the case that, due to the all-time high temperatures along the coast in September and October in 2020, large numbers of non-reproductive monarchs overwintered inland from the coast in small scattered clusters and that these monarchs survived to mate in late February and March. The high temperatures during September-October may also have resulted in many butterflies terminating reproductive dormancy both before and after reaching coastal areas. Some likely dropped out of the migration and failed to reproduce in areas with few or no milkweeds. Others, especially in urban areas where milkweeds were available, continued breeding through the winter months (James 2021, James et al. 2021). The overwintered females, perhaps joined by some breeding monarchs from urban areas, mated and moved inland. These females laid a sufficient number of eggs to produce a large first generation that then migrated eastward to the Sierras and the inner mountain West where, in the course of 3, and in some cases, 4 generations, a migratory population of more than 300K monarchs progressed toward the California coast in September and October. Unlike in 2020, these monarchs encountered October temperatures that were slightly below or close to the long-term average all along the California coast leading to a resurgence in numbers at the overwintering sites.

This interpretation is based on the behavior of diapausing monarchs that seek moderate (60-75F) to cool (40s-60F) temperatures to maintain a non-reproductive condition through the migration and the winter months. Additional support for this hypothesis comes from the decline in monarch numbers since 1998 at overwintering sites in the three southern-most California counties. These declines appear to be related to both increases in temperature from September through December and a general decline in monarch numbers (see Addendum). The estimations of the number of reproducing females required to create a large first generation that migrated to breeding areas are based on data and estimations of mean number of eggs per female and the probabilities of surviving from one generation to the next.

Interpretation

In the narrative that follows, I will outline an explanation for why the overwintering Western monarch population appeared to be at an all-time low in 2020 yet increased in a spectacular manner in 2021. I will explain why the low numbers at overwintering sites in 2020 and the relatively high numbers in 2021 are rooted in the monarchs’ response to temperature that is based on their need to remain non-reproductive during the winter months. In effect, I will set forth a hypothesis based on the response of adult monarchs to temperature. Support for the hypothesis is based on observations of monarch behavior in Kansas, Texas and Mexico and the decline in the numbers of sites and monarchs as temperatures increased over the last decades in San Diego, Orange and Los Angeles counties.

The scenario I envision was foretold by Fisher, et al., 2018 in their niche model paper that predicts the tendency for wintering monarchs to move away from the coast to winter in cooler locations in the coastal foothills if coastal temperatures continued to increase. As they pointed out, “Our results suggest that estimating the size of the western overwintering population in the future will be problematic, unless annual counts compensate for a shift in the distribution and a potential change in the number and location of occupied sites.”

Let’s start with what we think we know about the influence of extremely high temperatures on monarch migrations and on monarch physiology and behavior. Three things are likely to happen when the temperatures are extremely high (>90F) during the migration. First, the start of a migration can be delayed. Late recolonizations in May and June can result in later migrations that can be further delayed due to high temperatures that stop the migration or reduce the number of flight hours per day (Taylor, et al., 2019). Second, some of the monarchs will become reproductive since high temperatures increase the production of juvenile hormone and ovariole development. Third, monarchs become highly dispersed and do not form high density clusters under high temperatures. The latter point is supported by observations of clustering behavior when temperatures are in the 90s in Kansas and Texas and the numerous observations of temporary clustering sites that form from mid-October into early November in many areas along the California coast. Next, we need to know how monarchs respond to temperatures while at overwintering sites. Surviving from late October to reproduce in late winter or early spring requires that monarchs remain non-reproductive – a state of low endocrine production, low metabolism and low energy expenditure. To stay non-reproductive, monarchs appear to seek low temperatures that help maintain this condition.

Based on these considerations and data on the conditions that favor population growth, I have been trying to develop a stage specific model for the Western monarch population. The premise of the model is that an analysis of the effects of both physical and biotic factors within a stage allows us to predict the numbers entering the next stage. By tracking the results through all stages, it should be possible to a predict the relative size of an overwintering population. There are five stages in the West: overwintering, reproduction by overwintered adults, migration by first generation monarchs to the inner mountain West and Northwest, summer reproduction and fall migration. The temporal sequence of those stages is reasonably clear and is similar to the same stages in the East. What is less clear is where the monarchs breed in the summer months that return to the coastal areas each fall. Sometimes, the main production areas seem to be in the NW, and in other years (such as 2021) the conditions in the inner mountain west to the east of California seem to be the more favorable for monarch production. As part of the model, I’ve developed a large matrix of the monthly and stage specific time periods for the years 1998-2021. The matrix includes the temperatures and rainfall amounts for each interval. The idea was to use these data to determine if the physical factors could account for the decreases and increases in monarch numbers. The one and only monarch data base in the mix was the yearly monarch counts – the Thanksgiving counts coordinated by the Xerces Society. The underlying assumption has been that these totals (or the totals each year from the 20 sites measured most consistently) are reasonable representations of the population and therefore serve as a measure of the size of the monarch population. Indeed, comparing stage specific temperatures, and sometimes rainfall, seemed to work, and there is a reasonable but not entirely satisfactory explanation for the crash in the population from 2016-2019 based on these values. Yet, there were inconsistencies. There were years in which the Thanksgiving counts were lower (2001) and others when they were higher (2014) than expected based on the temperatures and rainfall amounts from one stage to the next. Overall, the explanatory power of the assembled data, once promising, was rather weak. Something was missing.

I was stuck but still hopeful. There was a mismatch between the metrics and the overwintering numbers. I was either using the wrong metrics or the numbers weren’t adequately representing the size of the population and both could be off. The metrics were certainly off being too course to reveal the details of population growth. But what about the Thanksgiving counts?

One approach to getting unstuck is to write down all your assumptions and to detail all you know about each assumption. In the case of the Thanksgiving counts, there are several assumptions. First, all the teams making the counts are equally skilled at “estimating” the number of monarchs in each cluster they examine. Second, all, or nearly all, roosting sites are located each year and that all clusters at each site have been identified and counted. Third, all, or nearly all, the migratory monarchs cluster at the coastal sites. Indeed, Schultz et al., 2017, assume that the monitoring constituted “an independent index of the total abundance in that year.” Human error is expected since making the counts is extremely difficult, demanding and tiring. Kudos to all those who have engaged in these counts and doing their best to be accurate. Given the conditions, the number of sites, the coordination required and more, these counts are a significant achievement. So, as counts go, due to the difficulties, we can expect some over-counts and under-counts at various sites. It follows that another assumption might be that these counting errors cancel each other out with the result that the counts are reasonable estimates of monarch numbers at the end of each season. Ok, if we accept that premise, we can ignore the third assumption – that all the monarchs cluster at the overwintering sites. The numbers don’t lie, except when they don’t represent what we think they do. It’s reasonable to expect that the counts are accurate measures of the population. After all, the monarchs in Mexico represent all areas east of the divide, with the exception of peninsular Florida, and are consolidated into a small number of colonies each year. So, why shouldn’t that be true in the West? In Mexico, the colonies occur on massifs at 10,500’ or higher where the temperatures for most of the winter range from the 30s at night to the 60s during the day with the temperature rarely reaching the low 70s at the sites themselves. In effect, these are low temperature islands that are surrounded by areas with much higher temperatures at the lower elevations. Thus, in the winter months, to stay non-reproductive, the monarchs need to remain on these islands. Freezing conditions are more likely to be a threat to monarchs in Mexico than high temperatures. That said, warmer winters are expected in Mexico in the future.

