284 



Fishery Bulletin 93(2). 1995 



However, the mean size at ocean entry backcalculated 

 from scales (173 mm FL) of the fish caught in Area C 

 was much larger than the mean sizes of naturally 

 produced fall chinook salmon reported by Nicholas 

 and Hankin (1988) in Oregon estuaries in the late 

 summer (100-140 mm FL by mid-September, de- 

 pending on the river system). 



Since the estimated average size at time of ocean 

 entry of the fish caught in Area C in September ( 173 

 mm FL) was much larger than the average size at 

 which age-0.0 wild smolts enter the ocean (Nicholas 

 and Hankin, 1988; Dawley et al. 1 ; Dawley et al. 4 ), it 

 is probable that these fish were released from a 

 hatchery. The only releases in the study area (ex- 

 cluding California) of large smolts in groups not rep- 

 resented by CWT's during the July-August period 

 were from the Oregon Aqua Foods (OAF) saltwater 

 rearing and release facility on Yaquina Bay, Oregon, 

 which is located within 5 km of the ocean. Almost 

 800,000 unmarked subyearling fall chinook and 

 55,000 unmarked subyearling spring chinook salmon 

 were released from this facility between 28 August 

 and 5 September 1983 10 , about three weeks before 

 our sampling in the ocean (22-24 September). The 

 mean weight at time of release for these groups 

 ranged from 47 to 57 g for the fall chinook salmon 

 and was 100 g for the spring chinook salmon. From 

 a regression of FL and body weight (this study) we 

 estimated that the mean lengths of these OAF-re- 

 leased fall chinook salmon at time of release ranged 

 from 152 to 162 mm, close to the mean size at ocean 

 entrance (backcalculated from scales) of the fish we 

 caught in Area C in September 1983 (173 mm). If 

 these OAF-released fish were the source of the un- 

 marked fish caught in the ocean in September 1983, 

 then they were in the ocean about three weeks be- 

 fore capture. If fish size at ocean entrance that we 

 backcalculated from scales was accurate, then their 

 growth rate since entering the ocean was slightly over 

 2 mmd -1 , considerably higher than the mean growth 

 rate (1.05 mmd" 1 ) estimated for age-1.0 Columbia 

 River fish downstream of rkm 75 (see below), and 

 about one and a third times that estimated for juve- 

 nile coho salmon, Oncorhynchus kisutch, caught in 

 the ocean in late summer (Fisher and Pearcy, 1988). 



Inshore-offshore distributions 



The inshore-offshore distributions of very small fish 

 (<130 mm FL), which were mainly age-0.0 fish, and 

 of larger fish (>130 mm FL) were similar; the me- 

 dian offshore distance of the catch of each size cat- 



10 Information obtained from the Pacific States Marine Fishery 

 Commission's salmon release data base. 



egory was about 13 km (Fig. 3A). However, small fish 

 were strongly associated with warm, brackish wa- 

 ters (mainly the Columbia River plume), whereas 

 larger fish were not (Fig. 3, B and C). Fully 38% of 

 the small fish, but only 2% of the large fish, were 

 caught in waters <17%<? (Fig. 3B). Over 40% of small 

 fish, but only 4% of large fish were caught where 

 sea-surface temperature was >15°C (Fig. 3C). These 

 data indicate that, at least over the depths sampled 

 (mainly >37m), the smallest chinook salmon juve- 

 niles were much more likely to be found in the warm, 

 low salinity waters of the Columbia River plume than 

 in the colder, more saline adjacent waters. 



Our largest catches of juvenile chinook salmon in 

 late summer were taken in September 1983. On the 

 basis of their scale characteristics we concluded that 

 these were hatchery subyearling chinook salmon that 

 had been in the ocean about a month (see section on 

 Size-Frequency Distributions). Almost all of these fish 

 were captured within 4 km of the shoreline, in depths 

 of <40 m (Fig. 4). This is a much more restricted in- 

 shore-offshore distribution than was found in general 

 for the juvenile chinook salmon (mainly age-1.0 fish) 

 caught during all months and years of our sampling 

 combined (Fig. 3A). Since daily upwelling indices dur- 

 ing September 1983 at both 42°N and 45°N were al- 

 most all positive (Mason and Bakun, 1986), the re- 

 stricted inshore distribution of juvenile chinook salmon 

 in this month does not appear to have been caused by 

 onshore transport of water. The difference in offshore 

 distribution between this group of large age-0.0 hatch- 

 ery chinook salmon and the mainly age-1.0 fish caught 

 in early summer may be due to behavioral differences 

 between these groups offish; the age-0.0 fish appear to 

 prefer shallower, more nearshore areas. 



Migration 



Several trends are apparent in the migrations of 

 CWT juvenile chinook salmon between ocean entry 

 and capture in purse seines (Fig. 5). In 1983, 1984, 

 and 1985 most fish originating in the Columbia River 

 basin were caught north of where they entered the 

 ocean. In May 1982, however, all Columbia River fish 

 were caught south of the river mouth, but by the fol- 

 lowing month most were caught to the north. All but 

 one of the ten tagged chinook salmon originating from 

 coastal Oregon hatcheries were caught north of 

 where they entered the ocean. 



Thirteen fish were caught more than 190 km north 

 of where they entered the ocean. Seven of these were 

 age-1.0 fish from the Columbia River drainage (three 

 Snake River fall, two Snake River summer, one URB 

 fall, one Deschutes River spring), one was an age- 

 1.0 fish from a coastal Oregon River system (Umpqua 



