FISHERY BULLETIN: VOL. 85, NO. 2 



count for the season (as observed at Ugamak Island 

 in 1985), the amount of error would be insufficient 

 to explain the low counts. In any event, the declines 

 in northern sea lion numbers observed in aerial 

 surveys since the 1970s have been confirmed by the 

 counts made from land at Ugamak Island and by 

 the pup counts taken in the eastern Aleutian Islands 

 and the Gulf of Alaska. 



Pup counts can provide a reliable index of popu- 

 lation change since almost all pups born can be 

 counted in surveys scheduled prior to the pups going 

 to sea. However, the index may be biased (Berkson 

 and DeMaster 1985). Briefly stated, if precount pup 

 survival (e.g., live birth rates) is density-dependent 

 the counts would overestimate the rate of decline. 

 If postcount pup survival (e.g., juvenile survival) is 

 density-dependent then the counts would underesti- 

 mate the rate of decline. Finally, if adult survival 

 is density-dependent there would be little bias. Few 

 data exist on density-dependent population regula- 

 tion in northern sea lions, so we cannot be sure 

 which, if any, of these mechanisms are operative. 

 However, available data suggest that while precount 

 survival in recent years is either unchanged or has 

 improved, postcount survival has decreased. In both 

 cases, the effect would be for pup counts to under- 

 state the rate of decline. 



It is unclear whether northern sea lion hauling 

 behavior has changed sufficiently to affect the 

 counts. The 1977-78 and 1985-86 Ugamak Island 

 data show that seasonal and daily hauling patterns 

 and the timing of critical events (e.g., the median 

 pupping date) have not changed. Similar data are 

 not available for the earlier surveys. Animals may 

 have dispersed to other sites, but they still would 

 have been counted in the regional surveys. There 

 is some evidence from Ugamak Island that females 

 may spend less time ashore now than before (e.g., 

 the high ratio of pups to adult females), which would 

 decrease the number of adult animals counted. The 

 low number of juveniles counted at Ugamak Island 

 in 1985-86 may simply reflect increased juvenile 

 dispersal away from the site due to changing prey 

 resources and earlier weaning. 



The decline in northern sea lion numbers onshore 

 in the region has not been due to the emigration of 

 animals to other regions because significant in- 

 creases have not been noted elsewhere. Numerical 

 decreases have been noted at the western extent of 

 the breeding range in the Kuril and Commander 

 Islands (Perlov 1982; Kuzin et al. 1984; Chelnokov 

 1984). Numbers began to fall in this region circa 

 1972, and by 1981-82 had fallen by 50% or more. 

 Abundance at sites in the western Aleutian Islands 



declined 34-61% from 1979 to 1985 (Kletti=^). Adult 

 northern sea lions at the Pribilof Islands have de- 

 clined in number from approximately 7,000 animals 

 in 1960 to 1,100 in 1981 (Kenyon 1962; Loughlin et 

 al. 1984). Most of these animals were located at 

 Walrus Island which was occupied by 4,000-5,000 

 northern sea lions in 1960, around 1,500 in 1975, 

 and only 600 in 1982 (Kenyon 1962; NMML fn. 11). 

 In 1960, 2,866 pups were counted at Walrus Island, 

 but this number had fallen to 334 by 1982 (Kenyon 

 1962; NMML fn. 11). Northern sea lions are regular- 

 ly seen farther north at St. Matthew and St. Law- 

 rence Islands during the ice-free season, but the 

 total has rarely exceeded 300 animals (Loughlin et 

 al. 1984). Frost et al.^* estimated 2,000 animals were 

 in northern Bristol Bay during the 1970s, with 1,500 

 observed in the summers of 1980-82. The only 

 rookery in Bristol Bay, Sea Lion Rock (Table 1), has 

 decreased in size by 81% since the counts of 1957 

 reported by Mathisen and Lopp (1963). The 1982 

 count of 7,962 animals in southeast Alaska was not 

 substantially different from the 1973 estimate of 

 8,430 (Calkins and Pitcher^^). The number of north- 

 ern sea lions in British Columbia and in the lower 

 United States (i.e., Washington, Oregon, and Calif- 

 ornia) do not appear to be increasing from 5,700 and 

 5,410 animals, respectively (Loughlin et al. 1984; 

 Bigg 1985). 



Proximate Causes of the Decline 



The declines of northern sea lion could either be 

 due to falling reproductive rates or reduced survival 

 of pups, juveniles, and adults (especially females). 

 There does not, however, appear to have been sig- 

 nificant declines in reproductive rates or in pup sur- 

 vival 1-2 mo postpartum. The pregnancy rate of 

 females taken in the Gulf of Alaska during April- 

 May 1985 was 62% (n = 62), which was not signif- 

 icantly different from the 67% (n = 102) found there 

 in 1975-78 (x^ = 0.002, P > 0.50; Pitcher and 

 Calkins 1981; Goodwin and Calkins^^). In addition. 



"E. Klett, U.S. Fish and Wildlife Service, Aleutian Island Unit, 

 Alaska Maritime National Wildlife Refuge, Box 5251 Naval Air 

 Station, Adak, AK 98791, pers. comm. March 1986. 



'"Frost, K. J., L. F. Lowry, and J. J. Burns. 1982. Distribution 

 of marine mammals in the coastal zone of the Bering Sea during 

 summer and autumn. Alaska Dep. Fish Game, Final Rep. RU613, 

 188 p. Alaska Department of Fish and Game, 1400 College Road, 

 Fairbanks, AK 99701. 



'^Calkins, D. G., and K. Pitcher. 1983. 1982 pinniped inves- 

 tigations in southern Alaska. Unpubl. rep., 15 p. Alaska Depart- 

 ment of Fish and Game, 333 Raspberry Road, Anchorage, AK 

 99502. 



'"Goodwin, E. A., and D. G. Calkins. 1985. Preliminary results 

 of ongoing investigations of San Miguel Sea Lion Virus, Lepto- 

 spirosis, and Chlamydiosis in Alaska Steller sea lions and their rela- 



360 