Monarch behavior during the winter months along the California Coast is different. Rather than staying in one place, monarchs are on the move when temperatures allow movement. Initially, monarchs reaching the coastal areas settle in small to modest numbers in a large number of scattered sites as well as the well-known wintering sites. Most of these temporary sites are abandoned with the occupants moving to more permanent sites as the temperatures continue to cool in late October and November. The presumption here is that monarchs leaving temporary sites seek clustering locations that are both cooler and better protected from variable weather conditions. These movements generally end by late November. The clusters/colonies are thought to be at their seasonal maximum at that time hence the timing of the Thanksgiving counts. However, the numbers generally begin to dwindle following (but perhaps even before) the Thanksgiving counts such that counts in the following January indicate numbers that are lower by roughly 35 to 45 %. Further, a large proportion of the sites that had monarchs during the first count no longer have monarchs during the second count. In addition, there are locations where the numbers increase from Thanksgiving to January and sites where the declines over that 5-6week interval are quite low. The reasons for the declines are not clear. Mortality could be one factor, and it could be that monarchs are still moving in an attempt to stay in reproductive diapause by seeking cooler and more protected sites. The assumption by Schultz et al., that counts are a useful index of abundance is difficult to justify. Due to the continuous movement of monarchs in the early fall and whenever the temperatures allow through the winter season, the numbers appear to be a moving target. While there is no doubt these counts are of value, the fact that a hundred and fifty more sites are counted now than as recently as 2011 makes it difficult to assess current and historical trends and to associate those numbers with changing physical and biological conditions.

While temperatures along the coast have historically been cooler than the interior as the monarchs arrive in mid-October, as the season progresses, and the interior cools, it can actually become cooler in many locations away from the coasts (Table 1). That the temperatures along the coast have been increasing during the winter months and that monarchs have been leaving the coastal sites earlier over the last 20 years or more seems clear (Addendum, Figure 1). Monarchs which used to stay clustered at some sites until early March are now leaving sites in large numbers in late January and early February according to many accounts although actual data on this point are scarce. Unfortunately, we don’t know the fate of monarchs that leave overwintering sites. There seem to be three options – they die, they become reproductive, mate and search for milkweeds that are scarce to non-existent at that time, or they move inland to sites which are cooler and more suitable for maintaining a non-reproductive condition. It’s probable that all three happened with the latter being the most frequent outcome for monarchs seeking cooler temperatures. I’m assuming that some of these monarchs survive from January through mid-March singly or in small scattered clusters well inland from the coast, perhaps at 1500-3000ft.

Table 1. Mean temperatures for months indicated for Tempest weather stations close to overwintering sites, nearest airports and counties for 2021 and county temperatures for 2020. The latter show the extreme temperatures for September and October in 2020. Temperatures are generally lower near overwintering sites for September and October but are warmer than county means during November and December (except for November Santa Barbara). Overwintering site temperatures are closest to airport temperatures in December. Counties generally represent large areas inland from the coast with lower temperatures than coastal areas from late fall through February.

Monarchs seek moderate temperatures. They need to stay non-reproductive for months. So, what happens if the coastal areas are too warm in mid-October? What if there are no temperature gradients, they are too weak or it is uniformly hot? It’s clear that there is an optimal temperature range for staying non-reproductive, and though we don’t know exactly what those limits are, monarch behavior appears to be telling us that they are seeking temperatures below 70F when migrating and need temperatures in the high 50s to 70s to reach and settle in to the overwintering sites to stay in those sites through the winter requires temperatures in the 40s to low 50s. There is a well understood relationship between temperature and the density of clusters that form during migrations and colony establishment. This relationship is key to what happens in California, and particularly what happened in 2020. First, two observations from Kansas are relevant. If the temperatures are in the 90s during the migration in eastern Kansas, most of the monarchs stop migrating and seek the shade and higher humidity found in the gallery forests along the rivers, streams and drainage areas. Once in these shaded and more humid sites, the monarchs roost in the trees often singly, but frequently in small and open, even scattered, clusters of 6 or so. Monarchs will stay in these locations for days if the temperatures remain in the 90s. The second observation involves clustering at the end of the day. When temperatures are in the low 80s, monarchs form clusters scattered loosely in trees from ground level to much higher. If the temperatures are in the 60s, with overnight lows in the 30s expected, the clusters are higher and denser in the trees with few or none of the monarchs settling near ground level. Similarly, in Mexico, the first clusters are loose and scattered and slowly consolidate as the temperatures become cooler. In effect, the colonies become smaller and smaller as the winter progresses usually reaching the maximum density and smallest area occupied by mid-December. That’s when the colonies are measured by WWFMX in collaboration with CONANP, UNAM and the MBBR. These observations indicate that cluster formation and density is a function of behavioral responses to temperature with loose or no clusters forming when it is hot to tight, cohesive clusters forming when temperatures are at the lowest that allow for flight.

These behaviors bring me to what happened in 2020 along the California coast. The temperature records for all the counties along the coast go back to 1895. That’s 127 years, and those years are ranked from coolest to hottest. I first looked at the mean temperatures for October for counties from San Diego to Marin. All averaged over 7F above the long-term means (7.0-7.6) and all but one (SB ranked 126) were ranked as the hottest in the record to date. The rankings were similar, but the deviations were lower when using 30yr means (5F-6.1F). I then looked at the mean temperature for September, a period during which the monarchs are moving toward the coasts. The results were the same, and when I combined September and October and then, September, October and November and finally added in December, each showed that those means ranked 125-127 (Tables 2a, 2b).

Table 2a. Temperature rankings for September through December separately and combined for 2020. There are 127 years in the record ranked from coldest to hottest with 127s being the hottest mean monthly temperatures in the record.

Table 2b. Deviations from long-term mean temperatures for the intervals indicated.

Given these extremes and the known effects high temperatures have on the initiation and progress of migrations, as well as the physiological and behavioral responses of monarchs to these conditions, what happened in 2020 that rolled over to the influx of monarchs in 2021? Here is the scenario that seems likely. The migration in 2020 started late and moved slowly due to the temperatures in the 90s along inland pathways leading to the coastal locations. For example, the mean maximum temperatures were the third highest in the 126year record for Fresno and the highest in the 75year record for Sacramento. Some monarchs became reproductive on the way to the coast, these monarchs dropped out of the migration altogether (James 2021, James et al. 2021). Other monarchs seeking cool temperatures stayed individually or in small clusters in the foothills and drainages inland from the coast and due to the continuing high temperatures stayed in these areas rather than moving to the coast. Some of the monarchs reaching the coast became reproductive and continued to reproduce through the winter where milkweeds were available. Others, a mere 1849, clustered at an all-time low number of sites along the coast. Santa Cruz, the county with the largest proportion (38.55%) of all the wintering monarchs in 2020, also had a high retention rate (91.2%) for the sites that were counted during the Thanksgiving and January surveys. The mean temperatures at Santa Cruz were the lowest among all counties in October, November, October-November and for September through December (Table 2). This result is consistent with the interpretation that monarchs seek the coolest locations or some combination of temperatures, protection from direct sun and strong winds.

