Grover: Feeding habits of pelagic summer flounder larvae 
249 
Materials and methods 
During fall and winter, pelagic larval summer 
flounder were collected in the NW Atlantic 
Ocean from 1985 to 1993 by the National Ma- 
rine Fisheries Service (NMFS) Marine Re- 
sources Monitoring, Assessment, and Prediction 
(MARMAP) surveys (Able et. al., 1990). These 
monthly to bimonthly surveys were conducted 
over the continental shelf from Cape Lookout, 
North Carolina, to Cape Sable, Nova Scotia, by 
the RV Albatross IV and the RV Delaware II. 
The vast majority (96.4%) of oceanic specimens 
were collected during October and November, 
and a few were collected in December, January, 
and March. Surveys used 61-cm diameter bongo 
frames fitted with 0.333- and 0.505-mm mesh 
nets (Sherman, 1980). The nets were towed ob- 
liquely through the water from the surface to 
depths of 22-75 m and back to the surface at a 
speed between 1.5 and 3.5 knots. Samples from 
50 stations were the focus of this study (Fig. 1). 
Collections were chosen for analysis on the ba- 
sis of number of summer flounder larvae that 
were present. All collections with 10 or more sum- 
mer flounder larvae were examined. Additional 
stations with fewer specimens were examined to 
balance time blocks so that each 2-h block included 
at least 10 specimens and at least 2 collections, 
and to include the full size range of larvae. 
Specimens were preserved in 10% formalin at sea 
and remained in formalin for approximately 12 
months. They were then transferred to 70% ethanol. 
Specimen shrinkage clearly occurred as a result of 
fixation. Within 24 hours of examination specimens 
were soaked in glycerin where they remained 
throughout their dissection (see Arthur, 1976; 
Gadomski and Boehlert, 1984). 
After standard length ( SL) (snout to notochord tip 
until full flexion, then to posterior edge of hypurals, 
Gadomski and Boehlert, 1984) of each larva was 
measured and morphological stage was determined, 
the digestive tract was removed. Contents of the en- 
tire digestive tract were evaluated. Gut contents were 
teased out and prey items were identified to the low- 
est possible taxon. 
Diet was analyzed in terms of numerical percent- 
age composition (%N), volumetric percentage com- 
position (%Vol), and percent frequency of occurrence 
(%FO). Prey that comprised <1% of the diet by num- 
ber and by volume were pooled into the “other” prey 
category. Prey volumes were estimated, generally by 
assuming a spheroidical geometry, from prey dimen- 
sions. The three analyses (%N, %Vol, and %FO) were 
combined to yield a more comprehensive assessment 
Figure 1 
Location of the oceanic sites in the NW Atlantic Ocean, 
off the NE United States, where pelagic larval summer 
flounder, P. dentatus, were collected. 
of prey importance, the index of relative importance 
(IRI = (%N + %Vol) x %FO) (Pinkas et al., 1971). 
A comparison of larval sizes with established defi- 
nitions of length at yolk absorption and length at 
flexion (Smith and Fahay, 1970; Martin and Drewry, 
1978) revealed that considerable shrinkage had oc- 
curred prior to the examination of these specimens. 
As a result, data were pooled across stations for 
analysis by morphological stage, rather than by size 
class. Morphological stages were defined as preflexion 
(PF): straight notochord and no indication of caudal- 
fin ray development; flexion (FLX): beginning of cau- 
dal-fin ray development (i.e. ossification), accompa- 
nied or not by an upturn of the notochord tip and 
ossification of ural bones (hypurals, epurals, para- 
hypurals); postflexion, premetamorph (PM): comple- 
tion of caudal-fin ray ossification, upturn of noto- 
chord, and ossification of the ural bones, accompa- 
nied by resorption of the notochord tip, such that it 
no longer extends beyond the edge of the hypural 
bones (Fahay 1 ); and metamorph (M): metamorphic 
1 Fahay, M. 1995. Northeast Fisheries Science Center, Howard 
Marine Sciences Laboratory, Natl. Mar. Fish. Serv., NOAA, High- 
lands, NJ 07732. Personal commun. 
