How do we know that most of the species that ever lived are extinct?

Jan 12 2010 Published by under Science

It's been a long time since I've responded to an Uncommon Descent post, and I'm starting to remember why. There's one that went up over there the other day on the fossil record that's really almost mind numbing - starting with the title, which is "Why Not Accept the Fossil Record at Face Value Instead of Imposing a Theory on it?"

Here's what seems to be the main argument:

Here’s a simple example – extinction estimates. Darwinists will say that 99.99% of species that have ever lived have gone extinct. Well, that’s actually a bunch of B.S. There are roughly 250,000 species that have been identified in the fossil record, and well over 1,000,000 species that exist today. Taken at face value, even if every species in the fossil record has gone extinct (which they haven’t), that means that 80% of species that ever existed ARE STILL ALIVE. That’s quite a stretch. So where do Darwinists get their number? By assuming that innumerable species existed in the transitional spaces. Why? Because they _must_ have existed there for their theory to be true.

You don't need to know anything about evolution to conclude that the fossil record is, by its very nature, spectacularly incomplete. You just need to know a little bit about sedimentary geology, a little bit about the process of fossilization, and a little bit about biology. When you take that knowledge and apply it to what we see on the planet, the conclusion that the fossil record is spectacularly incomplete is not just obvious, it's inescapable.

Let's look at just some of the major factors that lead to this conclusion.

Hard Part Bias:

Soft tissues fossilize poorly for a variety of reasons. They are more easily damaged, they decompose more rapidly, and they are often distorted by the weight of sediments that accumulate on top of them. Most of the fossils of soft-bodied organisms come from a very, very small number of sites. This is why the fossil record for jellyfish is very poor. It's why it took more than a century to figure out what kind of animal conodont elements came from. It's one of the reasons that there are entire phyla that have no fossil record at all.

Shells, teeth, and bones tend to stick around longer. They're tougher, and they're not as easy to eat. This means that the fossil record, as a whole, is very, very heavily biased toward preserving things that have hard parts.

If you doubt this, I'd suggest taking a walk along the beach. Bring a pad and pencil, and look at the shells you see there. How many are occupied by living organisms? How many are empty? And how many have something dead inside them. I'm willing to bet that barring really exceptional circumstances (like a very recent toxic waste spill) you'll find very few shells that contain the dead remains of their owners.

Environmental Bias:

Fossils are found almost exclusively in sedimentary rocks. This means that the organisms that live in areas where sediment accumulates are much more likely to be preserved as fossils than organisms that live in areas where sediment is eroding.

The bulk of sediments are deposited into shallow marine environments, and the fossil record for these environments is comparatively good. Large lakes, dry basins, swamps, mudflats, and river delta complexes are other environments that tend to accumulate sediments, and we've got a fair number of fossils of organisms that lived (or, more accurately, died) in these areas.

Mountains, hills, and high plateaus, on the other hand, are not places where sediments are deposited. They're the places that sediments come from. Things that live in these places do not get preserved.

In the Late Devonian, there was a mountain range to the east of the Appalachians. We know this because it left the record of its existence in the rocks that we see in the Appalachians today - rocks that were clearly formed in shallow sea and nearshore environments that lay to the west of mountains. In a sense, you could say that the formations of the Catskill Delta are the fossil remains of the older mountains - but whatever might have lived there did not fossilize.

Tectonic Issues:

The movement of the continents throughout geological time creates and destroys major geological features. Mountains have come and gone. Sedimentary beds have been transformed under immense heat and pressure to such an extent that any fossils that they contained are unrecognizable. Islands in ocean basins - which tend to contain many unique species - erode away and are sucked down into subduction zones.

The Devonian mountains I just mentioned are one example of these processes - some of the metamorphic rocks that lay at their core are now exposed outcrops. Current mountain ranges provide us another example. When we look at fossil-bearing rocks along a cliff face, we're looking at what erosion has left us. Anything that was embedded in the rocks that eroded away before we got there has been lost.

