404 
Fishery Bulletin 108(4) 
the behaviors exhibited by their potential prey have 
been explored (Ohman, 1986; Frost, 1988; Bollens and 
Frost, 1989; Bollens et al., 1992a; Bollens et ah, 1993). 
Field studies by Bollens and Frost (1989) indicated 
that abundances of actively feeding planktivorous fish 
(including Oncorhynchus spp.) are directly linked with 
the strength and timing of vertical migration exhibited 
by the copepod Calanus pacificus. Our results indicate 
that the adaptive response exhibited by species such 
as Calanus pacificus seems to be an effective mecha- 
nism for avoiding predation by species such as juvenile 
salmon. Thus, “available” prey items are not only those 
that are abundant or of the desired size, but those that 
are also available for visual detection. Availability may 
be affected by the prey’s presence or absence from the 
photic zone, or by the presence of pigmentation that 
makes the prey more detectable visually. Most of the 
prey items that were consistently consumed by salmon 
in this study (e.g., euphausiids, hyperiid and gammarid 
amphipods, and decapod larvae) possess characteristi- 
cally dark or large eyes. The ability of salmon to detect 
these potential prey items may be increased by heavy 
pigmentation, large body size, and their frequently not- 
ed association with the near surface layer (Lough, 1976; 
Peterson et al., 1982) where salmon typically feed. 
Diet overlap and potential interspecific competition 
among salmon species 
A variety of studies have relied on diet overlap as a pri- 
mary indicator of potential resource competition between 
co-occurring species. Although there is little consensus 
among studies of salmon, diet overlap has been most 
frequently observed between Chinook and coho salmon in 
Oregon and Washington (Peterson et al., 1982; Emmett 
et al., 1986; Brodeur and Pearcy, 1990; Brodeur, 1991), 
and to a lesser extent between chum, pink, and sockeye 
salmon ( Oncorhynchus nerka) in British Columbia and 
Southeast Alaska (Healey, 1980; Beacham, 1993; Land- 
ingham et al., 1998). Our results from Dabob Bay show 
the greatest spatial and temporal overlap between chum 
and Chinook salmon but also provide evidence only for 
resource partitioning (low diet overlap) between these 
two species, not necessarily competition (which would 
require resource limitation). 
In contrast, our data show significant diet overlap 
between Chinook and coho salmon (average PSI=77.9%), 
supporting earlier reports of potential resource com- 
petition between these two species. Although Brodeur 
and Pearcy (1990) did not see evidence for significant 
overlap between four salmon species when all observa- 
tions were combined, they observed significant overlap 
between Chinook and coho salmon during May and 
June, as well as during the 1983 El Nino. Likewise we 
found that significant overlap between Chinook and 
coho salmon occurred during June (of 1986 and 1987), 
largely because of the shared consumption of decapod 
larvae, which are visually conspicuous and seasonally 
abundant at this time of year. However, the co-occur- 
rence of juvenile Chinook and coho in Dabob Bay was 
less prominent than in coastal Oregon (Peterson et al., 
1982; Brodeur and Pearcy, 1990; Brodeur and Pearey, 
1992); however, our results are based on far fewer data. 
Despite the co-occurrence of different juvenile salmon 
species in Dabob Bay, and the occasional occurrence of 
significant diet overlap between these species, we did 
not see any indication of food limitation. That is, there 
was never a significant relationship between stomach 
fullness and zooplankton biomass, as might be expected 
if food was limited. However, just as with the electivity 
indices discussed above, we caution that our vertical 
plankton net hauls may not adequately sample the po- 
tential prey of juvenile salmon. Testing for food limita- 
tion by correlating salmon stomach fullness and the 
abundance of potentially more appropriate prey (e.g., 
macrozooplanktonic, micronektonic, and neustonic prey) 
would prove interesting, but was not possible given our 
sampling method. Similarly, our comparison of zoo- 
plankton dry weights with salmon stomach wet weights 
complicates the interpretation of food limitation and po- 
tential competition because conversions from wet-weight 
to dry-weight would be expected to vary between prey 
taxa (e.g., between gelatinous and crustacean prey). 
Resource limitation by juvenile salmon during their 
early marine transition may be influenced by several 
other factors not addressed in our study, including di- 
rect and indirect effects of hatchery production in the 
region (Quinn et al., 2005) and potential diet overlap 
with other zooplanktivores (Purcell and Sturdevant, 
2001). Furthermore, zooplankton dynamics in temper- 
ate marine waters are clearly influenced by interan- 
nual (El Nino cycles) and interdecadal (Pacific Decadal 
Oscillation) scales of climate variability (Mackas et 
al., 2001; Hooff and Peterson, 2006) and there are im- 
portant linkages to salmon survival during multiple 
life-history stages (Beamish and Bouillon, 1993; Loger- 
well et al., 2003). Although diet data from our three- 
year study could be averaged across years (as opposed 
to size classes, etc.), interannual climate factors can- 
not be overlooked. Indeed, based on a multivariate El 
Nino-Southern Oscillation index (MEI), 1987 ranks 
as a moderate to strong El Nino year (April-October 
MEI average = 1.91), and likewise represents the most 
anomalous year of our study for seasonal plankton com- 
position in Dabob Bay. This was particularly apparent 
at the deeper station, where the abundance of Oithona 
sp. and larvaceans seemed to be more characteristic 
of the shallow, nearshore station in 1985 and 1986. 
Mechanisms underlying the interannual variability of 
zooplankton composition in Dabob Bay warrant further 
exploration. 
Our combination of detailed analyses of prey fields 
and fish diets is clearly only one approach to under- 
standing juvenile salmon feeding ecology. Biochemical 
methods of studying energetics and feeding relation- 
ships (e.g., Johnson and Schindler, 2009) provide ad- 
ditional insight into juvenile salmon trophic dynamics. 
Studies across the variety of habitats encountered by 
salmon during their outmigration and early residence 
in marine environments will be necessary to fully un- 