The resurgence in the numbers at the overwintering sites in 2021 suggests that spring reproduction was extremely successful, so successful, that it seems likely that the resurgence was founded by a large number of mated females that began laying eggs in late February, a process that continued into April. Given the behavior I described with respect to clustering and the need for monarchs to seek out temperatures that enable them to stay non-reproductive, it seems likely that many of the overwintered females that began the breeding in 2021 overwintered singly or in small clusters at sites interior of the usual overwintering sites. Monarchs originating from winter breeding populations in urban areas may have also moved inland in the spring adding to the reproduction that produced a new generation from late April through May (James 2021, James et al. 2021).

Monarch demography

Part of the challenge in explaining what happened during the 2021 breeding season is not only figuring out the origins of females that started the first generation, but estimating how many survived to reproduce. That reproduction had to result in a large first generation that subsequently colonized the summer breeding areas, mostly east of California, followed by additional reproductive success that yielded an initial migratory population of at least 300,000 monarchs.

For the purpose of illustrating the potential reproduction of a cohort of females, I’ve assumed constant rates of mortality from one generation to the next. However, the incidence of parasitism by tachinid flies and infection by the protozoan Ophryocystis elektroscirrha (O.e.) tends to increase as the season progresses. Such increases would reduce both the proportion of larvae reaching the pupal stage and those surviving to the adult stage. In the case of the surviving adults with O. e., their fitness to reproduce would be compromised. We also need to recognize that realized fecundity, i.e., the mean eggs per female per generation, varies with temperature, nectar availability, and the distribution, abundance and quality of milkweeds. These factors also affect population growth. In other words, the calculations below represent a best-case scenario that is most often approached early in the season.

Calculations

The calculations that follow are based on estimates taken from the literature (Grant et al., 2020, Nail, et al., 2015, De Anda and Oberhauser, 2015, Oberhauser, et al., 2017). I used constant rates of egg laying and mortality for a cohort of 10,000 females for three generations. The results look promising but are not realistic.

10,000 females x 250 eggs = 2,500,000 eggs x 0.03 proportion surviving to pupal stage =75,000 x 0.76 = 57,000 new adults x 0.85 proportion that reproduces = 48,450 x 0.45 proportion of females = 21,803. These calculations therefore yield a 2.18 rate of increase in females per generation. If we multiply 21,803 x 2.18 for three generations, we get 225,884 females or a total potential migratory population of about 510,000. That’s more than enough monarchs to yield an overwintering count of 246K. However, it is unlikely that these rates of increase would be the same from one generation to another. They could be lower due to an increase in O.e. and tachinid parasitism or the losses due to other predators, all of which tend to increase as the season progresses. That could decrease the proportion surviving the pupal stage and the number of new adults substantially. As an alternative, we might imagine conditions that would allow females to lay more than 250 eggs followed by higher survival through the larvae stages that would lead to an increase in the population. The later scenario seems less realistic to me. It should be noted that the proportion of eggs (0.03) surviving to the pupal stage used in these calculations is probably too high by a factor of 2. If true, and all other estimates were the same as above, it would take about 20,000 females to produce a migratory population of 246K wintering monarchs. That number would have to be even greater if the incidence of O.e. and tachinid parasitism increased thus reducing the size of the reproductive population. The point of making these calculations is to show that to produce an overwintering population of 246K required an initial female population numbering in the 10s of thousands, probably 30,000 or more. While a portion of these females may have originated from scattered areas of winter reproduction, the majority were probably overwintered monarchs that survived in small, highly dispersed clusters in the coastal foothills (Fisher et al., 2018) that went undetected during the surveys for overwintering monarchs.

If these interpretations are correct, it is likely that the Thanksgiving counts may underestimate the size of the overwintering population during years when the October – November temperatures are substantially above the long-term average. The increases in temperature along the coast during both the fall and winter suggest that southern counties in California are likely to see fewer and fewer monarchs in the coming decade. In addition, the tendency of monarchs to seek cooler locations will lead to an increasing dependency on the more northerly overwintering sites. However, if the rapid increases in temperatures during October and November along the California coast (Table 2) continue at a rate similar to that of the last decade, monarch overwintering along the coast (Table 3) is likely to decline. Should this be the case, it will become increasingly important to maintain and improve on the structural features of the overwintering sites that offer the best protection throughout the winter.

Table 3. Increases in mean temperatures for October-November for the intervals indicated. Note the exceptional rate of change in the last decade for all coastal counties.

Table 4. Mean December through February temperatures for coastal counties in California.

The first column represents the long-term mean followed by means for the last four decades.

The long-term change, the increase from the previous decade and the projected increase are shown in the following columns. Most of the increase in mean temperatures occurred in the last decade. Monarch numbers at overwintering sites trend lower as mean December-February temperatures increase beyond 51.0F in San Diego, Orange and Los Angeles counties (Addendum Figure 1). The estimations for the next decade were based on the increases during the previous decade. Given the recent trends, the means for the next decade could be higher than shown.

Analysis

The analysis I’ve used to support the interpretations outlined in the text is based on monthly means for counties. While the counties are all coastal, their areas vary with some extending significantly inland, with the result that there is some loss of equivalency. Further, the means can only be seen as proxies for the conditions at the colony sites and should not be taken as exact representations of the temperatures which are likely to be lower at specific overwintering sites (Table 1). Means can also conceal events such as short intervals with weather extremes that can have a significant impact on the monarch population. For example, after an exposure to a 3-4day interval with temperatures in the mid 70s or higher, many overwintering monarchs are likely to break diapause. The same is likely to occur during the migration. Mean temperatures also don’t capture mortality due to other weather events such as high winds or extreme rainfall. All the same, there are patterns associated with increasing temperatures that appear to explain much of what has been happening at overwintering sites along the coast from 1997 (when the counts began) to the present. Mean temperatures have risen from 1.8F to 1.0F per decade from San Diego to San Francisco. The range of mean maximum temperatures is even greater being from 2.7F to 1.3F per decade. While these changing conditions affect the overwintering monarchs, and likely their movement and survival, a deeper dive into the weather data is needed to understand what determines the number of monarchs that arrive at the overwintering sites each winter.

Retention

In addition to the Thanksgiving Counts, counts are made for most counties again in January. Retention refers to percentage of monarchs remaining in each county once the January counts are completed. These data are summarized by county, for the 5 years (2017-2021) during which both Thanksgiving and January counts are available in Table 4. The numbers of monarchs in both San Diego and Ventura counties were too low for meaningful estimates of retention. Two factors, temperature and site quality, probably account for much of the retention. But high winds and other weather events could account for some of the decline. The overall retention doesn’t appear to be associated with differences in temperature among years for colonies collectively, e.g., warmest year (2020) and the coolest year (2021) have similar rates of retention (63.8 vs 62.8). However, within years in the 4 core colonies (SB, SLO, MT, SC), the highest retentions are often associated with the counties with the lowest temperatures, such as Santa Cruz in 2020, Table 6. The retention rates for both Alameda and Marin counties are both low, but there are no indications those rates are associated with temperatures.

Table 5. Mean temperatures for September – December and percentages of monarchs remaining at all overwintering sites in each county at the time of the January counts for the years 2017-2021.

The temperature records for 8 counties are summarized for the years 2017-2021 in Tables 6-10. The data represents deviations from the long-term means for each county. September represents the interval during which most of the long-distance migration occurs. October and November temperatures correspond to the period of arriving at and settling in at overwintering sites. December represents a period of decline in number of occupied sites and numbers per site. These tables also include the original and January counts.

