reasons: most could be increased by >30% to ac- 

 count for the higher catch rates of nets with com- 

 mercial size meshes (Table 2 ), and our estimates of 

 commercial catch rates did not account for birds 

 that dropped from the net before its retrieval. On 

 the July cruise of //o.vo Maru No. 67, 3 shearwaters 

 (of 6 caught) and 1 puffin (of 12 caught) dropped 

 from the net when it was stretched taut in the 

 retrieval process. On that cruise, 13% of the dead 

 birds would not have been counted if only figures 

 on the number of dead birds reaching the deck had 

 been used as in other estimates of catch rates. 

 Similarly, on cruises of the Hokusei Maru in 1978 

 and that part of the Oshoro Maru cruise south of 

 the Aleutian Islands in 1979. 1 of 9 (11%) and 3 of 

 66 birds (5% ), respectively, dropped out during net 

 retrieval. 



A valuable result of this study is the more 

 realistic estimate of bird catch rates, compared 

 with previous estimates (Sano 1978; King et al. 

 1979; Japanese Fishery Agency 1977 in DeGange^; 

 DeGange footnote 3). Prior estimates agreed but 

 were too low because they were based on research 

 gear and the assumption that catch rates were the 

 same regardless of mesh size, and they did not give 

 enough attention to geographic differences in 

 catch rates. They derived an overall mean value 

 for birds caught per tan fished which dilutes con- 

 siderably the high catch rates in certain areas. 

 DeGange's (footnote 3) analysis, representative of 

 earlier mortality estimates, derived a total of 

 about 112,500 birds caught annually in the 

 mothership fishery. We propose that a figure of 

 205,000, an increase of about 82%, is more realis- 

 tic (Table 3). If our estimate were increased by an 



■'DeGange, A. R. 1978. Observations on the mortality of 

 seabirds in Japanese salmon gill nets made from the OSHORO 

 MARU and HOKUSEI MARU, summer 1978. Unpubl. Rep., 

 37 p. U.S. Fish Wildl. Serv., Off. Biol. Serv., Anchorage, 

 Alaska. 



Table 3.— A comparison of bird catch in the 1978 mothership 

 salmon fishery using two methods of catch rate estimation; the 

 DeGange ( text footnote 3 1 method is typical of all previous types 

 of estimation. 



' Source of data: International North Pacific Fisheries Commission (text foot- 

 note 4). 



additional 309f to adjust for the high catch rates of 

 commercial meshes, the resultant figure of 

 266,500 birds is 136'/, higher. 



The demonstration that bird catch rates in- 

 crease logarithmically as distance to the Aleutian 

 Islands decreases, and are generally higher in 

 productive waters, is especially important. More- 

 observations are needed on catch rates in commer- 

 cial nets to clarify the critical distance, but fishing 

 at some distance within 50-75 nmi of the islands 

 would severely reduce breeding populations of cer- 

 tain diving birds. This is precisely what happened 

 when a salmon gill net fishery off Greenland was 

 concentrated too near to murre breeding sites 

 (Tull et al. 1972). In that fishery, 88% of the esti- 

 mated 350,000-500,000 thick billed murres 

 caught per year were entangled in nets set -30 

 nmi from the coast. Other species, however, 

 showed different distance-to-coast relationships; 

 for example, 36% of the black guillemot, Cepphus 

 grylle, were entangled <12 nmi from the coast, 

 80% of the greater shearwater. Puffin us gravis. 

 were caught between 12 and 30 nmi, and 75% of 

 the dovekie, Plautus alle, were caught between 12 

 and 60 nmi (50% at 30-60 nmi; Christensen and 

 Lear 11976)). 



In the North Pacific where the salmon driftnet 

 fishery is much larger than the one in Greenland 

 was, it is likely that gill netting has also been 

 concentrated near bird breeding sites. In a sample 

 of years for which data on the number of tans 

 fished were available to us ( n =12 during 1955-69 

 International North Pacific Fisheries Commis- 

 sion'*), 44.4% of effort east of long. 170° E, or sev- 

 eral million tans annually, was concentrated in 

 the six 2° x 5° blocks (24 total blocks fished in that 

 area) containing the western Aleutians. Such con- 

 centration of effort, coupled with marked geo- 

 graphic differences in bird catch rates, indicates 

 the limitations in estimating the total seabird kill 

 by using statistics averaged over broad areas as 

 attempted by King et al. ( 1979). Their estimate of 

 5.0 million birds killed by the entire mothership 

 fishery between 1952 and 1974, based on an aver- 

 age annual mortality of 250,000, is extremely low 

 and should be at least doubled. After all, in the 

 fishing area east of long. 170- E alone, we estimate 

 that at least 4.1 million birds were killed in just 

 the 12 yr mentioned above. 



international North Pacific Fisheries Commission. 1955- 

 79. Catch statistics of Japanese mothership gillnet and land- 



based driftnet fisheries. Int. North Pac. Fish. Comm. Doc. 



805 



