POWLES: DESCRIPTIONS OF LARVAL SILVER PERCH. BANDED DRUM, AND STAR DRUM 



tially account for development rate discrepancies. 

 Jannke (see footnote 6) illustrated 2.0 mm and 

 5.0 mm larvae identified as B. chrysoura. The 2.0 

 mm specimen, smaller than my smallest larva, 

 does not resemble Kuntz's larvae at similar sizes 

 nor my earliest specimen in pigmentation. Jann- 

 ke's 5.0 mm specimen is probably correctly iden- 

 tified: it has the fin counts and the characteristic 

 cleithral pigment swath of S. chrysoura but lacks 

 characteristic pigment of the ventral midline. 



Larimus fasciatus 



My material agrees fairly well with the only 

 published description, that of Hildebrand and 

 Cable (1934). The pigmented pectoral fin, em- 

 phasized by Hildebrand and Cable (1934), would 

 appear to confirm identification of all specimens of 

 their series. The drawings are not in good agree- 

 ment with the description in their text; for exam- 

 ple, their drawing of a 4.5 mm specimen shows 

 none of the pigment described. The pigment of the 

 brain which I have found characteristic of L. fas- 

 ciatus larvae was not mentioned by Hildebrand 

 and Cable; perhaps fading due to preservation was 

 responsible. The few body proportions given by 

 Hildebrand and Cable (particularly the position of 

 the anus at >50% SL) are characteristic of L. 

 fasciatus. 



Stellifer lanceolatus 



Hildebrand and Cable's (1934) description of a 

 series identified as S. lanceolatus was based on a 

 mixture of that species and L. fasciatus. Early 

 larvae ( =s3.5 mm SL) had pigmented pectoral fins 

 and developing pectoral rays characteristic of L. 

 fasciatus. Later larvae (^4.5 mm SL) lacked pec- 

 toral fin and brain pigment and showed pectoral 

 ray development only at >5.6 mm. Body depth and 

 preanus length values of the early larvae are 

 closer to those of L. fasciatus than of S. lan- 

 ceolatus. Characters given by Hildebrand and 

 Cable (1934) for separating early L. fasciatus and 

 S. lanceolatus were preopercular spination, mouth 

 shape, maxillary length, and amount of space 

 around the brain. However, my observations indi- 

 cate that none of these characteristics are suitable 

 for separating these species. 



As with L. fasciatus, there are discrepancies 

 between text and illustrations in the description of 

 Hildebrand and Cable (1934), particularly in the 

 early stages. My material agrees fairly well with 



that of Hildebrand and Cable (1934) at ^4.5 mm 

 SL, although I observed a dorsal midline melano- 

 phore dorsal to the termination of the anal fin in 

 specimens s=7.0 mm SL, which was not indicated 

 by those authors. 



Spawning Seasons and Areas 



Bairdiella chrysoura 



Spawning is reported by various authors to 

 occur in late spring and summer on the east coast 

 of the United States and the Gulf of Mexico, and 

 year-round in South Florida. The season appears 

 to begin later and to be shorter at higher latitudes: 

 June to August off New Jersey (Welsh and Breder 

 1923), May to July in Delaware Bay (Thomas 

 1971) and Chesapeake Bay (Hildebrand and 

 Schroeder 1928; Joseph et al. 1964), April to Au- 

 gust at Beaufort, N.C. (Kuntz 1915), April to May 

 off Georgia (Dahlberg 1972), and April to Sep- 

 tember off Louisiana (Sabins 1973). Year-round 

 spawning with peaks in January to February and 

 April to June is reported in South Florida (Jannke 

 see footnote 6). Spawning occurs at least April to 

 July in South Carolina waters, according to data 

 presented in the present study. 



Spawning reportedly occurs primarily in es- 

 tuarine and coastal waters, and this is indicated 

 by my data also. Hildebrand and Cable (1930) 

 reported captures of eggs and early larvae in es- 

 tuaries and to 19-24 km offshore off Beaufort, 

 N.C, but the reliability of their identifications is 

 uncertain. Their descriptions of eggs and early 

 larvae of 5. chrysoura are insufficiently detailed 

 to ensure separation from those of other sciaenid 

 and perciform fishes. Jannke (see footnote 6) and 

 Sabins (1973) judged from the small size of larvae 

 caught in tidal passes that spawning must have 

 occurred nearby in estuarine or coastal waters. 

 Specimens I examined were mostly taken in South 

 Carolina estuaries and tidal passes, with only one 

 specimen coming from continental shelf waters. 



Larimus fasciatus 



The information I have presented and the lim- 

 ited literature reports available indicate a long 

 spawning period, extending at least from May to 

 October, for L. fasciatus. Although larvae were 

 more abundant in MARMAP tows made from Au- 

 gust to September than tows made from April to 

 May, larvae were abundant in NMFS Sandy Hook 



133 



