Landingham et al.: Feeding habits of juvenile Pacific salmon 
295 
Inside water (IRI) 
(partitioned into inlets and passages) 
Outside waters and 
outer coast inlets (IRI) 
I J Decapod larvae I — I Euphausiids E23 Fishes t— J Hyperiid amphipods i * Tunicates 
Figure 7 
Index of relative importance (IRI) of principal prey of juvenile coho salmon in inside and outside waters and outer 
coast inlets of southeastern Alaska in 1983 and 1984 and outside waters and outer coast inlets of northern 
British Columbia in 1984. 
We examined patterns of prey selectivity by ap- 
plying Strauss’s linear index of food selection to the 
same sets of predator-prey samples that had been 
analyzed for similarity. Pink and sockeye salmon 
selected (L>0.10) neustonic prey more often than 
planktonic prey, and neustonic hyperiid amphipods 
were the most frequently selected organism. Chum 
and coho salmon selected neustonic and planktonic 
prey in nearly equal frequencies (Table 9). All salmon 
species avoided (L <— 0. 10 ) neustonic decapod larvae. 
Selection patterns for planktonic prey varied more 
than for neustonic prey; however, salmon generally 
selected planktonic decapod larvae, hyperiid amphi- 
pods, euphausiids, and fishes and avoided the pro- 
portionately more abundant calanoid copepods. 
Discussion 
This study of the food habits of five sympatric spe- 
cies of Pacific salmon during their first summer in 
the marine waters of southeastern Alaska and north- 
ern British Columbia is the first detailed study for 
this geographic area. Feeding patterns were dynamic, 
with shifts in the important prey categories between 
salmon species, years, and areas. 
We analyzed the importance of two major prey cat- 
egories, zooplankton and teleosts, and examined data 
for temporal and spatial shifts in feeding patterns 
among salmon species. First, in pooled samples (all 
periods and habitats), zooplankton were much more 
important in pink, chum, and sockeye salmon diets 
(60-70% IRI) than in coho and chinook salmon diets 
(36% and 24% IRI; Table 3). When the two years of 
data were analyzed separately, however, contrasting 
patterns emerged. In 1983 (August), pink, chum, and 
sockeye salmon were mostly planktivorous (83-97% 
IRI ), and zooplankton IRI was over 50% even for coho 
salmon. In 1984 (July and August), the proportional 
number, weight, and FO of teleost prey increased in 
diets of all salmon species; however, whereas zoo- 
plankton remained most important in the diets of pink, 
