30 

 1979 80 



81 82 83 84 



85 86 1979 80 81 82 83 

 YEAR 



84 85 86 



Figure 3.— Relative population number (RPN) for the Gulf of Alaska and the Shumagin, Chirikof, Kodiak, Yakutat, 

 and Southeastern areas, 1979-86. Dashed lines (— -) signify that the annual change was statistically significant. 



to 1982, and 1984 to 1985 (Fig. 3). Differences be- 

 tween other years were not significant. 



The RPN for each of the statistical areas of the 

 Gulf of Alaska generally increased from 1979 to 

 1986 (Fig. 3); the differences between 1979 and 1986 

 were statistically significant for all areas (Table 3), 

 showing that sablefish abundance has increased 

 throughout the Gulf of Alaska. The sharp jump in 

 the RPN for the Gulf of Alaska from 1980 to 1982 

 was caused by significant increases in four areas, 

 Shumagin, Chirikof, Kodiak, and Yakutat. The 

 sharp jump from 1984 to 1985 was caused by signif- 

 icant increases in two areas, Kodiak and Yakutat. 



DISCUSSION 



As noted earlier, sometimes not all strata were 

 sampled at a station (Table 1). This shortcoming was 

 a consequence of the length of the sampling gear 

 and the topography at a station. At stations where 

 the bottom gradient was slight for all or part of the 

 station, the 16 km of longline gear sometimes was 

 not long enough to sample all strata. As a result, 

 sometimes no stations were sampled within certain 

 strata. For these strata, catch per hachi values were 

 estimated using combined data from the current and 

 previous year's data. The resultant values are esti- 



449 



