Although the differences may be en- 

 tirely an artifact of the sampling method 

 and consequently must be approached 

 with caution, they are nonetheless of 

 considerable interest. Several of the 

 itenns seenn to appear more often in 

 one year than in the other. Widow 

 rockfish were found in 9.2 percent of 

 the coho salmon stomachs collected in 

 1958 but did not appear in the 1957 

 collections. Frequency of occurrence 

 of herring in king salmon was less in 

 1957 (40.0 percent) than in 1958 (56.7 

 percent). Squid occurred in 1.3 per- 

 cent of the king salmon stomachs in 

 the 1957 samples but accounted for 

 13.8 percent in 1958. Euphaus ids, found 

 to be taken by salmon frequently in 

 other areas (Heg and Van Hyning, 

 1951), were fovmd in none of the South- 

 eastern Alaska 1957 collections and 

 in only a small percentage of the 1958 

 collections. Several other food items 

 show considerable differences in per- 

 cent of occurrence between the 1957 

 and 1958 seasons (table 3). 



By Geographical Area 



Table 4 shows differences between 

 species in the percentage frequency of 

 occurrence of major food items by 

 numbered area (fig. 1). Area IV pro- 

 duced the lowest percentage occurrence 

 of herring and the highest percentage 

 occurrence of sand lance. Area III had 

 the most frequent occurrence of her- 

 ring inking salmon and also the greatest 

 variety of organisnns, and area I had 

 the most frequent occurrence of her- 

 ring in coho salmon. Figure 3 illustrates 

 the relative frequency of five major 

 food items for both species of salmon 

 in the four areas. This figure indicates 

 considerable variation in the diet of 

 salmon between geographical locations. 

 In areas I, II, and IV, sand lance was 

 an important item of the diet, but in 

 area III, it was of minor importance. 

 Capelin, a major food item in areas 

 II and IV, was tadten incidentally in 

 area III and did not occur in area I, 

 Although herring appeared as a signifi- 

 cant item in all areas, considerable 

 differences in its use existed between 



AREA I 



AREA n 



AREAUr 



AREA m 



ALL AREAS 



I I Herring 

 ^ Sand lance 

 S Squid 

 H Capelin 

 ^ Other smelt 

 M other 



Figure 3.- -Frequency of occurrence of major items found 

 in salmon stomachs, Southeastern Alaska, 1957-58. 

 [Based on total occurrence of food items.] 



areas; the greatest of these was the 

 disparity between areas III and IV. 



By l^ecies of Salmon 



A much larger percentage occur- 

 rence of herring was foiind in king 

 salmon (50.5 percent) (table 4) than in 

 coho salmon (14.9 percent). This ap- 

 parent difference in food preference 

 was also found by Senter (1940) and 

 Pritchard and Tester (1944). The sec- 

 ond most utilized food item in the 

 king salmon diet was squid (8.8 per- 

 cent). This organism was found in 32 

 of the king salmon stomachs and in 

 only 1 coho stonnach. In coho salnnon, 

 the occurrence of samd lance (12.6 

 percent) was slightly less than that of 

 herring. Capelin was also frequently 

 taken by coho salmon (10.9 percent) 

 but occurred in only 5.1 percent of the 

 king salmon stonnachs. Figure 4 indi- 

 cates the percentage of food items 

 taken by each species of salmon by 

 frequency of occurrence in stomachs 

 containing food. 



