400 
Fishery Bulletin 108(4) 
A Deep station (190 m) 
100 
80 
60 - 
40 - 
20 - 
0 
ZZ 
to 100 
C 
O 
B Shallow station 
80 
60 -I 
40 
20 1 
0 
£ 
1 
(60 m) 
7 
— - 
ZZ 
•S 
1 
/// 
z* / / ,<Z 
off z <z z 
z" / z 
^ /VV 
Z z 
<<? ^ 
1985 
1986 
1987 
; Calanus pacificus 
Metridia lucens 
D Pseudocatanus spp 
\/ \ Oithona sp. 
Other Copepoda 
| ' ■ Copepod nauplii 
[\Z Euphausiacea 
I ] Larvacea 
Pteropoda 
DU E 99 s 
E3 other 
Figure 5 
Composition of zooplankton in the upper 50 meters at night at a deep station (A) and a shallow station 
(B) in Dabob Bay, Washington. Zooplankton collections were made with duplicate vertical hauls of 
a plankton net at each station during each of the four seasons during 1985, 1986 and 1987. Dashed 
vertical lines separate years. 
in summer and autumn. In contrast, Oithona sp. and 
larvaceans dominated the shallow station in summer 
and autumn. In 1987, however, zooplankton commu- 
nity composition between the two sites was remarkably 
similar (Fig. 5) and was dominated by Oithona sp. and 
“other Copepoda” (predominantly unidentified stages 
1 and 2 calanoid copepodites) throughout the entire 
sampling period. 
Electivity indices for juvenile salmon 
All salmon species consistently and strongly selected for 
insect, arachnid, and cephalopod prey (£ ; >0.75) and rou- 
tinely selected against Calanus pacificus, copepods other 
than C. pacificus, ctenophores, larvaceans and pteropods 
(£,•<-0.25). All salmon species except pink salmon typi- 
cally selected for decapod prey (0.25< £,<0.75; both cari- 
dean shrimp and brachyuran crab larvae were present 
in the spring, but brachyurans dominated the decapod 
prey in summer). Hyperiid amphipods and teleosts were 
generally selected for by chum and Chinook salmon 
(0.25<£ ; <0.75) but were consumed in proportion to their 
relative abundance or were selected against by coho and 
pink salmon (-0.75<£ ( <0.25). Euphausiids were gener- 
ally neutrally selected (-0.25<£ ; <0.25). However, elec- 
tivity scores varied substantially within a given salmon 
species. For example, in 26 Chinook salmon diet samples, 
electivity indices for euphausiid prey ranged from -1.0 to 
0.99 (data not shown), but were not consistently related 
to predator size, season, or prey abundance. 
Important ontogenetic changes in prey-selection 
behavior were revealed by size-specific electivity val- 
ues (Fig. 6). For instance, smaller size chum salmon 
strongly avoided larvacean prey, whereas larger chum 
(>75 mm) showed roughly neutral or positive selectivity. 
Similarly, small (75-99 mm) Chinook salmon tended to 
avoid gammarid amphipods, whereas larger individuals 
(>100 mm) tended to select gammarids. Also, smaller 
Chinook salmon individuals tended to consume Calanus 
pacificus at rates proportional to their abundance in the 
environment, but the larger Chinook salmon strongly 
avoided these prey. In general, it seemed that larger fish 
