Fine-scale characterization of dispersed-village landscapes: data challenges and methodological advances for rural landscape mapping and assessment in Pomerania, Poland

Dispersed-village landscapes in Pomerania are shaped by local environmental conditions and long-term farming practices, yet standard spatial datasets generalize many fine-scale landscape elements. In this pilot study, we developed and tested a fine-scale mapping and assessment approach at the 64-ha Knyki-1 reference site using a custom 3-level land-use/land-cover (LULC) scheme, an uncrewed aerial vehicle (UAV)-derived digital surface model (DSM), field mapping, plant-cover inventories, and GIS analysis. We compared custom mapping with CORINE Land Cover (CLC) and the Topographic Objects Database (BDOT10k) using landscape- and class-level metrics. At the Knyki-1 site, standard datasets strongly generalized or omitted small patches, linear elements, woody structures, and structural transition zones that may correspond to ecotonal structures. The custom mapping represented 21 LULC classes and 148 patches, versus 2–6 classes and 6–28 patches in the standard datasets, showing how dataset choice affects landscape representation and structural inference. Field inventories documented ancient-forest indicators, archaeophytes, and protected vascular-plant taxa across multiple landscape components. The resulting fine-scale baseline may support future repeated monitoring and, after multi-site validation, regional landscape assessment, conservation planning, and management. These natural–cultural rural landscapes remain absent from national landscape classifications and are increasingly vulnerable to pressures from the abandonment of traditional agriculture.