Background

Acoustic telemetry is a commonly used technology to monitor animal occupancy and infer movement in aquatic environments. The information that acoustic telemetry provides is vital for spatial planning and management decisions concerning aquatic and coastal environments by characterizing behaviors and habitats such as spawning aggregations, migrations, corridors, and nurseries, among others. However, performance of acoustic telemetry equipment and resulting detection ranges and efficiencies can vary as a function of environmental conditions, leading to potentially biased interpretations of telemetry data. Here, we characterize variation in detection performance using an acoustic telemetry receiver array deployed in Wellfleet Harbor, Massachusetts, USA from 2015 to 2017. The array was designed to study benthic invertebrate movements and provided an in situ opportunity to identify factors driving variation in detection probability.

Results

The near-shore location proximate to environmental monitoring allowed for a detailed examination of factors influencing detection efficiency in a range-testing experiment. Detection ranges varied from < 50 to 1,500 m and efficiencies varied from 0 to 100% within those detection ranges. Detection efficiency was affected by distance, wind speed and direction, wave height and direction, water temperature, water depth, and water quality.

Conclusions

Performance of acoustic telemetry systems is strongly contingent on environmental conditions. Our study found that wind, waves, water temperature, water quality, and depth all affected performance to an extent that could seriously compromise a study if these effects were not taken into consideration. Other unmeasured factors may also be important, depending on the characteristics of each site. This information can help guide future telemetry study designs by helping researchers anticipate the density of receivers required to achieve study objectives. Researchers can further refine and document the reliability of their data by incorporating continuously deployed range-testing tags and prior knowledge on varying detection efficiency into movement and occupancy models.