Extremely high temperatures in September, as in 2020, Tables 1 and 6, probably reduced the numbers of monarchs reaching coastal sites by causing some to break diapause and drop out of the migration. Attrition due to the lack of nectar resources under these conditions could also be a factor. Unfortunately, there are no means to assess the impact of September temperatures on the number of migrants reaching the coast. October and November temperatures are key. They can be too extreme as in 2020 or favorable by being close to the long term means as in 2021, Tables 6 and 5. December temperatures were included to determine if they influenced the declines within and among years. While there is no clear association of these temperatures with the declines, more analysis might be helpful.

Table 6. Deviations from the long-term mean temperatures for 10 counties surveyed for monarchs in 2021. Original refers to the number of monarchs in the Thanksgiving count and remaining refers to the numbers counted at the same sites in the following January.

Table 7. Deviations from the long-term mean temperatures for 8 counties surveyed for monarchs in 2020. Original refers to the number of monarchs in the Thanksgiving count and remaining refers to the numbers counted at the same sites in the following January.

Table 8. Deviations from the long-term mean temperatures for 8 counties surveyed for monarchs in 2019. Original refers to the number of monarchs in the Thanksgiving count and remaining refers to the numbers counted at the same sites in the following January.

Table 9. Deviations from the long-term mean temperatures for 8 counties surveyed for monarchs in 2018. Original refers to the number of monarchs in the Thanksgiving count and remaining refers to the numbers counted at the same sites in the following January.

Table 10. Deviations from the long-term mean temperatures for 8 counties surveyed for monarchs in 2017. Original refers to the number of monarchs in the Thanksgiving count and remaining refers to the numbers counted at the same sites in the following January.

Summary

Monarchs are an enzyme. That’s the title I used for three posts to the Monarch Watch Blog to introduce the concept that monarchs operate most effectively at a limited range of temperatures for each activity. We can imagine an enzyme activation curve for each behavior. These optimal ranges have not been defined analytically or experimentally, but we know enough to approximate their limits. For example, temperatures ranging from 65-75F favor migration. Migratory flight appears to diminish either side of this range. Overwintering monarchs appear to do best when temperatures range from 30-60F with monthly means in the high 40s. Both of these conditions apply to the fall and overwintering conditions along the California coast where it is getting too warm both during the fall migration (September flights and October-November clustering at overwintering sites) and the wintering interval from December-February. These changing conditions have resulted in a decline in the numbers of monarchs overwintering in Southern California counties. The tendency for diapausing monarchs to seek temperatures within the range that maintains a non-reproductive condition leads to the progressive abandonment of warmer and less protected sites during the period of cluster formation and throughout the winter months. The result is both consolidation at some overwintering sites and the movement away from many sites as early as December. In the case of 2020, when few clusters formed, it is likely that the fall population overwintered inland in small scattered clusters as suggested by the niche model of Fisher, et al., 2018. Breeding populations that carried over from summer breeding in urban areas may have been augmented by fall monarchs that became reproductive. Offspring from these populations may have moved inland (James et al. 2021) along with overwintered monarchs to produce a large first generation. The numbers of females (10s of thousands) appeared to have been sufficient to produce a first generation that colonized summer breeding areas, mostly in the inner mountain west from May to mid-June in 2021. The multi-generation reproduction that followed produced a migratory population of 300,000, or more, monarchs that populated the overwintering sites under the near average temperatures that occurred from September through November 2021. The rates at which temperatures are increasing during the fall and winter along the California coast suggest that overwintering numbers are likely to decline in the coming decade. While the climate projections in Fisher, et al., 2018, were well into the future, the predicted outcomes seem to be partially represented by the distribution and abundance of wintering monarchs in 2020.

Acknowledgments

Exchanges with Paul Cherubini about Western monarchs on the WesternMonarchs@groups.io discussion list reminded me of the basics of how monarchs respond to temperatures during the migration and when clustering at overwintering sites. Jay Diffendorfer, Peter Ipsen, David James and Patrick Guerra offered their perspective on a number of points and Janis Lentz read the text and kindly pointed out missing commas, run on sentences and some non-sentences. Any errors or misinterpretations are mine.

References

De Anda, A., and K. S. Oberhauser. 2015. Invertebrate natural enemies and stage-specific mortality rates of monarch eggs and larvae. Pages 60– 70 in K. S. Oberhauser, K. R. Nail, and S. Altizer, editors. Monarchs in a changing world: Biology and conservation of an iconic butterfly. Cornell University Press, Ithaca, New York, USA.

Fisher, A.; Saniee, K.; Van der Heide, C.; Griffiths, J.; Meade, D.; Villablanca, F. Climatic Niche Model for Overwintering Monarch Butterflies in a Topographically Complex Region of California. Insects 2018, 9, 167. doi.org/10.3390/insects9040167

Grant, T. J., D. T. T. Flockhart, T. R. Blader, R. L. Hellmich, G. M. Pitman, S. Tyner, D. R. Norris, and S. P. Bradbury. 2020. Estimating arthropod survival probability from field counts: a case study with monarch butterflies. Ecosphere 11(4):e03082. https://doi.org/10.1002/ecs2.3082

James, D. G.; Schaefer, M.C.; Krimmer Easton, K.; Carl, A. First Population Study on Winter Breeding Monarch Butterflies, Danaus plexippus (Lepidoptera: Nymphalidae) in the Urban South Bay of San Francisco, California. Insects 2021, 12, 946. doi.org/10.3390/insects12100946

James, D.G.; Schaefer, M.C.; Krimmer Easton, K.; Carl, A. Reply to Davis, A.K. Monarchs Reared in Winter in California Are Not Large Enough to Be Migrants. Comment on “James et al. First Population Study on Winter Breeding Monarch Butterflies, Danaus plexippus (Lepidoptera: Nymphalidae) in the Urban South Bay of San Francisco, California. Insects 2021, 12, 946”. Insects2022,13,64. doi.org/10.3390/insects13010064

Nail, K. R., C. Stenoien, and K. S. Oberhauser. 2015. Immature monarch survival: effects of site characteristics, density, and time. Annals of the Entomological Society of America 108: 680– 690.

Oberhauser, K., R. Wiederholt, J. E. Diffendorfer, D. Semmens, L. Ries, W. E. Thogmartin, L. Lopez-Hoffman, and B. Semmens. 2017. A trans-national monarch butterfly population model and implications for regional conservation priorities. Ecological Entomology 42: 51– 60.

Pelton EM, Schultz CB, Jepsen SJ, Black SH and Crone EE (2019) Western Monarch Population Plummets: Status, Probable Causes, and Recommended Conservation Actions. Front. Ecol. Evol. 7:258. doi: doi.org/10.3389/fevo.2019.00258

Semmens, B. X., D. J. Semmens, W. E. Thogmartin, R. Wiederholt, L. López-Hoffman, J. E. Diffendorfer, J. M. Pleasants, K. S. Oberhauser, and O. R. Taylor. 2016. Quasi-extinction risk and population targets for the Eastern, migratory population of monarch butterflies (Danaus plexippus). Scientific Reports 6: 23265.

Schultz, C.B.; Brown, L.M.; Pelton, E.; Crone, E.E. Citizen science monitoring demonstrates dramatic declines of monarch butterflies in western North America. Biol. Conserv. 2017, 214, 343–346.

