Landingham et al.: Feeding habits of juvenile Pacific salmon 
291 
sumed by approximately 72% of the 
predators (Table 3). Identifiable fish 
(in order of importance) included 
smelt, Pacific sand lance. Pacific her- 
ring, and pricklebacks. Euphausiids 
were an important prey principally 
in inside inlets during July 1984. 
Incidental prey included decapod lar- 
vae, squid, and hyperiid amphipods. 
Variation in principal prey 
Diets of all four species varied greatly 
(Table 4). Variation in the FO of the 
six principal prey (calanoids, decapod 
larvae, euphausiids, fish, hyperiid 
amphipods, and tunicates) was great- 
est when compared between years: 
the FO of most invertebrate taxa 
decreased significantly (P<0.05) in 
the diets from August 1983 to August 
1984, whereas the FO of teleost prey 
increased (Tables 3 and 4). Except for 
coho salmon, variation in the FO of 
fish was greatest for inshore-offshore 
comparisons; significantly more fish 
were consumed by the other species 
offshore (>37 km from the coast) than 
inshore (0-37 km from the coast) 
(Table 4; 12.69<x 2 <47.72; P<0.001). 
In 1984, the FO of two principal prey 
taxa differed consistently in opposite 
directions between areas: pink, 
chum, and sockeye salmon ate teleost 
prey significantly less often (6.94< 
X 2 <13.42; P<0.01), whereas all four 
species consumed euphausiids sig- 
nificantly more often (5.66<% 2 <17.13; 
P<0.05) in British Columbia waters 
than in Alaska. The diet of sockeye 
salmon varied more than that of the 
other species between years (August 
1983 and August 1984), months (July 
and August), and areas (British Co- 
lumbia and southeastern Alaska; 
Table 4). 
Diet overlap 
<1) 
n 
E 
D 
z 
900 
800 
700 
600 
500 
400 
300 
200 
100 
0 
Total stomachs 
Number of stomachs sampled 
to achieve full prey spectrum 
Pink Chum Sockeye Coho Chinook 
Figure 2 
Total number of stomachs sampled for five species of juvenile Pacific salmon 
and number and percentage of randomly selected stomachs, from samples 
pooled over habitats and time periods, which were required to achieve the full 
prey spectrum for each species as defined in this study. 
Samples pooled over time and 
habitat Significant diet overlap ( C A >0.60) was found 
between pink and sockeye salmon, pink and chum 
salmon, and chum and sockeye salmon when all 
samples were pooled. Coho salmon diet overlap was 
less than 0.60 in all comparisons. Diet overlap was not 
calculated for chinook salmon because of small sample 
sizes (Table 1). The greatest diet overlap was between 
pink and sockeye salmon juveniles ( C A =0.93; Table 5); 
both species ate fish and hyperiid amphipods. The sec- 
ond-greatest diet overlap ( C x =0.91) was between pink 
