Using Acoustic Telemetry to Expand Sonar Escapement Indices of Chinook Salmon to In-river Abundance Estimates

Published: 23 August 2019| Version 1 | DOI: 10.17632/c8jr8x9r4y.1


Acoustic telemetry was used to determine what proportion of adult Chinook salmon traveled within the sampling range of a sonar project at the Nushagak River, Alaska. These proportions were used to expand sonar indices of Chinook salmon to full-river escapement estimates. At this site, the river bottom topography does not allow full-river coverage by the dual-frequency identification sonars (DIDSON) located on either side of the river. Of the 300-m river width, 50 m is ensonfied along right bank (RB) and 30 m along left bank (LB). Chinook salmon are known to travel beyond the sonars’ detection range. Acoustic tags were inserted into Chinook salmon across a 4-year period (2011-2014). As tagged fish were detected at the sonar site by an array of acoustic receivers, they were classified as to whether they traveled through the sonar beam along RB, LB, or offshore of the sonars’ range (Os). Fish that traveled near the beam edges created uncertainty. To handle this uncertainty, we drew three probability regions around each beam footprint. The first extended from shore to 5 m short of the DIDSON end range. Tagged fish that traveled through this region were assigned a p of 1 or -1 depending on their direction of movement and a bank assignment, LB, RB, or Os. A tagged fish traveling upriver through the first RB region was assigned [0 LB, 1 RB, 0 Os], a potential DIDSON count of 1. The second region extended from the end of the first region to 5 m offshore of the DIDSON end range for a p assignment of 0.5. The assignment for a fish traveling along LB through this region would be [0.5 LB, 0 RB, 0.5 Os], an equal chance of passing inside or outside the beam footprint. The third region (p=0.25) extended from the end of the second region to 5 m farther offshore. Some fish traveled through both DIDSON beams during a single upriver trip and would have been counted as 2 fish, so their p value was 2. Many tagged fish made a single, upriver trip through the array (ST fish), but several made multiple up and downriver trips, (MT) fish. Each trip was assessed in the same manner as the ST fish except that downriver trips yielded negative probability counts. If a fish traveled upriver along LB, downriver along RB, and back up mid-river, the p value would be 1, [1 LB, -1 RB, 1 OS]. A special case was presented by implied trips. Tagged fish first detected moving downriver were assumed to have moved upriver first but were either not observed or the track was rejected by the filters. Also, a fish that made two consecutive upriver trips had to have made an unobserved (implied) downriver trip. We assumed these fish had an equal probability of traveling through each region so implied upriver trips were assigned [0.33 LB, 0.33 RB, 0.33 Os] and implied downriver trips were assigned [-0.33 LB, -0.33 RB, -0.33 Os]. Multiple assignments for a given fish were summed to produce a single p value. Assignment tables and fish tracks are included for each study year.



Acoustics, Salmonid Fish, Acoustic Imaging, Fisheries Science, Fish Tagging