Repost: Inside the Columbian Mammoth, Signs of a Woolly Cousin

[I’m out in Grand Staircase-Escalante National Monument for some late season fieldwork this week. Here’s an essay from the archive, originally posted November 28, 2011.]

I’m picky about my paleo t-shirts. If I’m going to lay down twenty five bucks for fossil-themed apparel from a museum gift shop, I want the long-extinct critter on the tee to look reasonably accurate, and geeky humor is always a bonus.

My standards led me to an indecisive moment at the Page Museum gift shop at the La Brea asphalt seeps during my visit to Los Angeles in September. One of the souvenir shirts had the line “What happens in the La Brea Tar Pits stays in the La Brea Tar Pits” with a fluffy mammoth stuck in the black morass. I loved the joke, but my inner scientist – nit-picky bastard that he is – chimed in with “But there aren’t any woolly mammoths here. They’re all Columbian mammoths!” I thought to myself, “Self, you have a point.” Despite being the iconic symbol of the Ice Age, woolly mammoths (Mammuthus primigenius) were the shaggier mammoths of the cold northern steppe that had not made it down to LA. The La Brea asphalt seeps were within the realm of a different, closely-related species known as the Columbian mammoth – Mammuthus columbi. What’s a nerd to do?

I bought the shirt. Accuracy be damned, I was just too tickled by the joke. I thought that was the end of it, but then I met with American Museum of Natural History paleontologist Ross MacPhee in Las Vegas, Nevada at the 71st annual Society of Vertebrate Paleontology meeting earlier this month. If new research being conducted by MacPhee and colleagues is correct, then the cartoon mammoth on my shirt may be a little bit of a woolly mammoth, after all.

Both the woolly mammoth and the Columbian mammoths were New World branches of an elephantine tree rooted in the Old World. The first mammoths – the earliest representatives of the genus Mammuthus – evolved between four and 5 million years ago in Africa. From there the early mammoth populations spread north out of Africa and into Eurasia around 3.5 million years ago. Sometimes the wandering populations were adapted into new species in their isolation from others of their kind.

Paleontologist Adrian Lister and colleagues outlined the big picture of Eurasian mammoth evolution in 2005. Contrary to previous hypotheses, the straight “ladder” model of one mammoth species gradually transforming into another – a sole chronological lineage through time – didn’t fit. Instead, the spread of the fossil evidence indicated that new mammoth species originated when populations of the parent species became isolated and were rapidly adapted into new forms. (This is the “punctuated equilibria” pattern paleontologists Niles Eldredge and Stephen Jay Gould championed.) Mammoth species sprung up from cordoned-off populations, co-existed with their parent species for a time, and then spread over the landscape just as their ancestors had.

Among all this splitting, Lister and collaborators hypothesized, around 200,000 years ago the woolly mammoth (M. primigenius) descended from a population of the larger steppe mammoth (usually called M. trogontherii, but changed to M. armeniacus by paleontologist Nancy Todd in an Anatomical Record paper last year). Woolly mammoths later moved to North America by way of the connection between prehistoric Russia and Alaska, but the steppe mammoth may have had another descendant. Although the older “ancestral mammoth” M. meridionalis has not been ruled out as a candidate for the rootstock of North America’s peculiar species, the Columbian mammoth may have been a descendant of a population of steppe mammoths which trundled their way to ancient Alaska.

Both the woolly and Columbian mammoth could have split from different populations of a common ancestral species. (And, even if they descended from different species, both were still mammoths that shared a common ancestor less than two million years ago.) Despite their differences in appearance and their preference of different habitats – the larger Columbian mammoths in the temperate savannas to the south, and the smaller woolly mammoths in the cold steppes of the north – the two species were actually quite closely related. According to the new research by MacPhee and his collaborators, the two may have been so close that they could still successfully interbreed.

MacPhee collaborated with ancient DNA expert Jacob Enk and a team of other paleontologists and paleo-geneticists to investigate the heritage of the Columbian mammoth. They did not discover what they had expected. The initial study found that the anatomical differences between the woolly and Columbian mammoths were not apparent in their genes. In fact, the Columbian mammoth sampled in the project fell neatly within the genetic diversity of woolly mammoths from North America. Genetics grated against what had seemingly always been apparent in the fossil data.