Taylor, O. R., Lovett, J. P., Gibo, D. L., Weiser, E. L., Thogmartin, W. E., Semmens, D. J., et al. (2019). Is the timing, pace and success of the monarch migration associated with sun angle? Front. Ecol. Evol. 7:442. doi: doi.org/10.3389/fevo.2019.00442

Data Citations

NOAA National Centers for Environmental information, Climate at a Glance: National Time Series, published June 2022, retrieved during June 2022 from ncdc.noaa.gov/cag/

Xerces Society Western Monarch Thanksgiving Count. 2022. Western Monarch Thanksgiving Count Data, 1997-2021. Available at westernmonarchcount.org

Addendum

Figure 1 below and supporting commentary can be found at:
monarchwatch.org/blog/2020/02/25/monarchs-and-climate-in-the-west

Figure 1. Both the long-term mean (upper) and the mean for the last 20 years (lower) for January–February for 4 counties from Marin in the north to San Diego in the south are displayed above. The differences between the long-and-short term temperatures range from +1.7F for Marin County to +2.4F degrees (.85–1.2F degrees/decade) for San Diego County. Along with these increases in temperature, the number of sites and individuals per site has been decreasing in the southernmost counties and while increasing in Marin County. Over the 20-year interval in San Diego, the number of sites declined from 18 to 3 while the count declined from 2,590 to 12. Overall, the data suggest that the numbers of sites and monarchs both decline as temperatures increase. Further, the data suggest monarchs are seeking cooler temperatures by moving northward along the coast. The alternative possibility is that they could be moving to cooler inland sites – if such exist.

Note 1. Although the number of females that survived the winter of 2020-2021 to reproduce is unknown, it’s possible that the number was as high as 30,000. If so, then, how many monarchs started the overwintering cycle in the fall of 2020? Generally, since females are usually about 45% of the population, the total surviving would have been approximately 66,500. If this number represented only 50% of initial population due to mortality, the starting number would have been roughly 133,000. While these numbers are speculative, they provide a glimpse of the possible magnitude of the population that overwintered in 2020-2021. It appears to have been much larger number than the1849 represented by the Thanksgiving counts.

Note 2. In theory, monarchs could return from Mexico to repopulate the West and perhaps a few do so. The entry to the West could involve monarchs entering as they followed the Rio Grande north through Las Cruces and then moved to the West into eastern AZ and western NM.

An alternative route would involve advancing through the mountains in Mexico with entry into AZ east of Nogales into southeastern AZ south of Sierra Vista. From there monarchs could advance northward into east-central AZ. Both routes are >300 miles longer than the routes that take monarchs to reach Texas. That would involve a minimum of 6 more days of flight to reach the border.

At present, there is no evidence monarchs use either of these routes to reach the US. To establish that such movements occur, we need evidence from isotopes and first sightings. To obtain the isotope evidence, worn monarchs suspected of having returned from Mexico need to be collected and analyzed. First sightings recorded by Journey North and photos from iNaturalist from March and April in the West could help as well.

Monarchs return from Mexico to TX in early March. Arrival continues until mid-April. Similar arrival times would be expected in AZ and NM. In 2020, the year of interest, there was one first sighting in Carlsbad, NM on the 20th of March along with a note that 5 others were seen nearby. There were no other reports of monarchs in NM until the first and second of May. One was described as old and tattered and another as fresh. All were sighted east of the divide. No monarchs were sighted west of the divide until 12 of May in Utah. There were no records on iNaturalist as well. So, there is no evidence from first sightings in 2020 that helps us understand the growth of the population that summer.

Overall, while it is possible that monarchs reach the US along these routes now and then, the probability that the numbers that return are sufficient to lead to substantial increases in western monarch numbers seems low.

Note 3. Ecological release is another possible explanation for the resurgence of the population in 2021. This term is used to explain explosive population growth following an event that eliminates factors that normally limit population growth. Such an explosion of butterflies occurred after the 7month drought in Texas in 2011. That drought was followed by 7months of above normal moisture starting in the fall of 2011. Evidently, due to the suppression of predators and parasites during the drought, 16 species of butterflies were able to produce large numbers which led them to expand their distribution northward into Kansas and beyond in April and early May of 2012. While precipitation in 2020 was the second lowest for the 127year record for the West Climatic Region, the more normal, but still below average, pattern of precipitation in 2021 has not been associated with explosive increases in the numbers of other butterfly species.

Monarch Watch Update April 2023

This newsletter was recently sent via email to those who subscribe to our email updates. If you would like to receive periodic email updates from Monarch Watch, please take a moment to complete and submit the short form at monarchwatch.org/subscribe/

Greetings Monarch Watchers!

A brief update this time around – hope you are starting to see monarchs in your area or see them soon! We spotted the first monarch in Monarch Waystation #1 in Lawrence, Kansas on April 18th; the worn female was busy finding all the emerging common milkweed shoots to lay eggs on 🙂

Included in this issue:
1. Monarch Watch Open House & Spring Plant Fundraiser
2. Monarch Population Status —by Chip Taylor
3. Monarch Tag Recoveries from Mexico
4. Monarch Calendar & Directional Flight Projects
5. Monarch Waystations
6. About This Monarch Watch List

1. Monarch Watch Open House & Spring Plant Fundraiser

It’s here! Our annual Spring Plant Fundraiser is now online at https://spring.monarchwatch.org and we have thousands of plants looking for good homes. We are once again offering online ordering and curbside pickup (or limited local delivery) for this event. 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. Don’t miss out and be sure to take advantage of our “Buy 10 Plants, Get 1 Free” offer!

A complete list of plants and online ordering is available via the link below and pickup appointments are being scheduled for May 10, 11 & 13.

Monarch Watch Spring Plant Fundraiser: https://spring.monarchwatch.org

We are also having an in-person component of our Spring Open House & Plant Fundraiser on Saturday, May 13th. This will be a primarily outdoor event and there will be tours of our gardens, games, activities, monarch butterflies, caterpillars, and lots of butterfly plants available for your own garden! Complete details at https://monarchwatch.org/openhouse

If you are not able to participate locally, we invite you to contribute to this annual fundraiser by donating to Monarch Watch via https://monarchwatch.org/donate

Thank you!

2. Monarch Population Status —by Chip Taylor

A few notes on the development of the eastern and western monarch populations as of 20 April 2023.

Eastern monarch population

The Journey North program ( https://journeynorth.org ) has been recording monarch first sightings for 23 years. It’s an awesome record with a number of uses. Out of curiosity to see how first sighting this year in NE Kansas compared with other years, I skimmed through those records. This year the first monarch to appear in the Topeka, Lawrence, Kansas City area was spotted on the 10th. That is early and by the 18th there were a number of sightings. Checking the records revealed that more than one sighting by the 18th of April has only occurred in 7/23 years. In other words, the number of monarchs to reach our area is so few in most years that we simply don’t encounter them, or they may not even be this far north.

While I have been watching the first sightings accumulate in space and time this year, I’m finding it difficult to assess what the data mean. The numbers for Texas are down. That could mean the population is down, with fewer returnees or it could mean that monarch activity was low due to cool weather so fewer were seen. The latter seems possible since I’ve seen a number of reports by observers that mention the sightings of many monarchs over time or on single outings. While some of the monarchs arrived ahead of the emergence of milkweeds in some areas, it appears that the majority of eggs laid by returning monarch were laid in Texas, where due to slightly warmer temperatures, the larvae will develop faster than if the eggs were laid further north. I’ll need to look more closely at the data, but overall, the colonization by the returning monarchs fits a common pattern and there is no reason for concern at this time.

