FISHERY BULLETIN: VOL. 82, NO. 2 



Seasonal Survey Methods 



Stomachs for the seasonal study were collected 

 during rockfish surveys conducted by the Oregon 

 Department of Fish and Wildlife (ODFW) on 

 Heceta Bank off the central coast of Oregon. These 

 surveys obtained hydroacoustic and environmen- 

 tal data along with the trawl catches. A total of 317 

 stomach samples was collected during seven sur- 

 veys conducted in 1980-81 (Table 2). All surveys 

 used trawling gear similar to that used in the 

 summer surveys. 



Locations of the tows were chosen on the basis of 

 high concentrations offish found during acoustic 

 surveys over the outside edge of Heceta Bank 

 between lat. 44°20'N and 44°00'N between the 

 128 m and 238 m bathymetric contours (inset, Fig. 

 1). The duration of tows was variable but averaged 

 < 1 h. No tows were attempted at night because of 

 the lack of acoustical targets near the bottom at 

 this time. Stomachs were collected as described 

 earlier. 



TABLE 2. — Number of rockfish stomachs analyzed from the 

 seasonal Oregon Department of Fish and Wildlife collections 

 on Heceta Bank. All dates are in 1980 unless otherwise noted. 



Analysis of Stomach Contents 



The stomachs were opened and their contents 

 transferred to 50% isopropyl alcohol in the labora- 

 tory. Contents were examined using a variable 

 power dissecting microscope. Individual stomach 

 fullness was estimated according to a subjective 

 rating ranging from (empty) to 5 (stomach fully 

 distended with food). The condition of the contents 

 was assigned a value from (well-digested, barely 

 identifiable to phylum) to 4 (fresh). 



Prey were identified to the lowest possible taxon 

 and enumerated. In stomachs containing many 

 small prey, such as euphausiids, any large or rare 

 prey items were removed first. The remaining 

 contents were then subdivided by means of a 



272 



Folsom plankton splitter (McEwen et al. 1954), 

 and the contents of one subsample were used 

 to estimate the stomach contents of small prey. 

 The digested state of the contents of many stom- 

 achs made precise counts of some prey difficult. 

 Some paired parts of prey animals (e.g., eyes of 

 euphausiids, otoliths of teleosts) were more resis- 

 tant to digestion and total counts of these parts 

 were halved to yield minimum counts of prey in- 

 gested. Total lengths or greatest dimensions of 

 intact prey found in the stomach were measured to 

 the nearest 0.1 mm for the total sample (or a sub- 

 sample of at least 15 individuals) using a stage 

 ruler or ocular micrometer. All prey were blotted 

 dry with absorbent paper and wet weights of each 

 taxon were recorded to the nearest milligram. 



Analysis of Food Habits 



The minimum number of stomach samples 

 needed to adequately describe the diet of a species 

 was determined for all five rockfish species, using 

 a cumulative prey species curve. A subset of 

 stomachs of a particular species was randomly 

 chosen and the cumulative number of unique prey 

 taxa were then plotted versus the number of stom- 

 achs which produced these taxa. The point on the 

 abscissa where the curve begins to level off is 

 considered the minimum number of stomachs nec- 

 essary to describe the diet of that species. An 

 example of the cumulative prey curves for the first 

 28 stomachs of each of the species in this study is 

 shown in Figure 2. Although the curves assume 

 different shapes, all approach an asymptote at 

 sample sizes less than those analyzed. 



The contributions of the different prey items to 

 the total diet of the rockfishes were expressed as 

 percent frequency of occurrence, percent numeri- 

 cal composition, and percent gravimetric composi- 

 tion. Breadth and overlap were calculated for the 

 five rockfishes from the summer surveys and for 

 S. pinniger and S. flavidus from the seasonal 

 surveys, using the pooled p ; 's (relative proportion 

 of the total number or biomass of resource i used 

 by each species) for the major taxa. These include 

 all taxa identified to at least generic level that 

 exceeded Q.\ c /< of the total weight or number of all 

 identified foods. Resource breadth was computed 

 for each species using the following formula: 



B = 



1 



-I Pi 



where B equals R (the total number of prey taxa 



