DAUBLE ET AL.: SPATIAL DISTRIBUTION OF JUVENILE SALMONIDS 



juvenile fall chinook salmon emerging from the 

 Vernita Bar area (Page et al. 1976) and with the 

 period of greatest catches of juvenile coho, 

 chinook, and sockeye salmon, and steelhead at 

 Priest Rapids Dam (Becker 1985). The late sum- 

 mer sampling corresponded to the period of 

 greatest catches of 0-age fall chinook salmon at 

 Priest Rapids Dam (Raymond 1967; Sims and 

 Miller 1977; Hovland et al. 1982). 



Sampling periods within each week were se- 

 lected by a stratified random process to give 

 equal weight to weekday and weekend intervals. 

 To differentiate between diel variations in mi- 

 gi-ation pattern, each day was divided into four 

 equal time blocks starting at 0400 PDT (i.e., 

 0400-1000. 1000-1600. 1600-2200, and 2200- 

 0400). The scheme provided one all-dark, one 

 all-light, and two transition (dawn and dusk) 

 periods. All barge stations were fished simultan- 

 eously at a predetermined depth. One net set of 

 approximately 2 h duration was taken at each of 

 the surface, mid-, and bottom depths according 

 to a random schedule during each 6 h time block. 

 Water temperature, sample and station depths, 

 duration of set, and flowmeter readings were 

 recorded for each sample. Secchi disc depth was 

 recorded daily at noon. Current velocity meas- 

 urements were taken at 1 m below the surface, 

 at middepth, at 1 m from the bottom at each of 

 the four stations, and over a range of flows using 

 a Bendix Model Q-15 current meter. 



Three people worked each shift. Generally, a 

 net could be raised, checked for catches, cleaned 

 of debris, and lowered to the next sampling 

 depth within 5 minutes. Nets at all four stations 

 would usually be tended and repositioned within 

 15 to 30 minutes. Some samples were lost be- 

 cause of water levels and high flows. Once, a 

 submerged log hit the midchannel net and broke 

 the spreader anchor cable. 



The shoreline fyke nets were used for five 24 h 

 periods each week from 30 April to 29 July 1984. 

 To differentiate between diel variations in catch, 

 each 24 h day was divided into four equal time 

 blocks of 6 hours each. Collection intervals coin- 

 cided with the four barge fyke net sampling pe- 

 riods. Nets were set parallel to the shoreline and 

 opposite the barges at depths of 1 to 2 m. 



and 1984. Each shoreline station was sampled 

 once per week during each of the four 6 h time 

 blocks when fyke nets were sampled. A single 

 pass with the electroshocker was conducted 

 through each 400 m transect at depths of 1 to 2 

 m. Stunned fish were collected with dip nets, 

 and all juvenile salmonids were measured and 

 released. Catch per unit effort was based on 

 duration of shock. 



Duplicate seine hauls were made at each of 

 four permanent stations near the barges with a 

 9.1 by 1.2 m net constructed of 3 mm (Ms in.), 

 heavy-duty, knotless nylon mesh. The stations 

 were sampled once per week during daylight 

 hours from April through June 1984. About 50 

 m~ of shoreline were sampled with each set. All 

 salmonids were enumerated and subsamples 

 (usually five fish per station) were retained for 

 measurements. 



Data Analysis 



Estimates of the proportional distribution of 

 fish gi-oups caught at various stations and depths 

 by fyke net were based on a multinomial distri- 

 bution of the fish caught among the various com- 

 binations of station and depth (Cochran 1977). 

 Relative catch per unit effort (CPUE) was calcu- 

 lated on the basis of unit time/cross-sectional 

 area sampled and on volume sampled. A log- 

 linear model was developed to evaluate propor- 

 tional distribution estimates of 0-age fall 

 chinook, spring chinook, and sockeye salmon 

 smolts using the CATMOD procedure in SAS 

 (SAS 1985). The model was then used to test for 

 two-way interactions among fish gi-oups, barge 

 location, and sample depths. A binomial test for 

 differences (Mainland et al. 1956) was also used 

 to evaluate patterns of distribution for some 

 species. 



RESULTS 



Estimates of cross-sectional distribution were 

 different for each of the six gi-oups of juvenile 

 salmonids collected. The differences are de- 

 scribed in terms of species, life stage, and migi-a- 



tion timing. 



Supplemental Sampling Gear 



A boat-mounted electroshocker (Smith-Root 

 Type VI Electrofisher), powered by a 240 volt 

 generator, was used to sample nearshore fish 

 populations near the barge stations during 1983 



Distribution of 0- Age Chinook Salmon 



Three groups of 0-age chinook salmon were 

 collected: 1) naturally produced (wild) fish orig- 

 inating from adults spawning in the Hanford 

 Reach above the study site, 2) hatchery fish 



779 



