Ancient ecosystem reconstructed using fossil DNA

December 9, 2022 • 10:30 am

The oldest DNA sequenced up to now was from a mammoth molar preserved in permafrost, and was dated about 1.2 million years ago. Now a group of scientists, excavating a 100-meter-thick layer of frozen soil in the “polar desert” of northern Greenland, not only found short stretches of DNA that identified the plants, animals, and algae present a long time ago, but also showed that that the time was at least two million years ago.

This is the oldest fossil DNA ever sequenced; it was preserved because it had been adsorbed to minerals in frozen soil. And although the stretches of DNA had degraded into short bits—about 50 base pairs long—they were sufficiently similar to modern taxa that they could identify the groups from which they came. In fact, they could reconstruct the whole ecosystem of that area 2 million years ago. It was much richer in flora and fauna than today’s polar desert, for at that time Greenland wasn’t covered with ice, it was much warmer (mean summer temperature about 10°C), and organisms could migrate to Greenland over land bridges. This might give us a hint of what kind of ecosystem could develop (minus the animals, which are largely gone) should global warming melt the ice presently in Greenland.

You can read the Nature paper for free by clicking on the screenshot below (the pdf is here, reference at the bottom). Below that is a clickable and short popular account of the findings, also published in Nature.

The News article for tyros (short; click to read):

Here’s the location of the area analyzed in northern Greenland, Kap København, where the layer of soil occurred (yellow star). The layer’s presence was already known, and some of the samples had been dug up in 2006 and had been sitting in a Copenhagen freezer for 16 years. Somebody had a bright idea to see if they could identify and sequence the DNA in that soil, and it worked!

(from the paper): a. Location of Kap København Formation in North Greenland at the entrance to the Independence Fjord (82° 24′ N 22° 12′ W) and locations of other Arctic Plio-Pleistocene fossil-bearing sites (red dots). b, Spatial distribution of the erosional remnants of the 100-m thick succession of shallow marine near-shore sediments between Mudderbugt and the low mountains towards the north (a + b refers to location 74a and 74b).

Small stretches of DNA were sequenced and compared to modern DNA as well as DNA inferred in ancestors of modern taxa. The DNA had of course degraded, but they found stretches about 50 base pairs long. Comparisons were mostly to mitochondrial DNA for animals and to conserved chloroplast or other plastid DNA from plants. (They also found ancient pollen that they used in conjunction with the DNA data.)

On the right you can see what animals were found, mostly identified to genus or family because there wasn’t enough DNA to do a finer analysis. I’ll put a list of what they found below this figure:

(from paper): Taxonomic profiles of the animal assemblage from units B1, B2 and B3. Taxa in bold are genera only found as DNA

Here’s what they found from the DNA; these were all organisms living roughly at the same time about 2 million years ago. And remember, that area now harbors very little life.

A mastodon! The figure below shows its placement on the phylogenetic tree of elephants.

70 genera of vascular plants, including sedges, horsetails, willows, hawthorns, spruce, poplars, yew, and birch. Some of these no longer grow in Greenland, but the mixture of plants includes those found in much warmer habitat. See the paper for a full list.

Algae, fungi, and liverworts

Marine phytoplankton and zooplankton

A hare

A caribou-like cervid (caribou are another name for reindeer). How did they get to Greenland? Presumably it wasn’t an island then, but we don’t know for sure.

A bird related to modern geese

A rodent related to modern lemmings

Reef-building coral

An ant

A flea

A horseshoe crab (identified as Limulus polyphemus, the modern horseshoe crab, regarded as a living fossil). These days Limulus doesn’t breed north of the Bay of Fundy (about 45° N), while the location of this site was 82° N. That shows how much warmer it was in Greenland then, though of course the crabs could have evolved in the last several million years to be acclimated to warmer waters.

There were no carnivores found; all the animals were herbivores. That doesn’t mean that there weren’t carnivores there, but I doubt it.


(From paper): b, Phylogenetic placement and pathPhynder62 results of mitochondrial reads uniquely classified to Elephantidae or lower (Source Data 1). Extinct species as identified by either macrofossils or phylogenetic placements are marked with a dagger.

