Mapping lost woodland. A modeling experiment based on place names and historical references aimed at inspiring reforestation

After many centuries much of Europe is now largely deforested. Today, plans for reforestation are ongoing in the Netherlands and elsewhere. These plans have the potential to significantly impact the landscape. As a contribution to the current debate on large-scale reforestation in the Netherlands we have therefore conducted a quick and overall scan to establish the location of woodland remnants in the high and late Middle Ages (AD 1000-1500). Both the scan and the subsequent verification and critical evaluation of the resulting reconstructions made extensive use of digital datasets and existing inventories (place names, field names, historical maps, charcoal production sites, ancient woodland, ancient woodland indicator plants). Two different period maps were produced and tested: period map 1 (AD 750-1350), based on woodland-related place names; and period map 2 (AD 1250-1650), based on historical woodland references. The distribution map for the most recent period proved to be essentially correct, but an assessment of the accuracy of the map for the oldest period was difficult. It was also possible to establish that the locations of medieval and modern woodland do not coincide for the most part. Together with the long-term spatial dynamism of ancient woodland, this observation is an argument for looking beyond existing woodland locations and types when placing and creating new woodland.