296 
Fishery Bulletin 96(2), 1 998 
chum, and sockeye salmon (56-65% IRI), teleosts 
were most important in coho salmon diets (68%IRI). 
Spatial factors that influence both predator and 
prey, especially latitude and proximity to shore, may 
account for similarities or differences in principal 
prey (Andrievskaya, 1970; Brodeur and Pearcy, 1990). 
Most species of salmon are opportunistic and feed 
on a wide variety of prey (Beacham, 1986; Brodeur 1 ), 
but their diets are commonly composed of a few taxa 
readily available at a given time and location. The 
importance of zooplankton in the diets of juvenile 
pink, sockeye, and chum salmon in our study is simi- 
lar to results reported in other studies from the east- 
ern Gulf of Alaska (Manzer, 1969; Jaenicke et al., 
Table 4 
Chi-square (^ 2 ) values for variation in frequency of occur- 
rence of principal prey in four species of juvenile salmon. 
Comparisons were made between years (August 1983 and 
August 1984), months (July 1984 and August 1984), area 
(northern British Columbia and southeastern Alaska), and 
distance offshore (0-37km vs. 46-76km). No asterisk indi- 
cates P > 0.05. * = P<0.05; ** = P<0.01; and *** = P<0.001. 
X 2 
Between 
inshore 
Between 
Between 
Between 
and 
Taxon 
years 
months 
areas 
offshore 
Pink salmon 
Calanoids 
35.29*** 
0.07 
2.47 
0.35 
Decapods 
18.41*** 
0.07 
1.82 
13.04*** 
Euphausiids 
11.35*** 
2.90 
17.13*** 
3.86* 
Hyperiids 
2.34 
0.62 
1.18 
8.60*** 
Tunicates 
7.13** 
19.48*** 
6.08* 
13.83*** 
Teleosts 
6.39* 
0.00 
7.33** 
47 72*** 
Chum salmon 
Calanoids 
8.62*** 
0.25 
6.08* 
0.37 
Decapods 
4.02* 
6.63* 
2.04 
5.94* 
Euphausiids 
9.06*** 
1.74 
5.66* 
5.15* 
Hyperiids 
4.93* 
0.01 
0.12 
15.30*** 
Tunicates 
3.78 
1.77 
3.43 
1.36 
Teleosts 
3.53 
0.53 
6.94** 
26.46*** 
Sockeye salmon 
Calanoids 
6.84** 
3.73 
3.72 
0.08 
Decapods 
14.77*** 
26.12*** 
1.74 
1.36 
Euphausiids 
19 47 *** 
13.21*** 
14.81*** 
0.68 
Hyperiids 
73 17*** 
57.10*** 
0.34 
0.04 
Tunicates 
0.72 
0 77*** 
5.45* 
4.51* 
Teleosts 
11.05*** 
0.09 
13.42*** 
12.69*** 
Coho salmon 
Calanoids 
9.26*** 
4.35* 
— 
— 
Decapods 
14.96*** 
20.50*** 
0.00 
1.46 
Euphausiids 
0.80 
0.07 
7.20** 
3.98* 
Hyperiids 
48.46*** 
0.11 
3.86* 
— 
Tunicates 
0.86 
7.00** 
0.33 
— 
Teleosts 
5.77* 
6.90** 
1.84 
0.59 
1984; Hartt and Bell, 1986), the Sea of Okhotsk 
(Andrievskaya, 1968, 1970), and other regions of the 
northeastern Pacific Ocean (Peterson et al., 1982; 
Brodeur, 1989; Brodeur and Pearcy, 1990). We found 
that hyperiid amphipods were especially important 
in pink and sockeye salmon diets, and less so in chum 
salmon diets. The principal hyperiid amphipod found 
in diets of these species in the northeastern Pacific 
Ocean, Iincluding our study, was Themisto pacifica. 
In the more southern latitudes of coastal Washing- 
ton and Oregon, another hyperiid amphipod 
( Hyperoche medusarum ) is an important component 
in chum, coho, chinook, and sockeye salmon diets 
(Brodeur and Pearcy, 1990). In that area, however, 
euphausiids are more important in juvenile chum 
diets and, to a lesser extent in juvenile coho and 
chinook salmon diets, than are hyperiid amphipods 
(Peterson et al., 1982; Brodeur and Pearcy, 1990). 
Similarly, we found that euphausiids occurred less 
frequently and were less important in diets of all 
juvenile salmon located off southeastern Alaska, a 
downwelling region, compared to diets of salmon from 
southern British Columbia to Oregon, where up- 
welling is more prevalent (Ware and McFarlane, 
1989). 
Although our study showed that chum salmon are 
primarily planktivorous, like pink and sockeye 
salmon, we observed another difference in their most 
important prey. Tunicates (larvacea and salps) were 
important only in the diets of chum salmon. The IRI 
for tunicates, especially the larvacean Oikopleura 
dioica, was greater than for any other taxon in chum 
salmon diets (30. 1%). Tunicates were most prevalent 
in samples from outer coast inlets and outside wa- 
ters (0-37 km). This finding is not unusual. 
Oikopleura was the dominant prey item in two other 
studies of the diet of juvenile chum salmon from 
northern British Columbia: 62% FO in fish 32-106 
mm FL from Chatham Sound (Manzer, 1969) and 70- 
76% FO in fish 105-158 mm FL from Hecate Strait 
(Healey, 1991). 
In addition to interspecific differences in the most 
important prey, shifts in diet differed among the four 
salmon species both spatially and temporally. We 
observed spatial differences in July 1984, when sam- 
pling included waters of northern British Columbia, 
and in August 1984, when sampling was extended to 
76 km offshore. Pink, chum, and sockeye salmon con- 
sumed fish prey about twice as often offshore as they 
did inshore, and more often in British Colombia than 
in Alaska. The more piscivorous coho salmon, how- 
ever, consumed fish just as frequently inshore as off- 
shore and just as frequently in marine waters of 
Alaska and British Columbia. Diet varied by season 
and distance offshore for the highly similar pink, 
