How scientists reconstruct ancient food webs

An ecosystem is not a simple ladder with plants at the bottom and one predator at the top. It is a network of feeding relationships, scavenging, decomposition and competition. Fossils preserve only parts of that network, so palaeontologists combine direct evidence of feeding with anatomy, chemistry and knowledge of the rocks in which remains were buried.

Direct evidence of a meal

The clearest clues are preserved food in a digestive tract, prey bones with bite marks, or a fossilised mass of digestive material. Coprolites are fossil faeces; cololites preserve material in an intestine; regurgitalites record expelled contents. Together they are sometimes grouped as bromalites. Their contents can identify plants, fish, insects or bone fragments that an animal consumed.

Direct evidence still needs interpretation. A coprolite may be difficult to assign to its producer unless it is found with an animal or has distinctive shape, chemistry and contents. Bone fragments may have passed through a predator, been scavenged, or accumulated by another process. Synchrotron scans and micro-CT can reveal inclusions without destroying the specimen, allowing researchers to compare many samples and document the evidence.

Teeth and anatomy suggest likely diets

Tooth shape, wear and jaw mechanics can narrow the range of foods an animal could process. Broad grinding surfaces suggest different actions from sharp slicing teeth, while beak shape and neck motion may constrain browsing. These are functional clues, not a menu. Many living animals eat foods that their teeth alone would not predict, and ancient animals could change diet with age or season.

Plant fossils help reconstruct the resources available in a habitat. Pollen, leaves, seeds and wood can reveal vegetation and environmental change. A herbivore’s teeth can then be compared with the plants found in the same geological unit. This does not prove that a particular dinosaur ate a particular plant unless direct evidence connects them, but it builds a grounded picture of possible food sources.

Chemical evidence and network models

Stable isotopes in fossil tissues can record aspects of diet, water sources and trophic position. Lipid biomarkers may preserve clues about plants, microbes or digestive environments. These chemical signals must be checked for alteration after burial; a fossil’s chemistry can be changed by minerals and groundwater. Researchers compare multiple tissues, local sediments and modern reference materials to test whether a signal is biological.

A food-web model links potential consumers and resources using evidence strength and ecological constraints. It can show which relationships are supported directly and which are inferred from anatomy or co-occurrence. It is not a photograph of an ancient ecosystem. Models are most useful when they make assumptions visible and can be revised as new fossils appear.

A Triassic example from Poland

A 2024 study examined more than 500 bromalites from fossil sites in the Polish Basin and combined their contents with body fossils, trace fossils and geochemical analyses. The samples preserve fish, insects, plant material and bones. Across the studied Late Triassic and earliest Jurassic communities, these records helped track changes in feeding patterns as ecosystems and plant communities shifted.

The study provides an unusually detailed regional food-web reconstruction, not a universal account of every Triassic ecosystem. Producer assignments were sometimes uncertain, and the fossils come from particular localities and depositional settings. The Jurassic food-web overview shows how one ecosystem can be reconstructed from its own evidence without treating all prehistoric communities as identical.

Biases in the fossil record

Hard parts preserve more often than soft bodies; river channels, lakes and marine deposits sample different habitats; and collecting effort varies between regions. A predator may be well known from bones but rarely leave identifiable gut contents. Small organisms, fungi and microbes can be essential to food webs yet almost invisible in the body-fossil record.

Absence is therefore difficult to interpret. A missing prey fossil may mean the animal was rare, lived elsewhere, had soft tissues, or simply has not been found. Taphonomy – the path from death to discovery – filters every reconstruction. Researchers should distinguish observed interactions from plausible links and avoid filling every gap with a generic predator–prey story.

From remains to relationships

Ancient food webs become stronger when independent evidence agrees: a digestive inclusion identifies food, tooth wear supports the same feeding behaviour, and plants from the locality show that resource existed. Each link carries a confidence level. Multiple lines can reveal ecological change over time, including how organisms responded to climate shifts and extinctions.

The goal is not to draw a complete web at any cost. It is to recover the relationships that the fossils support and to say where uncertainty remains. Coprolites and other humble traces can be as informative as skeletons because they preserve interactions rather than just body shape.