1
Archaeological Data Management in an Interdisciplinary Environment Péter Szabó, Petr Kuneš, Jan Kolář, Helena Svitavská Svobodová, Jana Müllerová, Eva Jamrichová, Radim Hédl Institute of Botany, Czech Academy of Sciences, Brno and Průhonice, Czech Republic, [email protected] Study area: Moravia and Silesia (27 000 sq.km with more than 3500 settlements) THE LONGWOOD PROJECT (2012-2016) Long-term woodland dynamics in Central Europe: from estimations to a realistic model integrates ecological (palaeoecology, vegetation ecology) and historical (archaeology, history) data in a GIS environment Archaeological coverage: all available records of human activities (living, burying, mining etc.) in various forms of community areas (settlement, burial ground, quarry, pottery workshop, hillfort, etc.) from the Mesolithic-Neolithic transition (ca. 7000 BC) to the 13th century AD. Linked to parishes. The database after the first year of the project. Historical coverage: archival records on forest extent, tree species composition, type of management, abiotic disturbances etc. from the 12th century AD to the present. Linked to parishes. The database after the first year of the project. Palynological coverage: existing pollen stratigraphical data from the PALYCZ database and ca. 20 new sites. Point layer of the samples with metadata attached as attributes. Information on woodland cover as well as tree and herb species composition throughout the Holocene. The database after the first year of the project. Ecological coverage: tree and herb species composition. Historical forest vegetation records from the 20th century (stored in the database as point layers with metadata as attributes) are repeated to assess shifts in species composition related to the abandonment of traditional management. The database after the first year of the project. GEODATABASE woodland cover, species composition and human activities over the past 7500 years ARCHAEOLOGICAL COVERAGE HISTORICAL COVERAGE PALYNOLOGICAL COVERAGE ECOLOGICAL COVERAGE The research leading to these results has received funding from the European Research Council under the European Union’s Seventh Framework Programme (FP7/2007-2013) / ERC Grant agreement no 278065 AIMS to construct a multidisciplinary geodatabase focusing on long-term (Neolithic to present) patterns of woodland cover, structure, management and species composition to construct a spatio-temporal forest and landscape model based on the geodatabase to assess changes in woodland and the main drivers of change to relate spatio-temporal dynamics of the tree canopy composition and structure to environmental conditions and understorey herbaceous vegetation to compare the model and analytical results to approaches currently employed in forestry and nature conservation and to contribute to better informed guidelines for conservation and management COMPARISON AND INTEGRATION OF DATA ON THREE LEVELS 1.data – data (e.g. current oak distribution vs archival records on oak as a point layer) 2.model – data (e.g. REVEALS landscape model vs number of archaeological components) 3.model – model (e.g. kernel estimation based GIS reconstruction of 18th century forest composition vs potential natural vegetation model vs LOVE vegetation model for the 18th century) Plus external databases as GIS layers www.longwood.c z A PRELIMINARY EXAMPLE: landscape changes in southern Moravia The pollen site and the source area of archaeological data The most intense phases of deforestation correlate with the introduction of large scale cereal cultivation in the Roman Period and with high-medieval colonization (the latter known from written sources) Despite considerable fluctuation in the number of archaeological components, the forest/open land ratio is relatively stable from the Neolithic to the Roman Period. Does this refer to the incomplete nature of the archaeological record or to differing ways of landscape utilisation by variously-sized group that led to the same pattern of landscape openness? Increases in oak follow increases in the number of archaeological components, suggesting that oak might be dependent on management. Data and methods: 1)We estimated regional vegetation composition and land-use types based on pollen record from a lake basin using the REVEALS model. Five radiocarbon dates served to obtain age-depth relationship for dating using smoothing spline interpolation. Pollen counts within each 500-year interval were used to run the model. The radius of the regional area was set to 25 km and wind speed to 3 m.s-1 under neutral atmospheric conditions. Pollen productivity estimates for 27 pollen taxa were selected according to model validation against modern vegetation. 2)Archaeological data include all stratified finds and features within the area (settlements, burial grounds, hoards, hillforts and specialized production places and their fragments/parts). We excluded from the analysis all single finds (e.g. coins, hammer-axes). Data come from the Archives of the Archaeological Institute of the Czech Academy of Sciences in Brno, which is responsible for collecting information from all excavations in the region as well as from published papers.
