Miliacin from broomcorn millet – a fossil molecule of multiple interests for archaeologists and paleoenvironmentalists

Miliacin (olean-18-en-3β-ol methyl ether) is a fossil molecule that was recently detected in Lake le Bourget (LlB, French Alps) sediments. Comparison of potential plant producers with plants thriving around the lake at time of sedimentation concluded that Panicum miliaceum cultivated in the catchment was its most probable source. Since this discovery, it became possible to track the former presence of P. miliaceum over its area of extension and diffusion by detecting miliacin in sediments. This first allowed us reconstructing the history of millet cultivation around LlB, characterized by a beginning around 1700 B.C. and a maximum at 800 B.C. before a drastic drop at the Bronze Age/Iron Age transition. Examination of miliacin in sediments from Lake Lledro (Italy) and Paladru (France) revealed distinct dates of its first occurrence, attesting to a complex pattern of diffusion through the Alps. A twin high resolution record of miliacin in LlB Bronze Age sediments evidences periods of low and high concentrations that could result from varying levels of consumption potentially related to climate changes. While trying tospotpaleo-millet fields cultivated during the Bronze Age around LlB, we have analyzed soils in its surroundings. Miliacin is present but not abundant in soils and there is no logical pattern in its spatial distribution. Insufficient miliacin concentrations prevented any 14 C measurement that would have ascertained a molecular legacy of soils from Bronze Age millet cultivation. Reversely, we found miliacin in paleosoils from Ukraine that attested to the in situ cultivation of millet by Iron Age nomad populations. Miliacin is not solely produced by P. miliaceum. Previous studies indicate it can be found in a large spectrum of organisms (gramineae but also, surprisingly, molds), of which Setaria italica, the Italian millet. We analyzed seeds of several strains of P. miliaceum, several Panicum sp. and other gramineae such as S. italica. We found miliacin very abundant in any cultivar of P. miliaceum, present in all Panicum sp., in somePennisetum sp. but absent fromthree varieties of S. italica. In addition, we were unable to recover miliacin from Chaetomiumolivaceum, a mold, although operating under the