Using species-habitat models to predict bird counts from urban development plans

If urban environments are designed from the outset to be biodiversity-friendly, they can support resilient wildlife communities, improve the wellbeing of inhabitants, and reduce overall costs due to a reduced need for buying compensatory land. Urban planners might best design productive green infrastructure through predictive tools that output rapid, quantitative, and reliable wildlife statistics from digitized plans. Here, we demonstrate how species-habitat models could provide the foundation for such a tool. We took plans for nine housing developments across England and Wales and integrated them with high-resolution digital maps to derive the urban landscape measures used to parameterise previously developed species-habitat models. We produced predictions of abundance for 55 common bird species per development, which we summarised into bird trait groups and overall diversity. We compared predictions between planned developments and designs with alterations to greenspace, woodland, and housing configuration in five scenarios. We found that biodiversity predictions for actual plans were lower than at least one scenario per development, but that the scenario that achieved the highest species diversity varied between sites. This indicates that optimal solutions for bird di-versity depend upon the biophysical context of each planned development. We believe that this framework could provide a cost-effective and data-driven approach to support green infrastructure and be of particular importance for urban planners in jurisdictions where such considerations are mandatory.

Automated summary

If urban environments are designed to be biodiversity-friendly from the beginning, they can benefit wildlife communities, improve residents' well-being, and reduce costs by minimizing the need for compensatory land. Urban planners can use predictive tools based on species-habitat models to design productive green infrastructure efficiently. By integrating digitized plans with digital maps, we predicted bird abundance and diversity for nine housing developments in England and Wales, comparing different scenarios with alterations to greenspace, woodland, and housing configuration. The results showed that biodiversity predictions for actual plans were lower than at least one scenario per development, emphasizing the importance of considering the biophysical context for maximizing bird diversity. We believe this framework can provide a cost-effective and data-driven approach to support green infrastructure planning, especially in jurisdictions where biodiversity considerations are mandatory.