OLIVER ET AL.: WALRUS FEEDING IN BERING SEA 



dant in the area during the spring (Lowryetal. 1980), 

 but gray whales are seen infrequently (pers. obs.). 

 The biomass of benthic animals is dominated by 

 bivalve molluscs and echinoderms with large num- 

 bers of a few sedentary polychaete worms (Stoker 

 1978). The number and biomass of crustaceans are 

 much lower in the study area compared with the cen- 

 tral and western parts of the northern Bering Sea 

 (Stoker 1978). 



Field work was done from 22 May to 7 June 1981. 

 Remnants of shore-fast ice moved away from Nome 

 several days before our arrival. Well- developed pack- 

 ice and large groups of walrus were observed in the 

 general study area during the preceding month. 

 These animals probably fed in the region for at least a 

 month before our arrival. Therefore, the benthic 

 feeding record was likely to be quite recent. No wal- 

 rus were seen in the study area after the sea ice 

 moved offshore around 15 May 1981. 



(area= 0.0075 m 2 ; depth = 12-15 cm). Samples were 

 washed over a 0.5 mm screen, and preserved in a 

 solution of 4% formaldehyde. Animals were iden- 

 tified to the lowest possible taxon and counted. Al- 

 though juvenile bivalves and small species were 

 adequately sampled by the corers, larger individuals 

 were not, particularly the major walrus prey, Macoma 

 spp., Mya truncata, andSerripes groenlandicus. How- 

 ever, the siphons and siphon burrows of the deep- 

 burrowing clam, M. truncata, were counted in 1 m 2 

 areas to estimate the abundance of these large in- 

 dividuals. Sediment consolidation was measured 

 with a simple penetrometer, which was a weighted 

 rod (0.5 kg) dropped through a cylinder resting on the 

 bottom. Penetration was estimated as distance of rod 

 penetration into the surface sediments (e.g., Ronan 

 1975). 



RESULTS 



METHODS 



Thirty-three dives were made south and west of 

 Nome (Fig. 1). At each site, divers using scuba thor- 

 oughly searched the bottom for traces of walrus feed- 

 ing activity. The benthic feeding record was quan- 

 tified at only several sites where feeding traces were 

 found. Here, discarded bivalve shells were collected, 

 and the distance to the nearest pits and furrows was 

 recorded. Pit and furrow dimensions were measured, 

 as well as the number of pits in a patch, and the area of 

 bottom containing each distinct patch of pits. A 

 patch of pits was considered distinct when no ad- 

 ditional pits were found within 5 m of the group. 

 Water clarity of <1 m limited the patch size obser- 

 vations in all areas except Cape Nome. Shell lengths 

 and breaking strengths (using a hinged plate that was 

 calibrated to pounds of pressure) were measured at 

 the laboratory. 



Small benthic infauna that are not walrus prey were 

 sampled directly in feeding excavations, and in adja- 

 cent undisturbed bottoms with hand-held corers 



Benthic Feeding Record 



We observed two basic types of excavations, fur- 

 rows and pits. In both cases, shells of the excavated 

 clams were discarded close to excavations. There- 

 fore, different pits and furrows were easily linked to 

 the species of excavated bivalve prey. There were 

 three principal bivalve prey, Mya truncata, Serripes 

 groenlandicus, and Macoma spp. (mostly M. cal- 

 carea). These groups are recognized as major prey by 

 Vibe (1950) and Fay (1982). Feeding records gen- 

 erally contained one or two types of excavations: 

 Furrows, Mya pits, or a mixture of Mya and Serripes 

 pits. The relative abundance of discarded shells cor- 

 responded to the primary type of excavation in each 

 area. For example, 92 to 100% of the shells were Mya 

 truncata at stations that primarily hadMya pits; 76 to 

 83% of the shells were Macoma spp. at stations that 

 primarily had Macoma furrows; and Mya truncata 

 and S. groenlandicus shells were both abundant at 

 stations that had mixed pits (Table 1). Bivalve prey 

 thus were identified by discarded shells and by ex- 



Table 1. — Excavation type and percentages of discarded shells from the three major prey 

 found at the main feeding sites near Nome, Alaska. Percentages are based on the number of 

 reconstructed whole clams. 



'Areas contained either mostly Mya pits, mixed pits of Mya and Serripes. or mostly furrows with Macoma 

 2 Primarily Macoma calcarea. 



503 



