by Hari Balasubramanian
A selection of facts, research and personal impressions.
In February this year, I traveled to the small town of Angangueo in Central Mexico. A 4-hour bus trip from Mexico City, Angangueo is in a rural part of the state of Michoacan, in the mountainous trans-volcanic belt. Here, in a few select high elevation forests with oyamel (fir) and pine trees, millions of monarch butterflies from east of the Rocky Mountains congregate each year, after an astonishingly long journey – over 3500 km – from Canada and the northern reaches of the US.
The monarchs stay in Mexico from November to March. When the sun is shining, the butterflies – which otherwise huddle together in tightly packed and well camouflaged clusters on the branches and bark of oyamel trees – take to the air, like a beehive that has been stirred. If the sun stays up, as it did the day I visited, the monarchs quickly fill the sky and everything around you; you can even hear the faint flutter of their wings. I was in the very thick of things when this photograph was taken [1]. Every speck in the picture below, however faint, is a butterfly.
Monarchs have fascinated me for years now. In Massachusetts, a few months prior to my Mexico visit, I'd seen the odd monarch or two flying unhurriedly, seemingly without a purpose. So languid was their flight – the classic flap, flap and glide – that there was no way to tell that each butterfly, following some mysterious signals – its reproductive system having been put on pause, allowing the organism to focus on the rigors of the coming journey – was leaving for a distant forest in Mexico.
It's worth reiterating this: each butterfly that migrates south starts the journey alone and has never made the journey before. When birds make long journeys, there are often older individuals that guide the young. In the case of monarchs there is no guide. The recent discovery that the thin, seemingly inconsequential antennae of monarchs house circadian clocks that help in orientation only deepens the wonder [ref. and figure]. Traveling by day and over land, a monarch, “with a mass 20% that of a penny” [2], covers thousands of kilometers in a two month period.
The congregations I saw in the mountain forests of Mexico were the culmination of millions of independent journeys: a coming together of millions of evolutionarily hardwired wills. When they began, the monarchs were dispersed over a wide swath east of the Rocky Mountains. By the end of fall, they are precariously concentrated in a few hectares in Mexico. Butterflies that began in Canada, the Great Lakes region, northeastern US now huddle together for warmth.
Such migration is clearly a response to the cold winters of North America. But then why travel all the way to Mexico when other insects – butterflies such as the Mourning Cloak and even other monarchs which do not migrate at all – have simpler, less arduous overwintering routines? And why overwinter in the 10,000-feet high oyamel forests in Mexico where the temperatures are still pretty cold? In a research article from 2004, I found this explanation: “successful overwintering depends on a narrow climatic window where temperatures are benign enough for the monarch to avoid death by freezing, but cool enough to maintain a state of metabolic and reproductive diapause [dormancy] until spring arrives” [3]. That sounds reasonable, but it's very unlikely that anyone could have predicted the exact details of the migration.
In March, when the temperatures rise, the monarchs begin their journey back north. They get only so far as Texas where they perish. But the next generation starts from latitudes around Texas and then makes another leap north; and the next generation even farther north and wider. The website Journey North keeps track of this reverse, multi-generational migration each year. As of June this year, adult monarchs have been spotted as far north as Canada and Maine – see the blue dots on this map.
By the time it is late summer or early fall the monarchs are once again spread out in their northern haunts east of the Rockies. The annual cycle is complete and it is time again for the special generation of Mexican migrants to emerge. All generations except the Mexican one have a typical life span of 2-6 weeks. The Mexican migrants can potentially live up to 8 months but most of those months are spent in a semi-dormant overwintering state. Such variations in a single species!
Tracing the Migration: Tags and Isotopes
Surprising as it may sound, for most of the 20th century important details of the monarch's migration – such as their Mexican overwintering location – were unknown. How the pieces of the puzzle were put together is in itself a fascinating story.
