AHRENHOLZ: RECRUITMENT AND EXPLOITATION OF GULF MENHADEN 



100 



10- 



> 



o 



u 



< 



o 



o 



z 



ua 

 U 



1 

 100 



loH 



1 



100 



10- 



Area 5 

 n= 226 



Area 7 

 n= 624 



Area 8 

 n= 142 



AGE IN YEARS 



Figure 3. — The frequency of recoveries for a standard 100 tags 

 for each juvenile Gulf menhaden release area (ordinate 

 logarithmically scaled) and the adjusted number of recoveries 

 from each release area. Gulf of Mexico. 



an annual fall-spring, offshore-onshore migration 

 (Roithmayr and Waller 1963). 



The lack of precise recapture location informa- 

 tion and the limited range of the fishery relative to 

 the range of the species prevents the formulation 

 of a more detailed and possibly more accurate de- 

 scription of dispersal patterns of Gulf menhaden. 

 For instance, it is possible that fish from nursery 

 areas outside the fishing grounds disperse at a 

 more or less equal rate towards and away from the 

 fishery. However, the relatively strong showing of 

 tags from outside the fishery 2 and 3 yr after re- 

 lease and the relatively few schools of Gulf men- 

 haden that are sighted at the extremes of the 

 range indicate that dispersal is probably stronger 

 towards the center of the range, where the fishery 

 is heaviest, rather than away from it. 



MORTALITY RATE ESTIMATION 

 FROM ADULT TAG RECOVERIES 



To estimate the age specific exploitation rate of 



Gulf menhaden from different geographic nursery 

 areas, I needed an estimate of the instantaneous 

 natural mortality rate (M) and an estimate of the 

 exploitation rate (u) of fully recruited fish. I used 

 the tag recovery data from the adult tagging study 

 reported by Pristas et al. (1976) to estimate these 

 parameters, because these fish (being obtained 

 from commercial purse seine sets) were assumed 

 to be fully recruited when tagged, and the adult 

 tagging study is independent from the juvenile 

 tagging study which will be used as the data base 

 for the area and age specific exploitation esti- 

 mation. Before mortality rates were estimated, 

 however, adjustments were made to the adult 

 tag-recovery data to remove the potential for sys- 

 tematic errors in the results. 



A, B, and C errors (Kicker 1975) are commonly 

 associated with tagging studies attempting to es- 

 timate rates of survival and fishing. Type A errors 

 can occur when tagged fish die as a result of mark- 

 ing or shed their tag shortly after tagging and also 

 during the recovery phase when reporting of re- 

 coveries is incomplete (in this case loss of tags 

 within a reduction plant). Type A errors are 

 characterized by affecting estimates of rate of fish- 

 ing (hence rate of natural mortality) but not rates 

 of total mortality when these parameters are es- 

 timated by rates of recovery over several time 

 intervals. 



Adjustments of numbers recovered for incom- 

 plete recovery of tags from harvested tagged fish is 

 straightforward and is outlined in the Tagging 

 Methodology section. Adjustments of numbers re- 

 leased entailed adjusting for any significant dif- 

 ference in recovery rates associated with different 

 taggers and then determining a realistic estimate 

 of rate of loss due to death and shedding shortly 

 after tagging, which is applicable to all taggers. 

 Since only three taggers were employed in this 

 study, and one only tagged 200 fish in 1971, a re- 

 covery comparison was conducted between the two 

 taggers who tagged the greatest number (99.7%) 

 of the fish. Because the recovery rate for fish 

 tagged by one individual (tagger B) was much low- 

 er than the rate for fish tagged at the same time by 

 the other (tagger A) (Table 2), the number of tags 

 applied during 1969 by this tagger was adjusted 

 downward. The adjustment was made by multi- 

 plying tagger B's number released for each area by 

 the ratio of the recovery rate of tagger B to tagger 

 A for each area. All of the remaining releases 

 (1970 and 1971, with the exception already noted 

 above) were tagged by tagger A. Adjusted recov- 



329 



