AHRENHOLZ ET AL.: ATLANTIC MENHADEN POPULATION AND FISHERY 



1983). These authors inferred that menhaden 

 spawn in waters north of Long Island from May to 

 September, in the Middle Atlantic Bight south of 

 Long Island from March through May and again 

 in September and October, but primarily spawn 

 in the South Atlantic Bight from October through 

 March. 



Menhaden are believed to spawn in neritic 

 waters over most of the continental shelf, as well 

 as in bays and sounds in the Long Island waters 

 and northward (Reintjes and Pacheco 1966; Nel- 

 son et al. 1977; Ferraro 1980). Higham and 

 Nicholson (1964) concluded that menhaden do not 

 C'pawn inside Chesapeake Bay. The buoyant eggs 

 normally hatch in about 2 days, and the larvae 

 have absorbed their yolk sac in about 4 days at a 

 length of about 5 mm (Kuntz and Radcliffe 1917). 

 The larvae subsequently enter estuarine systems 

 when they are 10-34 mm long (June and Cham- 

 berlin 1959; Reintjes and Pacheco 1966; Lewis 

 and Mann 1971). The seasonality of larval immi- 

 gration varies among geographical sites within 

 years and also among years at the same site, 

 owing at least, in part to environmental condi- 

 tions. Immigration of larvae has been observed 

 from November through April in the South At- 

 lantic Bight, October to June in the Middle At- 

 lantic and Chesapeake Bay areas, and May to 

 October in waters of Long Island and northward 

 (see Reintjes and Pacheco 1966 for literature 

 summary; Lewis and Mann 1971). 



Following entry into estuarine waters, larvae 

 progress to lower salinity waters and are fre- 

 quently found in high abundance in smaller trib- 

 utary estuaries, where they metamorphose into 

 juveniles at a length of about 34 mm (June and 

 Chamberlin 1959; Wilkens and Lewis 1971; 

 Lewis et al. 1972). Young of the year generally 

 remain in estuaries until the fall when most mi- 

 grate downstream to larger rivers, bays, sounds, 

 and into the ocean. The range in length of juve- 

 niles in the fall has been observed from 40 to 185 

 mm (Kroger et al. 1974). After their estuarine 

 emigration in the late fall and early winter, juve- 

 nile menhaden from New England to the north- 

 ern portion of the South Atlantic Bight migrate 

 southward in dense schools often very close to 

 shore (surf zone) (Kroger et al. 1971; Kroger and 

 Guthrie 1973). 



There are debates whether the Atlantic men- 

 haden population is composed of more than one 

 stock (June 1958, 1965; Sutherland 1963; June 

 and Nicholson 1964; Nicholson 1972, 1978; Ep- 

 perly 1981). However, the apparently similar 



movement patterns make what potentially may 

 be genetically different spawning groups insepa- 

 rable in the fishery. Hence, the menhaden popu- 

 lation is currently treated as one exploited and 

 managerial stock. 



FISHERY DATA BASE 



Records of landings of Atlantic menhaden 

 taken by purse seine and the level of effort ex- 

 pended (vessel weeks) have been collected and 

 maintained back to 1940. Early summaries 

 (through 1971) are available from Nicholson 

 (1971a, 1975) and more recently (through 1984) 

 from Smith et al. (in press) (Fig. 1). 



Fishery landings have been sampled and num- 

 bers at age landed estimated since 1955. The 

 basic sampling methodology used was described 

 by June and Reintjes (1959), with some recent 

 modifications of sample size (Chester 1984). The 

 efficacy and statistical design of the sampling 

 methods are treated in detail by Chester (1984). 

 The aging technique is described by June and 

 Roithmayr (1960). Estimates of numbers at age 

 landed for the 1955-64 seasons are available by 

 fishing area on an annual basis only and are sum- 

 marized by Nicholson (1975). Estimates of num- 

 bers at age landed used in this report for the 1965- 

 81 seasons were available on a subseasonal basis 

 (weekly) but varied slightly from some published 

 annual summaries (Nicholson 1975 [through 

 1971]; Schaaf 1979 [through 1976]; and ASMFC 

 1981 [through 1980]) because of rounding differ- 

 ences, biostatistical error corrections, and use of a 

 different estimation methodology^. Smith et al. 

 (in press) gave landings at age values for 1965-84. 

 Values we used were similar except that the 

 1970-81 values given in Smith et al. (in press) 

 reflect minor corrections to the data file which 

 were made subsequent to these analyses. 



Size at age data were available from the port 

 sampling data. Summaries of length and weight 

 by age and geographic area are in Nicholson 

 (1975) and more recently, Smith et al. (in press). 

 Detailed treatment of seasonal length and age 

 distributions by geographic area are contained in 

 Nicholson (1971b, 1972). 



eSome of the landing estimates (1965-1976) used in the devel- 

 opment of ASMFC 1981 were obtained from an estimation pro- 

 cedure which used standardized, noncalendar weeks. While this 

 procedure had some advantages relative to accuracy, it was not 

 fully adaptable to the NMFS port sampling design. It was 

 deemed more desirable to maintain a continuous set relative to 

 estimation procedure. The differences are not large enough to 

 alter analytical conclusions. 



571 