In 1938, Toronto-based zoologist Fred Urquhart and his wife Norah Roden Urquhart wanted to figure out where the monarchs migrated. At that time it was known was that the butterflies disappeared somewhere but no one knew exactly where. At the other end, the people of Central Mexico – at least those that lived the mountains around Angangueo – surely saw masses of monarchs arrive and stay each year but had no idea where they were coming from. (Why did it take so long to put the pieces together, you might wonder. But then consider that only in 2009 was it hypothesized that dragonflies might be making a journey from India to East Africa across the ocean, stopping en route at Maldives – this is the new longest recorded insect migration in the world, surpassing the monarchs' journey.)
To follow individual butterflies, the Urquharts used tags. If live butterflies were tagged with a unique code, released and then later recovered someplace else (the yield is low: not all tags are recovered) then the path of the migration could be deduced. But this was easier said that done since the early tags – “a printed label affixed to the butterfly's wing with liquid glue”, “labels printed on gummed stock, like postage stamps” [2] – either affected the flight of the butterflies or wouldn't stick. What eventually worked was “a pressure adhesive label used for price tags on glass merchandise”.
Nowadays tags look like the one in the image above. I attended a capture and tag event at a wildlife sanctuary in central Massachusetts in September last year. Events like this are held all over the country, wherever monarchs are found. Rangers and kids, butterfly nets in their hands, went in search of monarchs. Once a monarch was in the net, the butterfly had to be carefully held while the tag was affixed. I couldn't muster the courage to do any of this; I feared a simple twitch of my fingers could affect the delicate butterfly's wings. Once tagged, the beautiful orange and black butterfly suddenly looked like a supermarket product.
Still, it was tags that revealed the first pieces of the migration puzzle in the 1960s. Urquhart enlisted the help of numerous volunteers in the US and Canada who participated in the tagging and retrieval effort. Here, for example, are some results from 1964-1971 showing where the monarchs were tagged and where the same tagged monarchs were retrieved. For most of the butterflies there is a “diagonal flight path, from northeast to southwest”.
But the trail was “lost in Texas” and although it was clear the butterflies moved into Mexico, no one had an idea what happened there. There were “years of frustration” with no further clues. Norah Urquhart wrote to Mexican newspapers “asking for volunteers to report sightings and help with tagging”. It was 
Imagine what this must have meant to the Urquharts who had given their whole lives to the search – what a reward it must have been to actually visit the forests and see the monarchs as they “swirled through the air like autumn leaves and carpeted the ground in their flaming myriads” and even accidentally find a monarch in a dense cluster that, according to the method the Urquharts pioneered, had been tagged in far away Chaska, Minnesota – conclusive proof of the length of the migration. Seldom do long-drawn out research projects have such a dramatic conclusion!
§
Tags, as I mentioned before, have low yields; few are recovered and especially few from the Mexican overwintering sites. There is, however, another interesting technique to trace the origins of monarchs which I'll summarize next.
In 1997, two researchers from Cornell University, Leonard Wassenar and Keith Hobson, collected a sample of 597 dead monarchs from 13 different overwintering colonies in Mexico (it's easy, as I discovered, to find dead butterflies near colonies). Wassenar and Hobson then analyzed the wing membranes of these butterflies to determine their stable hydrogen and carbon isotopic composition. This composition, it turns out, is an intrinsic marker that can be linked back to the birthplace – the so called natal origin – of the monarch:
“The isotopic composition of adult monarch wing membranes closely resembles, and permanently records, the isotopic composition of its natal (larval) food source. The isotopic composition of the monarch larval milkweed host plant (Asclepia sp.) is, in turn, controlled by continental isotopic patterns in rainfall for stable hydrogen and other climatic and physiological factors for stable carbon.” [ref.]
