Paper on “living fossils” finds recent radiation, but misses the point

October 23, 2011 • 7:15 am

Cycads are a group of plants that resemble tree of ferns or palm trees.  But they’re not closely related to either: rather they constitute an ancient group of gymnosperms (naked-seeded plants) that originated around 300 million years ago.  They reached their peak of abundance during the Jurassic, the age of the dinosurs, when they were abundant throughout the world.  Now, though, they are restricted to tropical and subtropical areas, and most of them are rare and endangered, though some are grown as ornamentals.

Cycads are one group of organisms known as “living fossils,” because the living species highly resemble ancient ones (other “living fossils” include the coelocanth, horseshoe crabs, and the tuatara of New Zealand).  In other words, living fossils show the “morphological stasis” so beloved of Steve Gould and other advocates of punctuated equilibrium.

The term “living fossil”, by the way,was coined by Darwin in Chapter 4 (“Natural selection”) of The Origin (my emphasis):

And it is in fresh water that we find seven genera of Ganoid fishes, remnants of a once preponderant order: and in fresh water we find some of the most anomalous forms now known in the world, as the Ornithorhynchus and Lepidosiren, which, like fossils, connect to a certain extent orders now widely separated in the natural scale. These anomalous forms may almost be called living fossils; they have endured to the present day, from having inhabited a confined area, and from having thus been exposed to less severe competition.

Note that Darwin attributed their long-term morphological stasis to a lack of competition: Darwin always thought that most natural selection resulted from competition between individuals of a species or between members of different species.

Other explanations for the stasis, have, however, included a lack of genetic variation (you can’t evolve if there’s no variation in your genes), or the possibility that the group either lives in a constant environment or seeks one out, so there is no selection to produce change.

Both of these explanations have had problems. The first—lack of genetic variation—is almost certainly wrong, for surveys of genetic variation in “living fossils” like horseshoe crabs show that they’re just as variable as species that have changed more over time.

The “constant environment” explanation, which seems more plausible to me, suffers from the fact that some living fossils seem to live in the same environments as species that have evolved more rapidly. And some, like cycads, can’t behaviorally seek out the environments to which they’re adapted.

We should not, however, think that just because “living fossils” look like their ancient relatives, that they haven’t changed.  When we compare extant with ancient species we see largely the external morphology: the hard parts that are on the outside and replaced by minerals. Perhaps some “ancient fossils” have changed extensively on the inside since their origin—either in the internal anatomy or physiology, or simply via wholesale DNA changes that were largely caused by random genetic drift.  Still, the constancy of external morphology over hundreds of millions of years is an evolutionary puzzle.

In our book Speciation, Allen Orr and I note that “living fossils” such as gingko trees and horseshoe crabs not only haven’t changed much on the outside, but never were particularly speciose: that is, the unchanging lineages also didn’t produce a large number of species over their evolutionary history. (It is possible for a group to produce lots of species but not change very much in external appearance. That just doesn’t seem to have happened.)

Here’s a cycad showing its “cone,” which resembles that of a pine tree (another gymnosperm):

Cycas circinalis (from Wikipedia)

A new paper in Science by Nagalingum et al. sheds some light—or purports to shed some light—on the puzzle of living fossils.

There are now about 300 living species of cycads in three families, and the authors used DNA from both the nucleus and chloroplasts to a) make a family tree of 199 of the species, and b) estimate when the living species diverged from each other.  Estimates of divergence time are made by using calibrated “molecular clocks,” which assume (and there is support for this), that divergence in DNA sequence is proportional to the absolute time in the past when species diverged.  With calibration, then, we can get a decent handle on when living species of cycads diverged from each other.

The “family tree” of cycads yielded no surprises: the families, and relationships between species within families, were pretty much related in the same way that previous data based on DNA and morphology had indicated.

But the surprise was this:  all living cycad species formed recently: within the last 5 to 12 million years. That means that although the group itself hasn’t changed much in appearance in 300 million years, the species we have today (i.e., the lineages that are reproductively isolated from one another) are relatively new.  That is, reproductively isolated lineages have arisen much more recently than previously thought, and the species we see today are not identical (in the sense of being reproductively compatible) with their ancient ancestors. Nor are they the products of a single unchanging lineage that extends back to ancient times.

That is a surprise, but doesn’t, I think, have much bearing on the question of “living fossils,” that is, the question of why some lineages don’t change in their appearance over long periods of time.  

And it’s clear that cycads don’t change much over millions of years, at least judging by the authors’ own statement highlighting the “morphological conservatism” of the group.   So that mystery remains.