The Hawaiian Islands provide another example of these processes. The current islands are the latest in a chain that's been around for a long time. The coral reefs of Midway atoll, way up the chain, sit on top of the remains of what was once a large volcanic island. The current large islands have large numbers of species that are found nowhere else in the world. When we look at other oceanic islands, we see that this is a common feature of isolated islands, so it's likely that Midway, and all the other islands that have eroded away as they've moved away from the hot spot, had their own unique species. But they, too, have left no fossil remains.

Access Issues:

We know, for a fact, that there are enormous quantities of fossil-bearing sedimentary rock that we can't get to. Some of these rocks are located in places we can't reach. Most of it's buried underground.

When I was a paleontological tech, I would occasionally work with samples that came from deep wells. In some cases, the rock that I was looking at had been pulled from over a mile underground. I could look at the fossils that were in the core that the drillers recovered, but not at anything that was even a millimeter beyond the edges of the drill. There are tens of thousands of cubic kilometers of fossiliferous sedimentary rock under the surface of the Allegheny Plateau alone. If those rocks - which span a decent chunk of the Paleozoic - contain just one currently unknown fossil species per thousand cubic meters of rock, the total number of species just in that area would exceed the number of extant species by an enormous margin.

And let's not forget Antarctica. Any undiscovered fossil in the rocks that are currently buried under the ice there is going to remain undiscovered for a very long time.

Search Effort:

This is something I actually explored the other day. As I pointed out then, ten thousand paleontologists working nonstop for 150 years at a rate of 2 cubic meters of rock per day are capable of thoroughly searching a miniscule percentage of what's out there.

This is why paleontologists are still naming and describing species. In the latest issue of the journal Palaeontology alone, there are a number of articles describing and naming new species from fossil remains. Just skimming through it quickly, I see two new trilobite species, a new bivalved arthropod, four new decapods, a new bryozoan, and three new bivalve species. Each of the papers describing the new species is the product of dozens, if not hundreds, of person-hours of work. And there's absolutely nothing special about the journal or issue I picked.

To sum all this up, we know that the fossil record is spectacularly incomplete. Based on the number of fossils that have been discovered, our understanding of the various biases favoring the preservation and of some forms of life over others, our knowledge of how much rock is out there that we have not and cannot examine, and our awareness of just how much effort it takes to find, identify, and describe just one new species, we can safely assume that the fossils that have been named represent only a very, very small fraction of the species that ever lived.

41 responses so far

  • Flint says:

    Another very rough approach to this problem is to get a statistical measure of (apparent) species lifespans from the fossils we do have. I've seen estimates that species (as identifiable from their fossils) might endure a mean of maybe 10 million years, with a standard deviation of several million at least.
    If the current number of living species is a very rough reflection of how many have been alive at any given time in the past, and if we're just looking at the last 600 million years or so (when fossils are found), that's time for 60 species lifespans . That gives a ballpark estimate of about 60 species that have gone extinct for every species alive today. Give or take a couple orders of magnitude, of course.

  • jackstraw says:

    From above
    "You don't need to know anything about evolution to conclude that the fossil record is, by its very nature, spectacularly complete."
    I'm thinking you mean spectacularly incomplete.

  • waldteufel says:

    None of the posters at UCD are very deep thinkers, but Barry is one of the dumbest.
    None of them know anything about geology.
    None of them know anything about biology.
    None of them know what the scientific method is.
    None of them know or care.
    Actually, I don't know why you bother posting about any of their scientifically illiterate drivel.

  • Toast says:

    Just a typo(?) :"spectacularly [in]complete, shirly?

  • Wheels says:

    I love how Arrington essentially advocates giving up on science. Instead of formulating hypotheses and then testing them by looking at the evidence to see if it fits, there's to be no testing of anything! Just standing back and admiring the pretty fossils.
    But I also like that he's trying to use ID as the basis for something other than writing books and lobbying school boards!

