Abstract
Solar farms offer an opportunity for habitat creation for wildlife, including insect pollinators, potentially simultaneously contributing to both low‐carbon energy and nature recovery. However, it is unknown whether cobenefits would persist under future land‐use change given that habitat value is context dependent. For the 1042 operational solar farms in Great Britain, we predict their ability to support bumblebee populations (both inside and outside the solar farm) under three different socioeconomic futures. These futures represent alternative 1 km scale landcover projections for the year 2050 with accompanying narratives. We downscale these to 10 m resolution, spatially allocating crop rotations, agri‐environment interventions and other habitat features consistent with the scenario narratives, to realistically represent fine‐scale landscape elements of relevance to bumblebee populations. We then input these detailed maps into a sophisticated process‐based model that simulates bumblebee foraging and population dynamics, enabling us to predict bumblebee density in and around Great Britain's solar farms, accounting for the effects of their changed habitat context and configuration in these different future scenarios. We isolate the drivers of bumblebee density change across scenarios and scales and show that solar farm management was the main driver of bumblebee density within solar farms, with ~120% higher densities inside florally enhanced compared to turf grass solar farms, although the exact figure was influenced by wider landcover changes. In foraging zones immediately surrounding solar farms, landscape changes had a greater impact on bumblebee densities, suggesting a single solar farm in isolation generally did not counteract the influence of wider land‐use changes expected under future scenarios. In addition to providing insights into the potential future value of pollinator habitat on solar farms, our methodology demonstrates how combining process‐based modelling with landcover projections that are downscaled to ecologically relevant resolutions can be used to better assess future effectiveness of habitat interventions. This represents a step change in our ability to account for species' interactions with socioeconomically driven futures, which can be extended and applied to other taxa and land‐use interventions.