FISHERY BULLETIN: VOL. 87, NO. 1 



tion would likely be substantial (Huntsman and 

 Manooch 1978a). 



A second potential source of error is the use of 

 a deterministic model to represent recruitment. 

 Changes in population size and CPUE are due not 

 only to the impact of fishing but also to fluctuations 

 in year-class strength. Because the estimates of 

 virgin recruitment were based on changes in CPUE, 

 fluctuations in recruitment that increase (decrease) 

 the decline in CPUE would result in a lower (higher) 

 estimate of virgin recruitment. Based on size-fre- 

 quency data from the MAB tilefish fishery, Turner 

 et al. (1983) suggested that fluctuations in tilefish 

 year-class strength may be substantial. Using size- 

 and age-frequency data collected off Georgia, Har- 

 ris and Grossman (1985) found little evidence for 

 strong fluctuations in year-class strength. Tilefish 

 are somewhat difficult to age, however, so differ- 

 ences in year-class strength could be hidden by age- 

 ing errors. 



A third source of error is the unknown number 

 of fish caught off South Carolina or Georgia, but 

 landed in Florida. The impact on the assessment 

 would depend on the magnitude of the catches and 

 the years in which the catches occurred. If we ar- 

 bitrarily assume that actual annual removals were 

 25% higher than combined South Carolina-Georgia 

 landings, recommended Fs would be unchanged, 

 whereas estimates of virgin recruitment and recom- 

 mended yield would increase by about 25%. Thus, 

 if Florida removals could be accounted for, estimates 

 of stock size would be more accurate, but the in- 

 crease in recommended yield would be offset by the 

 increased catches, and would not result in increased 

 overall landings. 



CONCLUSIONS 



The results of this study provide estimates of the 

 relationship between yield and fishing mortality and 

 of the recommended level for F. Results obtained 

 using commercial CPUE estimates indicate that sus- 

 tainable harvests from the fishery are quite low, and 

 would be obtained at Fs considerably lower than 

 observed in the developing fishery. We obtained 

 higher estimates of population size and sustained 

 yield using research CPUE data; however, we be- 

 lieve that the commercial CPUE data better reflect 

 population trends. We estimate that current stock 

 size is about 200-600 t, compared with a recom- 

 mended level of 400-800 t. If the stock could be 

 rebuilt to the recommended level, it should support 

 an annual harvest of about 50 t. A rebuilding stra- 

 tegy is feasible for tilefish because catches are low 



except when directed fishing occurs (G. Ulrich^). At 

 present, however, there are no restrictions on the 

 tilefish fishery and despite reductions in effort, 

 catches are probably large enough to prevent the 

 stock from rebuilding (G. Ulrich fn. 5). 



Apparent limitations on sustainable yield of the 

 tilefish fishery probably can be attributed to the 

 demographic characteristics of the stock. In a typical 

 fishery for a long-lived, slow-growing species, a few 

 years of high catches are followed by a sharp decline 

 and a subsequent period of low yield (Huntsman and 

 Manooch 1978b; Leaman and Beamish 1984; Fran- 

 cis 1986). Long-lived, sedentary species, such as reef 

 fishes, may be particularly vulnerable to overfish- 

 ing, even though fishing intensity may be low or the 

 method inefficient (Huntsman and Manooch 1978b). 

 Because tilefish are long-lived, slow-growing, and 

 sedentary (due to their dependence on the availabil- 

 ity of shelter), a similar pattern of exploitation can 

 be expected for the tilefish fishery off South Caro- 

 lina and Georgia. 



Leaman and Beamish (1984) recommended that 

 conservative harvest strategies be developed for 

 long-lived species untO the evolutionary implications 

 of longevity are better understood. They suggested 

 that extreme longevity (>50 years) may be an adap- 

 tive response to ensure population persistence under 

 reproductive uncertainty. For example, a long re- 

 productive life might enable a species to inhabit 

 deeper water (200-1,000 m) where few competitors 

 or predators are found, even though recruitment 

 into such areas may be highly variable (Leaman and 

 Beamish 1984). If variability in recruitment has a 

 significant effect on tilefish stocks, a conservative 

 management strategy emphasizing maintenance of 

 a range of age classes may be appropriate. 



ACKNOWLEDGMENTS 



Funding for this project was provided through 

 contract NA8DAA-D-000918 from the Georgia Sea 

 Grant Program, whose support is gratefully ac- 

 knowledged. We thank the crew of the RV Georgia 

 Bulldog for support in obtaining CPUE data, as well 

 as C. Barans, M. J. Harris, R. A. Low, Jr., M. V. 

 Rawson, S. G. Rogers, G. F. Ulrich, and two anony- 

 mous reviewers for comments on an earlier version 

 of the manuscript. We thank R. D. Methot for pro- 

 viding a sex- and age-structured model upon which 

 our model was based. We also thank C. Barans, 

 R. A. Low, Jr., M. V. Rawson, and G. F. Ulrich for 



'G. Ulrich, South Carolina Wildlife and Marine Resources 

 Department, Charleston, SC 29412, pers. commun. July 1987, 



186 



