How to build a biodiverse city: environmental determinants of bird diversity within and among 1581 cities

Cities are novel environments compared with the evolutionary history of the species that reside within them. Collectively, cities and their fauna can be thought of as ecosystems, recognized as playing a critical role in supporting global biodiversity, but they are fundamentally a combination of 'old species' surviving or thriving in a new environment. We aimed to understand-at a broad macroecological scale-how biodiversity responds to urban ecosystems both among and within cities. We integrated > 5 million eBird citizen science observations with remotely sensed landcover products throughout 1581 cities within the continental United States. We first investigated the species-area relationship as it pertains to cities and compared the slope of this relationship to randomly sampled polygons (i.e., among cities). Second, we investigated how biodiversity responds to an urbanization gradient at the level of localized bird observations (i.e., within cities). We found strong support for the longstanding species-area relationship: geographically larger cities had greater species richness. Surprisingly, the species-area relationship was stronger (i.e., steeper slope) in cities when compared to the species-area relationship for randomly sampled polygons in the study region. Our findings suggest that diverse and heterogeneous cities play a significant role in supporting biodiversity. But we also found that there is a consistent threshold where the level of urbanization begins to profoundly and negatively affect biodiversity. Critically, urban planning at the city-scale and at a local-scale (e.g., neighborhood) should focus on preserving attributes of water-cover and tree-cover for increased biodiversity to keep as much of the city as possible above this threshold value.