Geographic variation in urbanization filter effects on birds in China

We conducted field surveys in the capital cities of all 34 provincial administrative regions in China (Figure 1). During the bird breeding season of 2023 (June-August), bird surveys were conducted using a standard point-count method (Bibby et al., 1992). Survey points were set in well-vegetated urban areas such as schools, parks, resident community and road. Each survey point was monitored for 10 min, during which all avian species observed or heard within a 50 m radius were recorded. To reduce the probability of duplicate records, adjacent survey points were spaced at least 150 m apart. Each survey point was surveyed only once, with at least 40 survey points set in each city (some cities like Macau and Hong Kong had only 30 points due to their smaller size). Geographic coordinates and altitude were recorded for each survey point. The point-count surveys were conducted by 29 surveyors, each with over 2 years of birdwatching experience and demonstrated the skill to accurately identify the bird species in their respective cities. In total, we recorded the distribution and abundance of 231 bird species across 1,364 survey points in the 34 capital cities (Table S1).To obtain the potential regional species pool for each province, we collected the distribution data of all bird species across the Chinese provinces (Zheng, 2023). We excluded species belonging to the orders Phoenicopteriformes, Phaethontiformes, Gaviiformes, Procellariiformes, Ciconiiformes, and Suliformes, as these species inhabit coastal wetlands or open oceans and are rarely recorded in urban areas. Additionally, we excluded nocturnal species belonging to the order Strigiformes because our field surveys were conducted during the day only. We aligned the taxonomic framework of the species pool based on the work of Jetz et al., (2012) for subsequent phylogenetic analyses. Ultimately, we obtained distribution data for 1,186 bird species across the 34 provincial administrative regions in China (Table S2).In the present study, we define urban utilizers as bird species recorded at more than one survey point during our field surveys. Other species from the regional species pool were defined as urban avoiders. This approach was devised to avoid the misclassification of rare species that could have been recorded accidentally, thus ensuring a more accurate distinction between urban utilizers and avoiders. We also conducted a sensitivity analysis, employing a higher threshold (n = 2 survey points) to examine whether it would yield different outcomes.Here, we conducted field surveys of breeding bird communities in the capital cities of 34 Chinese provinces and categorized species as either urban utilizers or avoiders based on their occurrence rate among regional species pools. Bayesian phylogenetic generalized linear mixed model was used to investigate potential trait associations between urban utilizers and urban avoiders. To investigate the relative effects (measured by the model coefficients) of urbanization filtering across different environmental gradients, we incorporated the model coefficients of each trait as the response variables with city-specific environmental characteristics as predictors to constructed linear mixed models. We found that urban utilizer birds exhibited greater behavioral innovation and broader habitat ranges than urban avoider birds. Additionally, we observed that the filtering effects of urbanization on different traits exhibited significant variations, showing gradients in latitude, altitude, and artificial night light intensity. This variation along geographic gradients could potentially elucidate the inconsistent findings reported in previous work on bird adaptation to urban environments. Our findings highlight the context-dependent nature of urbanization filtering effects and contribute to a better understanding of the impact of urbanization on urban bird communities.