Mastodons, Mammoths And More: DNA Shows Elephants And Their Ancestors Mixed Over Time

Using powerful new techniques for analyzing DNA, researchers have sequenced not just the genes of modern elephants but of their extinct ancestors and relatives as well: mastodons, woolly mammoths and straight-tusked elephants.

Dr. David Reich of the Broad Institute and Harvard helped lead the consortium that has just published its findings — the most sweeping analysis yet of elephants and their extinct relatives — in the Proceedings of the National Academy of Sciences. He says the findings show that elephant species have often mixed and mingled over the millennia. But also that current elephants in Africa have been isolated for so long that they should be considered two separate species, which could affect conservation efforts.

Our conversation, edited:

How would you sum up what you did?

This study started as the Elephant Genome Project more than a decade ago. We sequenced a high-quality genome of the iconic elephants — the savanna and African elephant — and it took us a long time to publish it. We decided eventually to publish it by not just presenting the genome of the elephant, but also presenting the genomes of all diverse living elephants, and also diverse extinct elephants from which we obtained ancient DNA.

This study ended up being a big ancient DNA study as well, where we report the first ancient DNA from the extinct Mastodon, which is a distant cousin of the elephant. Also, ancient DNA from the first Columbian elephants, which are an American temperate mammoth; and also a very high-quality genome of an ancient European straight-tusked elephant that was more than 100,000 years old. So it was really an amazing set of data, and we were able to analyze it to obtain a qualitatively different picture of elephant population history than we had before.

And what did you find?

The biggest take-home message is really that mixture of very different elephant populations occurred repeatedly over elephant history.

Before this work, I was involved in studies that took the very thin data we had and worked out the average relationships amongst elephants, figured out which groups were closest to which other groups. But with this high-resolution, whole-genome data, we can figure out that great mixture events have occurred in the past.

For example, the European straight-tusked elephant is a mixture of a minimum of three very divergent lineages: one related to mammoths and Asian elephants; another related to forest elephants from western Africa; and another related to the common ancestor of forest and savannah elephants both from Africa. So a huge mixture at multiple time depths is characteristic of the elephants.

But the mixing stopped?

In some elephants it stopped, and in some elephants it continues. I think that combined with this mixture are long periods of extreme isolation. And that's the pattern.

So for example, in Africa today there are two species of elephants that are genetically extremely distinct from each other. They hybridize at their boundaries and there are known hybrid groups. But if you look at the forest elephants from western Africa and the savannah elephants from South Africa and East Africa, there's been essentially no gene exchange amongst their ancestors for the last half million years.

That's one of the big implications of the study. Conservation laws usually try to focus on preserving at least some populations of the species, and African elephants were often, by some people, argued to be the same species because of these hybrid zones. So conservation was not being done everywhere in Africa but only to conserve some groups.

But we now know that the forest elephants in Western Africa, which are half the size of the savannah elephants, are very distinct and special, and need to be conserved just as much as the savannah elephants. In fact, they are even more threatened.

So they're mainly isolated now — but elephants and their ancestors also mixed a lot....

The real big message for us is the ubiquity of population mixture and isolation.  We see it not just in the straight-tusked elephants but we also see it in the Columbian and woolly mammoths in North America, where there's a gene exchange over a long period of time, with low proportions of Columbian mammoth ancestry in Alaska, and more toward the south where the Columbian mammoths actually lived. So there's a gradient of mixtures.

We see lots of evidence of mixture of this type. So I think what's really becoming clear is that in elephants, as in humans, and as in polar bears, we now know that mixture is ubiquitous in the population structure. It's an important, not a peripheral, part of our history, and the history of many animals and plants.

I think that this study, in addition to illuminating the iconic elephants, who are such a wonderful and amazing and special and isolated species group, also says something broader about Nature.

And current-day DNA technology let's you see it...

It's because of this boom of ancient DNA and genome sequencing that we can actually do this. And one of the things that's in the study is that these ancient elephants are really very special.

Elephants are a very isolated species group. Their closest relatives who are still alive are sea cows and manatees, which are many tens of millions of years separated from them. So they're kind of like leftovers from evolution — but they're so amazing. And most of the species actually have died out in the last 15,000 years: mammoths, Columbian mammoths, woolly mammoths, straight-tusked elephants — in the last 50,000 years, all of these important elephant species that were special in their own right are gone. But they are within the time that we can recover their genome sequences using this technology of ancient DNA, so we can get a look at them.

What would you say to those of us who are very eager to see a mammoth or mastodon cloned?

We now have genome sequences of about 70 or 80 percent of the mammoth genome. The parts that we couldn't sequence are in parts of the genome that are too broken up and confused to be able to figure out where the sequences we're obtaining map to. My personal feeling is that if we tried to create a cloned mammoth, it would really be very unfair to the poor individual, because they would certainly die of errors in the DNA sequence that were stuck into their genome.