Why fossils shouldn’t really exist

I’m writing a book about ants. It’s still many months from completion. A chapter on ant evolution must mention fossils, so I thought I’d throw this suggestion into the internet maelstrom to see how it should be honed. Or binned.


Recognizable ants have been around for at least 100 million years. We have exquisitely preserved fossils in amber to prove it. A quick internet search shows that ants preserved in amber are widely available from dealers. I was tempted by several in the £15–£30 price bracket. The ants were all tiny, 2–4 mm, but with low monetary value they were unlikely to be fakes. However I was not temped by the mating ‘queen and drone’ for sale from a US website at $49,500. If memory serves, this is the same item that I came across over 10 years earlier when I was researching ‘the most expensive insect’ (Jones, 2010, Extreme insects, HarperCollins), and which was then on offer for $100,000. No matter how rare such an event might be to get fossilized, this seems to be a fairly arbitrary sum, even if now knocked down by 50%.

But it at least got me thinking.

Dinosaurs have captured the popular imagination, but this las led to some completely silly and imaginary ideas about fossils. Fossils are real, but they are so rare that they do, actually, verge on the mythical. I casually wondered out loud what might be the chances of something getting fossilized. This seemed like quite a straightforward question to me, but I soon found out that nobody was willing to offer up any calculations. Taphonomists (fossil palaeontologists) openly mocked my question, claiming that ‘putting odds on it is impossible’. Having thought about it a bit more I’m tempted to agree with them. Nevertheless, having set the metaphorical ball of fiery curiosity rolling through my brain, I here offer some back-of-the-envelope calculations.

Firstly — how many ants have ever been available for fossilization? We have to start with a calculation from over half a century ago, by C.B. Williams (1960, The range and pattern of insect abundance, The American Naturalist, 94: 137-151). Extrapolating from bird counts, insect trapping surveys, log–normal biodiversity indices and world-wide species lists, he estimated 1018 insects on the planet at any given time, of which say 1% were ants; this would give us 1016. Assuming some sort of conformity with modern faunas, and allowing each ant to live 1 year, the last 100 million (108) years would therefore have produced 1016 x 108 = 1024 ant individuals having crawled on planet Earth.

Secondly — how many fossils are known? It turns out that about 750 fossil ant species are described from, say, 10,000 ant fossils. At the final count this does not seem like very much at all. 

Finally — what proportion of fossils have we humans found? Of course humans have only discovered a tiny fraction of the fossils buried in the soil. This is where the calculations and guesswork start to get a bit fanciful. For the sake of argument, I’d suggest that humans have only discovered 0.001% of ant fossils;  this is a figure I admit to having arbitrarily plucked out of the air. I dare somebody to disagree with me.

So, if 10,000 known fossils represents 0.001% of the total, that total number might be 109. This would still give the ratio of all ants to fossil ants at something like 1024 : 109  or in more conventional usage 1015 : 1. These are incomprehensibly small odds.

Given that the chance of winning the UK’s National Lottery is reportedly 4.5 x 107 : 1 (45 million to one), the chances of an ant becoming fossilized are about on a par with the same person winning the lottery two weeks running. Given that 32 million people are reputed to do the lottery each week, my increasingly tenuous grip on mathematics suggests that it might take 613, 278 years for them all to have just a single win. For someone to win twice in a row seems mathematically (and morally) untenable. For an ant to become fossilized must be equally dubious. 

It’s that unknown how-many-more-fossils-are-left-in-the-ground part of the equation that gives me most anxiety. But even if my guesstimate is out by a million-fold (even a billion-fold) the odds are still vanishingly small. Really, in any sensible world, fossils shouldn’t exist — each and every one of them is a tiny miracle of geological chance. Perhaps I was being a bit dismissive of that mating ant pair in amber.

2 responses to “Why fossils shouldn’t really exist

  1. David Sargan

    Hi Richard, I think the problem with your calculation is that in plucking 0.001% (or 1 in 100,000) out of the air you have been ridiculously generous to fossil hunters. Amber ranges up to >3.2 x 10e8 years old. If you consider fossil bearing shales they go much further back, but in general we could assign fossil making to the period from 10e8 years ago, with, as you suggest, ants having the possibility to become fossilised throughout that period. Human fossil collection as an activity that has gone on only for shall we say 5000 years, but it relies on exposure of fossils by erosion combined with fallible and largely unsystematic human searching or potentially and to a degree on human mining activity. So where does that get us: over the period of fossil collection , only 1/ 20,000 of the erosive exposure of the last 100,000,000 years has taken place. Generously, for amber, we might hope that 1% of recently exposed amber fossils have been spotted before being eroded to nothing. But even at our least generous we could presume that the vast majority of fossil amber, as of other sedimentary stones, lies beneath the ground. Shall we say 90%? So on the amber side of the balance sheet we might end up with a multiplier of 20,000,000: 2 ex7. But then there is the shale side. Many ant species nowadays live below ground, and because when clay soils flood and are then compressed, the compression gives rise to shales, so one might expect ant fossils to be fairly common in shales. But they can only be seen when exposed by erosion and being small and thin, will then erode away rather quickly. So here I would suggest that the ratio of undiscovered to discovered ant fossils at the erosion surface will be much higher. Again a figure plucked out of the air, but 0.001% seems a generous detection rate. Probably an even larger portion of shales are still beneath our feet, so I would suggest the multiplier of 20,000 again for shale beds that are not accessible for fossil hunting. That leaves an overall multiplier of 20,000×100,000= 2ex9. i will also guess that because of there ease of detecion amber ant fossils outnumber stone ones by 20 to one. So amber side: 9500 x 10ex7, shale side 1000 x (2x10ex9); total fossilised ants = 2.1 x 10ex12. So the chances of fossilisation come way down: to something like 2 in 10 ex12. And that my friend is 20,000 every year! David Sargan.

  2. Pingback: Pick & Mix 34 – a very mixed bag | Don't Forget the Roundabouts

If something is bugging you, leave a reply here

Please log in using one of these methods to post your comment:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.