top of page

How reliable is the fossil record?

Whenever an organism is found within the rock record, many questions are raised. They're raised by scientists, such as 'what is it?', 'what is it related to?', 'what can it tell us about evolutionary history?', or by creationists, such as 'how do we know it is as old as you say it is?' or by the general public, like 'what was it like to see the animal in real life?'.

Answering these questions and many more is met with varying degrees of success, because there are simply some things that we can never know for sure. So let's take a look at just how reliable the fossil record is.

A slab of trilobite fossils

Some problems...

So, what problems can we face? Surely if you find some bones, that's it, case closed, right?

Well, the first problem we face is whether or not those bones actually belong together. Contrary to popular belief, fossils are not found in a neat pile of articulated skeletons in the exact pose they died in like you see in Jurassic Park. At least, not any more often than someone buying a winning lottery ticket.

fossil dig scene from Jurassic Park
Property of Universal Pictures

99.99% of the time, a fossil of say, a dinosaur, consists of a couple of bones from its entire body an not in their original position. This can lead to some confusion as to whether all of the bones belonged to the same organism, with chimera dinosaurs being the potential result.

The Archaeoraptor chimera fossil
The infamous chimera 'Archaeoraptor'. Image credit:

The next problem we can run into is what is implied by fossils being found together. If, let's say, we find several Velociraptors together, many would assume that they lived in packs. However, this implication could be very different from the truth. Each one could have been drawn to the same area for a drink and been caught in the same bit of quicksand, but were actually several days apart. Perhaps they were working together to take down prey, but only worked together with strangers until they eat and then go their separate ways, like komodo dragons.

3 komodo dragons sat by water
Image credit:

This kind of 'death assemblage' can be caused by so many factors that, on its own, it doesn't actually give us much conclusive evidence.

But sometimes the reliability of the fossil record is less about what is preserved and more about what isn't preserved. Despite how you can go to certain areas an be practically tripping over fossils (whether that's in the UK or anywhere else), fossilisation is actually an incredibly rare process.

If an organism dies, it has to avoid being picked at too much by scavengers and micro-organisms and then be buried relatively quickly by sediment. Even the, it is likely that only the hard parts of an organism will fossilise. This is actually why you find more animals as fossils than you do any other type of organism, such as plants.

With this in mind, now think of something like the Cambrian explosion. Did many diverse groups of organism rapidly evolve an suddenly appear in the fossil record, or was there actually lots and lots of different types of animals that were thriving but lacked any hard parts such as external skeletons that could readily fossilise? We don't know and, without a time machine or miracle, likely never will. This particular type of problem in the fossil record is known as preservational bias and really brings into question what else was stomping around during the Mesozoic that we don't know about.

The next problem is one that faces cladistics specifically and that is convergent evolution. Convergent evolution is when two organisms evolve the same feature independent of each other. For example, giraffes are not descended from sauropods (long-neck dinosaurs), but they have still evolved a long neck for (likely) a similar use.

Painting by CHarles R. Knight showing a sauropod reared up on its hind legs

This can become a problem when we look at organisms that are closely related but their affinity isn't fully explored. For example, the placement of turtles within the reptilian group has long been a confusing one. We know that reptiles first diverged into two main groups; the synapsids and diapsids.

diagram showing the difference between a diapsid and a synapsid
A: Synapsid and B: Diapsid

However, the presence of turtles has called into question whether or not a third group existed, the anapsids, where no temporal fenestra is present. From the fossil record alone, turtles and tortoises have appeared to either be within the same group as other lizards and snakes or a group in their own right. It wasn't until a genome computational study was performed in 2013 by Zhuo Wang et al. that they concluded that turtles are likely archosaurs, being a sister taxon to crocodiles and dinosaurs. This is still argued (perhaps a discussion for a different post), but genome studies are very different to fossil record reliance alone, therefore it can paint a very different taxonomical picture.

Another problem that is often brought up by some creationists and some conspiracy theorists is that of temporal placement. In other words, is it actually obvious when the organism was alive? Dating the rocks that an organism is found in is a technique that has been practiced almost since the start of the science and there are two ways; relative dating and absolute dating. Relative dating is when two or more features of a rock are said to have happened before or after each other based on how rocks are actually deposited. Rocks are always laid out horizontally, one on top of the other, getting younger as you go up the sequence. This means that, even if you don't know the exact age of a fossil you find at the bottom of a rock sequence, you know that it is older than a fossil you find at the top.

Absolute dating is also known as carbon dating. This is when geologists look at the radioactive decay of certain elements within a rock. Since radioactive decay is constant, the extent of such decay will give you an approximate age of the rock.

Lu–Hf isochron diagram
Image credit:

These two methods of dating are reliable, especially at a big scale. However, if we want to increase the resolution (i.e. try and get more specific than a million years) certain factors can start to mess with this reliability. For example, if a certain animal was killed by the K-Pg mass extinction event but appears above the characteristic clay layer that this event created. This could make it seem like the animal died way after the event, but what if a scavenger decided to pick it up and pick at the organism for a few days? What if afterwards it dropped into an estuary system and transported miles upon miles before finally settling somewhere? This kind of discrepancy can change the entire apparent context and makes looking at vital information very difficult if it's the case of only a few months.

How the problems are worked around

The most common and simple way around most of these issues is to simply acknowledge that they exist. In cases like these, paleontologists will never commit to any definitive conclusions, since it would simply be unscientific if other possibilities are present.

If there are questions about whether or not the bones belong together, paleontologists need to think a little more outside the box. Comparisons with other specimens is the first port of call, making sure everything matches enough. Other times it might be a case of figuring if the basic biomechanics would work for an organism, or if the bones show ontogenic traits (i.e. are they juvenile?).

When fossils are found to be together, further context is often searched for within the sedimentary rocks themselves. If a death assemblage is present, it will often show itself in the rock record, such as anoxic shale or tar pits that trapped animals. to be honest, this one is on an individual basis, so the things that paleontologists may look at are endless...or nothing at all and they won't confirm anything.

When it comes to what isn't preserved...I've got nothing. There isn't much reliability in estimating the kind of beasties that existed without fossil evidence, since the truth has often been stranger than've seen all the weird stuff the Earth has produced so far, right?

That last problem of phylogeny is one that has made many mistakes but has only recently begun to be addressed through the advent of computational genetic modelling, mapping out higher resolution possibilities using modern descendants to take the guess work out of it rather than getting confused by similar features on fragmentary fossils.

In conclusion...

As much as it pains me to say, the fossil record is alarmingly fragmentary and leaves a lot to imagination and, as great as imagination is, it isn't very...scientific.

This might sound defeatist, but it's quite the opposite, because when we start questioning things beyond what they appear, we get better at finding out the right answer and inquisitiveness is the secret to any science.

Until next time.

15 views0 comments

Recent Posts

See All


bottom of page