OLIVER ET AL.: WALRUS FEEDING IN BERING SEA 



FIGURE 4. — Shells of Mya truncata (A), Macoma calcarea (B), and Serripes groenlandicus (C) discarded by 



feeding walrus. 



low air supply. Nevertheless, this incomplete record 

 is the most accurate estimate of the number of clams 

 it is possible for a walrus to excavate during a sin- 

 gle dive. 



Effects on Benthic Communities 



Walrus undoubtedly have a significant effect on 

 abundance and size distribution of bivalve prey. Be- 

 cause we did not obtain adequate samples of the 

 large individuals in living bivalve populations, we 

 could not compare availability of various sizes of prey 

 with the sizes of prey consumed. Nevertheless, dis- 

 carded shells indicated that walrus consumed rela- 

 tively large individuals from the three principal 

 groups of bivalves (Fig. 5). 



It is important to determine whether walrus disrupt 

 different nonprey populations while feeding in dif- 

 ferent local habitats. Core samples were taken to 

 document the species composition and relative abun- 

 dance of the smaller infauna, which could be dis- 

 placed, injured, or killed during excavation of the 

 large bivalve prey. These samples indicate that ben- 

 thic infaunal communities were strikingly different at 

 the major feeding sites near Nome (Table 5). The 

 western areas were near Cape Rodney and were nu- 

 merically dominated by tube-building polychaete 

 worms, Myriochele oculata, and Polydora flava flava. 



TABLE 5. — Abundant infauna found at the three major feeding 

 sites of walrus. Mean numbers per 0.0075 nr with standard 

 deviations in parentheses. P = polychaete; C = crustacean; O 

 = ophiuroid; T = tunicate; Pr = protozoan. 





We located an excellent record of Mya pits and shells 

 there (Table 1), and observed the siphons and si- 

 phon-burrows of large, living M. truncata. At Sledge 



507 



