Abstract
BACKGROUND: To understand life-history strategies in migratory bird species, we should focus on migration behaviour and possible carry-over effects on both population and individual level. Tracking devices are useful tools to directly investigate migration behaviour. With increased use of tracking devices, questions arise towards animal welfare and possible negative effects of logger on birds. Several studies were conducted to address this question in birds that were tagged and tracked for one complete non-breeding season including migration but with mixed results. To detect individual-based decisions regarding migration strategy, we need to track the same individuals several times. So far, there are no studies investigating effects of repeatedly tagging on reproduction and life-history traits in individual migratory birds, especially in small birds. METHODS: We used long-term data of 85 tagged common swifts (Apus apus), a long-distance migratory bird, of a breeding colony in Germany to test whether carrying a geolocator or GPS logger once or repeatedly during non-breeding season affected return rate, apparent survival, and parameters determining reproductive success. Additionally, we checked for individual differences in arrival date and breeding parameters when the same individuals were tagged and when they were not tagged in different years. Further, we calculated the individual repeatability in arrival at the breeding colony and date of egg laying in repeatedly tagged swifts. RESULTS: Once and repeatedly tagged birds returned to the colony at a similar rate as non-logger birds and arrived earlier than non-logger birds. We found no effect of logger-type on return rate in logger birds. We detected no differences in apparent survival, time lag to clutch initiation, date of clutch initiation, clutch size, number of chicks and fledglings between logger and non-logger birds. We found neither an effect of loggers nor of logger-types on the arrival date and breeding parameter on individual-level. Arrival date was highly repeatable and date of clutch initiation was moderately repeatable within repeatedly tagged individuals.