Inferring seabird activity budgets from leg-mounted time–depth recorders

Leg-mounted loggers are increasingly used in seabird activity studies, but few studies have validated the information obtained about bird behaviour with independent data. Using Brunnich’s Guillemot Uria lomvia as a study species, we show by comparing interpretations of time–depth recorder (TDR) data with visual observations that activity budgets inferred from leg-mounted TDRs provide reliable information on colony attendance, and validate information on flight time by comparing periods interpreted as flight based on TDR data with periods interpreted as flight based on GPS speed information. Yet, special attention is needed because auks resting at sea occasionally withdraw one leg and/or foot into the plumage (leg-in-plumage). During this behaviour, the TDR may be warm and dry, potentially leading to spurious identification of colony visits. In our case study, spurious identification of colony visits would have resulted in mean trip duration being underestimated by a factor of 4, and number of trips being correspondingly overestimated. We therefore urge great care when deriving activity budgets from leg-mounted TDRs, but nonetheless recommend using leg-mounted TDRs to infer activity budgets from diving seabirds, particularly for longer deployments.