Hunt et al : Movement of Gadus morhua in the Gulf of Maine area 



845 



was used to assess exchange between management 

 areas. Observed numbers of returns were used for 

 this purpose because exploitation rates and commer- 

 cial fishery landings, required for weighting, were 

 not as readily available for areas outside of the Gulf 

 of Maine area. All releases and recaptures in divi- 

 sions 4W, 4V, 4T, 4R, 4S, and subarea 3, which were 

 geographically east of the division 4X-4W boundary, 

 were grouped for this part of the analysis; we refer 

 to this combined geographic area as "east of 4X." 



Because recaptures from tagging studies are usu- 

 ally dependent on commercial fisheries, the distri- 

 bution and abundance of recaptures is a function of 

 the distribution of fishing effort as well as of fish 

 movements. In an attempt to account for the effect 

 of fishing effort, we weighted recaptures using the 

 commercial fishery exploitation rates for cod stocks 

 in the area. Landings, effort, and exploitation rates 

 of cod by unit area were obtained from the published 

 literature (NEFSC, 1994; Clark, 1996; Hunt and 

 Buzeta, 1996). No provision was made for potential 

 discarding, misreporting of catches, or allocation to 

 unit area. The extent of these problems in landing 

 statistics are thought to be substantial in some years 

 and areas but have not been quantified (Clark, 1996; 

 Hunt and Buzeta, 1996). A review of both the qual- 

 ity and quantity of effort data indicated that a low 

 proportion of landings, particularly for fixed gear 

 components, were represented. Therefore, direct 

 measures of effort could not be used to make adjust- 

 ments to returns on the basis of distribution of ef- 

 fort. However, trends in landings by unit area and 

 fleet composition showed a relatively stable spatial 

 and seasonal pattern. In some years a considerable 

 proportion of total Canadian landings from division 

 4X was not allocated to a specific unit area. On the 

 basis of preliminary work by Clark ( 1996 ), an algorithm 

 was applied to allocate this unspecified proportion to 

 unit area. The reported annual exploitation rate for 

 each stock was then partitioned according to annual 

 percent of unit area landings and used as a year- and 

 area-specific index of the probability of recapture: 



A„,,, = ICy(C„,,x£^,), 



where A = adjustment factor; 



C = percent of annual reported landings by 



management area; 

 E = reported annual exploitation rate by 



management area; 

 ^ = unit area; and 

 V = year. 



Within a year, tag recaptures from areas with low 

 landings were therefore given higher relative weight- 



ing than those from areas of high landings. Annual 

 differences in exploitation rate were used to weight 

 for between-year effects. The derived adjustment 

 factors were standardized to the unit area and year 

 with the lowest landings and exploitation rate. 



Summaries of tag recaptures within an area by 

 area of release required standardisation to reduce 

 the impact of large numbers of releases from some 

 areas. For example, the large number of releases from 

 the Browns Bank area and subsequent number of 

 recaptures would tend to overshadow recaptures 

 from small-scale releases. Therefore all recaptures 

 for this part of the analysis were standardized to the 

 equivalent number of recaptures from 1000 releases. 



Seasonal effects of cod movements were evaluated 

 for the two Browns Bank and Georges Bank release 

 sites. Recaptures from these sites were aggregated 

 by quarter with recaptures in the first month after 

 release excluded. Results were summarized by unit 

 area within the release division. 



To evaluate temporal effects, data were partitioned 

 into groups representing recaptures from 0-12 

 months, >12 months after release, and total recap- 

 tures. The influence of size at release was also in- 

 vestigated but initial examination showed no sub- 

 stantial difference in either direction or distance 

 among size groups. This may be due to the fact that 

 most fish were greater than the L^q mature reported 

 by Hunt ( 1996) for Georges Bank cod at the time of 

 release and would be expected to diminish the po- 

 tential impact of size on movement. 



Recapture information that included latitude and 

 longitude was used to summarize the tag recoveries 

 by 10-min latitude and longitude squares for the 

 Browns and Georges Bank release sites. Recaptures 

 in each square were weighted by the adjustment fac- 

 tor index associated with the unit area in which the 

 square was located. The Browns Bank and Georges 

 Bank areas support substantial commercial fisher- 

 ies and are thought to be centers of spawning activ- 

 ity. Tag releases were made during the spawning 

 season and therefore recoveries should represent 

 movement of postspawning fish. 



The minimum, maximum, and average time at 

 large after tagging was calculated for releases from 

 each unit area. For recaptures with latitude and lon- 

 gitude location, the straight line distance between 

 release and recapture site was calculated to deter- 

 mine the minimum, maximum, and average distance 

 travelled. 



A small proportion of recaptures included size at 

 the time of recapture and these were used to esti- 

 mate individual specific growth rates. The increase 

 in length between release and recapture was adjusted 

 to an annual value ((increment x 365)/days at large) 



