FISHERY BULLETIN: VOL. 85, NO. 3 



Table 2. — Newman-Keuls multiple range tests on the slopes and intercepts of the 

 linearized length-fecundity regressions from four areas in the Bering Sea (following 

 the notation of Zar 1974). 



Comparison 



q(0.05,35,p) 



2 vs. 1 



3.6720 



2.88 



Table 3. — Newman-Keuls multiple range tests on the slopes and intercepts of the 

 linearized length-weight regressions from four areas in the Bering Sea (following 

 the notation of Zar 1974). 



SLOPES 



Ranks of slopes 



Ranked slopes 



Area 



1 



2.7927 



Aleutian Basin 



2 



2.8883 

 SE shelf 



3 



3.1635 



SE slope 



4 



3.3161 



NW slope 



Comparison Difference SE 



q(0.05,107,p) 



4 vs. 1 



0.5234 



1.7581 



0.2977 



3.70 



ity showed a significant difference (F = 3.608, 

 0.01 <P < 0.025); however, the multiple range 

 test on intercepts was inconclusive and showed 

 overlaps in the equality of intercepts by area. The 

 curves in Figure 7 all appear similar. A nonlinear 



regression on the combined data from all areas 

 explained a high percentage of the overall varia- 

 tion (99.63%) (Table 1 ). It seems reasonable to use 

 one equation to describe the length-weight rela- 

 tionship for all areas combined. 



492 



