Reliability of VHF telemetry data for measuring attendance patterns of marine predators : a comparison with time depth recorder data

Abstract

Very high frequency (VHF) radiotelemetry data has been used for over 30 yr to monitor the behavior patterns of otariid seals. These data have been used in a wide variety of ways, from characterizing the reproductive and foraging ecology of these species to inferring ecosystem changes based on variation in attendance patterns. Yet the accuracy of VHF data has never been appropriately evaluated. Our study compares VHF data collected on 16 lactating Antarctic fur seals to assess onshore attendance with concurrently collected time depth recorder (TDR) data used as the ‘true’ measurement of time spent onshore. Within the retrieved datasets, 25% of the VHF data could not be interpreted with any reliability. Additionally, there were significant differences in the number and duration of attendance bouts between the 2 instrument types, with VHF data overestimating attendance bout duration by approximately 8.9 h on average. Importantly, the magnitude and direction of errors between VHF and TDR measurements were not systematic, suggesting that VHF data is an inappropriate method for collecting attendance data. Modelling the raw VHF data in a state-space framework elicited mean attendance durations that were indistinguishable from TDR-derived measurements, suggesting this approach may provide a means to re-examine historic VHF data. Moreover, given the evolution of electronic tags in terms of sophistication, miniaturization, longevity and decreasing cost over the last 30 yr, TDRs are a more appropriate means of collecting attendance data on centrally foraging marine mammals such as otariids.