The upshot: Well, we know how that DNA sequences can be preserved for twice as long as we thought, though it has to be under very special circumstances. More important, if you find areas (and they’ll have to be in cold regions) where you can extract even small sequences of fossil DNA, you might be able to reconstruct whole ecosystems. What we’ve found are animals and plants that weren’t expected to be there (reindeer, horseshoe crabs, hawthorns) and so on—species adapted to warmer habitats or now found in areas not in Greenland.

There are two explanations for this: the related today have lost their adaptations to cold habitats when they were forced out of Greenland as the ice caps formed, or the climate was simply warmer. (Of course, both could apply.) But know the latter is surely a contributing factor from independent evidence about climate. Still, there could have been some evolutionary change in thermal tolerance as well, something for which we can’t really get evidence.

But these different explanations aren’t that important: what is important is that we’re able to reconstruct entire ecosystems from fossil DNA—DNA twice as old as previously known. I’ll let the authors have the last word (from the paper):

No single modern plant community or habitat includes the range of taxa represented in many of the macrofossil and DNA samples from Kap København. The community assemblage represents a mixture of modern boreal and Arctic taxa, which has no analogue in modern vegetation. To some degree, this is expected, as the ecological amplitudes of modern members of these genera have been modified by evolution. Furthermore, the combination of the High Arctic photoperiod with warmer conditions and lower atmospheric CO2 concentrations made the Early Pleistocene climate of North Greenland very different from today. The mixed character of the terrestrial assemblage is also reflected in the marine record, where Arctic and more cosmopolitan SMAGs of Opistokonta and Stramenopila are found together with horseshoe crabs, corals and green microalgae (Archaeplastida), which today inhabit warmer waters at more southern latitudes.

. . . In summary, we show the power of ancient eDNA to add substantial detail to our knowledge of this unique, ancient open boreal forest community intermixed with Arctic species, a community composition that has no modern analogues and included mastodons and reindeer, among others. Similar detailed flora and vertebrate DNA records may survive at other localities. If recovered, these would advance our understanding of the variability of climate and biotic in

Will northern Greenland be like this again should global warming continue? I doubt it, for many of the species, like caribou, can no longer get there, and some, like mastodons, are simply extinct. But it’s enough to know what was there two million years ago.


Kjær, K.H., Winther Pedersen, M., De Sanctis, B. et al. A 2-million-year-old ecosystem in Greenland uncovered by environmental DNANature 612, 283–291 (2022).

17 thoughts on “Ancient ecosystem reconstructed using fossil DNA

  1. Magnificent, fascinating!

    This is the danger (hear me out) of Nature – instead of reading research I’m _supposed_ to be focusing on, if I notice just one article like this while browsing – there goes my “productivity”!

  2. Very cool! One would normally sift thru the sediments to collect and identify pollen and spores and whatever bits of debris. But this adds to the toolkit, certainly.

    1. I read the paper. Only very few species/genera (IIRC, the mastodon was one of them) could be added via DNA to what was already known from macrofossils and pollen.

  3. Hang on – not a mammoth surely – rather a mastodon?

    “Most notably, we found reads in unit B2 and B3 assigned to the family Elephantidae, which includes elephants and mammoths, but taxonomically not mastodon (Mammut sp.)—which are, however, in the NCBI taxonomy, and therefore our analysis reads classified to Elephantidae or below therefore include Mammut sp. A consensus genome of our Elephantidae mitochondrial reads falls on the Mammut sp. branch (Fig. 4b) and is placed basal to all clades of mastodons. However, we note that this placement within the mastodons depends on only two transition single-nucleotide polymorphisms (SNPs), with the first one supported by a read depth of three and the second by only one …
    Furthermore, we attempted dating the recovered mastodon mitochondrial genome using BEAST49. We implemented two dating approaches, one was based on using radiocarbon-dated specimens alone, while the other used radiocarbon- and molecular-dated mastodons. The first analysis yielded a median age estimate for our mastodon mitogenome of 1.2 Myr (95% HPD: 191,000 yr–3.27 Myr), the second approach resulted in a median age estimate of 5.2 Myr (95% HPD: 1.64–10.1 Myr)…”

    Somewhat confused!
    Can you clarify?