The farther north the monarch's birthplace is, the stronger the presence of the carbon isotope and the weaker the presence of the stable hydrogen isotope (full details in this figure). Using such markers Wassenar and Hobson estimated that 50% of the overwintering butterflies come from the dark shaded region in the map above – it's a large area just south the Great Lakes covering 9 American states – and 95% from the larger region, shaded light gray. Another important discovery was that the colonies, with a couple of exceptions, were panmictic or largely mixed – that is, it didn't matter where you were born, you could still roost and huddle together with butterflies that had been born elsewhere.
(The same isotopic technique has been used to determine the probable origin of dragonflies migrating across the ocean from India to East Africa; the results, if correct, are actually far more astonishing.)
A Counting Challenge
Since monarchs are tremendously popular – butterflies have always had it easy: congregations of flies, mosquitoes or cockroaches don't interest anyone – there's constant concern about the number of overwintering butterflies. Weather extremes, the reduction of the milkweed habitat that monarch larvae and caterpillars absolutely need, illegal logging in the oyamel forests of central Mexico (though this seems to have become less of an issue): all are potential threats to the migration. If the numbers fall below some critical threshold, the iconic journey of the monarchs might suddenly disappear.
Still, I was under the impression that there must be an accurate number estimate. Before my visit, I'd read on the internet that the 2014-15 estimate was around 57 million butterflies. At the El Rosario sanctuary – where the greatest masses of butterflies are and where most tourists go – I ran into Eduardo Rendon-Salinas, a Mexican researcher with the World Wildlife Fund. He was being interviewed by CNN. The camera-man wanted Rendon-Salinas to answer questions with thousands of swirling monarchs in the background. It was the closest you could get to a fairy tale setting.
After the interview ended, I chatted with Rendon-Salinas for a few minutes and mentioned the 57 million estimate. The careful scientist that he was, Rendon-Salinas smiled and shook his head: “You can never give a number like that so easily.”
Rendon-Salinas went on to explain how difficult counting the butterflies was. The only reliable estimate so far, he said, was an indirect one: the number of hectares occupied by the butterflies across all the colonies. Every winter since 2004-05, Rendon-Salinas and his colleague Omar Vidal, enclose the perimeter of each butterfly colony — there are 10-12 potential colonies in the region that the monarchs use — using trees at the periphery as markers [5]. The total enclosed area across all colonies is then calculated. The best time to make these measurements is between end of December and early February, the coldest time of the year when the monarch congregations are relatively stable.
A different team used a similar methodology to calculate the area occupied from 1994 to 2003 [6], prior to Rendon-Salinas' measurements. Putting these area estimates together (ref.), we see that the number of hectares occupied by monarch colonies in Mexico was at its lowest ever in 2013-2014 – 0.67 hectares. The values for the last three years in succession are by some measure lower than previous lows. The good news is that in 2014-15 the number has gone up slightly to 1.13 compared to 2013-14. But consider the 1996-97 value of 18.19 or even the 2006-07 value of 6.87 – the forests must have been aflame with orange!
Acknowledgements and References
1. A retired Canadian couple, Pat McDonell and Hank Wiechel, both experts in photography, happened to visit the butterfly sanctuaries the same time as I did. They were kind enough to let me use their high resolution images: the first picture and the second to last picture in this essay.
2. Oberhauser, K. S., & Solensky, M. J. (2004). Monarch butterfly biology & conservation. Cornell university press.
3. Chapter 18 of Oberhauser, K. S., & Solensky, M. J. (2004). Monarch butterfly biology & conservation. Cornell university press.
4. Quotes in the section on tags are from Fred Uruqhuart's National Geographic essay.
5. Vidal, O., & Rendón-Salinas, E. (2014). Dynamics and trends of overwintering colonies of the monarch butterfly in Mexico. Biological Conservation, 180, 165-175.
6. Garcia-Serrano, E. and J. Reyes, B. Alvarez. 2004. Locations and Area Occupied by Monarch Butterflies Overwintering in Mexico from 1993 to 2002. In Oberhauser, K. and J. J. Solensky, eds., The monarch butterfly: biology and conservation, pp. 167-176. Ithaca, NY, Cornell University Press.