16 
Fishery Bulletin 110(1) 
the shelf break throughout much of the northern Gulf of 
Mexico from Texas to southern Florida. Similarly, most 
of the Straits of Florida larvae were collected from sta- 
tions closest to the coasts of Florida and the Bahamas 
(Fig. 3). A similar affinity for shelf edge habitat has 
been observed among adult grouper (Koenig et al., 1996; 
Brule et al., 1999; Sadovy and Eklund, 1999; Brule et 
al., 2003), and most spawning occurs inshore of or along 
the shelf break (Collins et al., 1998; Brule et al., 2003; 
Nemeth et al., 2007). Further, a higher specimen-to- 
sample ratio was observed in the Straits of Florida (665 
individuals in 384 MOCNESS stations) than that from 
the Gulf of Mexico (544 individuals in 16950 bongo and 
neuston stations). Sampling gear (MOCNESS vs. bongo), 
sampling strategy (discrete depth vs. oblique), and loca- 
tion of sampling all contributed to the wide differences 
in the numbers of grouper larvae collected during the 
two sampling programs. The MOCNESS sampled more 
water per tow than the bongo nets, and proportionately 
more of the sampling occurred at depths likely to con- 
tain grouper larvae (< 50 m). In addition, more of the 
Straits of Florida (including the area upstream from 
the sampling area) includes shelf edge habitat than the 
basin-wide sampling area of the Gulf of Mexico. This was 
especially the case during the spring SEAMAP survey 
(season of highest grouper occurrence) when the target 
sampling area is deep offshore water within the Gulf of 
Mexico. Thus, a higher percentage of grouper habitat 
(subsurface waters over shelf edge) was sampled along 
the transect through the narrow Straits of Florida than 
in the broad SEAMAP survey area within the Gulf of 
Mexico and likely accounted for many of the differences 
in catch rates between the two sampling programs. 
Analysis of the larval data supported the conclusion 
that most Gulf of Mexico grouper species depend on 
shelf-edge habitat for spawning. Juveniles of many of 
these species move inshore to coastal and estuarine 
nursery habitats (Eggleston, 1995; Ross and Moser, 
1995; Lindemann et al., 2000; Fitzhugh et al., 2005). 
However in the Straits of Florida, flexion and postflex- 
ion larval graysby were collected farther offshore than 
were preflexion larvae of the species (Fig. 3). At least 
two scenarios could explain this pattern in distribution. 
The offshore flexion and postflexion larvae collected 
in the straits could have been carried by the Florida 
Current into the sampling area from spawning sites 
as far away as the Gulf of Mexico or Caribbean Sea. 
Transport from upstream spawning locations explains 
the high diversity of grouper species collected in the 
area and is corroborated by genetic analysis of Gulf of 
Mexico and southeast United States populations (Zatcoff 
et al., 2004; Cushman et al., 2009). The fate of larvae 
carried away from coastal and estuarine habitat in the 
Loop-Florida-Gulf Stream currents is variable (Hare 
and Walsh, 2007; Richardson et al., 2009). Some indi- 
viduals arrive at suitable habitat along the U.S. east 
coast far from spawning sites (e.g., bluefish [Pomatomus 
saltatrix]\ Hare and Cowen, 1996), but many are carried 
too far north for survival (e.g., gray snapper [Lutjanus 
griseus]', Denit and Sponaugle, 2004) or never reach the 
coast (Hare and Walsh, 2007). Similarly, these later- 
stage larvae may have been advected offshore from 
nearby spawning sites and rely on regularly occurring 
oceanic events (e.g., gyres and meanders: Porch, 1998; 
frontal eddies: Sponaugle et al., 2005) or periodic events 
(e.g., wind storms: Shenker et al., 1993) to move them 
onshore toward nursery habitat. This second scenario 
would provide for some degree of self-recruitment. These 
scenarios may apply to other species of grouper; how- 
ever, most species were collected too infrequently or in 
too narrow a size range to detect differences in distribu- 
tion patterns between early life history stages. Further 
research is needed to determine the most likely pro- 
cesses driving the distribution patterns observed among 
Straits of Florida grouper larvae. The results of such 
an analysis, the identification of recruitment pathways 
and survival rates, would have major implications for 
the management of populations spawning in the area. 
Specimens identified as either E. itajara or Mycte- 
roperca spp. were collected during spring (majority) 
and fall surveys (Fig. 4E). The fall contingent was col- 
lected on the southwest Florida and Louisiana shelves 
and represents evidence of fall-spawning Mycteroperca 
spp. or a previously undocumented spawning location 
for E. itajara. Most species of Mycteroperca are known 
to spawn in the winter and spring months in the Gulf 
of Mexico and Caribbean (Hood and Schlieder, 1992; 
Bullock and Murphy, 1994; Brule et al., 2003; Fitzhugh 
et al., 2005), and there were no large Mycteroperca 
spp. larvae collected in the fall survey to confirm a fall 
spawning population (Fig. 4F). However, the spawn- 
ing seasons of many species of Mycteroperca are un- 
known, and at least one species (M. bonaci) is believed 
to spawn year-round (Brule et al., 2003), and therefore 
fall-spawned Mycteroperca spp. are possible. E. itajara 
are known to spawn in fall (Sadovy and Eklund, 1999). 
Although they are believed to occur throughout the 
coastal Gulf of Mexico (Heemstra and Randall, 1993), 
no E. itajara spawning sites have been recorded in 
the northwestern Gulf of Mexico (Sadovy and Eklund, 
1999). These specimens could indicate an undocument- 
ed spawning site for E. itajara in the northwestern Gulf 
of Mexico, but targeted sampling in the area and mo- 
lecular identification of larvae would be needed to verify 
and locate a new spawning site. Genetic confirmation of 
a northwest Gulf of Mexico population may be possible 
because Brazilian, Belizian, and Florida populations 
of E. itajara are genetically highly separated (Craig et 
al., 2009). 
Influence of environmental factors 
Interannual variability in the occurrence of grouper 
larvae was influenced by hydrography. The variables 
involved and the extent of that involvement varied by 
subregion and season (Figs. 5 and 6). Surface tem- 
perature and salinity were significant factors in the 
fall west Florida shelf model, which together with year 
and water depth, explained over 90% of the deviance 
in the data. Surface salinity was also significant in 