    One thing which I think ID can contribute to any historical aspect of earth history is shaving off hypothetical creatures. While there are certainly many creatures which haven’t yet been found, and I’m sure many of these creatures include chimeras of existing features in existing creatures, there is no reason to believe that there must be creatures where none have been found or evidenced. Darwinism has a bad habit of perpetually adding dashed lines in-between creatures for where it expects to find relationships. Instead, ID says that, perhap we can just take the fossil record as we find it.

    Is he, in fact, saying that ID makes the positive prediction that there are only a certain amount of life forms to find based on his statistical calculation? Is he putting forth the ID-based prediction that common descent cannot be usefully ascribed to the fossil record? Is he (gasp!) using the "theory" of ID to make a prediction about what we can/can't find in the fossil record?

  • Barry knows little about anything scientific, certainly nothing about sampling rates. Gee, Barry, are we really supposed to think that we have found anything like the numbers of species that we could find, or that anything like the total number of species could ever be found at all?
    But of course the genetic record alone indicates that every species alive today must have many ancestral species in its past. While earlier species are shared by many crown species, the total number of ancestral species represented by crown species must be a large multiple of present species.
    And knowing a bit about radiations, clearly extinct phyla and other taxa must add greatly to that number.
    For magical bypasses around known evolutionary processes this is not a numerical problem (Goddidit), but there is the little problem of explaining how the geologic column is roughly as expected from evolutionary theory, and not at all what any intelligent creator has ever produced, according to our observations.
    We're extrapolating from our small samples represented from the past, of course, and also from what we know about radiations in Hawaii, the Galapagos, and those after great extinctions.
    Only a pseudoscience insists on calling a small sample a complete record, and also ignores all of the genetic evidence pointing to huge numbers of extinct species.
    And, does he really think that the Permian extinction killed off only the species that we know from fossils prior to that extinction? That we have a complete record of Ediacaran life? If so, let us hope that he deals far better with evidence in the courtroom than he deals with it on UD.
    Glen Davidson

  • Nomad says:

    Is anyone else struck by how intellectually dishonest that quote actually is?
    If we start from the basic information that so far we have identified the fossil records of 250,000 species and currently one million species exist (I'm not interesting in the factual accuracy, just that Barry tried to use these two concepts as starting assumptions to base his commentary on).. surely the logical conclusion is that more species have ever existed than have been found as fossils. Surely that's the one inescapable conclusion? That, AT A MAXIMUM according to the founding assumptions that he presented no more than a quarter of all life that has ever existed has been found fossilized. But since he's ignoring the potential of any life existing but not existing (and being counted today) or being fossilized, a situation that is pretty much inevitable, the actual number is far worse, far fewer of the total species that have existed have ever been counted by humans.
    If Barry was actually interested in accepting the fossil record at face value that would seem to be the very first conclusion he must come to.
    Nowhere in those starting assumptions is any way to estimate the percentage of species that have ever existed which must still be alive today.
    Unless of course you're trying to make the fossil record fit your preconceived notions and are not in fact interested in taking anything at face value.
    You wouldn't be doing that, would you Barry?

  • Bob O'H says:

    Nice post, although I can't see it helping the UD regulars (or at least the regulars who aren't socks). I wonder how many fossils have been burned as coal. and, of course, the Danish Blue was discovered in stone destined to be cat litter.

    And let's not forget Antarctica. Any undiscovered fossil in the rocks that are currently buried under the ice there is going to remain undiscovered for a very long time.

    No, let's forget Antarctica - we don't want some bumbling fool digging up a shoggoth.

  • Mike Dunford says:

    @jackstraw & @Toast:
    Fixed the typo. Thanks for pointing it out.

  • Shane McKee says:

    Oh noes! Does Barry's logic mean that by the Devonian, humans and other primates will all be extinct? Quel horreur! Excellent post, but Mike, this will wash over the ID-iots. Forgive them, father, for they know not what they spew.
    Part of the problem is that they tend to be verbal thinkers, rather than conceptual thinkers. It's an interesting (though fallacy-ridden, and hence dangerous) cognitive style, and I don't think it's remediable. Indeed, I don't even think scientists are capable of being so fallacious; sadly many "philosophers" are not so blest. Watch out for these!