Western monarch population

The population rebound over the last two years in the West has been remarkable. The numbers counted at all known overwintering sites last November exceeded 335K. This was the largest population recorded since 2000 – a mere 22 years. How is that possible, and what does it mean for the future? Can we expect similar numbers or even more monarchs in November 2023? Probably not. The population is certain to be lower for a number of reasons, but how much lower and where will the monarchs originate from that reach the overwintering sites?

The western population development, like that of the east, is largely driven by the weather. So, the immediate question is are the weather patterns from last November to the present similar to any year in the past and the answer is yes. The winter and spring of 2006 in California was cold, though not as wet, as that of 2023. The temperature patterns were quite similar. Statewide the three-month interval from January through March was the coldest since 1955 and one of the wettest as well with many strong storms. These conditions had to take a toll on the overwintering monarchs and certainly reduced the number of females available to begin egg laying as the weather warmed, but when did that begin? In most years, especially recently, mating and reproduction begin by mid-March and sometimes earlier. This year the mean temperatures for March were 44.2F vs 44.0F for 2006. These were the 4th and 5th coldest March temperatures in the record that goes back to 1895. April started out cool in both years with temperature increasing in the last 10 days of the month (based on the 10-day forecast for this year). So, given these extreme conditions, how much reproduction has been possible to date this year? There are no data on this point.

Again, looking to 2006, at the end of the migration the count was over 200K. Will the numbers reach 200k this year. Maybe, but I’m skeptical. In most years, first generation monarchs would begin emerging in the next week and would begin moving into the Sierra foothills and beyond into Nevada and the inner mountain west. Some would begin moving in May toward the NW as well. Movement out of California looks to be limited for some time due to cold conditions in both the Sierras and to the north. If these conditions continue, most of the overwintering monarchs in November will have originated from California. This was first suggested by Paul Cherubini, a long-term observer of monarchs in the west, in a post to the Western Monarchs email list ( https://groups.io/g/WesternMonarchs ) on 5 April. Paul maintains that the population could “produce an overwintering population roughly as large as the past two winters”. That would surely be an interesting outcome.

3. Monarch Tag Recoveries from Mexico

More than 360 Monarch Watch tags were recovered from monarch overwintering sites in central Mexico during the 2022 tagging season. All of the tags have been examined and the “Tag recoveries from central Mexico” list has been updated. By default, this list is sorted by the report season then by tag code and now includes over 21,000 records.

Get out your tag codes and check out the updated list 🙂

Monarch Watch tag recoveries: https://monarchwatch.org/tagrecoveries

As a reminder, it is never too late for data so if you have not yet submitted your records, please do so at your earliest convenience via https://monarchwatch.org/tagging

Thank you to everyone who tagged monarchs in 2022 and also those who assisted with the recovery efforts!

4. Monarch Calendar & Directional Flight Projects

For those of you that are participating in our Monarch Calendar or Directional Flight projects, an email will be sent to everyone who has registered at the end of the observation periods with instructions to submit data. If you have not yet registered or would like more information about these projects, please see the links below.

Monarch Calendar Project: https://monarchwatch.org/calendar

Directional Flight Project: https://monarchwatch.org/directional-flight

5. Monarch Waystations

To offset the loss of milkweeds and nectar sources we need to create, conserve, and protect monarch butterfly habitats. You can help by creating “Monarch Waystations” in home gardens, at schools, businesses, parks, zoos, nature centers, along roadsides, and on other unused plots of land. Creating a Monarch Waystation can be as simple as adding milkweeds and nectar sources to existing gardens or maintaining natural habitats with milkweeds. No effort is too small to have a positive impact.

Have you created a habitat for monarchs and other wildlife? If so, help support our conservation efforts by registering your habitat as an official Monarch Waystation today!

Monarch Waystation Program: https://monarchwatch.org/waystations

A quick online application will register your site and your habitat will be added to the online registry. You will receive a certificate bearing your name and your habitat’s ID that can be used to look up its record. You may also choose to purchase a metal sign to display in your habitat to encourage others to get involved in monarch conservation.

As of 26 April 2023, there have been 42,680 Monarch Waystation habitats registered with Monarch Watch! Texas holds the #1 spot with 3,558 habitats and Illinois (3,270), Michigan (3,128), California (2,790), Ohio (2,226), Florida (2,213), Pennsylvania (1,877), Virginia (1,864), Wisconsin (1,842), and New York (1,384) round out the top ten.

You can view the complete Monarch Waystation Registry and a map of approximate locations via https://monarchwatch.org/waystations/registry

6. About This Monarch Watch List

Monarch Watch ( https://monarchwatch.org ) is a nonprofit education, conservation, and research program affiliated with the Kansas Biological Survey & Center for Ecological Research at the University of Kansas. The program strives to provide the public with information about the biology of monarch butterflies, their spectacular migration, and how to use monarchs to further science education in primary and secondary schools. Monarch Watch engages in research on monarch migration biology and monarch population dynamics to better understand how to conserve the monarch migration and also promotes the protection of monarch habitats throughout North America.

We rely on private contributions to support the program and we need your help! Please consider making a tax-deductible donation. Complete details are available at https://monarchwatch.org/donate or you can simply call 785-832-7386 (KU Endowment Association) for more information about giving to Monarch Watch.

If you have any questions about this email or any of our programs, please feel free to contact us anytime.

Thank you for your continued interest and support!

Jim Lovett
Monarch Watch
https://monarchwatch.org

You are receiving this mail because you were subscribed to the Monarch Watch list via monarchwatch.org or shop.monarchwatch.org – if you would rather not receive these periodic email updates from Monarch Watch (or would like to remove an old email address) you may UNSUBSCRIBE via https://monarchwatch.org/unsubscribe

If you would like to receive periodic email updates from Monarch Watch, you may SUBSCRIBE via https://monarchwatch.org/subscribe

This e-mail may be reproduced, printed, or otherwise redistributed as long as it is provided in full and without any modification. Requests to do otherwise must be approved in writing by Monarch Watch.

Monarch numbers: dynamics of population establishment each spring

Introduction
The short narratives that follow describe the outcomes of the timing and number of monarchs arriving from Mexico in March. Each cohort encounters different temperature scenarios in March and April that largely determine the size of the population the following winter season. To be sure, there are additional factors that determine the size of the migratory population and the overwintering number of hectares, but it is largely what happens in March, April and May that set the stage for the rest of the season. A key to population growth is both the number of first-generation offspring produced in this period and their age at first reproduction which is an outcome of their developmental history.

I’ve chosen 4 years – 2012, 2013, 2017, 2018 – to illustrate the dynamics associated with the interactions of the returning migrants and the weather each cohort encounters. The first pair represent both the earliest and most rapidly advancing migration and the smallest and latest cohort to advance north of 40N since 2005. The second pair were selected to show the consequences of advancing northward too soon as in 2017 or being limited to reproducing largely in Texas and southern Oklahoma as in 2018 due to colder weather further north. The latter two show that similar starting conditions can lead to different outcomes. The mean temperatures in March for Texas for three of these years were significantly above the long-term average (+5.4F, +6.9F, +7.3F). Elevated temperatures in this range are expected to become the norm in the near future.

