794 
Fishery Bulletin 95(4), 1 997 
Species identification 
According to the reported ranges of body depth-stan- 
dard length ratios (Horn, 1970; Ditty and Truesdale, 
1983), our larval specimens could have been any one 
of the Atlantic or Gulf coast species of Peprilus. How- 
ever, two points must be considered when interpret- 
ing these data. First, we found a strong allometric 
relation between body depth and standard length for 
individuals smaller than 15 mm SL. Because the col- 
lections of Horn (1970) include fish ranging in size 
from 6 to 222 mm SL, this allometric relation could 
have confounded his conclusions. Ditty and Truesdale 
(1983), however, apparently recognized this problem 
and therefore presented their data by size class. Sec- 
ond, our other analyses (caudal vertebrae and mel- 
anophore counts) supported the contention that P. 
triacanthus was the dominant species of Peprilus in 
our samples. Our sample was predominantly com- 
posed of individuals with 19 caudal vertebrae, which 
was consistent with findings of previous authors for 
P triacanthus (Collette, 1963; Horn, 1970; Ditty and 
Truesdale, 1983). The ventral midline melanophore 
counts were also similar to those used for P triacan- 
thus by Ditty (1981). 
The overall results of the above morphometric, 
meristic, and pigment analyses lead us to conclude 
that P triacanthus is the dominant and probably the 
only species of Peprilus collected in our samples. The 
two most definitive characters for identification were 
