618 



Fishery Bulletin 88(4). 1990 



This paper describes the diet composition and feeding 

 relationships of juvenile coho, chinook, chum, and 

 sockeye (0. nerka) salmon collected from coastal waters 

 off Washington and Oregon during a series of summer 

 purse-seine collections made from 1980 to 1985. Ocean- 

 ographic conditions and biological productivity were 

 highly variable during this period, both seasonally and 

 interannually (Fisher and Pearcy 1985, Brodeur and 

 Pearcy 1986), and thus provided a fortuitous situation 

 in which to examine the effects of environmental 

 variability on juvenile salmon feeding habits. In addi- 

 tion to examining the interannual variation in feeding, 

 the broad seasonal and geographic coverage of the 

 sampling permitted more detailed intra-annual and 

 areal resolution than was previously possible. Par- 

 ticular detail will be provided for coho and chinook 

 salmon juveniles, which were the dominant salmonids 

 caught (Pearcy 1984, Pearcy and Fisher In press). 

 Although the importance of other potential sources of 

 variability in the diet, such as those attributable to the 

 time of day and predator size, is recognized, these 

 additional factors are discussed in detail elsewhere 

 (Brodeur and Pearcy 1987, Brodeur In press). 



Methods 



Collection of stomachs 



Juvenile salmon were collected with small-mesh (32 mm 

 stretched), large volume (up to 1 million m^) herring 

 purse seines (Pearcy and Fisher 1988). Sampling was 

 conducted in known areas of salmon concentration in 

 1980; in following years, predesignated stations were 

 sampled along parallel transect lines spaced approx- 

 imately 37 km apart (Fig. 1). Along each transect, sta- 

 tions were located every 9.3 km (5 miles) beginning 

 onshore at the 37-m (20 fm) isobath and continuing out 

 to 56 km from the coast or until no juvenile salmon 

 were caught. Data from a total of 1 5 cruises have been 

 included in this study, with at least one cruise taking 

 place during every month from May through Septem- 

 ber (Table 1). June was the most intensely sampled 

 month and was sampled during each of the 6 years of 

 the study. 



Most collections were made during daylight or 

 twilight, although several complete diel series were 

 made (Brodeur and Pearcy 1987). Circular, quantitative 

 haul sets were made at most stations, although a small 

 proportion of the fish used in this study were collected 

 from sets which were held open for up to 45 minutes 

 or from non-quantitative sets (Pearcy and Fisher 1988). 

 After pursing the net, the catch was then concentrated 

 in the bunt of the seine, and either dip-netted out from 

 the seine or hauled onboard the stern of the sampling 

 vessel. Total sampling time for most sets was approx- 



126° 



125° 



124° 



48' 



47° 



I I II I 



I I 1 I 1 InM I I M M\ 1^ 

 I ^CAPE FLATTERY — 



46' 



45' 



44° — 



43' 



^OREGON- 



45° 



\XOOS 



BAY 



[:CAPE BLANCO 

 I '^1 T' I I I I I I I 



43° 



126° 



24° 



Figure 1 



Location of sampling transects and geographic subareas defined in 

 this study. 



imately 20 minutes. Ancillary physical (temperature 

 and salinity) and biological (chlorophyll o and some 

 zooplankton) data were collected before or after the fish 

 collections. A detailed listing of the station locations 

 and many of the physical and biological measurements 

 are given by Fisher and Pearcy (1985) and Brodeur and 

 Pearcy (1986). 



