Low‐intensity urban development delays the spring departures of nonbreeding migratory bird populations

Abstract Urban areas alter the timings (phenologies) of seasonal processes for plants and animals, yet the effects on bird migration, particularly pre‐breeding migration, are not well understood. Higher temperatures, higher levels of artificial light at night, and earlier vegetation emergence in urban areas can advance spring seasonal processes, including departure dates for pre‐breeding migration. Using occurrence data from eBird for six songbird species wintering in the southern United States, we examined how the probability of observing a species changed over time (from November to June) across an urbanization gradient (approximated by impervious surface area). We used changes in species' observation probabilities to estimate population‐level departure dates, defined as the dates when the probability of observing a species had dropped to 90%, 50%, and 10% of the maximum observation probability. Contrary to the earlier plant phenologies and avian breeding phenologies typically observed in cities, population‐level departures occurred later in areas with low‐intensity urban development containing intermediate levels of impervious surface area (20%–40%). These results could be a consequence of different individual‐level behaviors occurring on the nonbreeding grounds, including delayed departures or local movements. Determining the exact mechanism requires further study, but our findings suggest urban development is broadly affecting the behavior of these species as they depart their nonbreeding grounds, which could subsequently affect survival and reproductive success.