FISHERY BULLETIN: VOL. 84, NO. 3 

 Table 2.— Analyses of covariance for fecundity-length relationships. 



Slopes 



Intercepts 



2888- 



AUGUST.1974 



NOVEMBER, 1974 



22 



23 



24 



25 



LENGTH CMM5 



28 



Figure 3.— Comparisons of northern shrimp fecundity between 

 seasons for the Hawke Channel, based on predicted values from 

 equations in Table 1. 



close to the periphery and loosely attached may be 

 simply "crowded out". 



The evidence of egg loss described in previous 

 studies is sufficient to suggest that combining data 

 from different times of year is not appropriate. The 

 two samples compared in this study produced incon- 

 clusive results in that average fecunity was not con- 

 sistently lower over the complete size range in 

 November compared with the August sample. 



Annual variation in fecundity-length relationships 

 occurred in two of five areas sampled in different 

 years. The rate of increase in number of eggs with 



Table 3. — Paired comparisons for area differences when k (no. 

 of samples) >2. 



P values for H : Mean, = mean y 



Date/sample 



No. 



September 1982 



Hudson Strait 1 



North Labrador Sea 2 



Hopedale Channel 3 



Cartwright Channel 4 



August 1978 



Baffin Island 1 



Hopedale Channel 2 



Cartwright Channel 3 



0.0002 

 0.0001 

 0.0001 



0.5709 

 0.0121 



0.2141 

 0.0695 



0.0002 



0.4525 



increasing size only differed significantly in one case, 

 however. The reasons why fecundity differs between 

 years are not known but could be related to changes 

 in environmental conditions and/or egg disease 

 (Stickney 1981). In support of the latter, it is noted 

 that the proportion of nonviable eggs in the 1982 

 Hopedale Channel sample was higher than in the 

 1978 sample by an order of magnitude (D. G. Par- 

 sons unpubl. data). Fecundity was significantly 

 higher in the 1978 data. 



Teigsmark (1983) found that variation within a 

 population during successive years is as great as the 

 variation between populations in a single year and 

 was unable to make a conclusive statement about 

 fecundity of different populations of P. borealis in 

 the Barents Sea. He speculated that such differences 

 could be related to availability of food and popula- 

 tion density. 



554 



