HIGHTOWER and GROSSMAN: TILEFISH FISHERY 



c 

 o 



0.1 0.2 0.3 0.4 



Fishing mortaiity rote 



Figure 4.— Estimated equilibrium yield at four rates of natural 

 mortality (Af ), two assumptions regarding the stock-recruitment 

 relationship (A = 0.889 - recruitment dependent on spawning 

 stock, A = 1.000 - recruitment constant), and two values for the 

 female and male (F/M) selectivity parameter LSO^,, . 



Evidence that M is less than 0.25 was provided 

 by predictive models used to estimate M from the 

 growth rate {k) and maximum age (Alverson and 

 Carney 1975) or from k, L„, and mean water tem- 

 perature (Pauly 1980). Estimates from the Alverson- 

 Carney model were 0.107 (female) and 0.118 (male), 

 whereas estimates from Pauly's method were 0.175 

 (females) and 0.163 (males) (Harris and Grossman 

 1985). Furthermore, Hoenig (1983) provided a model 

 that predicts total mortality rate (Z) as a function 

 of maximum age. Maximum observed age from May 

 1982 to August 1983 samples was 32 for female and 



33 for male tilefish (Harris and Grossman 1985). 

 Using Hoenig' s model, Z would be 0.13 for each 

 sex; therefore, that would be a maximum estimate 

 of M. 



A potential source of error in this assessment is 

 the assumption that the selectivity pattern is con- 

 stant over time. Recent observations (R. Low, fn. 

 4) indicate that the size at first vulnerability to fish- 

 ing has decreased from about 1 kg (Harris and 

 Grossman 1985) to about 0.45 kg. The decreasing 

 size at recruitment increases the Hkelihood of re- 

 cruitment overfishing because fish are being har- 

 vested well before the size at maturity (about 2-3 

 kg for females). Thus, the current model probably 

 underestimates the impact of fishing. Higher Fs 

 could be sustained if small tilefish were not vulner- 

 able to fishing; unfortunately, it is difficult to regu- 

 late age at entry for hook-and-line gear (Myhre 1974) 

 and discard mortality under a minimum size regula- 



185 



