We observed patterns in echograms of data collected with a dual-frequency identification sonar (DIDSON) that were related to the tail beats of fish. These patterns reflect the size, shape, and swimming motion of the fish and also depend on the fish's angle relative to the axis of the beam. When the tail is large enough to reflect sound of sufficient intensity and the body is angled such that the tail beat produces periodic changes in the range extent covered by the fish image, then the tail beat becomes clearly visible on echograms that plot the intensity maximum of all beams. The analysis of DIDSON echograms of a mix of upstreammigrating Chinook salmon Oncorhynchus tshawytscha and sockeye salmon O. nerka resulted in the separation of two groups: (1) fish of sockeye salmon size that swam with a tail-beat frequency (TBF) between 2.0 and 3.5 beats/s and (2) fish of Chinook salmon size with a TBF between 1.0 and 2.0 beats/s. There was no correlation between TBF and fish size within each group, which suggests that the observed difference in TBF between the two groups was species-specific rather than an indirect effect of the groups' difference in size. The technique of extracting TBF from DIDSON echograms may also be useful for bioenergetics studies. Compared with electromyogram telemetry, it offers the advantages of being nonintrusive and faster to set up and analyze and therefore is suitable for analyzing larger sample sizes. The disadvantages are that the technique's potential is limited to relatively large fish, it can cover only relatively small areas, it cannot be used to follow individual fish over long distances, and some environments are too noisy to produce DIDSON images of sufficient quality.

Abstract.-The purpose of this study was to explore the extent to which a computer-driven process can be used to classify sonar images. The data we present come from a feasibility study for a hydroacoustic monitoring system aimed at the automatic detection of downstream-migrating adult American eels Anguilla rostrata in the intake canal of a small hydroelectric station. The images were collected by a dual-frequency identification sonar with sufficient resolution to show the distinct shape and swimming motion of eels, and thus to allow confident visual identification. The goal was to find a set of image processing, tracking, and pattern recognition techniques that would reproduce the results of the visual classification. Of the three classification methods that we tested with our example data set, neural network analysis had the lowest misclassification rate for eels (7% of the eels being misclassified as debris) and the second-lowest misclassification rate for debris (5% of the debris being misclassified as eels). Discriminant function analysis misclassified 12% of the eels as debris and 4% of the debris as eels. A K-nearest-neighbor analysis initially provided the poorest results (17% misclassified eels and 12% misclassified debris). However, after applying an algebraic correction, K-nearest-neighbor analysis yielded an accurate estimate of the number of eels in the data set. We discuss the value of flagging cases of uncertain classification, how image processing and feature selection can affect the results, and how the numeric ratio of the targets present determines what error rates are acceptable. We conclude that, depending on the application, different degrees of automation may be achieved, ranging from a relatively high degree of human supervision in the classification of all potential targets to a fully automated process that requires only periodic quality control and adjustments of the classification model.

No abstract

Abundance estimates of out‐migrating sockeye salmon Oncorhynchus nerka smolts are used to prepare preseason forecasts of adult returns and to set escapement goals. Here we describe a method for estimating smolt flux and abundance that uses side‐looking sonar. This method more efficiently covers the river cross section and is logistically easier to deploy than up‐looking sonar systems. To account for the skewed vertical distribution of smolts, we used a recently developed model for correcting echo integration bias associated with nonuniform target distribution. The correction is based on adjusting the integrated beam pattern for a given distribution of targets relative to the transverse cross section of the beam. We compared the results with estimates derived from video data and modeled the effect of three vertical distributions of smolts and three transducer pitch angles. The model estimated correction factors that ranged from 0.6 to 2.6. Correction factors were greater than 1 (which indicates negative bias in conventional echo integration) for smolt distributions skewed toward the edge of the beam and less than 1 (which indicates positive bias in conventional echo integration) for distributions skewed towards the center of the beam. For the scenarios modeled, the effect of the transducer pitch angle was small between the horizontal and −0.6° but increased nonlinearly as the angle increased. We conclude that, in the given application and at shallow transducer pitch angles, the bias in conventional echo integration is small and predictable enough to be corrected.

This study assessed the feasibility of three sonar technologies to estimate eel abundance, determine distribution, and describe approach behavior to advance strategies for providing safe downstream passage of out-migrating American Eels at hydroelectric facilities on the St. Lawrence River. A Simrad EK60 split-beam echosounder (120 kHz), Sound Metrics ARIS Explorer multibeam sonar (1100/1800 kHz), and Mesotech M3 multi-mode multibeam sonar (500 kHz) were deployed at Iroquois Dam for experimentally testing their capabilities in detecting and identifying known numbers and sizes of live adult eels tethered to surface floats released upstream of the sonar beams and allowed to swim through at known locations and times. In addition, sonars collected data continuously to monitor wild, out-migrating eels during July 15-22 and September 17-19, 2015. Results highlight several challenges in acoustically monitoring eels in a large, fast-moving river with a few orders of magnitude higher abundance of other targets that can lead to a high false positive error rate. The ARIS multibeam sonar, operating with 48 beams, holds the most promise for correctly identifying eels out to 16-20 m in range, but the M3 multibeam sonar has some value for tracking previously identified targets over larger areas.