Evaluating Mesolithic settlement patterns in mountain environments (Dolomites, Eastern Italian Alps): the role of research biases and locational strategies.

The wealth of Mesolithic evidence in the Alpine environments makes it possible to attempt a reconstruction of highland settlement patterns based on the distribution of known sites. However, just how representative this site distribution is has not yet been fully tested and the impact of research biases on the spatial organisation of Mesolithic findspots is not clear. In order to tackle these issues the locational pattern of Mesolithic sites recorded in an upland area of the Venetian Dolomites (North-Eastern Italy) was analysed. Point pattern analysis was used to correlate site distribution with two sets of covariates mirroring research biases and prehistoric settlement preferences. Point-process models were created and compared using both standard Akaike and Bayesian Information Criteria. Results indicate that both factors equally influence the current site distribution. The low number of statistically significant variables – slope and land-use – suggests the existence of additional variables that were not considered. An additional model helped us explore the importance of alternative variables and provided new perspectives for future investigation of high-altitude Mesolithic landscapes, with particular attention to highland mobility.