  • Amadan says:

    Barrogant in short:
    "Too much thinking makes your head hurt and disrupts the prayer meeting. Now smell the the nice flowers and get back to your Bible."


  • Frank J says:

    Will the day ever come when most American nonscientists see the word "Darwinists," and know better than to read further, because only radical anti-science activists and their hopeless, clueless cheerleaders use that term?

  • @ yum install Jesus:
    You say "nobody knows" how fossils are formed, and then you asserts that you do know? ("Fossils got there because of Noah's flood.") How contradictory.
    And "for all we know", the rocks where the fossils lie are from very, very different ages, so they could not have been formed by a mythic single, relatively brief worldwide flood.

  • a lurker says:

    We can certainly test the "Hard Part Bias" bit.
    Most of the known species today are insects. How many fossil species are insects are there? I don't know the answer but I know it is not all that great and certainly not a significant portion of the number of species. When the number of species fossilized are far smaller than the number of species alive today then I think we can conclude that most such species have not been found in the fossil record. The fossil species of insects for say over a hundred million years ago are not living species. Thus be inference there must have been huge number of insects not discovered in the fossil record.
    Of course we can look at things like large mammals that do fossilize somewhat reasonably (though no where near as well as shelled shallow-sea creatures). There are certainly far more types of fossil horses than exist today. There are certainly far more types of humans in the fossil record than exist today. Ditto elephants. So on and so forth. And as Mike points out, more are found all the time.
    Someone with access to the actual numbers are encouraged to post.

  • J-Dog says:

    Ha! I thought "yum install Jeebus" was a Poe Sock, but I clicked on the link and was treated to a mind-numbing Pile O'Tard! Evidently Mr. Jeebus is an equal opportunity denier - and hates all science. Please join me in wishing him good luck in his "The Earth Doesn't Rotate" fantasy world!

  • NJ says:

    As if the all caps weren't enough of a clue, if you head off to the nutcase's website (with html direct from 95!), you'll find him to be a geocentrist.

  • FastEddie says:

    Mike, all your explanatins are very satisfying to me. However, there's one piece that's missing to completely squish the UD dunderhead's complaint. Could you walk us through the math on how we get to the 99% figure for all-time extinction? Is it based on the fraction of species today which could potentially fossilize in the future, and this serves as a multiplier of something? Thanks.

  • I like this from yum's site:


    I think he's an amazingly dedicated Poe, pushing all of the buttons, complete with ugly website. Regardless, it's sure a mish-mash of freaky stuff on his site.
    Glen Davidson

  • Foggg says:

    Barry's career, before he 'took up' taphonomy:

  • beaglelady says:

    Real scientists are out there digging for fossils. It takes commitment, time, money and painstaking work that is sometimes dangerous and often discouraging. ID creationists keep their hands in their pockets and simply blabber foolishness. That's what's really going on.

  • eric says:

    Arrington's argument is, ironically, an argument against YECism. If, for the sake of argument, we accept his premise that the fossil record accurately represents the species which lived in the past, then the lack of modern human fossils must either mean:
    (a) no humans existed in the past, or
    (b) the species which fossilized are older than humans.
    Either way, its game over for YECism.

  • KeithB says:

    Mike Wrote:
    "And let's not forget Antarctica. Any undiscovered fossil in the rocks that are currently buried under the ice there is going to remain undiscovered for a very long time. "
    Well if climate change denialists have there way, maybe sooner rather than later!
    (Though I like Bob O'H's comment better!)

