Fishery Bulletin 110(1) 
low-sample-size bias-correction term (AICc; Burnham 
and Anderson, 2002) evaluates the trade-off between 
the number of covariates in a model and the likelihood 
of the model accurately predicting new data (Akaike, 
1973) — therefore reducing the chances of a model with 
redundant covariates appearing to better explain the 
data. The best model was indicated by the lowest value 
of each evaluation score, and in all cases, both tech- 
niques yielded the same results (data not shown). AIC c 
scores were also used to calculate a relative likelihood 
of each model being the best model (Burnham and An- 
derson, 2002). GAMs were created with the MGCV 
library (vers. 1.6-1) in R software (for Mac, vers. 2.11.0; 
R Development Core Team, 2008). 
Results 
Grouper larvae were generally collected in low num- 
bers during both the fine-scale sampling in the Straits 
of Florida and the broad-scale sampling in the Gulf 
of Mexico. A total of 665 individuals (521 individuals 
identified to species) were collected in 384 stations (both 
MOCNESS frame sizes and all depths combined) from 
the Straits of Florida. A total of 544 individuals were 
collected in 16,950 samples from the Gulf of Mexico (433 
individuals in 7848 bongo samples; 111 individuals in 
9102 neuston net samples). 
Seasonal occurrence 
Grouper larvae were collected during all months of sam- 
pling in the Straits of Florida (Table 4) and in all months 
except March and December in the Gulf of Mexico (Table 
1). Most Straits of Florida larvae, specifically preflex- 
ion larvae whose presence indicate recent spawning, 
occurred during February through May. A second, less 
diverse and less numerous group of larvae was present 
from July through October (Table 4). The high apparent 
abundance of larvae in February was due to a single 
collection of >150 specimens of preflexion Epinephelus 
guttatus (red hind) at one station. The lowest occur- 
rence and species richness (number of species captured) 
of larvae were observed during the months of January, 
July, August, and December. 
Spatial occurrence 
Straits of Florida Larvae were distributed in two dis- 
tinct assemblages across the Straits of Florida: an east- 
ern assemblage and a western assemblage (Fig. 3). Five 
species occurred significantly more frequently within 
the eastern 10 km of the transect (Fig. 3, A-E). For 
most of these species, the pattern was the same for all 
developmental stages. Cephalopholis cruentatus (gray- 
sby), one of the more abundant species, was collected 
across the transect across the straits, but was collected 
most frequently on the eastern side at the preflexion 
stage, whereas flexion and postflexion stage larvae were 
collected across the transect, occurring at no stations 
significantly more or less frequently (Fig. 3, F-G). Four 
species were collected significantly more frequently 
on the western side of the transect (Fig. 3, H-K). The 
remaining six species were not collected at high enough 
frequencies to analyze statistically (<5 specimens) or 
were collected evenly across the transect ( Paranthias 
furcifer , Fig. 3L). 
Gulf of Mexico Distribution patterns of grouper larvae 
of all sizes were categorized into five subregions of occur- 
rence in the northern Gulf of Mexico (Fig. 4). Small 
(1.3-4. 3 mm NL) preflexion larvae without prominent 
dorsal and pelvic spines are indicative of recent spawn- 
ing and were collected in 18 of 23 years of SEAMAP 
surveys. Repeated occurrence of these earliest stage 
larvae gave evidence of three of the subregions as areas 
of spawning activity: the Texas-Louisiana Shelf west of 
the Mississippi River (TX-LA; north of 27.5°N and west 
of 90°W; Fig. 2 subregion b), the north-central Gulf shelf 
off the coasts of Mississippi, Alabama, and northern 
Florida (MS-AL-nFL; north of 27.5°N and between 
90° and 85°W; Fig. 2 subregion c), and the west Florida 
shelf (wFL; north of 23° N and between 85-81°W; Fig. 2 
subregion d). These three subregions accounted for the 
vast majority of grouper larvae collected (n=314) and 
were the only subregions containing early-stage larvae. 
Later-stage larvae were also collected farther offshore 
in two additional subregions (Fig. 4, A-C and F). 
Sampling was conducted off the coast of the Yucat- 
an Peninsula in the southern Gulf of Mexico during 
January and February 1990. Although grouper larvae 
were collected (n = 14), the data were not included in our 
analyses because of the timing (winter) and infrequency 
(a single year) of sampling in the area. 
Graysby were the most abundant group of larvae 
collected in the Gulf of Mexico that could be identified 
to species. Graysby larvae occurred during both spring 
and fall surveys. Most specimens were collected during 
July-October and were distributed primarily along the 
west Florida shelf (n = 35). A few larvae were collected 
on or near the Louisiana shelf and Mississippi-Ala- 
bama-north Florida shelf during the fall survey (zz=2) 
and in deep offshore waters off southwestern Florida 
during the spring survey (n = 5; Fig. 4 A). 
Small Mycteroperca spp. larvae (i.e., specimens with 
dorsal-ventral tail pigment; Table 1) were collected 
during April-June and September and November in all 
three presumptive spawning subregions of the northern 
Gulf identified in this study (Fig. 4D; spring: all three, 
fall: TX-LA, MS-AL-nFL). These specimens were pri- 
marily collected along the shelf break. A similar spatial 
and temporal distribution pattern was observed for sev- 
eral slightly larger larvae identified as either E. itajara 
or Mycteroperca spp. based on the presence of pigment 
on the cleithral symphysis, standard tail pigment, and 
broad-based, long and curved spinelets (Marancik et 
al., 2010). These specimens also were collected dur- 
ing April-June and September-November along the 
shelf break of all three presumed spawning subregions 
(spring: all three, fall: TX-LA and wFL; Fig. 4E, Table 