Mitochondrial DNA – the genetic material located within cell organelles called mitochondria, distinct from DNA in the nucleus of the cell – was the focus of the work by Enk and co-authors. Columbian and woolly mammoths lived close enough to us in time that such genetic data can be extracted, analyzed, and compared – the genetic mammoth library is constantly growing. For this particular study, the scientists drew out the entire mitochondrial genome of the Huntington mammoth – a roughly 11,200 year old Columbian mammoth recovered from Utah in 1988 – and compared the beast’s DNA sequence to the mitochondrial genomes of woolly mammoths in an attempt to see how the different species might be related. Most of the previous genetic work focused on woolly mammoths from Siberia and the area of the Bering Strait. There was the need to see how mammoths from the interior of North America fit into the emerging tangle of relationships.

On an anatomical basis, woolly and Columbian mammoths would be expected to be cousins which diverged from a common ancestor sometime between one and two million years ago. This is not what the genetic investigation found. “[T]he Huntington mammoth mitogenome is largely indiscernible from those of endemic North American WMs [woolly mammoths]”, Enk and co-authors wrote. The genetic readout of the Utah mammoth fell deep within the genetic diversity of woolly mammoths previously sampled from Alaska. This did not appear to be a case of contamination or mistaken identity – at a genetic level, researchers could barely distinguish a Columbian mammoth found in Utah from Alaskan woolly mammoths. What could this mean?

A sample size of one isn’t very good for drawing out large-scale conclusions. Enk and co-authors were tentative about what their result might indicate. Perhaps Columbian mammoths somehow retained mitochondrial DNA similar to that in woolly mammoths despite how different the two species looked. Some sort of genetic retention or convergence is a possibility. Then again, maybe the Huntington mammoth hinted at some inter-species amorousness.

Columbian mammoths and woolly mammoths were ecological neighbors, much like the modern savanna and forest elephants of Africa (Loxodonta africana and L. cyclotis, respectively). We know that these modern elephants sometimes interbreed where their territories meet – why should mammoths be any different? Individuals of one species could have taken a stroll into the typical habitat of the other and taken up with the locals. But it wasn’t just sex. One mammoth species was successfully contributing genetic material to the other, indicating that the brief unions were resulting in fertile mammoth offspring. The Columbian and woolly mammoths were closely related species that had not been diverging from each other for a very long period of time. Successful interbreeding between the two is not an unreasonable hypothesis.

In one scenario, Enk and collaborators suggest, some Columbian mammoths could have made themselves at home within woolly mammoth populations – their mitochondrial genomes may have become established within woolly mammoth populations through interbreeding and subsequently spread. This would explain why the Columbian mammoth had mitochondrial DNA so much like that of an Alaskan woolly mammoth. The reverse could have also happened, with woolly mammoths meeting up with Columbian mammoths along the boundary between cold steppe and temperate savanna. Connecting lines are easy to draw when data points are few.

As an indirect argument in favor of such previously-undocumented hanky-panky, the researchers cite problematic mammoth specimens that seem to combine traits of both Columbian and woolly mammoths. These specimens have often been given their own distinct species names such as Mammuthus jeffersonii. Some of these are found along the hypothesized geographic boundary between woolly and Columbian mammoths – such as the Great Lakes region – and may turn out to be hybrids. Even the Huntington mammoth, with its distinct anatomy but perplexing genetic makeup, might be a hybrid. The trouble is quantifying variation within a species and determining how many species are truly present in that sample. Biologists argue enough about this when they have the living animals to study – the problem is doubly difficult when studying extinct animals. Identifying between a variant of a species, a member of a different species, or a hybrid of the two is a thorny task.

Enk and collaborators note the need for further work to sort out what was actually going on. One animal is not enough to figure out whether the genetic similarities between the mammoths were the result of interbreeding or some phenomena relating to how genes are lost or conserved in lineages over time. Rather than making the picture of North American mammoth relationships clearer, the results obtained from the Huntington mammoth raised a slew of perplexing questions.

