Marancik et al.: Spatial and temporal distribution of grouper larvae in the Gulf of Mexico and Straits of Florida 
3 
fin base and principal ray formation (Moser, 1996): the 
preflexion stage occurs when the notochord is straight; 
the flexion stage occurs when the notochord is obviously 
flexed and caudal rays are forming; and the postflexion 
stage occurs when the notochord tip is aligned vertically 
with the caudal base plate (hypural) elements. Owing 
to a single grouper larva collected in neuston samples 
(1 individual in 383 neuston stations), only MOCNESS 
samples were used in analyses. 
Gulf of Mexico Grouper larvae were collected from the 
Southeast Area Monitoring and Assessment Program 
(SEAMAP) resource surveys conducted in the United 
States territorial waters of the Gulf of Mexico by the 
National Marine Fisheries Service (NMFS) Southeast 
Fisheries Science Center. All SEAMAP plankton sam- 
ples included in our analyses were collected from 1982 
through 2005 with either a bongo net consisting of a 
61-cm frame and 335-pm mesh nets towed obliquely from 
2-5 m off the bottom or to a maximum depth of 200 m, 
or with a neuston net with 1x2 m frame and a 950-pm 
mesh net towed at the surface. Samples were collected 
throughout the day and night depending on when the 
ship reached each station. Environmental data consis- 
tently collected over the entire SEAMAP time series 
were surface temperature, surface salinity, and water 
depth and therefore these were the only environmental 
variables considered in analyses. Plankton samples were 
initially fixed in either 5-10% unbuffered formalin (the 
majority of samples) or 95% ethanol. Formalin-fixed 
samples were transferred to 95% ethanol after 48 hours, 
and samples initially fixed in ethanol were transferred 
to fresh 95% ethanol after 24 to 36 hours. All fish larvae 
were removed from samples, identified to the lowest taxo- 
nomic level possible, and measured at the Sea Fisheries 
Institute, Plankton Sorting and Identification Center in 
Szczecin, Poland. Grouper larvae were further identified 
on the basis of morphological characters (Marancik et 
al., 2010). 
Plankton collections were made in all months of the 
year during the 23 years of SEAMAP surveys included 
in our analyses. The greatest effort was conducted in 
May (2419 neuston and 1529 bongo samples) and Sep- 
tember (2167 neuston and 1904 bongo samples); the 
least effort occurred in February (40 neuston and 41 
bongo samples) and March (50 neuston and 178 bongo 
samples; Table 1). The most complete sampling cover- 
age of the continental shelf of the northern Gulf of 
Mexico began in 1986 and continues to the present. 
Unfortunately, the months of November through March, 
likely the peak spawning season for many grouper spe- 
cies (Hood and Schlieder, 1992; Coleman et al., 1996; 
Nemeth et al., 2007), were rarely and inconsistently 
sampled during SEAMAP. Grouper larvae have been 
re-examined and identified from collections through 
2005; therefore only data from SEAMAP surveys from 
1986 through 2005 were statistically analyzed. 
The most temporally and spatially consistent sam- 
pling effort was conducted during two dedicated 
SEAMAP plankton surveys: the spring and fall surveys 
(Fig. IB). Within these two annual surveys, sampling 
coverage was fairly consistent from 1986 through 2005. 
The percentage of stations sampled gulf-wide, roughly 
representing the area covered, ranged from 28.7 % to 
54.0% (mean = 45.1%) in the spring and from 26.1% 
to 76.7% (mean=61.3%) in the fall, and the targeted 
survey area was usually represented over its entire 
north-south and east-west extent (Table 2; Lyczkowski- 
Shultz and Hanisko, 2007; Muhling et al., 2010). The 
most consistent sampling occurred in April-May and 
September-October except for three years during which 
sampling began late (spring 2003, spring 2004, fall 
2005) and one year which finished early (fall 1997). 
The spring and fall surveys targeted different bathy- 
metric zones with overlap at the shelf edge. During the 
spring plankton survey, conducted in April and May 
(1982-present), stations were sampled from the shelf 
edge to the United States Exclusive Economic Zone 
(EEZ) within a 0.5°x0.5° (56-km) grid. The second dedi- 
cated plankton survey, called the “fall plankton sur- 
vey,” was conducted from late-August through October 
(1986-present) from the coast to the continental shelf 
edge (10-200 m water depth) and from south Texas 
to south Florida. Additional specimens and data came 
from plankton sampling conducted by the National 
Marine Fisheries Service (NMFS) Southeast Fisher- 
ies Science Center during SEAMAP summer and fall 
trawl surveys, winter plankton surveys, squid-butterfish 
surveys, Alabama summer and fall plankton surveys, 
and the fall pelagic fish survey in the Gulf of Mexico 
(Table 3; see Lyczkowski-Shultz and Hanisko, 2007, 
for details). 
Analyses 
Seasonal and spatial occurrence The spatial consistency 
and monthly frequency of sampling in the Straits of 
Florida makes these data the best suited for determin- 
ing the seasonality of larval grouper occurrence and, in 
turn, presumed seasonality of spawning. Only specimens 
identified to species were used in analyses. We used 
quotient analysis to define potential and peak season 
of occurrence and cross-transect distribution, using the 
Straits of Florida data. With this analysis, the ratio of 
the proportion of larval occurrence to the proportion of 
observations was determined within environmental (spa- 
tial or temporal) bins in order to discover when or where 
larvae were collected with higher (or lower) frequency 
than would be expected if larvae were evenly distrib- 
uted. Quotient values >1 indicate a relatively higher 
occurrence of larvae (based on the number of observa- 
tions) than expected, whereas values <1 indicate lower 
than expected occurrence (van der Lingen et al., 2001). 
Significance of the quotient values (above or below the 
null of 1) was determined by a bootstrapping technique 
similar to that used in Bernal et al. (2007). Quotient 
analysis is relatively robust for data sets containing 
many zero values, allowing analysis of the complete data 
set and cross-transect relationships despite the rarity of 
grouper larvae in collections. Analyses were conducted 
