672 



Fishery Bulletin 97(3), 1999 



more variability in the ratios from year to year 

 than in the fall survey, but high values predomi- 

 nated except from the 1968-71 period when the 

 ratios were lower Ratios for age 1 were variable 

 with no obvious tendency but most ages 2 and 3 

 ratios were greater than 0.5. The short length 

 of the DFO survey series precluded investiga- 

 tion of this feature for these data. 



The effect of total 5Zj,m abundance on the 

 spring ratios of relative abundance was exam- 

 ined for this period because the greatest 

 amount of dispersion onto the bank occurs 

 during spring. This was evaluated by plotting 

 the NMFS spring survey ratios of relative 

 abundance against the beginning of year popu- 

 lation numbers aggregated into three age 

 groupings: age 1, ages 2 and 3, and ages 4-8. 

 Population numbers were obtained from the 

 1995 eastern Georges Bank haddock assess- 

 ment (Gavaris and Van Eeckhaute, 1995). No 

 relation between abundance and ratios of rela- 

 tive abundance was apparent. The DFO time 

 series was not investigated for this effect be- 

 cause of its short time span. 



The distribution of age 4+ haddock deter- 

 mined by using the catch-per-tow data from 

 the DFO spring survey pooled over 1986 to 1996 and 

 the NMFS spring survey pooled over 1968-71, 1972- 

 84, and 1985-95 are shown in Figure 6. During the 

 1972-84 and 1985-95 period ( 1986-96 for the DFO 

 survey) there were aggregations throughout the Ca- 

 nadian side of 5Zj,m, especially in the area of the 

 northeast peak, whereas the U.S. side showed lower 

 abundance. This was especially evident for the 1985- 

 95 period, but there were a number of occurrences of 

 haddock along the southern flank of the bank in this 

 period also. The spring distribution from 1968 to 1971 



Age 



■) 



-0.25 



0.25-0.50 



(1.50-0.75 



0.75- 1.00 



—I — ■—■ 



JIM 

 i I 'HI 



Figure 7 



Age by age correlation of haddock ratios of relative abundance from 

 the NMF.S spring survey. 



was markedly different from the recent distribution 

 pattern. The northeast peak aggregations were not 

 seen, and haddock seemed to be distributed uniformly 

 throughout 5Zj,m. The distribution pattern during 

 this period was strongly influenced by the exception- 

 ally abundant 1963 year class. 



Instantaneous rates of net migration 



Haddock landings by the U.S. and Canadian fisher- 

 ies have been affected by the spawning closure area 



