Lazarus plant: 30,000-year-old flower resurrected from naturally frozen seeds

February 22, 2012 • 9:11 am

I won’t go on about this cool new paper at length, for it’s already been described by Ed Yong at Not Exactly Rocket Science as well as in a piece at The New York Times.  Still, it behooves us to know about it. The upshot is that a group of Russian scientists recovered from the Siberian permafrost a cache of seeds and fruits stashed by ancient squirrels, and managed to use tissue culture to regenerate plants from immature fruits.  The estimated age of the seeds is 32,000-30,000 years old, so this is clearly the most ancient organism ever “revived.”

The plant, a species that still exists (at least the morphological similarities suggest conspecific status), is Silene stenophylla. Silene (of which there are several species) is also known as “catchfly” or “campion”  (literate readers will recognize it as the flower with which Mellors the gamekeeper bedecked Lady Chatterly’s pubic hair in Lady Chatterly’s Lover).  It’s often used in evolutionary studies because some species have separate sexes while others do not, and ditto for sex chromosomes.  It could thus tell us something about the evolutionary origin of gender and gender-specific chromosomes.  Some species are also gynodioecious (i.e. some plants are “female” [male parts sterile], while other plants are hemaphroditic), and this could also give us a clue to how “male” versus “female” plants arose.

But I digress.  In a new paper published in the Proceedings of the National Academy of Sciencesby Svetlana Yashina et al., the authors describe finding a group of fossil burrows, 20-40 meters below the surface, in the permafrost of northeastern Siberia.  Some of these burrows contained as many as 600,000 seeds/fruits!  These were stashed by equally ancient ground squirrels of the species Urocitellus parryii:

The Arctic ground squirrel Urocitellus parryii, photographed by Ianaré Sévi.

Permafrost provides the dry and cold conditions needed to preserve seeds; that’s how they’re preserved in special seed banks.  Attempts to germinate the seeds failed, but they managed to grow one species of Silene by dissecting out the “placental tissue” (special tissue in the fruit to which the seed is connected), culturing it in nutrient media and then adding hormones (auxins, etc.), to induce formation of roots and shoots. And they got the plant to grow, flower, and, after cross fertilization with other ancient plants, set seed.  Here are two specimens of the plant grown from cultured ancient tissue:

The two articles cited above will give you more information.  What interests me most about this is that the “species” still exists, and this allows us to see how much evolutionary change has transpired in 30,000 years. (This is similar to the way that bacterial evolutionists can freeze an ancestral culture and then, after reviving it, compare it to its descendants that have undergone many generations of evolution).

The plant appears to have actually changed during those 30,000-odd generations.  (This is probably genetic rather than environmental change because the differences between ancient and modern plant are seen in the second generation of cultured ancient plants which have been produced by cross-mating them.)  The authors note the differences:

Thirty-six ancient plants (12 from each fruit) and 29 extant plants were morphologically tested. All ancient plants were morphologically identical. During vegetative development, the ancient and extant plants were morphologically indistinguishable from one another. However, at the flowering stage they showed different corolla shape: petals of extant flowers were obviously wider and more dissected (Fig. 3). Moreover, all flowers of the extant plants were bisexual (b) (Fig. 3A), whereas the primary flowers (two to three in number) of each ancient plant were strictly female (f) (Fig. 3 B and C, f), and then bisexual flowers were formed on each ancient plant (Fig. 3C, b).

For those botany geeks among us, here’s Figure 3 showing the differences (click to enlarge):

The one thing I really wanted to know, and which the authors didn’t study, is whether the ancient plants are reproductively compatible with the modern ones.  They crossed ancient plant with ancient plant, and showed that they cross readily, as of course do modern plants crossed with modern plants.  But they didn’t cross ancient plants with modern ones!  They need to do that.

If they found reproductive incompatibility in those crosses, that would suggest incipient (or full) speciation between ancestor and descendant, something that we rarely get to study because ancestor and descendants never get the chance to meet and mate (this would be like mating the 750,000 year old ancestors of Homo sapiens with modern H. sapiens, since a comparative number of generations have transpired).   And even if reproductive isolation didn’t evolve, one can still study the genetic basis of differences in petal shape and appearance of different kinds of flowers.  Ten to one the Russian team is doing this, and I look forward to the results.

h/t: Dr. John H. Willis III, esq.


Yashina, S., S. Gubin, S. Maksimovich, A. Yashina, and E. Gakhova. 2012. Regeneration of whole fertile plants from 30,000-y-old fruit tissue buried in Siberian permafrost. Proc. Nat Acad Sci. USA, published ahead of print February 21, 2012, doi:10.1073/pnas.1118386109

34 thoughts on “Lazarus plant: 30,000-year-old flower resurrected from naturally frozen seeds

  1. Maybe I’ve been reading too many creationists tracts, but I would wager that “All ancient plants were morphologically identical” will be quote-mined in the very near future and taken completely out of context. Any takers?

