Background: Biotelemetry has become a key tool for studying marine animals in the last decade, and a wide range of electronic tags are now available for answering a range of research questions. However, comparatively, less attention has been given to attachment methods for these tags and the implications of tag retention on study design, especially when designing a comparative study looking at multiple species. Here, we reported our findings on acoustic tag retention rates for juveniles of two species of marine turtle: the green sea turtle (Chelonia mydas) and the hawksbill sea turtle (Eretmochelys imbricata). We captured both species twice annually (spring and fall) from 2012 through 2017, as part of a capture-mark-recapture study at Buck Island Reef National Monument, St. Croix, U.S. Virgin Islands. We assessed tag retention rates using physical recaptures of turtles previously outfitted with an acoustic tag.
Results:We deployed 72 acoustic tags on 60 juvenile greens and 37 acoustic tags on 29 hawksbills. We estimated the half-life for tags on greens to be 150 days (95% CI 117-188 days), whereas the half-life for tags on hawksbills was 1077 days (95% CI 870-2118 days), a marked difference. We observed that tag attachment holes, drilled into the posterior marginal scutes, migrated laterally towards the outer edge of the marginals in both species. Green turtles tended to exhibit tear-outs, as their attachment holes wore and/or tags grew near the edge of their scutes, whereas hawksbills tended to maintain the structure of these holes and did not exhibit these tear-outs.
Conclusions:We conclude that hawksbills can be tagged with long-battery-life acoustic tags for long-term studies of habitat use and movement patterns, whereas greens are likely to shed their tags in the 1st year, making long-term studies difficult. This study is the first clear evidence that tagging protocols should vary between species of hardshelled turtles. Furthermore, shed tags on the seafloor continue to be detected by acoustic receivers, creating a challenge in data filtering before analysis. We encourage future research into an efficient method for filtering these data points prior to analysis.