Abstract
Barley is one of the main cereals found in archaeological sites in the north-western Mediterranean basin, over the last 8 millennia. Grains are preserved in archaeological sediments by charring after or before dehusking. Morphological criteria for distinguishing 2-row from 6-row barley, but also hulled barley from naked barley, can be affected by charring and this can complicate identification of barley subgroups. In the last decade, geometric morphometrics applied to uncharred barley grains has shown the possibility of identifying barley subgroups, but its applicability to archaeological remains still needs to be ascertained. We used 3985 raw and charred grains of 113 current varieties to (1) assess charring effects on the shape of barley caryopses, depending on their subgroup and whether they were charred husked or dehusked and (2) select the best dataset for identifying barley subgroups. We also used 700 archaeological grains, from the Neolithic period to the end of the Middle Ages, to (3) compare the morphometric taxonomic assignment of the grains with their carpological identification based on discrete anatomical features in a series of 7 archaeological samples. Our results indicate size and shape differences between barley grains when charredhusked or dehusked. Although the charring process results in greater morphometric homogeneity, it allows subgroups of barley to be identified. For the 2-row vs. 6-row identification, more than 71% of the charred modern grains can be correctly identified. For the hulled vs. naked identification, the correct cross-validation percentages range from 70 to 73%. Finally, the good correlation between carpological and morphometric identifications of archaeological grains suggests that it is possible to identify archaeobotanical samples in the future.