FISHERY KILLKTIN: VOL. 8«. NO. 4 



male and female Caulolatilus microps (Ross and 

 Merriner 1983) and female Lopholatilus chamaele- 

 onticeps (Erickson et al. 1985). Like Ross and Mer- 

 riner (1983) and Erickson and Grossman (1986), we 

 also found a few adult males (2 of 50) with ovigerous 

 tissue (previtellogenic residual oocytes) in the testic- 

 ular mass. Histological sections of testes revealed 

 no gross structural features that indicated prior 

 functional female status (e.g., remnants of an 

 ovarian lumen (Sadovy and Shapiro 1987)). Further- 

 more, no ovary was ever observed in transition to 

 a testis (transitional ovotestes), however transition 

 can occur within a matter of weeks. Similar to other 

 branchiostegids (C microps, Ross and Merriner 

 1983; Brayichiostegus wardi and B. seTratus, Dooley 

 1978; L. chamaeleonticeps , Erickson and Grossman 

 1986), sex ratios were skewed in favor of males at 

 large sizes. However, both sexes were present at 

 most sizes and ages (45-95 cm FL and 5-10, 16, and 

 19 years) and only females were present at ages 

 29-36 years, ruling out protogyny. Disparate sex 

 ratios at size are apparently due to differential 

 growth and mortality rates between sexes (Turner 

 et al. 1983; Harris and Grossman 1985). 



Our histological examination of juvenile gonads 

 suggest that some L. chamaeleonticeps may undergo 

 prematurational sex reversal. All juvenile gonads 

 examined (63 fish, 146-400 mm FL) appeared to be 

 female, based upon gonad cell structure (i.e., pres- 

 ence of only oogonia and previtellogenic oocytes). 

 We are very tentative about the determination that 

 all of these small fish (<400 mm FL) were females, 

 because undifferentiated gonia, oogonia, and sper- 

 matogonia are very similar in appearance (Yama- 

 moto 1956; Hoffman 1963; Hyder 1969; Ross 1978). 

 Once a gonad has developed gross structure such 

 as spermatogenic tubules and crypts, or an ovarian 

 lumen, determining sex is straightforward. We 

 found residual oogonia in 2 of 50 histologically ex- 

 amined testes. We found no juveniles with truly 

 intersexual gonads, nor were we able to observe a 

 lumen. Ross and Merriner (1983) suggested that the 

 confamilial C. microps underwent prematurational 

 sex reversal, and that gonochorism in C. microps 

 might be a regression from monandric protogyny. 

 Their conclusions were based upon findings from 

 four juveniles (one specimen with a totally ovarian 

 gonad and the remaining three with gonads that 

 contained substantial amounts of testicular tissue), 

 and adult males (8 of 41 examined) with residual 

 oocytes in the testicular mass. 



That prematurational sex reversal has been ob- 

 served among several families and species of fish 



suggests that either prematurational sex reversal 

 is more common among fishes than suspected, or 

 that early gonad development in fishes involves an 

 all-female or female-appearing stage. For example, 

 the salmonid Salmo gairdneri (Mrsic 1923); the 

 cyprinidsBrac^danio rerio (Takahashi 1977), Bnr- 

 bus tetrazona (Takahashi unpubl. data cited in Taka- 

 hashi 1977), Rodeus ocellatus (Shimizu 1979), 

 Cyprinus carpio (Davis and Takashima 1980), and 

 Carassius auratus (Stromsten 1931; Takahashi and 

 Takano 1971); and the anabantids Mac ropoc/us eon- 

 color and M. opercularis (Schwier 1939) are hatched 

 as all females having ovaries with no testicular char- 

 acteristics. About one-half continue normal devel- 

 opment to mature females and about one-half under- 

 go a transitory intersexual stage before becoming 

 adult males. The female juvenile cyprinids Brachy- 

 danio rerio, Barbus tetrazona, Rhodeus ocellatus, 

 and Carassius auratus all develop an ovarian lumen, 

 as well as oogonia and previtellogenic oocytes. Pre- 

 maturational sex change is reportedly common 

 among the hermaphroditic Sparismatinae (Scaridae) 

 that spend their entire life as males (Robertson and 

 Warner 1978). 



Social Control 



We interpret the data on sexual maturity and sex- 

 ual dimorphism to suggest that some sexually 

 mature males delay participation in spawning for 

 up to 3 years and 10-15 cm, and that this mating 

 system is socially mediated. Histological assess- 

 ments of sexual maturity revealed that both males 

 and females produced mature gametes by about 50 

 cm FL and 5-6 years. However, visual inspection 

 of gonads to determine maturity gave the same 

 result as the histological evidence in females, but 

 indicated that males were not mature until attain- 

 ing 65-70 cm FL and 7-8 years. Ross and Merriner 

 (1983) also reported that some Caulolatilus microps 

 males that were visually assessed as immature were 

 later shown by histological methods to be produc- 

 ing mature gametes, and Erickson and Grossman 

 (1986) found L. chamaeleonticeps off Georgia that 

 showed active, yet incomplete spermatogenesis. 



Sexual dimorphism in tilefish is manifested in the 

 size of the adipose flap which is conspicuously larger 

 in males than females and larger ultimate body size 

 in males; size of the adipose flap may be an indica- 

 tion of male breeding status. The adipose flap of 

 males became larger at 65-70 cm FL (7-8 years) 

 when males were judged sexually mature by visual 

 inspection (i.e., had developed large testes), not 



758 



