376 
Fishery Bulletin 109(4) 
Figure 2 
Frequency of occurrence (FO) of primary prey (prey with total FO values>0.05) identified from Steller sea 
lion (Eumetopias jubatus) scat collected in Oregon and northern California from 1986 through 2007. FO is 
presented by collection site and season (breeding season = May-August, nonbreeding season = September- 
April) in descending order of overall pooled FO. Error bars indicate exact 95% binomial confidence intervals. 
Scientific names for prey types can be found in Table 1. P=Pacific; N=northern; s = staghorn; T=threespine. 
How important a specific type of prey is to the sur- 
vival of an opportunistic marine pinniped predator is 
unknown. Abundance of prey may have more impact 
on survival when a predator feeds on schooling fish 
rather than on more solitary types of prey, such as 
flatfish and sculpin. Furthermore, a diversity of prey 
types may be important in sustaining populations and 
help buffer the effects of ocean climate changes. For 
example, analyses by Merrick et al. (1997) and Trites 
et al. (2007b) showed a strong positive correlation be- 
tween diet diversity and rate of population change. 
Our data are consistent with this finding; we identified 
17 primary (>5%) prey types (Fig. 3) and the popula- 
tion has been growing at approximately 3% per year 
(Pitcher et al., 2007). 
Perhaps as a reflection of their diverse diet, we found 
a surprisingly high number of statistical differences in 
diet composition between months (Table 3), years (Table 
4), and sites (Table 5). Although simulation exercises 
(not presented) indicated that the Mantel test was not 
overly sensitive (e.g., it did not reject a null hypothesis 
simply due to a large difference in a single prey type), 
the procedure is only a hypothesis test and does not 
lend itself to estimation of effect sizes or biologically 
