Functionality and habitat-dependent performance of a 3-D wireless acoustic telemetry system in a small natural lake (Kleiner Döllnsee, Germany)

Positioning telemetry offers a broad range of opportunities to study fine-scale movement of aquatic animals but also poses challenges in data processing and analyzes. Therefore, a better understanding of the underlying technology and knowledge of the strengths and limitations of applied telemetry systems are required. A wireless acoustic telemetrysystem consisting of 20 receiver stations was installed in a natural lake. 155 stationary tests were conducted to assess performance of processing software (e.g. two options of partial symbol reconstruction, PSR) and performance in four differently structured habitats: pelagic, emerged macrophytes, submerged macrophytes and sublitoral. Preliminary acoustic-telemetry tracking data of pike were evaluated and an effect of the transmitter on daily increment of pike was assessed. Processing data without PSR yielded significantly greatest accuracy (AC; 21.1 ± 23.7 m) and precision (PR; 48.9 ± 33.3 m). Applying a Hidden Markov Model significantly improved AC (7.1 ± 15.2 m) and PR (5.1 ± 14.1 m). Highest position solution (PS) was achieved in the pelagic (69.3 ± 19.7%) with significantly better AC (2.2 ± 2.4 m) and PR (1.5 ± 5.1 m) than in other habitats. Lowest AC (20.8± 22.4 m), PR (20.5 ± 28.5 m) and PS (7.9 ± 13.8%) were achieved in emerged macrophytes. Performance in submerged macrophytes and the sublitoral was equal (AC: 7.7 ± 10.2 m vs. 7.7 ± 18.9 m; PR: 4.4 ± 9.6 m vs. 4.2 ± 12.9 m) with significantly higher PS in the sublitoral (23.5 ± 29.6% vs. 57.5 ± 24.3%). Scattering in positions induced false movement instationary tests (0.5 ± 0.9 m/min). Geometry and time synchronization of receiver stations also affected performance. Estimates of built-in depth sensors of transmitters were comparable to manual measurements. 18 moving tests yielded comparable results to stationary tests but moved distance was underestimated by 12 ± 15%. Movement data of pike were marked by periods of non-detection lasting up to several days. Radio-telemetry showed that pike were mostly inactive in complex habitat structures, thus justifying the assumption of pike being inactive in periods of non-detection in acoustic telemetry. Pike also displayed activity bursts when leaving complex habitats determined by radio-telemetry, thus terminating periods of non-detection in acoustic telemetry. A correction of pike movement estimates was therefore suggested and validated. Comparing daily increments of tagged and untagged pike, impaired growth was verified for tagged pike. These results suggest that wireless acoustic telemetry systems can be successfully applied in natural lakes, but data quality depends critically on the use of appropriate software, user settings and on habitat structure.