The Power of Many: Utilizing Citizen Science Data in Species Distribution Models to Forecast Urban Avian Biodiversity in the Metropolis Ruhr, Germany

In the last couple of years, the Citizen Science (CS) approach became more reliable, generating a huge amount of environmental data (‘Big Data’) considering the help of volunteers and amateur researchers. In the same time, Remote Sensing capabilities and resolutions improved significantly. Especially in urban environments such as the Ruhr Metropolis, where private lands and high diverse landscapes are predominant and large concentrations of people as potential volunteers are available, combining CS data and Remote Sensing techniques with the predictive power of Species Distribution Models can play an important role to comprehensively investigate and evaluate avian biodiversity representing keystone species in urban ecosystems. However, spatially modeled habitat suitability for multiple avian species in dense and fragmented urban environments are still lacking.

An ensemble of different machine learning algorithms, CS datasets of multiple avian species expected to react differently to urban conditions and environmental predictors consisting of bioclimatic variables, digital surface models and land use derived information were applied to forecast avian biodiversity patterns and distributions. These comprehensive predictions of habitat suitability enable policymakers to make sophisticated decisions in landscape planning and conservation taking into consideration present and future land use changes due to urban densification and urban sprawl, especially within the context of an ongoing historical unique transformation process in the Metropolis Ruhr.