The old charcoal kiln sites in Central Italian forest landscapes

Production of wood charcoal in the Mediterranean countries started over two millennia ago and vanished almost completely only in the last century. The legacy of this activity are thousands of abandoned charcoal kiln platforms, in which soil and vegetation characteristics are deeply affected. Understanding the consequences of such effects at the forest level demands a better knowledge of the density, distribution and morphology of these sites, as well as the influence of forest type and local geomorphological characteristics. We examined these aspects using field surveys and Airborne Laser Scanning (ALS) data in 1-ha sample quadrats distributed along an altitudinal gradient in three major forest types of Central Italy, namely evergreen sclerophyllous forest, oak-dominated thermophilous deciduous forest and montane beech forest. We found on average 5.5 kiln sites per ha. The highest overall surface proportion covered by charcoal platforms was recorded in oak-dominated forests, due to their generally larger size. In beech forests, kiln platforms were more numerous than in the other two forest types, but smaller. Density was intermediate in the sclerophyllous forests, where the overall proportion of surface was lowest. The charcoal-enriched soil layer was usually single and continuous (e.g. not interrupted by mineral layers). The thickness of this layer was similar in the three forest types, but increased with slope inclination. Several features of our kiln platforms such as density and shape were distinct from others in Central and Northern Europe, probably reflecting different forest histories and purposes for which they were built. Using ALS, we could detect all kiln platforms in beech forest on steep slopes and approximately 75% of the kilns in oak forests on hilly terrain. Hence, all further ecologically-or archaeologically-oriented study in our region at the landscape level will benefit from the use of hillshade and/or slope images from ALS data. (C) 2016 Elsevier Ltd and INQUA. All rights reserved.