Monitoring Swifts Apus apus by timed counts at breeding sites and effects of cloud cover on relative abundance

ABSTRACT A method for monitoring the abundance of Swifts flying close to nesting sites based on timed counts was developed and tested. Counts lasting up to 30 minutes were made between early May and late July each year during 2019–24 at over 50 sites across 680 km2 of East Kent. Numbers varied with site, date, time of day, duration of count, and ambient rainfall and wind speed. Temporal and environmental variables were statistically accounted for in a survey protocol that involved three 10-minute counts at each site made between 16 June and 23 July, two to seven hours after sunrise in dry weather and wind speeds under 21 km/h. Numbers of Swifts increased from early May to reach a peak in early July, when their behaviour suggested that most birds were non-breeders. Total mean site count for a sample of 10–30 sites   was compared with that in the previous year, to estimate the annual change in abundance. Overall, abundance fell by 57% during the study. Year-to-year changes, which varied from -47% to +17%, were strongly and very significantly negatively correlated with mean daily cloud cover in the period 1 May to 23 July. A model is proposed explaining how cloud cover may regulate food availability and the abundance of non-breeding Swifts at breeding sites, and may also prompt long-distance feeding movements. A residual, steady annual decline of 9–12% in Swift abundance remains unexplained.