Mapping Riverscape Connectivity : What Geospatial Routing Can Tell Us About Invasive Minnow Spread in Sub-Arctic Catchments

Sammanfattning

The rapid expansion of the invasive European minnow (Phoxinus phoxinus L.) across sub-Arctic catchments threatens pristine salmonid habitats. Effective management requires reliable landscape screening tools to predict invasion pathways and identify structural migration barriers. This study developed a digital elevation model (DEM)-based hydrological routing framework to map riverscape connectivity and predict minnow distribution in the Kuokkel area, Sweden. Using presence/absence data derived from environmental DNA (eDNA) and trap and gill net catches, three modelling approaches were evaluated under a nearest downstream lake routing assumption: standalone maximum elevation drops, stream channel slope thresholds, and combined multi-variable formulations. The single-variable 55 cm elevation-drop model emerged as the most parsimonious framework, achieving a peak classification accuracy of 0.704. Standalone slope-only models plateaued at a lower accuracy of 0.634 (at a 40% threshold), while combined multi-variable models (55 cm drop + slope thresholds) failed to outperform the baseline, peaking at 0.676 and suffering from degraded specificity due to structural noise. To demonstrate its utility for proactive environmental management, the optimized 55 cm model was extrapolated across the catchment using point-represented centroids to project migration barriers for all unsampled lake basins connected to lakes Vassijaure, Låktajaure, Paktajaure, and the Torneträsk drainage network. Despite limitations inherent to static terrain snapshots and some localized anthropogenic distortions, this point-based geospatial framework provides a resource-efficient preliminary screening tool for tracking and managing invasive species spread in complex, dynamic sub-Arctic aquatic networks.