426 



Fishery Bulletin 100(3) 



viduals consumed by harbor seals on the Columbia River 

 were extremely v'ariable, even among sample collections 

 fewer than two weeks apart. Effort and sample sizes were 

 unequal for season and years but we chose to include all 

 data to better describe harbor seal diet. More than 45 prey 

 tEixa were described in 1385 samples with identifiable 

 prey remains; however, most of the diet by number and 

 frequency was composed of about 17 prey taxa (Table 3). 



Seasonal effects were important for 15 of 17 harbor seal 

 prey with FO >5% (Table 3). Annual effects also were 

 important for FO of lamprey (Lampetra spp. ), Pacific hake 

 iMerluccius productus), and northern anchovy (EngrauUs 

 morda.x), and year-season interactions were included for 

 three prey taxa (Table 3). All taxa had variances of FO 

 greater than predicted by a binomial model (over-disper- 

 sion values, b>l; Table 3). 



