How Does Something That Light Become a Fossil? Fall Into the Right Mud

Fossilisation favours hard mineralised tissue, and insects have none, being built from a tough organic material rather than from bone or shell. They are also small, light and fragile, so a dead insect is normally eaten, broken up or decomposed within days. Against that, insects are produced in extraordinary numbers, which means that even an extremely low chance of preservation applied to an enormous population yields a substantial total. They also fall into water frequently, since flying animals land on it and are trapped by the surface, which delivers them into exactly the settings where fine sediment accumulates.

A small number of environments account for nearly all specimens:

• •Fine-grained lake sediment, which settles gently and buries without crushing
• •Lagoons with still oxygen-poor bottom water, which excludes scavengers
• •Volcanic ash falling into water, which buries rapidly and in fine particles
• •Tree resin, which traps and then hardens around the animal
• •Tar seeps, which preserve very recent material
• •Copal, which is resin not yet fully hardened and is much younger

Specimens in hardened resin are qualitatively different from those in sediment, which is why they dominate the popular image. The animal is preserved in three dimensions rather than flattened, with legs, antennae, wings and hairs in their original positions. Surface detail survives at microscopic scale, including the structure of compound eyes and the scales on a wing. Colour patterns are occasionally retained. Behaviour is recorded directly, with specimens trapped while mating, while feeding, while carrying prey and while emerging from a moult. And associations are preserved, so a flower and its visitor, or a parasite and its host, can be found together. That combination is unavailable from any other kind of fossil.

Material preserved in resin has attracted claims that outrun the evidence and the corrections are worth knowing. Recovering usable ancient DNA from such specimens was reported in the early 1990s and has not been replicated, with careful reinvestigation attributing the results to contamination, and the current position is that DNA does not survive the tens of millions of years involved. Burmese material, which is scientifically extraordinary, has been the subject of serious ethical objection since it is mined in a conflict zone and its sale has funded armed groups, and several journals have restricted publication of specimens collected after a stated date. Fakes are common in the trade, made by embedding modern insects in synthetic resin or in young copal.

The accumulated evidence has settled several questions about the history of the group. Insects appear in the record over four hundred million years ago and flight appears within the following fifty million, far earlier than in any other group. Giant forms occurred during the period of elevated atmospheric oxygen and disappeared with it. The great diversification of insects tracks the rise of flowering plants closely, which supports a long argument that the two radiations drove each other. Mouthparts preserved in detail show what particular species ate. And extinct orders with no living relatives appear repeatedly, which indicates that the surviving diversity is a fraction of what has existed.

Insects have no mineralised tissue and are small and fragile, so preservation is improbable, and their enormous numbers and frequent falls into water make it common anyway. Fine lake sediment, oxygen-poor lagoons, ash falls and tree resin account for nearly everything. Resin preserves three-dimensional specimens with microscopic surface detail and records behaviour directly, which no other fossil does.