The authors do note that cycads show “low levels of genetic diversity” compared to other groups, so it’s formally possible that their morphological stasis is due simply to a lack of genetic variation.  But that is contradicted by the fact that they’ve formed so many new species in the past few million years—for the formation of new species requires genetic variation.  No, the explanation for the morphological conservatism of cyads—and of other living fossils—probably lies elsewhere.

Another mystery remains: why was there a relatively sudden radiation of cycad species in the last 12 million years?  The authors think that climate may be involved:

The near-simultaneous initiation of diversification of six of the living cycad genera across the globe (in Australia, Africa, south-east Asia, and central America) indicates a single trigger may have been responsible. During the late Miocene, the global climate shifted as the world’s landmasses largely assumed their current positions (28). This closed the last of the equatorial seaways that had allowed warm tropical water to circulate the globe, leading to a shift from globally warm, equable climates to present day cooler, more seasonal climates (29). The majority of cycad species live in tropical or subtropical climates in regions of predominantly summer rainfall (2). Thus, it is possible that cycad diversification was largely driven by the global climate change that increased the geographic extent of those subtropical and tropical biomes that became marked by seasonality.

What is disturbing is that most cycads are endangered: the authors note that fully 62% of all cycad species are on the IUCN Red List of Threatened Plants, a higher proportion than in any other plant group.

So what the author have shown is that morphological conservatism of a group does not reflect the lack of ability of that group to produce new species.  What they haven’t shown is why morphological conservatism exists in the first place.  There are ways to rule out some possibilities. If genetic diversity for external characters were pervasive, you could rule out the lack-of-variation explanation. You could, for instance, perform artificial selection on living cycads to see if you could change them into plants that looked markedly different, as we’ve done with the ancestors of corn and broccoli. But ruling out alternative explanations doesn’t always tell us the real explanation unless only one remains.


Nagalingum, N. S., C. R. Marshall, T. B. Quental H. S. Rai, D. P. Little, and S. Mathews.  2011.  Recent synchronous radiation of a living fossil.  Science (published online, 20 Oct. 2011).

37 thoughts on “Paper on “living fossils” finds recent radiation, but misses the point

  1. Could it simply be a matter of their morphology being optomal, or at least at some type of plateau?

    Modern sharks closely resemble their Jurassic progenitors for good reason. Might something similar be going on with other living fossils?


    1. Modern sharks closely resemble their Jurassic progenitors for good reason.

      actually, not really. This is exactly what this mystery encompasses.

      While some shark species do resemble those representing the lineage 140 million years ago, others do not, and there are quite a few species of sharks that have evolved in the meantime.

      But, even so, there are also osteichthyians (bony fishes) that occupy all the same environments sharks do, and have radiated much more widely since.

      It is exactly like comparing cycads with the whole of gymnosperms or angiosperms, say.

  2. Thanks for commenting on this Jerry. The headlines in the BBC etc. science press are spinning this story to say that cycads are not really living fossils after all. As you point out, this is not so. To my knowledge no competent botanist claims that the modern species are old, only that the group is. And there is no question they have not changed much in morphology. They have very primitive features for seed plants including flagellated sperm cells. They also lack axialary branching. I have just two quibbles. Cycads were more widespread and abundant during the Mesozoic no question. However, I don’t think they were ever terribly common. Most of the cycad-like fossils were actually members of a different group – the Bennettitales. Finally, I never heard them called “tree ferns” by anyone.

    1. My quibble as well. Tree ferns are true ferns. The only thing I’ve ever heard cycads called, other than ‘cycads’ is ‘sago palms’.

      1. That’s also misleading, sago palm refers to _Metroxylon sagu_, a true palm. I’ve seen _Cycas revoluta_ being called sago cycad not sago palm. The mix probably stems from the fact that one can extract starch (sago) from the stem of cycads as it is done for the true sago palms.

  3. I have yet to read the article but I assume that they can be divided into groups based on geographical spread? I also assume that they spread to their present locations from an origin on a part of Gondwana land? Also their relationship with ‘predators’ (what is the right word – plant eaters?), pests, browsers etc – must be important. Are they stable because they have not had to form elaborate defences, or are they in fact just as tasty/nutritious as other plants? Is it that since the demise of dinosaurs they have not been eaten by mammals for some reason?

    1. Just checking The Secret Life of Trees by Colin Tudge (p.77-8), and he says possibly the first ever symbiosis between plants and insects was between Cycads & beetles. I think Umkomasia is right and that they are not tree-ferns – those are spore producers, the Cycads produce seeds. cCan anyone cornfirm or deny that?