        1. What do you have against anuses and genitals that you want to slander them so?

          I happen to be very fond of both, but AiG can go jump in Gehenna.

  2. If they found reproductive incompatibility in those crosses, that would suggest incipient (or full) speciation between ancestor and descendant, something that we rarely get to study

    That is fascinating. Biologists often talk of speciation because of spatial isolation, but we don’t really see discussion of speciation due to temporal isolation. And this possibility further complicates the notion of species itself — if organisms designated by the ostensibly same binomial can’t interbreed, what does that mean for species designation?

    1. I’m not sure one could jump readily to the conclusion these are ancestor and descendant. It’s possible that the population of the ancient plants and the ancestors of the modern plants had already diverged by then, with one of the branches later becoming extinct.

      One would need to track changes at say, every 1,000 year interval and look for either progression or population replacement.

    2. Talking about speciation via temporal isolation requires that we come up with a corresponding diachronic definition of “species”. Reproductive isolation as a criterion is only coherent in a synchronic context.

      Diachronic definitions of “species” AFAIK are in two camps:

      1) Morphology (this is what paleobiologists tend to use) – if individuals look “different enough”, they belong to different species. This approach is understandable in a paleobiological context, but what constitutes “different enough” is arbitrary, and this view is incompatible with synchronic views of species (which are based, in one fashion or another, on the concept of discontinuity between groups of individuals – a concept that is incoherent in the context of reproductive continuity within a group of individuals through time);

      2) Phylogenetic – branching points in a phylogenetic tree are taken to indicate speciation events in which: a) the two daughter lineages are different species; b) the ancestral species is taken to be a different species from either daughter. This is logically necessary (if one of the daughters is identical with the ancestral species… which one? and why? answers to these questions lead us to a morphological criterion that is inevitably both arbitrary and logically incompatible with synchronic conceptualizations of species, as mentioned above.

      In neither case is interbreeding a reasonable criterion. All organisms form a reproductive continuity through time due to common ancestry; there are no reproductive discontinuities there to look at. You could arrive at a putative observation of reproductive discontinuity only through crude and artificial subsampling of the data (e.g., only comparing the individuals from 30,000 years ago and those alive presently, but not any of those in the intervening portions of the lineage(s)). However, a complete sampling would not yield a discontinuity, so we would know any putative species boundary to be purely artifactual. Such breeding experiments might be -very- interesting in other contexts, but not as a way of saying something meaningful about taxonomy.

      1. “Such breeding experiments might be -very- interesting in other contexts, but not as a way of saying something meaningful about taxonomy.”

        If they cannot be crossed at all, that would tell us the modern plants and these resurrected ones are not the same species, under any definition. The morphological differences could support species separation too. But, maybe they’ll be fully interfertile, or fertile with some incompatibility. We’ll see.

        I look forward to the results of crossing experiments, and I’ll guess that there will be reduced fertility in the ancient/recent crosses.

        The fact that they’re finding thousands of seeds/fruits in these dens suggests the possibility of true population samples of these Pleistocene plants (incl. others besides Silene?). That’s exciting.

    3. And this possibility further complicates the notion of species itself — if organisms designated by the ostensibly same binomial can’t interbreed, what does that mean for species designation?

      Isn’t that more just a nomenclatural problem (i.e., time to change it!) than a problem of species definition?

      1. Not really — the issue is what counts as a species across time, regardless of how we name them. More specifically, the issue is whether it makes sense to talk of a lineage as the same species when the individuals at two different points in time can’t interbreed. Or, conversely, if one can talk sensibly about the “speciation” of one lineage turning into another, without any branching.

        1. Doesn’t seem that much more difficult to me than the problem of a species def across space; e.g., ring species. Either way, in many cases it’s basically an arbitrary line. Yet another manifestation of biology’s inherent “messiness.”

  3. This really is exciting for us botany geeks. I hope they have done or will do some sequencing. We can learn a lot about mutation rates of common markers (like ITS), not to mention the frequency of “natural genetic engineering”, by comparing the genomes of the ancient and modern plants.

  4. JAC: “But they didn’t cross ancient plants with modern ones! They need to do that.”

    Then of course sequence both genomes and do an alignment.

  5. Flipping interesting! Thanks. I had no idea this kind of thing was being done. Seeing a pic of a 30 000 year-old plant is quite an experience.

  6. Not that Jerry doesn’t know this but:

    The problem with comparing modern with ancient plants is that the latter are probably not the very individuals that contain the particular copies of the genes that were ancestral to the genes in the modern plant. The genes in the modern plants are descended from copies in other individuals in those ancient populations. Or maybe from copies in other populations of the ancient species.

    So you can’t just look at the modern sequence and the ancient sequence and count differences.

    Similar, a gene copy in a modern Egyptian and a copy of the same gene in a mummy probably last shared an ancestral copy about 100,000 years ago. They will have accumulated a lot more differences than we would expect to occur in a mere 3,000 years.