  • qetzal says:

    OK, I may get slammed for this, but I'm going to play Anti-Evolutionist's Advocate for a moment. (Insert joke about a devil's advocate defending a position popular with fundamentalist Christians here.)
    As AEA, I might respond to the posted argument by saying:
    "Sure, the actual number of organisms that ever lived must be far greater than the number of fossils we've recovered, for all the reasons you've described. But none of that proves that the number of extinct species is far greater. All those organisms that didn't fossilize could just be additional members of the species we've already found."
    Taking off my AEA hat, I think we have to add additional facts to address this objection. For example, we can see today that different environments always contain many different species. It's reasonable to think the same was true in prehistoric times. The fossil record even bears this out, at least in cases where we can compare distinct environments that were each conducive to fossilization. Accordingly, each distinct environment that was not conducive to fossilization would likely have contained its own distinct species. It would not likely contain the same species that are already known from fossils in dissimilar environments.
    We can also note the fact that we continue to find new fossil species, even in the face of the limitations noted here, which suggests we're still a long way for anything like a complete catalog of all prior species.
    I'm sure there are other arguments that can be added to address this.
    Now if you'll excuse me, I'm going to go burn my AEA hat. And wash my head.

  • Kate from Iowa says:

    I was hoping someone would bring up bugs and other squishie-crunchie things, a lurker! Save amber and a few very very rare finds in mud (and I think I remember hearing about one in a coprolite, poor bug!) there's virtually 0% bug record, if you look at just contemporary extant species.
    I just don't understand how anyone can say "just take the fossil record at face value" and be serious about it. But if they're reading, I need your help. You see, I'm a prince from Nigeria, and I need your bank account to help me get my money out of the country, and I'll even let you keep some of it if you'll help me.

  • topquark says:

    Arrington says that

    Taken at face value, even if every species in the fossil record has gone extinct (which they haven’t), that means that 80% of species that ever existed ARE STILL ALIVE.

    If that was so, you would expect that whenever you found a fossil, there would be an 80% chance of it being a living species, and a 20% chance of it being an extinct species. I don't know the exact value, but I'll estimate the odds are much more like 1% living, 99% extinct. Perhaps Barry should calculate the odds of an underlying distribution of 80/20 and a found distribution of 1/99 occurring in a sample size of millions of fossils - or perhaps the mathematical geniuses at the DI could do it for him. The stupidity of these people is truly staggering.

  • Alann says:

    When counting species there are approximately 60,000 vertibrates. In about the last 400 years about 240 vertibrate species went extinct. Invertebrates (insects are the bulk of species alive today) aren't tracked the same way. Looking only at this non-fossil record the extinction rate is actualy quite high. Even if we adjust it down by a factor of 10 (to one in a million per year) when we extend this over just 10 million years the odds of a species going extinct rise well over 99.99%. (only 1 in 20,000 remain).

  • John Harshman says:

    No, no, he's right. We need to take the fossil record at face value.
    Nematodes make a nice test case. Let's see. One species of nematode was created in the Jurassic, and quickly became extinct. (There are claims of a Carboniferous nematode, but it isn't clearly diagnosable.) The next species was created in the Eocene, and also quickly became extinct. A few more ephemeral species were created in the rest of the Tertiary. Then, perhaps at the beginning of human history, or perhaps mostly since we invented microscopes, around 100,000 species were created that still exist today.
    How does that sound?

  • snaxalotl says:

    taken at face value, the earth is flat, the sun goes around the earth, the bible is a straightforward record of witnesses to miracles without any layers of editing and redrafting, and Glen Beck knows what he's talking about. seems like the only people who should be taking things at face value are children, and the only people who think taking things at face value is a proud intellectual position would be slow learners

  • qbsmd says:

    "Why Not Accept the Fossil Record at Face Value Instead of Imposing a Theory on it?"
    A while ago, I learned that there's a phrase to describe that: naive Baconianism, which is a misunderstanding of an outdated philosophy of science that seems to involve only ad-hoc explanations rather than predictive theories.

