Genetically, Central American mammoths were weird
The species’s boundaries in North America seem to have been fairly fluid.
We tend to lump all mammoths together as big, hairy elephant-like beasts with enormous tusks. But there were a number of mammoth species, including less furry ones that inhabited temperate regions. And the furry ones include at least three species: the Eurasian steppe mammoth, the Arctic-specializing woolly mammoth, and the late-evolving North America-only Columbian mammoth.
Because these species inhabited the Arctic, it has been remarkably easy to obtain DNA from them, providing a genetic picture of their relations. The DNA suggests that the woolly mammoth is an offshoot of the steppe mammoth lineage, and was the first to migrate into North America. But the Columbian mammoth was a bit of an enigma; some genetic data suggested it was also a steppe offshoot, while other samples indicated it might be a woolly/steppe hybrid.
But all of that data came from animals living in colder environments. In contrast, the Columbian mammoth ranged as far south as Central America. And now, a group of researchers has managed to obtain a bit of genetic information from bones found in the Basin of Mexico, which includes Mexico City. And these mammoths appear to form a distinct genetic cluster, and are all more closely related to each other than to any other woolly or Columbian mammoths.
Getting ancient DNA
DNA does not survive well in hot environments, which is why most of our picture of Columbian mammoths comes from regions where the species likely overlapped with its woolly contemporaries. These painted a somewhat confused picture. Data from the nuclear genome suggests that they’re a hybrid of steppe and woolly mammoths. But the mitochondrial genome, which is inherited from the mother via the energy-producing organelles found in every cell, suggested they had a distinct origin from woolly mammoths.
This led a Mexican-European research collaboration to get interested in finding DNA from elsewhere in the Columbian mammoth’s range, which extended down into Central America. The researchers focused on the Basin of Mexico, which is well south of where any woolly mammoths were likely to be found. While the warmer terrain generally tends to degrade DNA more quickly, the team had a couple of things working in its favor. To begin with, there were a lot of bones. The Basin of Mexico has been heavily built up over the centuries, and a lot of mammoth remains have been discovered, including over 100 individuals during the construction of Mexico City’s international airport.
In addition, the team focused entirely on the mitochondrial genome. In contrast to the two sets of chromosomes in each cell, a typical cell might have hundreds of mitochondria, each of which could have dozens of copies of its genome. So, while the much smaller mitochondria don’t provide as much detail about ancestry, they’re at least likely to survive at high enough levels to provide something to work with.
And indeed they did. Altogether, the researchers obtained 61 new mitochondrial genomes from the mammoths of Mexico from the 83 samples they tested. Of these, 28 were considered high enough quality to perform an analysis.
Off on their own
By building a family tree using this genetic data, along with that from other Colombian and woolly mammoth samples, the researchers could potentially determine how different populations were related. And one thing became very clear almost immediately: They were in a very weird location on that tree.
To begin with, all of them clustered together in a single block, although there were three distinct groupings within that block. But the placement of that block within the larger family tree was notably strange. To begin with, there were woolly mammoths on either side of it, suggesting the lineage was an offshoot of woolly mammoths. That would make sense if all Columbian mammoths clustered together with the Mexican ones. But they don’t. Some Columbian mammoths from much further north are actually more closely related to woolly mammoths than they are to the Mexican mammoths.
Drawing this all out on a map, you end up with a very strange situation. Rather than mitochondrial DNA being specific to a single species of mammoth, it appears to be linked to geographic location. At least based on the data we have, two mammoths are more likely to have similar mitochondrial DNA if they lived near each other than if they were the same species. Which, just to be clear, is not how genetics is supposed to work.
The researchers come up with two potential explanations for this. The first is that what we identify as the Columbian mammoth was the product of multiple hybridization events, each taking place at different locations and producing somewhat isolated Columbian populations. That would make Columbian mammoths less of a distinct species and more of a collection of hybrid populations that may have been kept somewhat isolated from each other by distance.
The alternative, which the researchers favor, is that the North American woolly mammoth population carried a lot of distinct mitochondrial lineages by the time any hybridization took place. As long as the hybridization event involved enough individuals, then some of these lineages would have ended up in the population that produced what became the Columbian mammoth.
Genetically, it’s a very weird situation, and would benefit from some nuclear DNA to give us a clearer picture of what this population looked like genetically. However, the level of success with getting much in the way of mitochondrial DNA was low enough that this is unlikely to happen. So, what may be needed is a more exhaustive look at the Columbian mammoths that remained further north, where DNA is more likely to have survived the millennia since their extinction.
Science, 2025. DOI: 10.1126/science.adt9651 (About DOIs).
John is Ars Technica’s science editor. He has a Bachelor of Arts in Biochemistry from Columbia University, and a Ph.D. in Molecular and Cell Biology from the University of California, Berkeley. When physically separated from his keyboard, he tends to seek out a bicycle, or a scenic location for communing with his hiking boots.
