Scale of Biotelemetry Data Influences Ecological Interpretations of Space and Habitat Use in Yellowtail Snapper

Abstract

As opposed to passive, broad‐scale acoustic telemetry arrays, acoustic positioning systems generate high‐resolution animal locations that provide information on long‐term, fine‐scale movement patterns and habitat preferences. However, limited comparisons have been made between more common broad‐scale acoustic data and fine‐scale positioning data and it is unknown whether differences exist in ecological inferences gained or lost between using either array configuration over the other. Broad‐scale movement and habitat use information was collected for eight Yellowtail Snapper Ocyurus chrysurus tagged within an array of 78 stationary acoustic receivers deployed in Buck Island Reef National Monument, a marine protected area located northeast of St. Croix, U.S. Virgin Islands. An additional 25 receivers were nested within the larger array as a VEMCO Positioning System and were used to assess fine‐scale habitat use for five of the eight tagged fish. Broad‐scale results inferred from network analysis revealed that all individuals had core use receivers along the shallow shelf break situated west of Buck Island, preferring an area that was coarsely characterized by sand and seagrass benthic habitats. Fine‐scale results using Euclidean distance analysis (EDA) suggested fish positions occurred randomly or independent of benthic habitat type. Further exploration of positioning data suggested that there were two contingents or groups of fish displaying unique movement patterns within the fine‐scale positioning array. Individuality in space and habitat use was thus masked when using an EDA approach at the study population level, as it was also missed during broad‐scale analyses. Discrepancies between broad‐ and fine‐scale habitat inferences suggest that positioning systems are necessary for interpreting habitat use in complex coral reef ecosystems. Nested positioning systems appear to add substantial information that is not obtainable using broad‐scale data alone, and caution is necessary in inferring habitat use when only coarse‐scale location data are available.