Validation of a novel satellite-transmitted data product to monitor long-term activity in free-ranging elasmobranchs

While satellite-transmitted measurements of fish depth, temperature, and location are possible over timescales of months to years, whole-animal activity levels are generally measured over much shorter time periods. Remote... [ view full abstract ]

While satellite-transmitted measurements of fish depth, temperature, and location are possible over timescales of months to years, whole-animal activity levels are generally measured over much shorter time periods. Remote transmission of activity data, derived from accelerometer records, is limited by satellite-tag data storage, sufficient battery for transmitting data, and restricted satellite bandwidth for receiving transmitted data. As such, studies of whole-animal activity are generally limited to relatively short deployments where logged data can be feasibly retrieved, or opportunistic retrievals of longer-term logged data. Here, we validate a new satellite-transmitted data product for remote measurement of fish activity levels, movement, and temperature, over a period of months, without requiring recovery of the tag. Accelerometer sensors from pop-off satellite archival tags (PSATs) measure fish activity, which is transmitted as an ‘activity time series’ (ATS) along with depth, temperature, and irradiance data for geolocation, to the Argos satellite system. We tested these new PSATs which generated ATS data on captive cobia (Rachycentron canadum) and dogfish (Squalus acanthias), each over a period of 5 days. Deployments were paired with video monitoring to compare archived tag data, the transmitted ATS data product, and visually observed fish behavior including tail beat frequency (TBF). We used these data sources to validate ATS behavioral state classification using a supervised Hidden Markov Model (HMM). For highly mobile species whose behavior precludes the retrieval of long-term archival records, the ATS product may present an alternative for field-based studies of activity level under variable environmental conditions, and across broad spatiotemporal scales.