The history of Castanea sativa (sweet chestnut) cultivation since medieval times has been well described on the basis of the very rich documentation available. Far fewer attempts have been made to give a historical synthesis of the events that led to the cultivation of sweet chestnut in much earlier times. In this article we attempt to reconstruct this part of the European history of chestnut cultivation and its early diffusion by use of different sources of information, such as pollen studies, archaeology, history and literature. Using this multidisciplinary approach, we have tried to identify the roles of the Greek and Roman civilizations in the dissemination of chestnut cultivation on a European scale. In particular, we show that use of the chestnut for food was not the primary driving force behind the introduction of the tree into Europe by the Romans. Apart from the Insubrian Region in the north of the Italian peninsula, no other centre of chestnut cultivation existed in Europe during the Roman period. The Romans may have introduced the idea of systematically cultivating and using chestnut. In certain cases they introduced the species itself; however no evidence of systematic planting of chestnut exists. The greatest interest in the management of chestnut for fruit production most probably developed after the Roman period and can be associated with the socio-economic structures of medieval times. It was then that self-sufficient cultures based on the cultivation of chestnut as a source of subsistence were formed.

Knowledge about the glacial refugia of the thermophilous European Castanea sativa Mill. (sweet chestnut) is still inadequate. Its original range of distribution has been masked by strong human impact. Moreover, under natural conditions the species was probably admixed with other taxa (such as Quercus, Fraxinus, Fagus, Tilia) and thus possibly represented by low percentages in pollen records. In this paper we try to overcome the difficulties related to the scarcity and irregularity of chestnut pollen records by considering 1471 sites and extending the palynological approach to develop a Castanea refugium probability index (IRP), aimed at detecting possible chestnut refugia where chestnuts survived during the last glaciation. The results are in close agreement with the current literature on the refugia of other thermophilous European trees. The few divergences are most probably due to the large amount of new data integrated in this study, rather than to fundamental disagreements about data and data interpretation. The main chestnut refugia are located in the Transcaucasian region, northwestern Anatolia, the hinterland of the Tyrrhenian coast from Liguria to Lazio along the Apennine range, the region around Lago di Monticchio (Monte Vulture) in southern Italy, and the Cantabrian coast on the Iberian peninsula. Despite the high likelihood of Castanea refugia in the Balkan Peninsula and northeastern Italy (Colli Euganei, Monti Berici, Emilia-Romagna) as suggested by the IRP, additional palaeobotanical investigations are needed to assess whether these regions effectively sheltered chestnut during the last glaciation. Other regions, such as the Isre DØpartement in France, the region across northwest Portugal and Galicia, and the hilly region along the Mediterranean coast of Syria and Lebanon were classified as areas of medium refugium probability. Our results reveal an unexpected spatial richness of potential Castanea refugia. It is likely that other European trees had similar distribution ranges during the last glaciation. It is thus conceivable that shelter zones with favourable microclimates were probably more numerous and more widely dispersed across Europe than so far assumed. In the future, more attention should be paid to pollen traces of sporadic taxa thought to have disappeared from a given area during the last glacial and post-glacial period.

No abstract