West Anatolia (WA) has been a center of human population through the Holocene period. Theories about a past tree-cover reduction of this historically and biologically important spot are divergent between natural climatic and anthropogenic impacts. While WA climate changed from wetter to drier conditions throughout the Holocene, the latter preserves its reliability considering the archeological records on settlement activity. Previous studies investigated WA vegetation by either using proxy records only or simulating this region as part of bigger domains. Nevertheless, WA has not been the focus of a regional modeling study linking predictions with findings from climatology, palynology and archaeology. Such focus is essential for constructing a compact understanding about the living conditions of past human civilizations, and for determining the sensitivity of the region to climatic and anthropogenic impacts. Here, we simulate the Mid- and Late Holocene (MH and LH, respectively) WA vegetation using a dynamic vegetation model (LPJ-GUESS coupled to an eco-informatics toolbox, PEcAn) to predict the natural potential vegetation change due to climate. In addition, we compiled pollen data from sites that cover the WA north-to-south for inferring the local vegetation history as recorded by pollen, and determining whether the changes in the region occurred synchronously.
Results/Conclusions
A comparison of simulated present-day vegetation at our pollen sites to modern vegetation maps of West Anatolia and surface pollen-data showed that the model simulated the present vegetation accurately on the basis of Plant Functional Types (PFTs). An agreement of the past vegetation as simulated by the model and the pollen data, and a more-or-less synchronous signal in all pollen records would indicate a climate-caused tree-cover change. All pollen records show a decrease in arboreal plants and an increase in herbaceous plants during 4-2 ka BP, however with differences in the onset of this transition which could be due to different settlement times at these sites. While there were changes in the community composition of woody-PFTs, such tree-cover reduction is not detected in the model outputs when comparing simulated present-day, LH and MH vegetation. Hence, our preliminary results suggest an anthropogenic impact on vegetation change during the LH. Our findings are also consistent with archaeological records from this region. A possible interplay of climate and human-induced vegetation change will be investigated with further past climate scenarios, and a region-specific parameterization of the vegetation model.