      1. Tree ferns are indeed true ferns – reproducing by spores- I have never heard of Cycads called “tree ferns” (they’re not) – who knows where this came from (random Journo perhaps?)

        With regard to species being recent – that’s not surprising – The occurrence of “modern species” within ancient lineages is common. In fact a similar situation occurs within the true Tree fern genus Dicksonia.

        Dicksonia has c. 25 to 30? species, with a centre of diversity in New Guinea. Molecular evidence suggests the current species are the result of relatively recent speciation – perhaps within the last 9 million or so . That is not to say that the lineage isn’t ancient (there are whole fossils back to the Triassic).

      2. Botanical nitpicking:

        Yes, ferns produce spores. However, seed plants also produce spores. Seeds are a modification of the process (megaspores are retained within an integumented megasporangium rather than being released into the environment), not a replacement.

    2. They are used as food by humans, in a way reminiscent of true palm trees although they have some sort of nasty associated toxin.

      Dogs are often poisoned by chewing on their roots or seeds.

    3. The oldest fossils are from the Permian of China, but based on the systematic data that place them at the base of the seed plants, they are even older. They spread quickly in the Triassic and are already know from nearly all continents including Antarctica by that time.

      1. What? Important fossils from China? I guess it’s possible, but there haven’t been many really good paleontological finds there in the last decade or two. China should be trying to emulate that country where they’re digging up all those feathered dinosaur/ early bird fossils. Or that country that found those fossils important for early mammal evolution. Or that one that found that massive orb weaving spider fossil. You know that one that rhymes with China, with the capital that rhymes with Beijing.

  4. Typo: “although the group itself hasn’t changed much in appearance in 300 years” – you left out “million.”

  5. Sorry to pop up again but this is one of my favorite group to study. It is not widely known outside of botany circles that cycads have remarkable, specialized root structures (coralloid roots)that house symbiotic bacteria that fix nitrogen for them. They are also insect pollinated, which is rare in gymnosperms. “Primitive” in systematics just means relatively basal – it does not mean simple or unsophisticated! Also, despite their decline in recent floras, many of them are tough. Here in Mobile Alabama, Cycas revoluta is in almost everyone’s yard because it so tough and easy to grow. So maybe the ancient morphology is simply a case of not messing with a winning setup and having the toughness to survive environmental perturbations in the past. By the way – Equistemum (horsetails)are another ancient group morphologically but with molecular evidence for recent speciation. Maybe this is actually a common pattern in

    1. I wonder how important is the fossil calibration in those studies. Nymphaeaceae, too, are an early divergent group with a quite recent radiation, according to molecular dating; but there are some beautifully preserved fossil of possible crown group Nymphaeaceae that could possibly take back the date of the radiation. So, for Equisetum, there are petrified specimens from the Jurassic that show all the characters of the crown group.

    2. Horsetails propagate very easily via bits of rhizome which is why gardemers hate them. Perhaps this means they radiated quickly in comparatively recent times when the chance arose but then isolated populations speciated. Perhaps – as plants are harder to fossilize (?) – they specieated like this before but those populations were wiped out.

    3. Yes, sago cycads are very popular in the South. I’ll give you a little gardening tip. Is your sago turning a little dull and getting some brown on the leaves? Set it on fire! I discovered this quite by accident when I started a backyard brushfire once. All the leaves of the sago were singed off and there was a small amount of charring on the stem. This happened in December, by the middle of February that plant had grown five inches taller and was sprouting a new head of the brightest, greenest leaves I ever saw. It even grew a cone later on in July.

      I figure sagos might actually need to be burned every few years to reach their full potential. I’m trying to figure out a safe way to replicate what happened by accident so I can start a fun side business that’s perfect for a mild pyro like me. Best I can figure out is to build a large “turkey fryer” type propane burner around the sago, and then light up that sucker for about five minutes or until all the leaves are burned off.

  6. Certainly the things that botanists call “tree ferns” are not closely related to cycads. Nor are regular ferns close relatives of cycads.

    However, it is possible that salespeople or commercial dealers of ornamental plants may call these “tree ferns”, as the finely-divided leaves look a bit like fern leaves, and they do have a trunk.

    A botanist would not condone this non-evolutionary use of the term, though.

  7. I’d wager the same general thing is true for most ‘living fossils’. There is no fossil record, for example, of the only extant genus of coelacanths. The living horseshoe crab genus Limulus is only about 20 million years old.
    ‘Living fossilness’ is not about species or genera, but about gestalt morphology of the type that defines traditional Families and Orders. Which is I think Dr. Coyne’s point.

