If you picture a hunter-gatherer meal as little more than roasted meat, a new archaeological study suggests you may be missing half the menu. Researchers led by the University of York analyzed charred food residue stuck to prehistoric pottery and found repeated combinations of fish with carefully chosen wild plants.
The headline finding is not just that plants were eaten, but that they were cooked on purpose, in particular ways, and in ways that varied by region. In practical terms, these “recipes” hint at deep ecological knowledge long before farming transformed European landscapes.
A kitchen written in soot
The evidence comes from “foodcrusts,” the carbonized layers that can form when stews or porridges scorch and cling to the inside of a pot. Anyone who has had to scrub a burnt pan knows how stubborn that layer can be, and it turns out it can also carry clues for thousands of years.
The team examined 85 pottery vessels from 13 archaeological sites across Northern and Eastern Europe, dating from the 6th to the 3rd millennium BCE (roughly 6,000 to 2,000 BCE). Under high-powered microscopes, identifiable plant tissues showed up in 58 of those 85 foodcrust samples.
These were not vague traces, but recognizable fragments from wild grasses and legumes, fruits, and underground storage organs like roots and tubers, along with leaves and stems from herbaceous plants.
Animal fats still dominated the chemical signatures, and the study says those fats were mainly from freshwater fish or shellfish. A small number of samples also showed dairy fats, which the authors say were most likely obtained from neighboring farming groups.
A clay pot simmers over an open flame, echoing how ancient Europeans prepared meals combining wild plants and other foods.
Why older tests missed the greens
For years, one of the most common tools for reconstructing ancient diets has been lipid residue analysis, which reads the fats absorbed into pottery. The catch is that this approach tends to spotlight animal foods because fats from fish and other aquatic foods can overwhelm weaker plant signals.
To get around that bias, the researchers paired chemical and isotope tests with high-resolution imaging, including digital microscopy and scanning electron microscopy.
They ran lipid tests on 74 foodcrust samples and bulk isotope tests on 64 samples when enough material remained, then used microscopy to look for diagnostic plant and animal tissues in the charred layers. In the chemistry alone, plant lipids showed up as minor components in 52 out of 85 samples.
They also did something refreshingly hands-on. In controlled cooking experiments, they simmered guelder rose berries (Viburnum opulus) in about 3.4 fluid ounces of water, sometimes mixing the fruit with freshwater carp in a 1:1 ratio, and kept some stews around 250–320°F while the crust formed.
Regional recipes, not one “paleo” menu
The most striking part of the results is how specific the plant choices were from place to place. In the upper and middle Don River basin, foodcrusts commonly held seeds from wild grasses and wild legumes, with some seed traits suggesting clover relatives, and these plant seeds often appeared alongside freshwater fish.
Move into the Upper Volga and Dnieper-Dvina area, and the pattern shifts. Pots from those sites frequently contained guelder rose berries and “green vegetable” tissues from plants in the amaranth family, including goosefoot types, again often cooked with fish.
Along the Baltic, the story becomes even more locally tuned. At the Polish coastal site of Dąbki, Viburnum berries showed up repeatedly in pots that also carried strong freshwater fish signals, while other sites preserved evidence of different plant parts such as roots and tubers.
The bitter berry that shows real know-how
Viburnum opulus deserves special attention because it is not a simple “grab-and-eat” fruit. The study notes that the berries are mildly toxic when raw, taste bitter, and can release a strong pungent smell during cooking, which means preparation choices matter.
So why keep putting it in the pot? At the site of Zamostje 2, the berries were preferentially cooked with freshwater fish, a pairing the authors suggest could make the fruit more palatable after heating.
Their experimental cooking and microscopic comparisons backed up the identification of Viburnum tissues embedded in the ancient crusts. It is a small detail, but it reads like practical knowledge passed along at the fire.
What this changes for archaeology and sustainability
One of the study’s clearest takeaways is methodological, and it matters beyond this one dataset. The authors warn that plant processing in pottery is likely “grossly underrepresented” when researchers rely on lipid tests alone, since plants may leave recognizable tissues even when they barely register in the fat chemistry.
The bigger picture is about how humans interacted with ecosystems before agriculture. By showing selective plant harvesting and repeated plant fish combinations, the findings add weight to the idea that forager “plant management strategies” may have helped set the stage for later cultivation.
Next time someone claims there was one timeless hunter-gatherer diet, it is worth remembering what these pots are really saying. Local landscapes shaped local menus, and the people cooking on those fires seemed to know their plants as well as their fish.
The study was published in PLOS One.
