328 
Fishery Bulletin 1 12(4) 
Mean sizes at age have also varied among these 
studies — mean sizes at age-1 were 288 mm in total 
length (TL) in South Carolina and 396 mm TL in Loui- 
siana (Wenner et al. 5 ; Fischer and Thompson, 2004). 
Emergent patterns from tagging studies of southern 
flounder also provide evidence that supports the pos- 
sibility of phenotypically distinct stocks. Several tag- 
ging studies initiated in North Carolina waters have 
arrived at the same 2 general conclusions. First, the 
majority of tagged southern flounder were recaptured 
very close to the tag site (although this result could be 
more of a reflection of sampling effort than of actual 
movement and distribution). Second, those fish that 
were captured away (>20 km) from the tag site were 
collected in locations exclusively south of the release 
location (Monaghan 7 ; Craig and Rice 8 ). 
These studies indicate that at least among younger 
age classes (i.e., age-1 and age-2 fish that dominate 
commercial and recreational landings), southern floun- 
der have restricted home ranges and may be isolated 
geographically from other stocks. Such site fidelity to 
certain locations and subsequent environmental condi- 
tions could contribute to phenotypic differences. Ad- 
ditionally, the documented migration south may oc- 
cur over considerable distances (individuals tagged in 
North Carolina have been recaptured in Florida), and 
even small numbers of migrating individuals could 
suffice to genetically homogenize basin populations 
(Palumbi, 2003). 
Genetic differentiation of southern flounder at the 
basin level has been established previously (Anderson 
et ah, 2012), and our objective was to examine varia- 
tion in otolith shape throughout the range of this spe- 
cies — in the South Atlantic, in particular — to identify 
possible phenotypic stocks at the within-basin scale. 
Although other methods of phenotypic stock identifica- 
tion (e.g., testing for spatial variation in growth rates 
or meristics) are also useful, powerful statistical meth- 
ods to evaluate variation in otolith shape have devel- 
oped rapidly in recent years (Stransky, 2013). Further, 
some of the variation in otolith shape is genetically 
determined, and therefore such variation should be 
comparatively less sensitive to short-term changes in 
environmental conditions. Three spatial scales were ex- 
amined for possible population structure — between-ba- 
sins (Gulf of Mexico and South Atlantic), within-basin 
(among sites within the South Atlantic and within the 
7 Monaghan, J. P., Jr. 1996. Migration of paralichthid 
flounders tagged in North Carolina. Study 2. In Life his- 
tory aspects of selected marine recreational fishes in North 
Carolina. Completion Report Grant F-43, Segments 1-5, p. 
2.1-2.44. North Carolina Department of Environment and 
Natural Resources, Division of Marine Fisheries, P.O. Box 
769, Morehead City, NC. 
8 Craig, J. K., and J. A. Rice. 2008. Estuarine residency, 
movements, and exploitation of southern flounder (Paralich- 
thys lethostigma) in North Carolina. Final Report Fishery 
Resource Grant 05-FEG-15, 39 p. [Available from North 
Carolina Sea Grant, NC State Univ., Campus Box 8605, Ra- 
leigh, NC 27695-8605.1 
Gulf of Mexico), and within-state (among sites within 
North Carolina). Any descriptions of geographic stocks 
(or lack of) will be useful both in current management 
of the stocks and for the design of studies to examine 
differences in stock production that could inform future 
regional management of the southern flounder. 
Materials and methods 
Data collection 
Southern flounder were collected in the South Atlantic 
from state waters (<3 nautical miles, or 5.6 km, from 
the coasts) of North Carolina, South Carolina, Geor- 
gia, and Florida (Fig. 1) and were identified by mor- 
phological features described in a U.S. Fish and Wild- 
life species profile (Gilbert 1 ). In the analysis for this 
study, 289 right otoliths from age-1 southern flounders 
were used. Sixty-five samples were collected from the 
Gulf of Mexico, and the remaining 224 came from the 
Atlantic states (Figs. 1 and 2). Because of low sample 
sizes from both Florida (Atlantic waters) and Georgia 
and the close proximity of the sampling locations from 
which fish were obtained in those 2 states (<161 km 
between locations), Florida and Georgia samples were 
pooled to represent the southern extent of the range in 
the U.S. South Atlantic. Additionally, where available, 
samples from the Gulf of Mexico (locations off both 
Florida and Texas) were included to enable between- 
basin comparisons. 
Given the previously documented genetic differences 
(Anderson et al., 2012), we expected to also detect suf- 
ficient contrast in otolith shape of southern flounder 
between the basins that would aid our interpretation 
of variation at finer spatial scales. We explored 3 lev- 
els of spatial resolution for possible stock differentia- 
tion. The between-basin scale was explored to compare 
differences in otolith shape with established genetic 
differences. Within each basin, we examined varia- 
tion among states because state boundaries delineate 
current “stocks” for management of southern flounder. 
We examined variation between 2 states within the 
Gulf of Mexico and 3 states in the U.S. South Atlan- 
tic. Finally, we investigated the possible existence of 
shape differences at a local spatial scale by examining 
fish from 3 distinct areas within North Carolina (Fig. 
2), from which it is unlikely that juveniles would have 
moved. 
Fish were collected in North Carolina during the 
fall of 2009 and 2010 as part of the North Carolina 
Division of Marine Fisheries (NCDMF) fishery-inde- 
pendent gillnet sampling program. Additional samples 
were purchased from licensed seafood dealers and were 
obtained through participation in directed commercial 
trips. Samples from South Carolina, Georgia, and Flor- 
ida were collected during the fall of 2010 and 2011 as 
part of existing fishery-independent sampling programs 
in each state. Additional samples were obtained from 