  8. Here’s an explanation for living fossils: chance.

    If you start with any given initial species, then look at the tree of descendants over millions of years, there is always the chance, and a pretty good chance (relatively speaking), that there will be at least one line that remains pretty similar to the initial species.

    So given all the species in the past, it is highly likely that, just by chance, a few will having living descendants virtually unchanged. Or looking the other way, it is highly likely that a few species living today will trace back to distant nearly identical ancestors.

    In other words, there’s nothing magical or surprising about the existence of “fossil species” — each is a chance special case where selection pressures and random genetic drift happened to be just right over the X million years to preserve morphology. The specific reasons why a specific species is a “living fossil” will almost certainly be different for each such species.

    What would be interesting is to look at statistics of a large sample of different living species and see how far back we have to go for each species to find a morphologically distinct ancestor. I predict (off-hand) that we would see something that looks roughly like a simple exponential distribution — perhaps exactly an exponential distribution — for P(t) where t is the duration of unchanged morphology.

    1. Probably depends on the formal definition of measure for “morphologically distinct”.

      “Why do we have living fossils as often as we do” looks like a question like there might be a valid controversy– in so far as the mechanism isn’t yet well nailed down to a precise mathematical description, not that I expect a serious overthrow of the contemporary evolutionary synthesis to result.

      1. Actually I don’t think it matter much what measure you use, whether looking at a single trait or many or all, as long as they are quantified in a consistent way, but basically we can only really consider traits preserved by fossilization.

        How often *do* we have “living fossils”? My impression is that they are fairly rare — a reason that they are so remarkable when found.

        My question is: is there any reason to think they are anything other than statistical flukes? Samples from the extreme tail of an exponential (or similar) distribution?

        And my point, or hypothesis is that it will turn out that there is no general principle or process governing “living fossils”. Each case will be a unique convergence of the right combination of selection pressures and initial species such that one lineage happens not to change over a very long time.

        Is there something I’m missing that makes living fossils stand out in a way that winning lottery tickets don’t?

  9. The wollemi Pine was known only from 200 million year old fossils until living trees were discovered in the 1990s. There are about only 100 trees in the wild.

    Wollemi Pine

  10. I’m sure the evidence in far more ancient sediments are No True Cycads.

    Since when was “there can be genetic change without morphological change” news? I also remain unconvinced that the common ancestors of any 2 species can be predicted by the modeling technique used. For me it’s a bit like taking radiocarbon dating and trying to apply its parameters to all nuclides. In general for a given species you may get X mutations per 1000 years but there is no known fundamental reason why we should assume the rates would be somewhat constant for all times in the past.

    1. In fact when we make phylogenetic trees for neutral alleles, we often find the branch lengths of one or two species are exorbitantly longer than all the others, even though all diverged recently (according to the tree-generating algorithm). This is proof that the molecular clock is not universal. But on the average, it does seem to be fairly consistent.

      1. Yup… when Jerry writes: “Estimates of divergence time are made by using calibrated “molecular clocks,” which assume (and there is support for this), that divergence in DNA sequence is proportional to the absolute time in the past when species diverged.”

        I can’t help thinking “yes, but…” There is certainly good support for divergence being proportional to time, but dating also relies on various assumptions (varying depending on the complexity of the analysis) about the rate at which mutations accumulate over time and how much that rate can vary across the topology… and there are always exceptions to the general validity of these assumptions. All the extant cycad species could be quite old, if the assumption that rates don’t change dramatically is violated. I haven’t read this particular paper, but this uncertainty is often represented by very large error bars, giving unhelpful estimates like “15 million years ago, plus or minus 12 million years”…

  11. There might be another reason for the apparent discrepancy between the recent diversification of Cycads and their morphological stasis: reproductive isolation. If the phylogenetic reconstruction is based, as it is usual, on neutral markers, one would expect diversificaiton in an ancient lineage that becomes fragmented in isolated populations (e.g. due to changes in geography and climate). The “ancient fosil” character refers to the lack of change in key adaptive traits. A useful complement to the analysis of neutral data would be to evaluate variation in quantitative traits that may confer adaptive value, and compare these to variation in morphological traits from extant and fosil species.

  12. And some, like cycads, can’t behaviorally seek out the environments to which they’re adapted…

    Well, yes and no. Plants have their ways. As evidence the floral assemblage shifts in latitude in response to advancing and retreating ice sheets.

  13. I highly recommend Oliver Sacks’s book, The Island of the Colorblind, which discusses people native to several tropical islands (S. Pacific, if memory serves) who enojy eating cycads, despite the fact that the eating induces vision loss in some individuals.

Leave a Reply