Extracting data for latitude, climate, and species’ intrinsic traits
In order to assess the relationship between survival and latitude, we recorded the geographic coordinates for each species in each study from information provided in the paper itself or by locating the study area on Google Maps. For 26 studies that measured survival over broad spatial scales, such as at the national or continental level (e.g., the MAPS dataset; DeSante et al. 2015), we calculated the centroid of the breeding range for each species within the area specified by the study with occurrence data extracted from eBird using the auk package (Strimas-Mackey et al. 2018) in R (v.3.5.3; R Core Team 2019). This allowed us to estimate a unique latitude and longitude for the centroid of each species’ realized breeding range rather than simply selecting an unweighted point in the study area itself. As latitude is often used as a surrogate for variation in climatic conditions between the north and south poles, we evaluated the predictive power of three key extrinsic factors that characterize the environment of a species and are hypothesized to influence avian survival: annual precipitation (Rockwell et al. 2017; Shogren et al. 2019), minimum winter temperature (Robinson et al. 2007; Salewski et al.2013), and temperature seasonality (Ricklefs 1980; Lloyd et al.2014). We also tested whether species’ intrinsic traits explained global patterns in avian survival rates by collecting data on body mass, clutch size, and species’ migratory habit, which we obtained from information contained in the paper, published reference databases (i.e., Jetz et al. 2008 for clutch size; Wilman et al.2014 for body mass; Barçante et al. 2017 for migration), or the Handbook of birds of the World Alive (del Hoyo et al. 2018). Further details describing our data compilations methods are available in Supporting Information S2.