Data and code for 'Resource use and associated energy expenditure of African crowned eagles (Stephanoaetus coronatus) in an urban-forest mosaic landscape'

Abstract Rapid urbanisation is a major driver of global change. Understanding how wildlife persists in highly transformed landscapes is increasingly important for conservation. Using insights from movement (GPS) and accelerometry (ACC) data, we aimed to understand the behavioural and physiological mechanisms associated with an apex predator’s successful exploitation of transformed landscapes. The African crowned eagle (Stephanoaetus coronatus), a subtropical forest eagle, breeds at unusually high densities in Durban, eThekwini Municipality, South Africa. Here, they inhabit an urban-forest mosaic environment - much of which falls under the Durban Metropolitan Open Space System (D’MOSS), a network of protected and semi-natural green spaces embedded within the city. Using data from five GPS-tagged crowned eagles, we assessed individual resource use patterns through continuous time Resource Selection Functions. We then combined high-resolution GPS and ACC tracking data to determine how energy expenditure, inferred by Overall Dynamic Body Acceleration, varied across habitat types and with distance from the D’MOSS boundary. We found that crowned eagles selected forested habitat relative to built-up habitats, given availability. Higher energy expenditure was associated with built-up environments and increasing distance from the D’MOSS boundary. Our findings suggest that habitat structure may influence energy costs but quantifying these relationships in heterogeneous urban landscapes is complex. We underscore the importance of protected habitat networks in cities and provide information on the physiological costs and benefits of navigating urban-forest mosaic landscapes. These insights can inform urban planning to conserve natural habitat networks and ecosystem features that support the persistence of urban biodiversity, including apex predators. Data: To support data transparency, we have included the location data and code to reproduce the analyses of one individual, STV, who has since died. This includes nest-related data, but STV occupied a very well-known nest within an established community and therefore the site is considered at low risk of disturbance. stv_gps_clean.csv Cleaned and processed GPS tracking data. These data were prepared for use in RSF analysis and for integration with ACC data. stv_acc.csv Raw ACC data for the individual STV used to calculate ODBA and subsequently merged with GPS data. master_daytime_clean.csv Final processed dataset used as input for statistical models. Lat/long coordinates were removed prior to deposition to protect sensitive nest information but are not necessary to reproduce the analyses. Code: stv_ctmm_rsf.R R script used to run an individual-level RSF, adapted from a script published by Björn Reineking on AniMove. odba_merge.R R script used to process ACC data, calculate ODBA and merge with GPS data. crownies_models.R R script with modelling workflow.