      1. Oh ta! I thought maybe they meant that the DNA was nearer the root of where Mastodons & Mammoth divide… I sometimes find cladistics very confusing, & that paragraph is hard to follow.

    1. PS There are Barren-Ground Caribou in West Greenland (the eastern & north-eastern ones are extinct), so as glaciers melt they could spread there. Those in Svarlbard got there when the sea was much lower. As it is, humans could move them today – we do it all the time, sometimes to good, sometimes ill-effect!

      It is a pity we have no predator DNA at present. I suppose as they are rarer, there is less DNA to be preserved.

  4. I do love this sort of thing. I would love to learn more about extinct ecosystems, not just an animal or plant fossil in isolation, or to know more about species interactions that have lost a link in the chain somewhere. Living in Missouri, the Osage Orange, Maclura pomifera, grows in my yard and it’s hard to see those big green fruits (alien brains to my 7yr old imagination) and not wonder about ground sloths and mammoths and who knows what else. Neat stuff.

  5. Lots of interesting finds from more recent times – entire animals such as a cave bear:

    and a puppy of some sort:,location%20he%20was%20found%20in.

    and cave lion cubs etc. And, of course, much plant material.

    Fossil amber yields fascinating stuff. About 18 months ago an article was published in the media, looking at amber from Costa Rica, between 40 million and 20 million years old. I will look for it and read again. Apparently, very little change in the ecosystem over the 20 million year gap, either in insect or in plant life. The major difference was the presence of hairs from predatory cats in the younger amber, whereas not in the older material.

    Why so little evolutionary change? I am not an evolutionary biologist but possibly the ecosystem of 40 million years back was characterized by great stability, and factors that drive evolution were unusually weak?!

    Any thoughts?

    Also – the angiosperms (flowering plants) are a subject of deep fascination! They are thought to have emerged in the deep Cretaceous, about 130 million years ago. At the start they were mainly freshwater aquatic, some of them resembling water lilies of today.

    However, fossils of angiosperm-like plants appear in the Triassic, perhaps as much as 270 million years into the past. Tantalizing!


  6. Fascinating stuff! This sort of information would have been very challenging to myself years ago when I was a creationist. It was learning about things like this that slowly over time eroded my creationist beliefs. Reality is so much more interesting than fairy tales. 🙂

  7. The extraction technique had to be developed to suite to the same polar bindings to silica et cetera that has preserved the DNA longer than previously seen.

    The irony, to me, is that this comes just a few days after the local media pushed the previous record claim of a 1.65 Myrs old mammoth DNA among three Siberian mammoths about 1 Myrs old. That team numbered Swedes as well, just in time with Svante Pääbo’s Nobel Prize award [ , ].

    1. Going back to the beautifully-preserved armored dinosaur (nodosaur) of my previous comment, it is fascinating that in a few cases we can infer color – in this case red.


      It says that the authors discovered chemical traces of pheomelanin, the same pigment that gives redheads their hair color, within the dinosaur’s fossilized hide. The nodosaur was darker red and brown on top than on the bottom. This pattern, seen today in deer and antelope, obscures a creature’s silhouette.

      From memory of prior reading of CNN and BBC science articles etc – in other cases, especially fossil birds, we can infer color from melanocytes (cells within the skin, eyes or feathers that produce and contain the pigment, melanin). Ancient birds and small dinosaurs may often have been highly colored. In birds, reds and blues may have been common (like some ducks) and, in small dinosaurs, oranges and blacks may have been common.

      It is fascinating to conjecture that colors may have evolved to be somewhat similar to today’s species. For example, could ocean-going pterosaurs have been grey on top and white below, like todays seagulls? Could carnivorous marine reptiles have been grey or blue on top and whiter underneath?

Leave a Reply