Laake et al : Pinniped diet composition 



441 



Figure 2 



Estimates of the average consumption of each prey group (P ) by harbor seals on the Columbia River between 1995 and 1997 

 for spring (A), summer (B), and fall (C). 



prey items weighing 10-20 g, SSFO predicted consump- 

 tion that was 10 times greater than BR. Likewise, for prey 

 items with a 1 kg mass or greater, SSFO predicted con- 

 sumption estimates that were less than one-tenth of the 

 BR estimate. For prey items near the median mass of 173 

 g, both estimators produced similar results. 



Discussion 



The consumption estimates for the Columbia River harbor 

 seals could be improved by incorporating differences in 

 energy density across prey and by measuring the sex- and 

 age-structure of the seal population through time rather 

 than using a life-table which may not be appropriate. 

 These are valid criticisms and they could be overcome by 

 collecting additional data. However, we believe these are 

 less important than the current inadequacies in measur- 

 ing diet composition that will likely affect any study that 

 attempts to estimate consumption based on scat analysis. 

 It is well recognized that digestion does not act equally 

 on all hard parts and is certainly species-specific for oto- 

 liths (Harvey, 1989). We included other hard parts such 



as bone to reduce bias due to differential otolith digestion; 

 however, that inclusion may not remove all of the selectiv- 

 ity bias and it certainly introduces several other problems 

 discussed below. Also, because hard parts are used for 

 diet composition, if seals are only eating the fleshy parts 

 of large fish, a significant portion of their diet may be 

 missed. 



