How Fossils Form: From Living Organism to Evidence in Rock

Rapid burial gives remains a better chance of surviving. Mud, sand, volcanic ash, or other sediment can cover a body and reduce exposure to scavengers, oxygen, and weather. Hard parts such as shells, teeth, and bones are more likely to persist than soft tissues, although exceptional conditions can preserve delicate material too.

One common route in how fossils form is mineralization. Groundwater carrying dissolved minerals can move through buried remains. Minerals may fill tiny spaces or replace some of the original material, gradually producing a stone-like record of the structure. The exact process varies, so not every fossil is simply a bone that turned into rock in the same way.

Fossils can also form as moulds and casts. If a shell is buried and later dissolves, it can leave a hollow impression called a mould. If minerals or sediment later fill that space and harden, the result can be a cast that copies the original shape. Compression can leave a thin film or flattened outline, especially for plants and other delicate organisms.

How fossils form is not limited to body parts. Footprints, burrows, nests, bite marks, and fossilised droppings can preserve evidence of behaviour. These trace fossils can show how an organism moved, fed, or interacted with its environment even when no skeleton is present. In some cases, a trackway tells a story that isolated bones cannot.

The fossil record is incomplete because preservation strongly favours some environments and organisms over others. Creatures with hard shells living near sediment-rich water have a better chance of fossilising than soft-bodied organisms on exposed land. Rocks can also be melted, buried too deeply, eroded, or changed by heat and pressure, removing fossils that once existed.

Scientists therefore treat fossils as evidence rather than a perfect catalogue of all past life. They compare fossils with rock layers, radiometric dates, living organisms, and other geological clues. A missing fossil does not automatically prove an organism was absent. It may simply reflect the many filters between a living population and the small fraction of remains that survive long enough to be discovered. Fossil age is another question, and it is separate from the preservation process itself. Scientists can compare the rock layer containing a fossil with layers above and below it, and some rocks can be dated using radioactive isotopes. Fossils can also help correlate rock layers between places when the same distinctive species appears in both. The strongest conclusions usually come from combining several lines of geological evidence rather than relying on one fossil by itself. Fossil evidence is strongest when its geological setting is recorded carefully. Removing a fossil without noting the surrounding rock layer, orientation, and nearby features can erase clues that help explain its age and environment.

A few places on Earth preserved far more than bones and shells. Palaeontologists call them Lagerstätten, a German term for exceptional fossil deposits. The Burgess Shale in the Canadian Rockies preserved soft-bodied animals from the Cambrian period, including creatures with no hard parts at all, because fine mud buried them quickly in conditions with little oxygen. Lake sediments and volcanic ash in north-eastern China preserved feathered dinosaurs so finely that the outlines of the feathers survive. Without deposits like these, whole groups of soft-bodied animals would be almost invisible to science.

Other routes to preservation do not involve rock at all. Insects, spiders and even small lizards have been trapped in tree resin that hardened into amber, keeping their shape in remarkable detail. Woolly mammoths have been found frozen in Siberian permafrost with skin, hair and stomach contents intact. The asphalt seeps of the La Brea Tar Pits in Los Angeles trapped animals such as sabre-toothed cats and dire wolves. Because these sites are so unusual, scientists use them to check how much ordinary fossil beds leave out.

Not everything that looks like a fossil is one. A classic example is the mineral dendrite, a delicate branching pattern that can look exactly like a fern or a moss pressed into stone. Dendrites usually form when water carrying dissolved manganese or iron seeps into a thin crack, and the minerals crystallise in a branching pattern as the solution spreads. Rounded concretions, which grow as minerals collect around a centre within sediment, can likewise resemble eggs, bones or even skulls.

Palaeontologists separate real fossils from these pseudofossils by looking for features that only living things produce. Genuine remains show consistent anatomy repeated across specimens, such as the same pattern of veins in a leaf or segments in a shell, and sometimes internal structure under the microscope. Purely mineral patterns tend to vary randomly and follow cracks in the rock. The question becomes serious with the oldest possible traces of life, where tiny structures in ancient rocks, or in meteorites, have been debated for years because chemistry alone can mimic biology.

Much of what palaeontologists know about fossilisation comes from taphonomy, the study of everything that happens to remains between death and discovery. Because nobody can watch a fossil form over millions of years, researchers study the early stages directly. They leave carcasses to decay in tanks of water or sediment, record which tissues disappear first and note how bodies fall apart. A dead animal in water, for example, may fill with gases from decay and float for a time, shedding parts as it drifts. That helps explain why some marine skeletons are found scattered or with pieces missing.

Decay experiments have also produced a warning. When researchers let simple living relatives of vertebrates, such as lancelets and young lampreys, rot under controlled conditions, their features did not vanish at random. Several of the anatomical traits most useful for classifying these animals tended to decay early. A partly rotted fossil can therefore look simpler and more primitive than the living animal really was, which could place it too low on the evolutionary tree. Palaeontologists now weigh the possibility that a missing feature was lost to decay rather than never present in the first place.

Trace fossils can record a single moment. At Laetoli in Tanzania, a team led by Mary Leakey uncovered in 1978 a trail of footprints pressed into a layer of volcanic ash around 3.6 million years ago. A nearby volcano had spread fine ash across the landscape, rain dampened it into something like wet cement, and early human relatives walked across it before it hardened and was buried by further eruptions. The prints show a big toe in line with the others and a heel strike much like ours, strong evidence that upright walking came long before large brains.

Footprints can also vanish almost as soon as they are found. In 2013, rough seas stripped sand from the beach at Happisburgh in Norfolk and exposed hollows in ancient estuary mud. Researchers recognised them as human footprints, probably left by a small group that included children, and later estimated them to be more than 800,000 years old, among the oldest known outside Africa. Within about two weeks the tide had eroded them away, so the evidence now survives only in photographs and 3D models. On an eroding coast, speed of recording matters as much as the luck of discovery.

How fossils form involves unusual preservation, usually beginning with burial and continuing through physical and chemical change. Remains can be mineralized, leave moulds or casts, or survive as trace fossils such as footprints. Because fossilisation is rare and selective, the fossil record has gaps. Each fossil is a piece of evidence, not a complete diary of ancient life.