Shulzitski et at: Population connectivity among Dry Tortugas, Florida, and Caribbean populations of Lutjanus analis 
507 
biogeographic boundary from populations along the 
coast of Brazil ( Vasconcellos et ah, 2008). Along the 
Pacific coast of the U.S., copper rockfish ( Sebastes cauri- 
nus) exhibit a pattern of isolation-by-distance over 1400 
km, and along the Oregon coast, populations appear 
to be further structured by habitat and oceanographic 
barriers (Johansson et al., 2008). Finally, significant 
genetic structure is present across Barbados, Panama, 
and Belize in the barred hamlet (Hypoplectrus puella) 
and the black hamlet ( Hypoplectrus nigricans) (Puebla 
et al., 2008). 
A second, alternative scenario indicates that mutton 
snapper across the Caribbean region do in fact repre- 
sent a single panmictic population. Genetic homogeneity 
across large distances is not an uncommon pattern in 
marine systems — even with the use high-resolution mi- 
crosatellite markers. Wreckfish (Polyprion americanus) 
lack genetic structure within both the North Atlantic 
and South Pacific basins, although interoceanic differ- 
ences are apparent (Ball et al., 2000). Weak structure 
across a major biogeographic boundary off the coast of 
northern Australia indicates high levels of gene flow 
for the mangrove jack ( Lutjanus argentimaculatus ) 
(Ovenden and Street, 2003). Heist and Gold (2000) 
found genetic homogeneity across 1500 km of the Gulf 
of Mexico for the red snapper (Lutjanus campechanus). 
And bluehead wrasse ( Thalassoma bifasciatum) lack 
genetic structure across the entire Caribbean basin 
(Purcell et al., 2006). 
Larval dispersal has long been recognized as a ho- 
mogenizing force in marine systems. Given a pelagic 
larval duration (PLD) of 27 to 37 days for mutton snap- 
per, it is reasonable to imagine dispersal by ocean cur- 
rents across large geographic distances. The PLD of 
mutton snapper is similar to that of the red snapper (28 
to 30 days), a species that exhibits genetic homogene- 
ity across geographic distances comparable to those in 
the present study (Heist and Gold, 2000). For snapper 
species the four- to five-week period of time spent in 
the water column may facilitate long distance dispersal 
and, thus, genetic homogenization. Although reef fishes 
shown to exhibit significant population structure tend 
to have shorter PLDs (beaugregory [Stegastes leucostic- 
tus\: 19-21 days, slippery dick: 22-30 days, and French 
grunt: 13-20 days), there are numerous exceptions. 
Early life history traits, such as PLD, are often unable 
to satisfactorily explain patterns of genetic structure 
found across species (e.g., Galarza et al., 2009). Al- 
ternatively, the contribution of the larval stage to the 
maintenance of genetic homogeneity may be due to the 
interaction of species-specific larval behaviors with the 
physical environment, rather than a direct function of 
PLD. The complexity of this interaction could explain 
the inconsistency in studies that relate early life history 
traits to broad-scale genetic patterns. 
In addition to the pelagic larval stage, dispersal in 
the adult stage is likely to be a route for gene flow 
among mutton snapper populations. In fact, annual 
migrations of tens to hundreds of kilometers to large 
spawning aggregations may serve to homogenize allele 
frequencies across a given region (e.g., Florida reef 
tract) and hence the adult stage may serve to homoge- 
nize genotypes at a regional scale, with larval dispersal 
maintaining genetic connectivity among regions. 
Management and conservation implications 
In the face of genetic homogeneity across large geo- 
graphic distances, it is crucial to differentiate between 
ecological and evolutionary time scales. Boundaries 
between populations on an ecological scale are not nec- 
essarily congruent with those on an evolutionary scale. 
Our ability to detect ecologically significant population 
structure with genetic techniques continues to improve, 
yet remaining challenges limit the inferences we can 
draw from such a data set. The present study indicates 
that mutton snapper populations across the study area 
may be extensively connected over evolutionary time. 
Yet very little effective migration per generation can 
prevent genetic divergence among regions (Slatkin, 1987) 
even while the number of fish arriving from a distant 
location may be insignificant from the management 
(i.e., ecological) perspective (Cowen et al., 2007). Thus, 
ecologically significant dispersal between populations 
could still be slight, even given the genetic homogeneity 
observed in the present study. One well-studied example 
comes from bluehead wrasse, where genetic homogene- 
ity of populations throughout the Caribbean (Purcell et 
al., 2006) contrasts with results obtained with natural 
elemental signatures recorded in otoliths (Swearer et 
al., 1999). Retention signatures found in 89% of sum- 
mertime recruits to the leeward side of St. Croix indicate 
high levels of local retention despite the maintenance of 
genetic homogeneity on a large scale. Thus, future work 
must continue to try and bridge the distinction between 
evolutionary and ecological time scales. 
The hydrographic environment of the Tortugas South 
Ecological Reserve, including the formation of the Tor- 
tugas Gyre, supports the potential for retention of lar- 
vae spawned on Riley’s Hump (Lee et al., 1994). Al- 
though there is evidence that oceanographic processes 
in this region can lead to the delivery of these larvae 
to reefs and nursery habitats of the Florida Keys and 
southeast Florida shelf (Limouzy-Paris et al., 1997), 
there is also evidence that some of these same process- 
es can actually advect larvae offshore (D’Alessandro 
et al., 2007). Thus, at the present time, the pathways 
of larval dispersal and patterns of connectivity among 
populations, and the relative contributions of larval 
sources to fisheries in the Dry Tortugas, Florida Keys, 
and southeast Florida, remain unclear. Domeier (2004) 
has provided indirect evidence of a recruitment path- 
way originating at the TSER that may deliver larvae 
to the Florida reef tract and to nursery habitats as 
far north as Palm Coast, Florida. Drifter vials were 
released over a mutton snapper spawning aggregation 
site in the TSER for two consecutive years. Based on 
vial returns, the overall range of dispersal was similar 
across years, yet the pattern of concentration of returns 
