858 



Fishery Bulletin 97(4), 1999 



4Xp 



Releases 



30 



>. lliillll-Jli-ii.j-.IL-«i-.ill-.— ■!■■-■ 



C 5 II 16 21 27 34 41 58 



5Zj Releases 



Jll I iL.lliI I......J....JJ. !.... 



5 11 16 21 27 36 42 50 59 



Elapsed months since release 



Figure 11 



Summary of adjusted cod tag recoveries for releases in the 

 Browns Bank (4Xp) and Georges Bank (5Zj ) areas by elapsed 

 time from release to recovery, in months. 



vey indices. They concluded that extensive move- 

 ment does occur with a strong seasonal pattern in 

 which almost 100% of the biomass is on the Cana- 

 dian side in the fall-winter period and about 65% 

 in the spring-summer period. Results of the present 

 study support the occurrence of seasonal movement 

 but also indicate that it is not a closed system. There 

 is evidence of immigration and emigration and an 

 apparent net loss from the Georges Bank area to 

 the Browns Bank and division 4X area. 



In division 4X, two substocks (Bay of Fundy and 

 Scotian Shelf) are already assumed (Clark, 1996) 

 with distinct growth characteristics. However, tag- 

 ging results show interchange between the Bay of 

 Fundy, the inshore areas of southern Nova Scotia, 

 and Browns and Georges Banks. The greatest ex- 

 change occurs between the offshore Browns Bank 

 area and the inshore areas and the Bay of Fundy. 

 Our results indicate that the extent of movement 

 by cod in the Gulf of Maine area is substantial and 

 that it crosses the present stock boundaries between 

 4X, 5Y, and 5Z. The interchange appears to be pri- 

 marily between the eastern part of Georges Bank 

 (the Canadian management unit ) and 4X. Although 

 movement between 4X and 5Y was observed, the 

 number of releases and recaptures was small and 

 the tagging location was close to the boundary be- 

 tween these areas. Exchange between 5Z and 5Y 

 also occurs but at a relatively low rate. 



Investigation of the potential impact of cod move- 

 ment between stocks on models used for population 

 estimates will require further work. Interpretation 

 of diagnostics for population models such as ADAPT 

 (Gavaris, 1988) from a stock concept might give 