As background, although some monarchs returning from Mexico enter Texas during the first week of March, the influx increases in the second week of March with monarchs often detected in areas of the Edwards Plateau as well as San Antonio, San Angelo and Austin. This progression continues N, NE as weather permits, with most of the returning monarchs dying by the 1st of May. In this analysis, I’ve totaled the first sightings from 30N, just north of the latitude (29N) at which milkweed diversity and abundance increases in Texas, northward beyond north of 40N. The longitudinal range is 80W (western PA) to 100W (mid Dakotas). These progressions each year are related to March and April temperatures for Texas, Oklahoma and Kansas. Throughout this treatment, 40N is used as the southern limit of the summer breeding habitat since tagging indicates that more than 80% of the monarchs reaching the overwintering area originate north of 40N.

2012
March in 2012 was the warmest recorded in the US for that month since record keeping began in 1895. The median arrival for first sighting in Texas was 13March, the earliest in the 24-year record of first sightings reported to and recorded by Journey North. The temperatures were significantly above the long-term averages for March and April from Texas through Kansas (Table 1). These conditions allowed monarchs to advance significantly from Texas into Oklahoma, Kansas and north of 40N by the 2nd of May, with 23% recorded beyond 40N. Overall, this recolonization was the earliest to date for all regions north of 40N with only 3.6% of the sightings reported in the last observation interval from 31May-9June (Table 2, see Appendix). There are significant consequences of advancing northward rapidly; getting ahead of emerging milkweeds and laying eggs in regions which are cooler results in a longer development time for immatures and therefore a delayed age to first reproduction. The latter can have the effect of lowering the rate of population growth. While it is likely that arriving too early at the northern latitudes had a negative effect on population growth, the summer temperatures were also extreme resulting in a substantial drought that also had a negative impact on the development of the generation destined to migrate. Together, these negative spring and summer conditions resulted in a decline in winter numbers from 2.89ha in 2011 to 1.19ha (-1.7ha). These developments were a precursor and likely a contributor to the crash in the population that occurred in 2013.

2013
The number of first sightings in 2013 was low relative to previous and following years, raising the possibility that many monarchs did not survive the winter in Mexico or during the return migration from the overwintering sites to Texas. It is also possible that the weather in Texas limited the activity of the monarchs and therefore the number of sightings. While the mean temperature for March in Texas (0.9F) was close to the long-term mean, the lower April temperatures for Texas through Kansas evidently limited northward movement, with the result that 83% of the first sightings were limited to the 30-35 latitudes (roughly Austin to Oklahoma City). The lower April temperatures for both Texas and southern Oklahoma evidently slowed the development of the first-generation offspring, resulting in the latest recolonization of the areas north of 40N in the Journey North first sighting records (55.7%, Table 2). Thislate arrival is in contrast to the arrivals in 2012 in which only 3% of the sightings occurred in the last observation period (Table 2). This delayed beginning to the production, together with the low numbers, were surely factors that led to the all-time low number of hectares at the overwintering sites during the 2013-2014 winter season.

2017
The mean temperature for March 2017 was 7.3F above the long-term average and the highest yet recorded. While the median arrival date (26March) was later than average (21March), arriving monarchs rapidly advanced beyond 35N such that only 28% of the first sightings were recorded for 30-35N and 56% recorded from 35-40N. The rapid advance beyond 35N resulted in two unusual events: massive clustering in northern Oklahoma beginning on 4Aprilwhich is described in detail in a Blog posting – Spring roosting: A rare event and a wind-aided surge of these monarchs northward from 7-9April that advanced the front 300miles – well into Nebraska. This event is described in another post to the Blog – Monarch Population Status (5/11/2017). In this case, monarchs were well ahead of emerging milkweeds, egg dumping was common on the few that were above ground and some eggs were subsequently frozen at the northern limits of this push.

We tracked the development from egg to adult for eggs laid 9-10April on tropical milkweeds in large containers maintained outside or our greenhouse in Lawrence, KS. The shortest time from egg to adult was 45 days – at least 15 days longer than if these same eggs had been laid in Texas rather than Kansas. It follows from these observations that distributing eggs too far north too soon contributes to competition among larvae due to egg dumping and perhaps greater predation on eggs. In addition, the loss of some eggs and larvae due to freezing temperatures and a delay in the development of the more northerly part of the distribution results in a higher mean age at first reproduction for the first-generation cohort. That in turn delays colonization of all areas north of 40N as shown by the late first sightings in 2017 (42.9%, Table 2). The overall result of the return migration and the colonization by the first generation was a modest decline in overwintering numbers from 2.91ha in 2016 to 2.48ha in 2017 (-0.43ha).

2018
The March temperatures in 2018 were also above the long-term average (+5.4F), but the outcome differed from that of both 2012 and 2017. In this case, 70% of the first sightings were in the 30-35N range. The monarchs were mostly confined to Texas and southern Oklahoma due to cooler temperatures in April from Texas north to Kansas. The result was the production of a large first generation that, on average, developed more rapidly than the first-generation cohorts in 2012 and 2017. The movement north of 40N was aided by warmer than average May temperatures (+7.0F) in the Midwest. Those elevated temperatures resulted in an earlier colonization of the northern breeding area with only 16.6% of the sightings being recorded in the last observation period (Table 2). As a consequence of these conditions, the population grew from 2.48ha to 6.05ha (+3.57ha). These conditions were similar to those that preceded the development of the population in 2006 (6.87ha).

Summary
These narratives demonstrate the complexities of the interactions of returning monarchs with the environmental conditions encountered during March and April. The development of the population is largely determined by the timing of the return together with the numbers arriving along with the weather conditions that restrict or enhance the expansion of the colonizing population. The limits of movement defined by weather conditions determine the latitudinal distribution of eggs, egg and larval mortality, the growth rate of immatures and the mean age of first reproduction for each returning cohort. In the future, elevated temperatures in March are likely to be associated with more rapid expansions of the breeding population resulting in smaller cohorts of first-generation offspring moving north. A later arrival of these monarchs in the summer breeding areas will also lead to smaller summer and fall populations and later migrations.

Acknowledgements
The interpretations in this text are largely based on Journey North’s first sightings maps and the data they are based on. The value of this data set, started in 2000, lies in the volume of first sightings reported through the winter to the 31st of July. Janis Lentz kindly organized a subset of the data for analysis.

Table 1. Deviations from the long-term (1902-2000) mean temperatures for the months and states indicated.

Table 2. Total first sightings from 30N to >40N latitudes from 1March to 2May with percentages sighted across the latitudes for each year. Median dates of first sightings in Texas, hectares measured at the end of the season and net change from the previous year are listed. The last column shows the percentage of first sightings recorded from 31May to 9June (see Appendix).

Appendix
The first sightings above 40N were sorted into four 10-day intervals from 1May-9June to assess the timing and number of first-generation monarchs arriving across the northern portion of the summer breeding range. Relatively few first sightings are recorded after 9June. That date is close to the end of directional flight by the first generation. The spring migration appears to stop on or close to the 12th of June as the difference in daylength from one day to the next drops below one minute. For a full discussion of this issue, see Monarch Puzzle Wrap Up.

Distribution maps of first sightings reported to and recorded by Journey North for 1January to 2May for the years indicated. The data in this analysis were based on sightings from 1March to 2May. Sightings were summarized from north of 40N from 1May to 9June as well. The line representing 40N latitude extends from the northern border of Kansas to Philadelphia.

Monarch numbers: trends due to weather and climate

The U.S. Fish and Wildlife Service will be making a decision as to whether to list the monarch as either threatened or endangered under the Endangered Species Act as opposed to the current finding which is “warranted [for ESA listing] but precluded [by other higher-priority listing actions]”. That decision will involve many considerations, most important of which will involve an understanding of monarch biology together with the perceived threats to the monarch population.

