FISHERY BULLETIN: VOL. 87, NO. 2, 1989 



would also be minimal since R/f does not depend on 

 exact distances. There was no way to control for 

 these potential difficulties other than discarding all 

 short-distance returns, and that would have resulted 

 in a loss of 1,173 of the 3,032 most accurate re- 

 captures. It was believed that the initial screening 

 of recaptures, which eliminated 2,607 of 5,639 recap- 

 tures, removed most of the inaccurate return data. 



Another source of error relates to the estimated 

 distribution of fishing effort into the grid system off 

 Mexico. Interview coverage was 100% of all vessels 

 coming into the port of Tampico, Tamaulipas, and 

 likely reached a majority of Mexican vessels fishing 

 off Tamaulipas. Captains were asked to specify fish- 

 ing areas and effort but not how much time was 

 spent in each grid. If more than one grid was fished 

 on a trip, it was assumed that effort was divided 

 equally among all grids fished. This would tend to 

 reduce the estimated effort over more favored fish- 

 ing grounds and increase it elsewhere, inflating R/f 

 values over favored grounds and decreasing R/f 

 values elsewhere. In Tamaulipas waters, effort ap- 

 peared to be most concentrated between 24°20'N 

 and 25°09'N off the Laguna Madre de Tamaulipas. 

 Six of 10 releases were made to the north (25°10'N- 

 25°59'N) and R/f values north of these 6 release sites 

 could have been artificially low. This would not af- 

 fect the already significant northward trend in pink 

 shrimp movement, but would increase the nonsig- 

 nificant trend in northward brown shrimp move- 

 ment toward significance. 



A different problem exists with the estimated dis- 

 tribution of fishing effort off Texas. Whereas the 

 captains of nearly 100% of the Tamaulipas fleet 

 were interviewed, only 3.1% of those of the Texas 

 fleet were interviewed (via logbooks) in enough 

 detail to estimate effort in grids. Regular port agent 

 interviews recorded 55% of total Texas effort. Com- 

 parison of the depth distributions of effort between 

 Texas interviews and Texas logbooks indicated no 

 significant differences. However, comparison of the 

 effort expended within one-degree quadrangles of 

 latitude and longitude determined by each method 

 indicated similar estimations of effort from 26°00'N 

 to 26°59'N (32% of total effort by logbook, 33% by 

 interview) but overestimation of effort from 

 27°00'N to 27°59'N by logbooks (38% vs. 25%). For 

 Texas releases with recaptures north of 26°59'N (9 

 of 12 releases), R/f values north of release sites were 

 probably underestimated. Since neither brown 

 shrimp nor pink shrimp R/f values indicated signif- 

 icant directional movement off Texas, it is unlikely 

 that the underestimation of northward R/f values 

 would affect the comparisons. 



Recaptures standardized by fishing effort are 

 rarely used to analyze movement patterns of aquatic 

 organisms. Bayliff and Rothschild (1974) and Bay- 

 liff (1979) reported movements of yellowfin tuna, 

 Thunnus albacares, in terms of recaptures weighted 

 by fishing effort in the eastern Pacific Ocean (Mex- 

 ico to Ecuador). Their recapture patterns were not 

 tested for directional movement after individual 

 releases, and interpretation of their results could be 

 confounded by long recapture periods (up to one 

 year) and the large ocean surface areas addressed 

 (0°-25°N, 80°-150°W). Gitschlag (1986) employed 

 the R/f index as a means of reducing bias associated 

 with nonuniform fishing effort upon apparent move- 

 ment patterns of pink shrimp in an area approx- 

 imating a quadrangle of one degree latitude and 

 longitude off Florida. The only other study report- 

 ing both recaptures and effort for marked shrimp 

 was conducted by Somers and Kirkwood (1984) on 

 tiger prawn {Penaeus esculentus and P. semisul- 

 catus) movements in Australia, but recaptures per 

 unit effort were not analyzed. In reality, most mark- 

 recapture experiments on penaeid shrimp and other 

 organisms have been more concerned with obtain- 

 ing estimates of fishing and natural mortality rates, 

 growth rates, or stock ranges rather than assess- 

 ing movements per se. We believe that employing 

 R/f values yields more accurate information on 

 shrimp movements than recaptures alone. 



ACKNOWLEDGMENTS 



The following scientists contributed significantly 

 to the success of this project: National Marine Fish- 

 eries Service— N. Baxter, D. Emiliani, G. Gitschlag, 

 G. Matthews, J. Nance, M. Renaud, E. Scott; Insti- 

 tuto Nacional de la Pesca— R. Cruz, E. Rosas, R. 

 Orta, M. Medellin, M. Gomez, J. Balderas. The cap- 

 tains and crews of the vessels Bip-IX, Chapman, and 

 Oregon II are commended for field operations. 

 Special thanks are extended to all port agents for 

 collecting recaptures and effort information and to 

 commercial fishermen for their cooperation. Ian 

 Somers, CSIRO, Cleveland, Queensland, Australia, 

 and two anonymous reviewers provided construc- 

 tive criticism of this manuscript. 



LITERATURE CITED 



Bayliff. W. H. 



1979. Migrations of yellowfin tuna in the eastern Pacific 

 Ocean as determined from tagging experiments initiated 

 during 1968-1974. Bull. Inter-Am. Trop. Tuna Comm. 

 17(6):445-506. 



310 



