Seyoum et al.: Genetically determined population structure of Lachnolaimus maximus in the southeastern United States 
451 
majority of larvae produced south of the Everglades 
region would be transported into the Florida Keys or 
Atlantic Ocean and there would be very little larval 
transport north along the west Florida shelf. Models of 
regional hydrodynamics combined with the increased 
information available regarding hogfish distribution 
and spawning could simulate larval dispersal and set- 
tlement around Florida. Elsewhere, biological data are 
lacking; for example, courtship in hogfish has been ob- 
served off North Carolina (Parker, 2000), but the tim- 
ing and extent of spawning in this region (cluster 3) 
have not been described. 
Adult behavior may also play an important role 
in determining connectivity and population structure 
(Frisk et al., 2014). Hogfish are protogynous hermaph- 
rodites (McBride and Johnson, 2007). Sex change does 
not occur until fish are at least 305 mm in total length 
(TL) (Davis 1976), and most fish remain female until 
they are at least 350 mm TL (McBride and Johnson, 
2007; Collins and McBride, 2011). A single, dominant 
male will spawn daily with multiple females in harems 
of up to 15 individuals (Colin, 1982; McBride and John- 
son, 2007; Munoz et ah, 2010; Collins and McBride, 
2011). For hogfish, stable harems and a strong associa- 
tion with reef habitats promote relatively strong site 
fidelity to specific locations for months or even years 
(Colin, 1982; Lindholm et al., 2006; Munoz et al., 2010). 
However, gradual ontogenetic movement offshore with 
growth is evident. 
Juveniles settle in shallow, inshore habitats (Davis, 
1976; Collins and McBride, 2011), and hogfish <2 years 
old have rarely been collected from habitats deeper than 
30 m (similarly, it is rare to collect individuals >8 years 
old from nearshore habitats). For those reasons, it is 
assumed that hogfish gradually move to deeper water 
farther from shore as they grow. Hogfish may live for 
23 years (McBride and Richardson, 2007), and females 
mature as early as an age of 1-2 years (McBride and 
Johnson, 2007). Therefore, gradual movement of indi- 
viduals across the shelf through time would increase 
the exposure of larvae to different environmental con- 
ditions and current regimes that may, as a result, af- 
fect larval distribution and gene flow. Finally, although 
there is no evidence of natal homing by hogfish, it has 
been documented for other reef fish (Paris et ah, 2013) 
and should be considered a possibility. 
Demographic differences in hogfish between or 
within regions of Florida have been noted (McBride 
and Richardson, 2007; McBride et ah, 2008; Collins 
and McBride, 2011). In the central-eastern Gulf of 
Mexico, adult size, longevity, and fecundity differed 
between deepwater and shallow-water fish across the 
shelf (Collins and McBride, 2011; Collins and McBride, 
2015), but these life history traits were unrelated 
to the genetic structure of hogfish (MERPDC, 2012). 
Similar demographic differences for hogfish have also 
been observed between the Gulf of Mexico and south 
Florida (Florida Keys) (McBride and Richardson, 2007; 
McBride et al., 2008). Although hogfish within these 
regions are now known to be genetically different, it 
seems unlikely that life history differences are solely 
the result of genetic differences. The effects of ontoge- 
netic behaviors, as well as the effects of fishing on size 
at age, maturity, fecundity, and harem stability, likely 
play a significant role in the demographics of hogfish 
throughout the range of this species (McBride et al., 
2008; Munoz et al., 2010). 
The close relationship between clusters 1 and 3 was 
surprising because both are geographically separated, 
and the genetic forces responsible for such a relation- 
ship are not readily apparent. The geographic isolation 
and modest sample size of cluster 3 may have led to 
a sample-specific association between clusters 1 and 
3, one that may not hold up once more samples from 
the Carolinas are examined and compared. Still, some 
specific genetic mechanisms could be operating and 
deserve consideration. Drift may be important within 
cluster 3, which had the smallest N e , and thereby led to 
a coincidental similarity between clusters 1 and 3. It is 
also possible that some environmental correlate along a 
latitudinal cline could be driving selection for particular 
genotypes, and the microsatellites, while neutral mark- 
ers, could be linked in some way to adaptive markers. 
A possible driver for selection is temperature, which is 
high (>25C°) and relatively homogeneous around Flor- 
ida during the summer ( June-September), but, during 
the rest of the year, temperature is high only around 
south Florida (cluster 2) (He and Weisberg, 2003). 
Drift and selection are possible, but the most test- 
able hypothesis is based on gene flow. Suppose that 
hogfish in each cluster are disproportionally connect- 
ed by hydrodynamically driven gene flow with a more 
southerly stock (McBride and Horodysky, 2004); in such 
a case, the most likely scenario is that cluster 2 alone 
is affected by introgression from the south, and the ef- 
fect of such gene flow does not reach farther up along 
either the east or the west coast of Florida. The re- 
sult would be as we report here: a disruption of the 
genetic continuity of hogfish around the Florida penin- 
sula. Such a disruptive pattern has been reported for 
European anchovy ( Engraulis encrasicolus ), explained 
by introgression of an African population with popu- 
lations around the Iberian peninsula (Zarraonaindia 
et al. 2012; Antoniou and Magoulas, 2014). This last 
hypothesis is particularly interesting because it would 
expand the currently understood sources of recruit- 
ment to the Florida Keys, but proving it would require 
examination of new collections from the Bahamas and 
Caribbean Sea. 
Application of the findings in this study to moni- 
toring, assessment, and regulatory actions regarding 
hogfish are relatively straight forward. Landings and 
fishing effort data in Florida are already collected at 
the county level (FWC 1 ), and the Florida Keys are con- 
tained within one county (Monroe). Federal fisheries 
data can also be separated by coast (e.g., NMFS 2 ), and 
it is likely that regional stock assessments will reveal 
further differences in fishing effort between regions. 
Fishing effort within the region of cluster 1 is signif- 
icant (NMFS 2 ), but pressure is less concentrated there 
