FISHERY BULLETIN: VOL. 87, NO. 1 



1978 1979 1980 1981 1982 1983 1984 1985 1985 



1978 1979 1980 1981 1982 1983 1984 1985 1986 

 Year 



Figure 3.— Estimated fishing mortality rates (F) from 1978 to 

 1986 at natural mortality rates of 0.10-0.25. At each level of 

 natural mortality, the virgin recruitment level was determined 

 based on the correlation between F and estimated fishing effort, 

 where estimates of effort were based on commercial vessel or 

 research vessel CPUE. Results for the alternative set of selectiv- 

 ity parameters (LSOg^,) were similar and are not shown here. 



to the lack of CPUE estimates for those years. 

 Virgin biomass levels were similar at different levels 

 of M, ranging from 1,130 t atM = 0.15 to 1,570 t 

 at M = 0.10 (Table 4). Estimates of 1987 biomass 

 were somewhat similar and relatively low, ranging 

 from247tatM = 0.15 to 636 t atM = 0.10(Table 

 4). Equilibrium yield curves differed in a predictable 

 way at different levels of M (Fig. 4). Except at low 

 Fs (<0.10), equilibrium yield increased as M in- 

 creased, due to the higher estimates of virgin re- 

 cruitment at higher levels of M. Recommended F 

 was higher at higher levels of Af, ranging from 0.10 

 atM = 0.10 to 0.48 atM = 0.25 (Table 4). Recom- 



mended yield increased from 41-52 t at M = 0.10 

 to 62-82 t at M = 0.25. 



These results demonstrate that an important 

 source of uncertainty in assessing tilefish stock 

 status is the estimate of M. Harris and Grossman 

 (1985) obtained 1982-83 catch curve estimates of 

 Z equal to 0.25 for both female and male tilefish. 

 Because the areas sampled by the RV Georgia Bull- 

 dog were thought to have received little fishing 

 pressure, M could be as high as 0.25. Catch curve 

 estimates of Z would be biased, however, if vulner- 

 ability to fishing increased with size. Turner et al. 

 (1983) reported a decline in size at recruitment in 

 the expanding MAB fishery and suggested that 

 when larger tilefish were present, smaller ones 

 either were less vulnerable to the gear or were 

 avoided by fishermen. Z (and M) could be under- 

 estimated if vulnerability to fishing increased with 

 size. Alternatively, Z (and M) could be overestimated 

 if significant catches of predominantly older fish 

 were made in the areas sampled by the RV Georgia 

 Bulldog. 



184 



