MERRICK ET AL.: DECLINE OF NORTHERN SEA LION 



on the island for 1969, 1977, 1985, and 1986 (Fiscus 

 fn. 12; Withrow fn. 10). The number of pups declined 

 by 72% between 1969 and 1977, with two rookeries 

 abandoned. A decline of 35% occurred between 1977 

 and 1986, with three more rookeries abandoned. 

 Pupping has probably declined on other parts of the 

 island, because the number of breeding animals 

 decreased at other rookeries (Table 4) where pups 

 were not counted in 1968 and 1977. Between 1985 

 and 1986, the only years with complete island 

 surveys, there was a decline of 18% in pup numbers. 

 Despite the decline in the number of pups born, the 

 ratio of pups to adult females has increased from 

 0.75 in 1968 and 0.73-0.81 in 1977-78 to 0.95-1.06 

 in 1985-86. The Ugamak Island pup mortality dur- 

 ing the first 1-2 mo postpartum in 1985-86 was 

 3.4-4.5%. Median pupping dates were similar in 

 1977-78 (13 June) and 1985 (12 June). 



Finally, the Ugamak Island data allowed an evalu- 

 ation of the times of peak abundance and variability 

 in the counting methods. Counts peaked in 1985-86 

 between the third week of June and the first week 

 of July, with 90% of the maximal breeding season 

 population onshore at midday during this period. 

 Hourly counts between 1000 and 2000 ADT during 

 this period were always within 10% of the day's 

 maximum. These occupancy patterns were the same 

 as those observed by Withrow (1982) at Ugamak 

 Island in 1977-78. 



The Ugamak Island data also were used to pro- 

 vide an estimate of the accuracy of the 10 June 1985 

 aerial survey of the island. There was only a 3% 

 difference between the aerial photo count and the 

 simultaneous ground count. Thus, as noted by 

 Withrow (1982), aerial photo counts accurately 

 reflect ground counts. Despite this accuracy, the 

 aerial photo count of 10 June 1985 was 24% lower 

 than the maximal ground count of 25 June 1985. 

 This bias was likely caused by the aerial survey oc- 

 curring slightly before the period of peak abundance. 



DISCUSSION 



The number of northern sea lions found at sites 

 within the study region, which included at least 

 140,000 animals circa 1958, totaled less than 68,000 

 in 1985-a decline of 52% (-2.7% per year). All in- 

 dicators (regional numbers, breeding animals on 

 rookeries, pup production, and the Ugamak Island 

 data) confirm this decline. The rates of decline are 

 probably underestimated because declines probably 

 did not begin until after 1958. For example, the 

 Ugamak Island population showed no decline be- 

 tween 1957 and 1969, and then declined at least 



10.1% per year between 1969 and 1986. This in- 

 dicates that the eastern Aleutian Island northern 

 sea lion population may have begun to decline in the 

 early 1970s rather than in 1958, the base year for 

 rate calculations. Data are insufficient to calculate 

 regional or area rates from 1969 to later dates. 



Declines may have occurred in two phases. The 

 first phase may have begun in the 1970s and been 

 confined to the eastern Aleutian Islands and west- 

 ern Gulf of Alaska. Numbers for the entire study 

 region fell by 25% (-1.6% per year) between 1958 

 and 1977. Numbers in the eastern Aleutian Islands 

 appeared to stabilize in the mid-1970s, while those 

 in the central Aleutian Islands and western Gulf of 

 Alaska may have increased. A second phase of the 

 decline may have begun during the late 1970s, with 

 all areas being affected and overall numbers falling 

 36% (-5.2% per year) between 1977 and 1985. 

 Results of the 1986 pup survey and 1986 Ugamak 

 Island study indicate that the decline is still con- 

 tinuing. 



Alternative Explanations for 



the Declines in 

 Northern Sea Lion Numbers 



Consideration has been given as to whether the 

 declines could be explained by counting errors or 

 biases, changes in northern sea lion behavior, or 

 emigration. However, errors in counting or, for that 

 matter, changes in technique do not explain the 

 decline in numbers. Because counts taken in the 

 1950s and 1960s were conducted in the spring 

 (before abundance had peaked on land) and because 

 sites were missed, we believe that they underesti- 

 mated sea lion numbers. Braham et al. (1980) esti- 

 mated that only 42% of the eastern Aleutian Island 

 sites were surveyed in 1957 (Mathisen and Lopp 

 1963). Animals present were probably accurately 

 counted in the pre-1970 surveys because observers 

 were all experienced. Kenyon and Rice (1961) com- 

 pared their visual counts to concurrent aerial photos 

 taken and found that their error was between 6 and 

 10%. Even if they had overestimated numbers the 

 error may have been counterbalanced by any under- 

 estimate from counting too early in the year. The 

 counts of the 1970s and 1980s were probably more 

 comparable because they were made during the 

 period of peak numbers onshore. Also there was 

 little variation during this period in methods and 

 personnel were experienced in these survey tech- 

 niques. Furthermore, the methods used were be- 

 lieved to have been the most accurate. Even if all 

 of the aerial counts were lower than the maximal 



359 



