Radiocarbon dating pitfalls: freshwater fish make medieval skeletons appear centuries too old

A skeleton dated far too early at Svendborg

The Danish town of Svendborg once housed a Franciscan monastery. As in other monasteries across the country, being buried close to the church, or even inside it, was considered a privilege. One such burial, known as skeleton III9B, now poses a problem.

Its radiocarbon date placed it between 1100 and 1155 CE. Yet the monastery was only founded in 1236, nearly a century later. The two pieces of information cannot both be correct.

Twenty-seven burials re-examined

Kaare Lund Rasmussen, emeritus professor at the University of Southern Denmark and a specialist in chemical analyses of heritage, re-examined the dating of 27 people buried in Danish monasteries, in their cemeteries or inside churches, together with Johannes van der Plicht of the Centre for Isotope Research in Groningen and Jesper Lie Boldsen of forensic medicine.

Five other skeletons, in addition to III9B, were found to be wrongly dated. Among them is an individual from the Our Lady's monastery in Viborg, whose dating range went back to 976, more than two centuries before the building was constructed. Three burials at the Cistercian monastery of Øm show the same problem.

When fish age the bones

Radiocarbon dating relies on the amount of this isotope present in organic matter, which steadily declines after death. The problem lies in diet. Fish from lakes and rivers sometimes contain less carbon-14 than food from terrestrial sources, because old carbon is released by sediments along the banks.

When a person eats large quantities of these fish, part of this old carbon becomes fixed in their bones. Centuries later, the measurement then makes these remains appear older than they really are. This phenomenon is known as the reservoir effect.

An effect that is hard to correct

The researchers stress one point: there is no universal correction. According to Kaare Lund Rasmussen, a fish from the Rhine can appear up to 1,300 years older than a contemporary terrestrial organism, but the effect varies enormously, from zero to 1,000 years. Subtracting a fixed number of years would therefore be an error.

Seawater produces a comparable but more predictable effect. In the North Sea it is estimated at about 400 years, which allows a correction based on a known reservoir effect. The authors call on archaeologists to take this "freshwater trap" into account and apply the necessary corrections.

A question that goes beyond Denmark

According to the authors, the problem is not limited to Denmark and probably affects a large part of Europe. It is especially relevant to studies of populations whose diet included a lot of fish, whether humans or animals.

For French archaeological sites, particularly abbeys and priories located on the banks of rivers or lakes, this warning underlines the importance of cross-checking radiocarbon dates with historical context and dietary analysis.