MacPhee – one of the authors on the paper – delivered a presentation about the research at SVP. I missed the actual talk, but I was fortunate enough to talk to MacPhee about the project at the conference’s press event.

MacPhee expressed how genetic studies are changing our understanding of mammoth biology and ecology. The nice, neat scenario of two distinct species living in entirely different environments doesn’t work anymore. Citing the previous work of co-author Hendrik Poinar and his lab, for example, MacPhee noted that genetic studies are finding “subclades [genetically-related groups] of woolly mammoths that you couldn’t find morphologically.” Anatomical and genetic data don’t always match the way paleontologists expected them to, and these discordances offer opportunities to investigate aspects of mammoth lives that were previously obscured.

The initial Columbian mammoth study hinted that there might have been something strange going on in Ice Age North America. Ongoing work has deepened the mystery. Since the time of the published study, MacPhee said, “The Poinar lab has done another six Columbian mammoths. All fall within the woolly mammoth subgroup.” Maybe these are all hybrids, or maybe there’s some other explanation dealing with the nature of genetic change in populations, but the fact that all the Columbian mammoths are coming up “woolly” raises another possibility. Maybe there wasn’t a distinct Columbian mammoth species.

“The idea that there were many mammoth species [in North America] gets harder and harder to maintain”, MacPhee said. On the basis of our present knowledge of mammoth genetics, there might be “one highly polytypic taxon” – or, in other words, just one species of extremely-variable mammoth which spread from Alaska to Utah, Wisconsin, and elsewhere within the North American interior. Studies like these may even change the way paleontologists categorize prehistoric species when both anatomical and genetic information is available. “We will have to alter our way of thinking to graded lineages rather than boxes,” MacPhee said. Conceptual boxes help with organization, but studying organisms requires a more flexible understanding which takes into account variation due to age, disease, geographic location, and other aspects of natural history.

Nothing is set in stone just yet. Additional research and analysis is required. Though MacPhee deemed the possibility unlikely, there’s a chance that the common mitochondrial DNA among the mammoths might be attributed to some quirk of genetics, and better genetic sequences from the DNA in mammoth cell nuclei are needed to see how the already-gathered mitochondrial data compare. Likewise, possible hybrid specimens – such as so-called “M. jeffersonii” individuals – have not yet been studied. We need a better understanding of woolly mammoth and Columbian mammoth populations through time and across space in North America. Paleontologists and paleo-geneticists are really just starting to put together a picture of how these animals lived and interacted.

For the moment, though, MacPhee, Enk, Poinar, and collaborators are most prominently promoting the interbreeding hypothesis. The genetic connection between the southern mammoths and the northern mammoths appears to be strong. What this means for me and my t-shirt isn’t quite clear yet. Maybe the Columbian mammoths of La Brea had some contributions from woolly mammoths deep within their mitochondrial DNA, or, perhaps, those big mammoths stuck in the asphalt seep should really be recategorized as a southern variety of woolly mammoth. I don’t know. At least, when I wear that shirt, I’ll have something geeky to puzzle over with anyone else pedantic and paleo-minded enough to listen.

[For another La Brea identity crisis, see my post on the American lion Panthera atrox.]

References:

AGENBROAD, L. (2005). North American Proboscideans: Mammoths: The state of Knowledge, 2003 Quaternary International, 126-128, 73-92 DOI: 10.1016/j.quaint.2004.04.016

Enk, J., Devault, A., Debruyne, R., King, C., Treangen, T., O’Rourke, D., Salzberg, S., Fisher, D., MacPhee, R., & Poinar, H. (2011). Complete Columbian mammoth mitogenome suggests interbreeding with woolly mammoths Genome Biology, 12 (5) DOI: 10.1186/gb-2011-12-5-r51

LISTER, A., SHER, A., VANESSEN, H., & WEI, G. (2005). The pattern and process of mammoth evolution in Eurasia Quaternary International, 126-128, 49-64 DOI: 10.1016/j.quaint.2004.04.014

Todd, N. (2010). New Phylogenetic Analysis of the Family Elephantidae Based on Cranial-Dental Morphology The Anatomical Record: Advances in Integrative Anatomy and Evolutionary Biology, 293 (1), 74-90 DOI: 10.1002/ar.21010