Designing nature networks for cities: combining multi-species modelling approaches.

Context: Urban wildlife habitats are often fragmented and of poor quality, yet cities hold potential to support biodiversity, particularly for small-bodied species like insect pollinators. Enhancing habitat connectivity is essential for improving biodiversity and increasingly prioritised in planning frameworks. Combining diverse approaches to assess habitat connectivity may yield the greatest overall success. Objectives: We compare two multi-species modelling approaches for assessing urban ecological corridors. The first species-specific approach uses combined habitat suitability maps of four insect pollinators and assesses connectivity using resistance modelling (circuit theory). The second landscape-level approach has been developed by urban environmental planners ("Green Network Development officers") and identifies core areas as species-rich habitat patches using spatial data, species records (of plant and pollinator indicator species), and local expertise, then models connectivity between these using least-cost paths. By comparing these two approaches, we aim to identify gaps and priority areas for habitat creation or management. Methods: Using biological records, we mapped habitat suitability for pollinators and applied circuit theory to assess connectivity and identify "pinch points"-bottlenecks to movement that can be targeted for corridor enhancement. Results: While both approaches showed 39 km Conclusions: Our species-specific modelling approach identified 31 pinch points outside of planner-defined corridors, highlighting important areas of movement constraint not captured by the current planning framework. Incorporating species-specific modelling into urban planning also helps identify key habitat variables impeding movement, enhancing the biological understanding of the system. We recommend urban planners adopt multiple, complementary approaches for corridor delineation and collaborate with researchers, ecologists, and citizen scientists. Supplementary Information: The online version contains supplementary material available at 10.1007/s10980-026-02315-0.