Multiscale pollen-based reconstructions of anthropogenic land-cover change in Karula Upland, south Estonia

Pollen-based quantitative vegetation reconstructions using multiple sedimentary basins from the same area, along with their quantified relevant pollen source areas, are a powerful means to study how long-term human impact has affected vegetation and shaped the currently protected heritage landscapes at different spatial scales. Our study presents the outcome of a palynological investigation in Karula Upland, south Estonia, for the last 6500 years. Centennial-resolution pollen records from one large (175 ha) and three small (5 ha) lakes, and one small bog (0.1 ha) were used to reconstruct the vegetation at different spatial scales using the Landscape Reconstruction Algorithm. The results are discussed in combination with archaeological sites and historical knowledge.

The first signs of small-scale forest clearings connected to local human settlements are already visible in the Middle Stone Age (3100–4100 BCE). The first finds of cereal pollen (2500 BCE) from Lake Ähijärv suggest that grain crops were introduced to south Estonia during the Late Stone Age. The evidence of local crop farming in Karula is traceable since the Bronze Age. The widespread practice of slash-and-burn agriculture led to a major shift in land-cover with replacement of old-growth forests with the early-successional birch, occupying long-term fallows, during Late Bronze Age and Early Iron Age (700–250 BCE). A notable regression in farming is visible during the second part of the Early Iron Age (100–600 CE), with the most prominent change taking place around 500 CE, roughly coinciding with the 6th century Northern Hemisphere climate cooling and Migration Period. Permanent fields gained importance alongside slash-and-burn cultivation, during the Late Iron Age, ca 600–700 CE, shifting the vegetation composition towards more open land-cover. The ∼50 % open mosaic land-cover of the heritage landscape, protected today in Karula Upland, was formed during the Late Iron Age.

The current study shows that sedimentary basins as close as ca 2 km from each other sometimes tell different stories, highlighting the need to quantify the size of the pollen source area to combine successfully archaeological, historical, and palynological evidence. Quantitative pollen-based vegetation reconstructions provide an environmental context for known, and possibly unknown, archaeological evidence within the pollen source area.