One threat we have been addressing is the loss of habitat through many new conservation efforts in public and private settings. These efforts, while impressive, need to increase to offset annual rates of habitat loss as well as past losses due to land management. Counter to these positive developments there are negative trends in weather and climate that must be considered. It is these weather and climate-related stressors which, if they continue, will diminish the monarch migration in the coming decades.

Monarchs are being pinched by changes both at the beginning and end of the breeding season.

Increasing temperatures in March and April in the South Region (Texas and Oklahoma) are allowing monarchs to move too far north too soon and September and October temperatures in the Midwest and South Regions respectively are delaying the migrations in a manner that diminishes the number of monarchs reaching the overwintering sites (Taylor, et al 2020). These relationships are summarized, along with the rate of change in temperature per decade, in Table 1. For reference, rates of change of more than 0.5F per decade are extraordinary and rates of 1.0F or higher are indications of truly worrisome trends that are unlikely to moderate or reverse.

Table 1. Rate of temperature change per decade, 1982-2022. Changes of less than 0.5/decade are considered to be inconsequential due to the impact of other factors on monarch numbers. Temperatures of more than 0.5/decade are likely to have either negative or positive effects on the development of the population but are known to have negative effects on both the spring and fall migrations. Blanks represent months when monarchs are absent or nearly so from the city indicated. The city temperatures are strongly correlated with state and usually regional temperatures. To provide readers with a sense of what the trend lines look like, I have included three images (see Appendix) obtained from the data available via Climate at a Glance. Color code: white – neutral, yellow – either, pink – negative, green – positive*.

March
High temperatures in March in Texas and Oklahoma along with SW winds allow monarchs to advance too far north too soon, i.e., beyond the emergence of milkweeds. Spreading eggs out to the north also has the effect of increasing the mean age at first reproduction. This delay in turn reduces the rate of population growth. Data supporting this interpretation is presented in another blog article: Monarch numbers: dynamics of population establishment each spring.

April-May
The increase in temperatures in the South in April and May could have both negative and positive effects. Positive outcomes could include more rapid development of immatures which would lessen exposure to predators and parasites. Negatives might include shorter lives for egg laying females.

June, July, August
The increase in temperature in June the Midwest stands out relative to the smaller increases in May, July and August. The conditions in June should have a positive impact on the development of the population since it will accelerate the development of larvae and shorten the mean age to first reproduction for the cohort(s) developing that month.

September
Higher temperatures in September in the Midwest and Northeast (Culbertson, et al., 2021, Ethier and Mitchell 2023) are having the effect of delaying the migrations through those regions. Delayed migrations have several causes, but all are associated with lower numbers of monarchs reaching the overwintering sites (Taylor, et al., 2019, 2020).

October
The high temperature effect applies to the South Region in October. Since this region is prone to droughts, high temperatures combined with droughts during October can have a significant impact on the lipid levels (Brower, et al., 2015, Hobson, et al., 2020) and are likely to affect adult survival since monarch numbers are lower at the overwintering sites following these conditions (Taylor, et al., 2020). The elevated temperature in 2019 may be one of the reasons the overwintering numbers declined from 6.05ha in 2018 to 2.83 in 2019. The mean temperatures during October 2019 in San Antonio were the highest in the 128-year record.

Other Threats
Other threats include El Niño and the ability of monarchs to enter diapause. Both constitute existential threats to the continuation of the monarch migration as we know it today. I’ll dive into these topics as time permits.

References

Brower, L. P., Fink, L. S., Kiphart, R. J., Pocius, V., Zubieta, R. R., and Ramírez, M.I. (2015). “Effect of the 2010-2011 drought on the lipid content of monarchs migrating through Texas to overwintering sites in Mexico,” in Monarchs in a Changing World: Biology and Conservation of an Iconic Butterfly, eds K. S. Oberhauser, K. R. Nail, and S. Altizer (Ithaca, NY: Cornell University Press), 117–129.

Culbertson, K. A., Garland, M. S., Walton, R. K., Zemaitis, L., & Pocius, V. M. (2021). Long‐Term Monitoring Indicates Shifting Fall Migration Timing in Monarch Butterflies (Danaus plexippus). Global Change Biology. doi.org/10.1111/gcb.15957

Ethier, D. M. and Mitchell, G.W. (2023). Effects of climate on fall migration phenology of monarch butterflies departing the northeastern breeding grounds in Canada. Global Change Biology. doi.org/10.1111/gcb.16579

Hobson, K.A., O.R. García-Rubio, R. Carrera-Treviño, L. Anparasan, K.J. Kardynal, J. McNeil, E. García-Serrano and B.X. Mora Alvarez. 2020. Isotopic (δ²H) analysis of stored lipids in migratory and overwintering Monarch Butterflies (Danaus plexippus): Evidence for critical late-stage southern nectaring? Frontiers in Ecology and Evolution 8:572140. doi.org/10.3389/fevo.2020.572140

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., & Grundel, R. (2019). Is the timing, pace, and success of the monarch migration associated with sun angle? Frontiers in Ecology and Evolution, 7, 442. doi.org/10.3389/fevo.2019.00442

Taylor, O.R. Jr, Pleasants. J.M., Grundel, R., Pecoraro, S.D., Lovett. J.P. and Ryan, A. (2020) Evaluating the Migration Mortality Hypothesis Using Monarch Tagging Data. Front. Ecol. Evol. 8:264. doi.org/10.3389/fevo.2020.00264

Appendix

Figure 1. Temperature trends in March for selected cities. The horizontal line represents the long-term average. Starting in 2004, most of the yearly means are well above the long-term average.

Monarch Population Status

The WWF-Telmex Telcel Foundation Alliance, in collaboration with the National Commission of Natural Protected Areas (CONANP), the National Autonomous University of Mexico (UNAM), 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 2.21 hectares, a 22% decrease from the previous season (2.84 ha):

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

Report (english): Areas of forest occupied by the colonies of monarch butterflies in Mexico during the 2022-2023 overwintering period

WWF story: Troubling news for monarch butterfly populations

Monarch population growth is largely about timing, numbers and weather. Monarch Watch follows the population closely from month to month throughout the year looking for changes that will help us predict the relative size of the population from one interval or stage to the next. This year, starting in May, the data indicated that the mid and late summer numbers would be low and that the migration itself would involve a smaller population than in recent years. This expectation was realized. The number of butterflies tagged during the migration was the lowest in 9 years. This was another sign that the overwintering numbers would be low since the number tagged is correlated with the size of the overwintering population. Two other factors, droughts in Texas and late arrival at the overwintering sites, are associated with low numbers and both occurred last fall.

All in all, monarchs had a bad year due to a sequence of unfavorable weather events. While low numbers are something of a concern, in recent years monarchs recovered from low numbers in 2012 (1.19 ha), 2013 (0.67 ha) and 2014 (1.13 ha) and they will do so again – weather permitting. For more information on the status of the population through the last season see monarchwatch.org/blog/2023/01/04/monarch-population-status-49

Note:
The WWF-TELMEX Telcel Foundation Alliance collaborates with CONANP to systematically monitor the hibernation of the Monarch since 2004, and they join the Institute of Biology of the National Autonomous University of Mexico (UNAM) to analyze changes in forest cover in the area core of the Monarch Butterfly Biosphere Reserve in order to have scientific bases that support the implementation of conservation strategies for the benefit of the species, ecosystems and human beings.