FISHERY BULLETIN: VOL. 86. NO. 4 



tine et al. 1980; Grimes et al. 1986). Several crusta- 

 ceans and fishes are concentrated in and near tilefish 

 excavations, which are formed by the combined 

 activity of tilefish and their associates (Able et al. 

 1982; Grimes et al. 1986). Several associates are im- 

 portant components of tilefish diets, suggesting 

 close trophic linkages within the burrow community 

 (Turner and Freeman MS^; Grimes et al. 1986). 

 Limited mark-recapture data suggests that tilefish 

 are sedentary (Grimes et al. 1983). Tilefish and their 

 associates have definite temporal activity patterns, 

 and male-female pairing may occur (Grimes et al. 

 1986). Habitat preference, contagious distribution 

 within suitable clay substratum, definitive activity 

 patterns, limited movement, and pair bonding pro- 

 vide opportunities for social interaction and organi- 

 zation within populations. 



Commercial landings of the Mid-Atlantic-southern 

 New England tilefish stock (Katz et al. 1983) have 

 varied widely since 1916 when 4,500 metric tons (t) 

 were landed in 10 months (Freeman and Turner 

 1977). In the early 1970s an important longline fish- 

 ery developed, centered in New York-New Jersey, 

 and landings increased. This fishery became one of 

 the most valuable finfisheries in both states during 

 most years since 1978 (Grimes et al. 1980; U.S. 

 Department of Commerce 1980a-c; D. J. Christen- 

 sen pers. commun.^). From 1977 to 1980, annual 

 landings were 2,061, 3,412, 3,840, and 3,575 t, but 

 catches declined to 3,200 and 1,900 t in 1981 and 

 1982 (U.S. Department of Commerce 1980a-c; D. J. 

 Christensen fn. 6). Effects of fishing on the stock 

 have been drastic, reducing stock size by one-half 

 to two-thirds from 1978 to 1982 (Turner 1986). 



The purpose of this paper is to describe the repro- 

 ductive biology of tilefish in the northern stock, and 

 present evidence that participation in spawning by 

 males is socially mediated. Furthermore, we show 

 that the mating system has been modified by com- 

 mercial fishing. 



MATERIALS AND METHODS 



Samples of tilefish from the United States Mid- 

 Atlantic-southern New England area were obtained 

 from domestic longline and New Jersey recreational 

 (headboat^) fisheries. Information gathered for each 



^Turner, S. C, and B. L. Freeman. Food habits of tilefish, 

 Lopholatilus chamaeleonticeps. Unpubl. manuscr. Southeast 

 Fish. Cent., Natl. Mar. Fish. Serv., NOAA, Miami, FL 07732. 



•^D. J. Christensen, Northeast Fisheries Center Sandy Hook 

 Laboratory, National Marine Fisheries Service, NOAA, Highlands, 

 NJ 07732, pers. commun. 1983. 



sample included the statistical nature of the sam- 

 ple (random or nonrandom), capture method, loca- 

 tion (LORAN C), date, and length (FL, cm), sex, 

 weight (whole and eviscerated), and height (mm) of 

 the adipose flap of each animal. When possible, 

 gonads were excised, weighed (0.1 g) and preserved 

 in 10% formalin. We assigned each fish to one of 

 three classifications (male, female, or unknown) and 

 one of six stages of sexual maturity following visual 

 examination (Nikolsky 1963). Some fish used for 

 reproductive studies were aged using thin sections 

 of their saggital otoliths (Turner et al. 1983). 



Routine histological sectioning and staining (hae- 

 motoxylin and eosin) techniques and light micro- 

 scropy (450 X and 1000 x) were used to examine 

 gonad structure of six small fish (<50 cm), and assess 

 the state of sexual maturity. Ovarian development 

 corresponded to that described by Moe (1969) for 

 red grouper, Epinephelus morio, and Yamamoto 

 (1956) for the flounder Liopsetta obscura. Sperma- 

 togenic development was identical to that of Tilapia 

 leucosticta (Hyder 1969) and toadfish, Opsanus tau, 

 (Hoffman 1963). Females were assessed as imma- 

 ture when ovaries contained only previtellogenic and 

 early vitellogenic oocytes (Yamamoto 1956; Moe 

 1969; Waltz et al. 1982; Ross 1978; Erickson and 

 Grossman 1986). Males were considered mature 

 when active spermatogenesis was occurring and 

 spermatozoa were present in spermatogenic tubules 

 (Ross 1978). Limited spermatogenesis occurred in 

 the testes of some immature males (Erickson and 

 Grossman 1986). 



A gonosomatic index (GSI) was calculated for 



GW 

 females according to the formula GSI = ^fr^ x 100, 



where GW = fresh gonad weight (g) and BW = 

 fresh eviscerated body weight (g) (Nikolsky 1963), 

 for describing spawning seasonality. The seasonal 

 progression of mean ovum diameters was also used 

 to establish the reproductive seasonality, and the 

 ovum-diameter frequency distribution for ripe 

 females was used to indicate spawning frequency 

 (isochronal vs. heterochronal) (Hickling and Ruten- 

 berg 1936). Separated ova from each fish were 

 placed in liquid and stirred; a random sample for 



measuring was then obtained by extracting three 

 aliquots with a large syringe. Diameters of 500- 

 1,000 ova from each female were measured (near- 

 est ocular micrometer unit) using a binocular dis- 

 secting scope and a filar micrometer. We assumed 



''Headboats are vessels which charge anglers for fishing on an 

 individual, thus "per head", basis. 



746 



