248 



Fishery Bulletin 89(2), 1991 



nition of only one species in the tropical western Atlan- 

 tic region. 



My examination of approximately 1000 specimens of 

 Symphurus possessing 12 caudal fin rays and a 1-4-3 

 interdigitation pattern, collected in inshore waters 

 from North Carolina, throughout the Gulf of Mexico 

 and Caribbean, to Uruguay, reveals that previous 

 studies failed to recognize the presence of multiple sym- 

 patric species among their material. Neither the 

 hypothesis of multiple populations within a single 

 polytypic species, envisioned especially by Jordan and 

 co-workers, nor Ginsburg's hypothesis of one wide- 

 spread polytypic species comprised of allopatric sub- 

 species, adequately explain the divergent morpho- 

 logical variation observed in the specimens and the 

 sympatric (sometimes syntopic) occurrences of speci- 

 mens with different morphological attributes. Instead, 

 the present study recognizes not one, but rather five, 

 western Atlantic members of the S. plagusia complex, 

 which are somewhat phenetically similar species that 

 differ in morphology and pigmentation. Four of these 

 species have largely sympatric, but not necessarily syn- 

 topic, distributions. The fifth species, S. civitatium 

 Ginsburg 1951, is completely allopatric to the others, 

 occurring along the southeastern and Gulf coasts of the 

 United States and northern Mexico. 



Three species in this assemblage, S. plagusia, S. 

 tessellatus (Quoy and Gaimard 1824), and S. civitatium, 

 were described previously. Two additional species are 

 described herein. Available taxonomic and ecological 

 information is summarized, differential diagnoses are 

 provided for each species, and a key to identification 

 of the five species is included. 



Materials and methods 



Methods for counts and measurements and general 

 terminology follow Munroe and Mahadeva (1989) and 

 Munroe (1990). Meristic data, exclusive of scale counts, 

 were taken from radiographs. ID pattern refers to the 

 pattern of interdigitation of dorsal pterygiophores and 

 neural spines. In species accounts, total ranges for 

 meristic features are presented first, followed by modal 

 counts when data were sufficient. 



Measurements less than 150 mm were taken to the 

 nearest 0.1 mm with dial calipers or ocular micrometer. 

 Measurements over 150 mm were taken to the nearest 

 mm with a steel ruler. Measurements are expressed 

 either as thousandths of standard length (SL) or thou- 

 sandths of head length (HL). 



Morphometric abbreviations 



ABL anal fin length 



BD body depth 



CD chin depth 



CFL caudal fin length 



DBL dorsal fin length 



ED eye diameter 



HL head length 



HW head width 



LHL lower head lobe width 



OPUL width of upper opercular lobe 



OPLL width of lower opercular lobe 



PA pelvic to anal fin length 



PAL preanal length 



PDL predorsal length 



PL pelvic fin length 



POL postorbital length 



SNL snout length 



UHL upper head lobe width 



UJL upper jaw length 



All descriptions of pigmentation are based on fishes 

 fixed in formalin and stored in ethyl or isopropyl 

 alcohol. 



Maturity was estimated by macroscopic examination 

 of stages of developing ova and extent of posterior 

 elongation of the ovaries (ovaries of mature females 

 are sometimes conspicuous through the body wall in 

 transmitted light; in immature females and large 

 females, ovaries are best observed by dissection). Since 

 no obvious differences in male testicular size were ap- 

 parent, estimates of maturity were based entirely on 

 females. Immature females were those with non- 

 elongate or only partially elongate ovaries. Mature 

 females had fully elongate ovaries. Gravid females were 

 those individuals with enlarged ovaries filled with 

 large, macroscopically visible ova. 



When available, depth-of-capture information (con- 

 verted to the nearest meter) was recorded and sum- 

 marized for specimens listed in the "Material exam- 

 ined" sections in each species account. If depth of 

 capture included a range of depths over which the nets 

 were towed, a mean depth for that particular trawl was 

 calculated. 



Synonomies are selective for S. plagusia and S. 

 tessellatus because of the numerous locality citations; 

 synonomies are presumed to be complete for the other 

 species. Because of their common occurrence, <S. pla- 

 gusia and S. tessellatus are listed in numerous studies, 

 beginning with the oldest literature dealing with 

 shallow- water marine fish faunas of the Caribbean and 

 temperate regions of eastern South America. Since 

 little descriptive or ecological information was pro- 

 vided in most original accounts of these tonguefishes, 

 it is often impossible, when reviewing the literature, 

 to determine accurately the species studied. For exam- 

 ple, in studies of tonguefishes occurring in the Carib- 

 bean and along the coasts of Central and northern 