    1. “The genes in the modern plants are descended from copies in other individuals in those ancient populations. Or maybe from copies in other populations of the ancient species.”

      You can overcome ascertainment bias by increasing sample size, no?

    2. Joe,

      What I’m assuming is that there would be fixed differences at some loci in the modern plants that arose via selection in the last 30,000 years and differ from all genotypes in the “fossil plants”; and, further, that those differences cause reproductive isolation. My suggestion is to look for the phenomenon of reproductive isolation (just like they saw phenotypic differences); I wasn’t suggesting that they do this by comparing DNA sequences. (That, of course, would still be interesting!)

  7. I’m surprised that squirrels would use Silene seeds as a food source, as they are quite small. Do modern squirrels still stash S. stenophylla seeds?

    1. Excellent point. I wonder if tundra vegetation tends to have fewer of the “meatier” type seeds, so that tundra ground squirrels have to cast a wider net? And it would seem the seeds would have to be gathered while still in the pod to make it worth the effort at all…

  8. Ha ha! In the midst of my wonder of a 30,000-year-old seed being discovered (and newly blossoming), my next thought was: “I wonder how the 6,000-year-old (dinosaurs co-habitating with Adam and Eve) Creationists are going to explain away this flower?”

    And one of the first comments I see on this thread is: “…I expect Answers in Genitals to throw a tantrum…”

    I love this site!

  9. Ha, I nearly missed that fascinating variation/speciation issue since I decided to forgo that press release.

    But I didn’t miss the equally fascinating example of sympiatric speciation in archaea from a hot spring:

    “The analysis revealed two distinct groups of S. islandicus among the 12 strains. The microbes were swapping genes with members of their own group more than expected, but sharing genes with the other group less than expected, Whitaker said. And the exchange of genetic material between the two groups was decreasing over time.

    This indicates that the two groups are already separate species, even though they share the same habitat, Whitaker said. The differences between the two groups were slight, but speciation was clearly under way, she said.”

    Apparently that has important consequences:

    “”That means there are orders of magnitude more species of microbes than we ever thought there were,” she said. “And that’s kind of mind-boggling.””

    Incidentally, the spring was (historical) Siberian too.

    1. Oh, I forgot: apparently the trick that makes it work is that the swapping is uneven over the chromosome.

  10. Gah. Science News quotes Jane Shen-Miller of UCLA as saying “This is like regenerating a dinosaur from tissues of an ancient egg.”

    Well, maybe, if you’re willing to ignore three orders of magnitude difference in timescale.

    The Science News piece also makes it sound like the squirrels actually dug burrows 38 meters deep. This reinforces my opinion that editorial standards have really slipped there in the past couple of years.

  11. Yeah but it is still a Silene. LOL.

    Interesting paper and quite a lot of work to regenerate those plants from cell cultures.

    What struck me is how deep those arctic squirrel burrows were. I didn’t assume the squirrels were digging 20-40 meters though. They must have been buried in newer sediments over the decimillennia.

    So the scientists must have dug holes or trenches 20 to 40 meters (60 to 120 feet) deep. That would be a major effort in itself even in civilized areas. In Siberia everything is always much harder.

  12. Definitely noteworthy work.

    Even given all the caveats, DNA sequencing is fast and easy these days and they really should look at mitochondrial, chloroplast, and genomic sequences.

    What would be really cool would be resurrecting extinct plant species. 30,000 years is a long time, an ice age came and went, and there might well be some in those burrows.

    They’ve done that with viruses. We have resurrected the 1918 flu virus from sequence data. We have also resurrected an ancient human retrovirus, Phoenix from molecular fossils in our genome. I forget the exact age of Phoenix, 5 million years or 15 million or some such.

  13. Unfortunately, the article is behind a pay wall, so I wasn’t able to find answers to some questions that occurred to me. First of all, how many different species of seed were found in those fossil burrows, and in which proportions? Did the authors try to regenerate any other species (if there was more than one)?

    Secondly, how morphologically variable is modern Silene stenophylla? From what I have seen so far, the authors are just comparing two individual clones. That doesn’t provide much information.

    In the Flora of the USSR, S. stenophylla is said to have pink flowers, and the illustration that goes with it looks rather different from the thing that is labelled as the extant S. stenophylla in the article. From this I gather that S. stenophylla may well be somewhat variable. In other words, there could exist modern populations that more closely resemble the resurrected plants than the one illustrated by the authors.

    As others have pointed out above, there is no reason to assume that the modern population on or near the site is derived from the one that lived there 30,000 years ago. The intervening glacial period may have wiped out the species locally, and the plants that are found there today may well have recolonised the area during more recent times. The nearest relatives of the resurrected plants should perhaps be sought elsewhere within the distribution area of S. stenophylla.

    1. For ‘two individual clones’ read ‘two individual populations’ (or perhaps more accurately, ‘one individual clone and one population’).

Leave a Reply