FISHERY BULLETIN: VOL. 87, NO. 3. 1989 



ratio at the inshore stations, either in the north- 

 ern and southern areas separately or in the com- 

 bined areas. Since we could not detect persistent 

 gradients by this test (nor by analogous ^tests), 

 we concluded that differences in the inshore/ 

 offshore placement of samples between two 

 groups should not preclude the testing of hy- 

 pothesis Ho2, which concerns ratios. 



In comparisons between periods or between 

 areas (Fig. 3), the most significant difference 

 was the greater biomass of the "small zooplank- 

 ton" in the north in 1966-69 than the northern, 

 1960-65 or the southern, 1966-69 biomasses. 

 These differences were significant {P < 0.05) 

 even when the inshore subsets of data only were 

 considered. Probably as a result of this, the ratio 

 of euphausiid to small zooplankton biomasses 

 was significantly lower in the northern, 1966-69 

 data set than elsewhere. This difference was also 

 significant by <-test. 



The biomass of small zooplankton was signifi- 

 cantly greater in the northern than in the south- 

 ern area in both periods, in fact (as in 1955-59). 

 The null hypothesis that the biomass of eu- 

 phausiids was the same in all sets of data could 

 not be rejected. Nor did the weight per eu- 

 phausiid in the northern area change. 



These results were supported by the compari- 

 son of biomasses at specific northern stations 



which had been sampled several times. The bio- 

 mass of small zooplankton was greater, and the 

 euphausiid/small zooplankton ratio less, in 

 1966-69 than in 1960-65 (0.05 < P < 0.1 by 

 two-tailed test in both cases), while the bio- 

 masses of euphausiids did not differ (P > 0.1). 



The significant increase in median biomass of 

 small zooplankton in 1966-69, relative to 

 1960-65, could have been due to the inclusion of 

 data from 1969, when samples were taken differ- 

 ently (see Methods) if the different method itself 

 resulted in increased catch. However, Smith 

 (1974) reported that the method used in 1969 

 resulted in a smaller biomass (per unit volume 

 filtered) than did the pre-1969 method. In our 

 data, the biomass of small zooplankton was 

 greater, and the euphausiid/small zooplankton 

 ratio less, in 1966 than in 1969 (rank sum tests). 

 Hence, the change in sampling in 1969 could 

 hardly have been responsible, in itself, for the 

 elevated biomass of small zooplankton in 1966-69 

 relative to the earlier years. 



Thus, euphausiid biomass could not be shown 

 to increase coincident with the onset of the whit- 

 ing fishery, either in absolute units or relative to 

 the small zooplankton. None of the null hypoth- 

 eses relating to absence of change of euphausiids 

 in the northern area at the time of the whiting 

 fishery could be rejected, and in fact the ratio of 



PERIOD 



Category 



1960-'65 



1966-'69 



units 



c 



0) 



c 



3 



o 



small zoopl. 

 euphausiids 

 euph. /small 

 wt. per euph 



small zoopl 

 euphausiids 

 euph. /small 

 wt. per euph 



Figure 3. — Overall and (inshore subset) medians for each space/time block in the 1960s, for 

 dry weight biomasses of small zooplankton and euphausiids, their ratios, and the dry 

 weights per euphausiid. Arrows connect medians which differed significantly (P < 0.05 for 

 Ho = no difference); a thick arrow indicates that the comparable medians of the inshore 

 subsets of data also differed significantly. All comparisons were "vertical" or "horizontal"; 

 no "obUque" comparisons were tested. 



638 



