28 
Fishery Bulletin 95(1), 1997 
the following criteria: 1) they were caught before the 
first release of Anadromous, Inc. spring chinook 
salmon on 19 June; 2) metacercarial cysts were 
present on their skin or fins; 3) they had one of the 
STEP fin clips; or 4) they were <100 mm FL. Fish 
were classified as spring chinook salmon if they were 
>101 mm FL and did not meet any of the criteria for 
fall chinook salmon. 
Stomach contents were examined and prey items 
identified to the lowest possible taxon from 116 fall 
chinook salmon and 65 spring chinook salmon col- 
lected between 29 June and 13 August 1987, the pe- 
riod of greatest overlap in the bay of the two groups 
(Fig. 2). Stomach contents from a single fall chinook 
salmon caught on 7 June were also examined. 
Individual prey taxa in each stomach were weighed 
to the nearest 0.001 g after removing excess mois- 
ture by blotting. Those taxa that were too light to 
register on the scale (weight <0.0005 g), were as- 
signed a weight of 0.0004 g. The estimated total 
weight of all food assigned this arbitrarily small value 
was only 0.05 g out of a total weight of 60.2 g for all 
taxa from all stomachs. 
In the analyses of stomach contents, juvenile fall 
and spring chinook salmon were grouped by FL, by 
two collection areas (“lower bay,” stations 1-3, and 
“mid-bay,” stations 4-5) and by two sampling peri- 
ods: 29 June to 17 July and 3-13 August. Within each 
class, the percent frequency of occurrence (FO) and 
percent by weight of each prey category in the diet 
was calculated. The percent by weight (p.xlOO) of 
each prey category in each class was calculated as 
100(p,-)= 100 
N n 
V <7=11=1 
( 1 ) 
where w iq is the weight of food category i in fish q, n 
is the number of food categories, and N is the num- 
ber of fish in the class. 
Dietary overlap between classes was calculated by 
using the Schoener overlap index (ro; Schoener, 1970; 
Wallace, 1981; Linton et al., 1981): 
where p tj and p ik are the proportions by weight of 
food category i (Eq. 1) in the diets of fish in classes j 
and k, respectively, and n is the number of food cat- 
egories. Dietary overlap was calculated by using 14 
categories of major prey and, because the overlap 
index is sensitive to the taxonomic resolution 
(Brodeur and Pearcy, 1992), it was also calculated 
by using the 86 lowest taxonomic levels identified 
(to genus or species in some cases). An overlap of 
>0.60 was considered significant (Zaret and Rand, 
1971; Brodeur and Pearcy, 1992). 
Results 
Stomach fullness 
The frequency distribution of stomach-content weight 
as a percentage of body weight (“stomach fullness”) 
was skewed for both fall and spring chinook salmon 
(Fig. 3); therefore, nonparametric ranks tests were 
used to compare stomach fullness among different 
classes of fish. The median stomach fullness was 
higher for fall chinook salmon than for spring chinook 
salmon (2.4% vs. 1.2%, respectively; Mann Whitney 
(Wilcoxon) W test, W=ll,630, P<0.0001). Stomachs 
were empty in a higher percentage of spring chinook 
salmon than of fall chinook salmon (16% vs. 1%), 
contributing to the difference in median stomach 
fullness of these two groups (Fig. 3). 
