FISHERY BULLETIN: VOL. 80, NO. 2 



to several days. In addition, allocation to treaty 

 Indian fishermen is required by Federal Court 

 rulings (the Boldt decision), which established 

 separate treaty and nontreaty quotas. Biologists 

 representing Washington State, participating 

 tribes, and the University of Washington agreed 

 that a target quota should be 20% of the total 

 estimated population biomass (Trumble 1980). 

 The application of hydroacoustic techniques 

 offered timeliness in the Strait of Georgia 

 herring stock assessment program (Thorne 

 1977a) when combined with midwater trawling, 

 analysis of catch records, and spawning ground 

 surveys. This paper presents results of applica- 

 tion of these techniques during 1976-79 to the 

 management of the fishery. 



METHODS AND MATERIALS 



The concept of the sac-roe herring stock 

 assessment program is to estimate the biomass of 

 adult herring in prespawning condition, and add 

 to this the biomass of adult herring removed by 

 spawning or being caught. Hydroacoustics pro- 

 vided estimates of total pelagic fish biomass, and 

 midwater trawl sampling provided species com- 

 position data to identify prespawning adults; the 

 combination of hydroacoustics and midwater 

 trawling will be referred to as "acoustic-trawl." 

 Spawning ground surveys provided estimates of 

 spawning escapement, and catch records 

 tracked the success of the fishery. Catch and 

 escapement estimates provide a postfishery 

 check on the accuracy of the acoustic data, and in 

 conjunction with acoustic-trawl surveys, a series 

 of in-season estimates of the total biomass 

 of mature herring. 



Acoustic Survey Equipment and 

 Methods 



The hydroacoustic data acquisition system 

 consisted of a modified 105 kHz Ross 3 200A 

 echosounder, an interface amplifier that re- 

 duced the signal frequency from 105 kHz to 5 

 kHz, a Sony TC-377 tape deck which recorded 

 the 5 kHz data on magnetic tape, and an 

 oscilloscope to monitor system operation. The 

 transducer produced a 7% degree full angle 

 beam at the half power (-3dB) points. The 

 modifications of the system include an internal 



'Reference to trade names does not imply endorsement by 

 the National Marine Fisheries Service, NOAA. 



calibration oscillator to monitor and measure 

 system gain (Thorne et al. 1972; Nunnallee 1973). 

 The echosounder transmitted a pulse length of 

 0.6 m s. The echosounder and transducer were 

 periodically calibrated at the Applied Physics 

 Laboratory, University of Washington. Normal- 

 ly, calibration occurred at the beginning and end 

 of each field season. 



The acoustic data from the magnetic tapes was 

 processed with a digital echo integration system 

 implemented on a PDP 11/45 computer (Thorne 

 1977 b). A mean target strength of -33 dB/kg 

 was used by the program to scale the integrator 

 data to estimates of fish density. This value was 

 originally established on the basis of both 

 comparisons with net tows and in situ target 

 strength measurements (Thorne 1977a) and the 

 value still appears to be reasonable. Although 

 considerable information has been obtained on 

 the dependence of target strength on fish length 

 (FAO 1978; Thorne in press), the variation in 

 mean fish lengths in the Strait of Georgia is 

 insufficient to warrant using a length-dependent 

 variable instead of a constant for the target 

 strength scaling factor. Herring typically range 

 from 18 to 24 cm SL, and compose 70-90% of the 

 biomass in the acoustic-trawl surveys. 



During 1976 and 1977, the University of 

 Washington's 12 m RV Malka was used as the 

 acoustic platform, with a hull-mounted trans- 

 ducer. Subsequently, the acoustic program 

 chartered a 10 m gill net vessel and used an over- 

 the-side pole-mount for the transducer. 



Acoustic surveys were conducted during April 

 and May (and to the first part of June 1976) in 

 order to bracket the spawning migration of the 

 herring (Lemberg 1978). The surveys were 

 conducted between Point Roberts and Lummi 

 Island on a standardized trackline which had 10 

 transects, each about 8 km in length (Fig. l).The 

 surveys were typically conducted at twice 

 weekly intervals around the peak of the 

 migration, and less frequently during the early 

 and late stages of the run. 



During the day, herring normally aggregate 

 in tight schools at depths of 40 m or more. At 

 night, the schools disperse and form widespread 

 layers 5-10 m thick at depths of 10-30 m; herring 

 density decreases such that many fish are 

 distinguishable as individual targets. Until 

 actively ready to spawn, herring remain in water 

 deeper than 20 m. The survey area encompassed 

 the prespawning holding area, bounded by the 8- 

 10 fathom contours on the inside, and the 50-60 



382 



