278 



Fishery Bulletin 93(2). 1995 



Washington (Fig. 1; Pearcy and Fisher, 1988, 1990). 

 The sampling area was divided latitudinally into 

 three sections (A, B, and C in Fig. 1 ). Additional sam- 

 pling in July 1984 occurred off northern California 

 as far south as 40°32'N and off the west coast of 

 Vancouver Island as far north as 50°26'N, and in May 

 1985 in a concentrated area off the mouth of the Co- 

 lumbia River. The months sampled in each year are 

 shown in Table 2. The latitudinal range of sampling 

 varied among years and months (Pearcy and Fisher, 

 1988; see also Fig. 5, this paper). Transects were gen- 

 erally 37 km apart and collecting stations along each 

 transect were 9.3 km apart starting at about the 37 

 m depth contour and continuing out to 37 or 46 km 

 offshore. Occasionally stations farther offshore were 

 sampled. 



For this study we considered all chinook salmon 

 less than 401 mm fork length (FL) to be juvenile fish. 

 This was based on the sizes of known age coded-wire 

 tagged (CWT) fish in our catches. This length range 

 included those fish that entered the ocean as year- 

 lings or subyearlings in the same year in which they 

 were caught in the ocean (age 1.0 or age 0.0, respec- 

 tively 9 ) and those that entered the ocean as subyear- 

 lings in the year prior to capture in the ocean ( age 0. I 9 ). 



9 The number before the period indicates winters spent in fresh- 

 water after hatching and before migration to the sea, and the 

 number following the period indicates winters spent at sea ( Koo, 

 1962). 



Fork length (FL) of most fish was measured at sea 

 to the nearest mm. CWT fish were measured at sea, 

 then frozen, and later weighed in the laboratory. Tags 

 were decoded by personnel at the Oregon Depart- 

 ment of Fish and Wildlife laboratory in Clackamas, 

 Oregon. 



Growth rates of CWT chinook salmon between re- 

 lease from hatcheries and capture in the ocean were 

 estimated by (FL 1 -FL Q )/d; where FL Q = mean length 

 of fish in the tag group at release, FL X = length of 

 the individual fish when caught in the ocean, and d 

 = the number of days between release and capture 

 (d >10). Often only mean weight at release was 

 readily available for a tag group. For these groups 

 we converted mean weight at release to mean FL by 

 using the geometric mean functional relationship 

 (Ricker, 1973) between average weight (g) and aver- 

 age FL (mm) for CWT groups released in California 

 and the Columbia River for which both mean weight 

 and mean length at release were measured: ln(FL) - 

 0.3122(ln(u;*))+3.8233, n=311, r 2 =0.94. Average in- 

 stantaneous rates of growth in weight (% body wt/d) 

 between release and ocean capture were estimated 

 by (ln(Wfj)-ln(Wf ))/d; where Wt 1 = weight at cap- 

 ture in the ocean, Wt Q = average weight at release, 

 and d = the number of days (>10) between release 

 from the hatchery and capture in the ocean (Ricker, 

 1975). 



Growth rates of chinook salmon in the Columbia 

 River estuary and in the ocean prior to capture in 



