Background In recent years, large-scale acoustic telemetry observation networks have become established globally to gain a better understanding of the ecology, movements and population dynamics of fish stocks. When studying a species that uses different habitats throughout its life history difficulty may arise where acoustically suboptimal habitats are used, such as shallow, vegetated areas. To test the feasibility of active tracking in these acoustically suboptimal habitats, we quantified detection probability and location error as a function of several environmental variables with two transmitter types in a shallow freshwater embayment. Results When placed in nearshore areas (< 1 m deep), the higher-powered transmitter (158 dB) had significantly greater detection probability than the lower-powered transmitter (152 dB). For both transmitter types, detection probability declined at 200 m; however, at the 100 m distance the higher-powered transmitter had greater than 50% detection probability per ping cycle (50.4%) while the lower-powered transmitter was substantially less (29.4%). Additionally, detection probability increased when the transmitter was deployed within sparse, senescent Phragmites spp. vegetation (14%). Estimated positional accuracy of transmitters deployed at known locations (location error) was variable (error range: 13–259 m), and was generally higher for the more powerful transmitter. Location error was minimized when the lower-powered transmitter was located near softened shoreline areas compared to near man-made armored shorelines (i.e., rip-rap). Conclusion While benefits exist for maximizing transmitter power (e.g., increased detection range in open-water environments), use of a lower-powered transmitter may be advantageous for active tracking specific locations of fish inhabiting shallow water environments, such as in estuarine tidal marshes and shallow wetlands. Thus, when planning acoustic telemetry studies, researchers should conduct site-specific preliminary detection probability/location error experiments to better understand the utility of acoustic telemetry to investigate fish movements in acoustically suboptimal conditions.

Restoration of wetlands in the Laurentian Great Lakes is designed to improve numerous ecosystem services including nutrient sequestration, support recreational uses, augment waterfowl habitat, and provide spawning and nursery habitat for fish. These projects often use Northern Pike Esox lucius, as a sentinel species for assessing restoration success because of their obligate use of emergent and submergent vegetation by the early life stages. However, invasive species such as Common Carp Cyprinus carpio also utilize these habitats and can have negative effects. Thus, fostering the use of these wetlands by Northern Pike and limiting access by invasive carp is a common restoration management concern. Using a dual frequency sonar system to observe fish migration in a recently connected Lake Erie coastal wetland, we were able to identify fine-scale environmental correlates to immigration by each species during early spring. Both species overwhelmingly migrated during daylight hours and against the flow of water that was generated through seiche-driven water oscillations. However, Northern Pike predominantly immigrated during the initial increase in water temperature immediately after ice-out; Common Carp also began to enter the wetland at low temperatures shortly after ice-out, but immigration exhibited a slow exponentially increasing pattern, such that the primary immigration timing differed by almost 30 d. Thus, we demonstrate common fine-scale spawning migration patterns (i.e., water oscillation effects and time of day) for both species and reinforce the temperature-correlated differences in spawning migration. Using these results could enable managers of connected wetlands to more precisely and dynamically manage habitat accessibility for desired species like Northern Pike while better regulating invasive Common Carp spawning migration and use of these systems.