TURNER ET AL .: TILEFISH IN MIDDLE ATLANTIC-SOUTHERN NEW ENGLAND 



FL) based on visual staging of gonads (Idelberger et 

 al. 1981). Both sexes of B . japonicus japonicus andC. 

 microps exhibited similar prematurational growth 

 rates, and the rates diverged in association with the 

 maturation of females (Ross and Huntsman 1982). 

 The divergence of growth rates of the sexes 

 simultaneous with female maturation presumably 

 reflects earlier and higher energetic costs of re- 

 production for females. The relatively slow growth of 

 theL. chomaeleonticeps of unknown sex suggests that 

 they may be among the smaller members of their year 

 class, and their more rapid growth than females at 

 ages 6-8 suggests that at least the older ones may 

 be males. 



Differential growth rates of males and females 

 should lead to skewed sex ratios at size as Wenner 

 (1972) had demonstrated. Idelberger et al. 7 have 

 shown that in 1978-80 the sex ratio of tilefish >55 cm 

 FL was disparate (Table 2). Fewer males than 

 females occurred between 56 and 65 cm FL (38-42% 

 males), males predominated between 71 and 85 cm 

 FL (64-84% males), nearly equalnumbers of each sex 

 occurred in the 86-95 cm FL range (50-58% males), 

 and only males were above 95 cm FL. Dooley (1978) 

 suggested that the skewed sex ratios of tilefish might 

 have been caused by either protogynous sex reversal 

 or differential growth. Idelberger et al. (footnote 7) 

 have examined histological sections of tilefish 

 gonads and reported that adult tilefish do not 

 undergo sex reversal. Additionally we tested sex ratio 

 at age (for ages 4-10 and all ages above 10 combined) 

 and found significant deviations from 1:1 only at age 

 7 (log-likelihood test: G = 5.32, P < 0.05). We con- 

 cluded that differential growth rates probably 

 caused the skewed sex ratios at length. We believe 



'Idelberger, C. F., C. B. Grimes, and K. W. Able, Rutgers Universi- 

 ty, New Brunswick, NJ 08903, unpubl. data, 1982. 



TABLE 2. — Percentage of male tilefish > 50 

 cm FL in 1978-80 (from Idelberger et al. text 

 footnote 7) and results of log-likelihood 

 tests of the hypothesis that sex ratio did not 

 differ from 1:1 (G). 



that more rapid growth of males out of the 56-65 cm 

 FL range causes the initial, smaller percentages of 

 males. The greater numbers of males in the 7 1-85 cm 

 FL range is a result of their younger age and thus 

 shorter exposure to mortality. The males in this size 

 range were 9- 1 2 yr olds, while the females were 12-23 

 yr olds. Continued rapid growth of males through the 

 86-95 cm FL interval (only males age 13-15 yr fall 

 within the range) and the much slower growth of 

 females, in conjuction with their greater longevity, 

 result in an accumulation of old females between 86 

 and 95 cm FL (50-58% males). Ross (1978) also 

 hypothesized that differential growth rates, not pro- 

 togynous sex reversal among adults, caused skewed 

 sex ratios in Caulolotilus microps. Clearly, mortality 

 rates influence sex ratio at length in L. chamaeleon- 

 ticeps and, if mortality is increased by fishing, the pro- 

 portion of females in the larger size intervals will 

 decrease. 



Size Structure and Mortality 



The prominence and progression of strong modes 

 through the length- frequency data from the longline 

 fishery indicate that strong year classes were present 

 and recruitment of tilefish has varied. Evidence of 

 weak year classes may be seen in the relatively low 

 frequencies of 61-65 cm FL fish in 1974 and of 56-60 

 cm FL fish in 1978, especially in the Southern New 

 England area (Fig. 5). Such fluctuations in year-class 

 strength will cause variations in population size and 

 thus create problems in estimating allowable catches. 



The truncation in size structure of longline catches, 

 which Grimes et aL (1980) attributed to the effect of 

 fishing, has continued. The proportion of fish > 70 

 cm FL declined from 71% in 1974 to 16-21% inl980 

 (Fig. 5). This was accompanied by an increase in the 

 longline fleet size from about five vessels in 1974 

 (Grimes et al. 1980) to about 30 vessels in 1980 and 

 increased effort per vessel at least through 1980 8 . 



The difference in size structure of the foreign trawl 

 and the longline catches (Fig. 8) shows that longlines 

 select for larger fish. In addition, the reduction in size 

 at full recruitment in the longline landings since 1974 

 suggests that when larger tilefish are present, smaller 

 ones are either less vulnerable to the gear or they are 

 avoided by the fishermen. If this is true, longline 

 length frequencies show higher relative frequencies 

 of large fish than actually exist in the population, 

 which would lead to an underestimation of 

 mortality rates. 



*P<0.01. 



"Grimes, C. B., K. W. Able, and S. C. Turner, Rutgers University, 

 New Brunswick, NJ 08903, unpubl. data, 1982. 



761 