  • amphiox says:

    And, given what we know about modern ecosystems and extrapolating back:
    If we found the fossil of one species of meat eating animal, all by itself, we know that there must have also have been at least one species (likely more) of plant eater, that it ate, and at least one species of plant (likely more) that it ate.
    And perhaps with a little less absolute certainty, at least one species of intestinal worm that parasitized it, one species of insect that processed it's dung, and another that sucked it's blood, and at least one species of insectivore that ate those insects, and for every one of the multicellular species, at least one species of bacteria that made it sick, and so on.
    Every isolated fossil find must have lived in an ecosystem with hundreds of other species. Few of them may have been preserved along with the fossil we find, but we know they were there.

  • mark says:

    We do know how fossils have formed (by various mechanisms, not just in one way)--we have even conducted experiments to better understand the processes.
    The title is not only dumb, as Mike says, but shows a complete lack of understanding of what science is all about. "just accept the fossil record at face value...?" What does that mean? How does doing that help us understand fossils, biodiversity, evolution, or other questions about biology? We don't impose our theory on the data; quite the other way around.

  • Mercedes says:

    Did or did not your precious Darwin take all of his findings from the Galapagos Islands and apply it to every species that has ever existed? You can trash all you want about those that believe in taking the fossil record at face value (which I am definitely not saying that I believe.) However, you must consider the fact that just because a sock is an item of clothing does not mean it belongs on your hand or your head. Furthermore, after nearly 150 years since The Origin of Species, the fossil record still is insufficient to support Darwin's claims, however prettily you might dress it up. You claimed at the beginning that soft tissues can become distorted because of the pressure that sediment exerts on them. So what fossils that have been "dug up" may have become warped over the years, yes? Please, check your sight before shooting yourself in the foot.

  • Stephen Wells says:

    @33: since the answer to your first question is "no, he didn't", you should stop whining and go get an education instead.
    re: "naive Baconianism": in fairness to Bacon, a thinker I greatly admire, he didn't say that we should not build models; he said that we should not be too swift to build models before we have enough information.

  • Ichthyic says:

    Did or did not your precious Darwin take all of his findings from the Galapagos Islands and apply it to every species that has ever existed?
    spoken by someone who obviously has never bothered actually reading any of Darwin's scientific publications, let alone actually reading the entire Origins.
    Dunning Kruger.
    look it up.

  • Ichthyic says:

    the fossil record still is insufficient to support Darwin's claims
    how close have you looked, really? i suspect you haven't even tried.
    which makes you a liar.

  • Rolf Aalberg says:

    "your precious Darwin" ...
    Intro says it all, no need to read the rest.

  • mark says:

    Aw..., somebody has upset Mercedes, who responded by saying things that are not true or, for that matter, that do not make sense.
    The fact is, since Origin was published 150 years ago, the number of described fossil species has vastly increased, as has our knowledge of genetics, development, and many other aspects of biology (and geology and chemistry). Careful observation and experimental ecology have provided enormous support for the science of evolution. Some hypotheses have been found in need of modification or even discarding, but no genuine challenge to the fact of evolution has been made.

  • Jim Harrison says:

    Somebody mentioned the fossil record for land invertebrates (aka squishy things): when I was a kid, circa 1965, the available English-language books said that little was known about the evolution of insects because of the scarcity of their remains. I think a fair number of folks still think that conclusion holds. In fact, it just happened that a lot of the work on insect evolution was done by Russian paleontologists whose work was not widely known in the West during the Cold War. Insect fossils are far more abundant than scientists in these parts realized. Grimaldi and Engel's splendid tome, Evolution of the Insects, has all the details.
    I don't think anybody who is at all familiar with the paleontological literature has any doubts about the huge number of extinct species. Creationists and ID folks are interested in theology, not biology, so it's no wonder they are hugely ignorant. Things are bound to look different if your basic motive is to know about the bugs and how they came to be.

  • Jim Thomerson says:

    I was a paleontologist through my MS, but became a neontologist for the PhD and subsequent career. I can state from experience that it is much more difficult to obtain, recognize, and adequately describe, a fossil than to do the same a living organism. This difficulty accounts for much of the under-description of fossil organisms.

  • Jim Thomerson says:

    There were some 50-odd new species of catfishes described last year. This suggests that estimates of actual number of living species might also be low.