tion of their stomach contents ( Good and Ainley 

 unpubl. data) indicated that entanglement oc- 

 curred as they attempted to eat fish caught in the 

 net. A much smaller percentage (ca. 5% ) of shear- 

 waters (pursuit plungers) were caught while eat- 

 ing entangled fish, but probably none of the diving 

 species (murres, puffins, etc.) scavenged from the 

 nets (confirmed by stomach contents). 



Rates of Entanglement 



We tabulated birds by the mesh size in which 

 they were entangled (Table 2). Catches in 112-115 

 mm meshes were combined as were those in 118- 

 121 mm meshes. Not included were birds caught by 

 commercial boats because we were not sure that 

 all specimens reached us. Catch rates for six mesh 

 sizes <82 mm were negligible, but rates for 106- 

 138 mm meshes, including those used in the com- 

 mercial salmon fishery (121-138 mm) (Fredin et al. 

 footnote 2) were statistically greater than for the 

 other categories. In 157-233 mm meshes, with or 

 without the 233 mm mesh — no birds caught and 

 only 27 tans set — the rate was similar to that for 

 82-93 mm meshes. Based on those results, we 

 excluded from further analysis birds caught with 

 s=82 mm meshes. 



The number of birds caught per tan was directly 

 proportional to the density of netable specie5 pres- 

 ent at the netting locality (r = 0.7154, P<0.01, 

 ^-test). The correlation was improved ( r = 0.7604) 

 when catch rates were adjusted for the actual 

 amount of time the nets were deployed (i.e., birds/ 

 tan per hour deployed). The correlation would 

 probably be improved even more if density indices 

 were adjusted for the detectability of birds. For 

 instance, tufted puffins on the water in the outer 

 third of the 300 m count zone would probably not 

 have been seen, especially in rough seas. The diffi- 

 culty in detecting puffins might account for the 

 lack of correlation between density indices and 

 catch rates in Table 1, where significant correla- 

 tions existed for more easily observed species. 



The facts that bird densities differed geographi- 

 cally (or oceanographically; Table 1) and that 

 catch rates were proportional to bird density led us 

 to look for geographic differences in catch rates. In 

 the northern North Pacific, the mean number of 

 birds caught per tan increased with latitude. 

 Rates were lowest, varying 0.00-0.03 bird/tan, be- 

 tween lat. 39° and 43° N (Subarctic Front and the 

 Transitional Domain); were slightly higher, 

 0.03-0.08, between lat. 44° and 48° N (Western 



Subarctic Domain); and were highest, 0.04-0.51, 

 between lat. 49° and 51° N (Alaskan Stream). In 

 that part of the Bering Sea sampled, the number of 

 birds caught per tan increased with decreasing 

 latitude. In the central Bering Sea (Bering Sea 

 Gyre) catch rates were low, ranging 0.01-0.05, and 

 were higher at the gyre periphery. Farther south 

 in the extreme periphery of the gyre, the Bering 

 Current, catch rates ranged 0.04-0.16. These rates 

 were consistent with the relative differences in 

 bird densities in these oceanographic regions (Ta- 

 ble 1). Greater bird densities in the Alaskan 

 Stream and Bering Current were due to the more 

 productive waters there and to closer proximity to 

 the Aleutian Islands where murres, puffins, and 

 several other netable species breed. In fact, as a 

 direct function of distance, within 200 nmi of the 

 Aleutians catch rates varied logarithmically 

 (P<0.01; Figure 2). Rates were especially high 

 within 50-75 nmi of the islands, where most of the 

 murres and puffins in net samples were breeding 

 adults (based on the presence of incubation 

 patches). Most of those caught in the Bering and 

 Western Subarctic Gyres and Transitional Do- 

 mains were immature. 



ijj 0.5- 



1 1 1 1— 1 1 I I 1 I I I I I I I 1 I I 



50 iOO 150 



DISTANCE (NMI) TO THE ALEUTIAN ISLANDS 



Figure 2. — Logarithmic relationships between bird catch rates 

 and distance to the Aleutian Islands within 200 nmi, 1979: 

 S(HMl = Hoyo Mam stations south of the islands ( r = 0.9167, 

 P 0.01: y = 0.74 - 0.13 In .r), S(NM) = Nojima Maru stations 

 .south of the islands ( r = 0.9186. P 0.01; y = 0.36 - 0.07 In .r), 

 and N(OM) = Oshoro Maru stations north of the islands (r = 

 0.9869, P < 0.01; v = 0.74 - 0.13 In .vi; see Figure 1 for station 

 localities. 



Overall Seabird Mortality 



The above estimates of catch rates, if applied to 

 the commercial fishery, are minimal for two 



804 



