424 
Fishery Bulletin 109(4) 
eastern Caribbean Sea and the Florida Keys revealed 
no evidence of either genetic heterogeneity or population 
subdivision. Shulzitski et al. (2009) found similar results 
in their study of mutton snapper from the west coast of 
Puerto Rico, the Florida Keys, and localities in Belize 
and Honduras. Because two of the localities (one in the 
Florida Keys and one along the west coast of Puerto 
Rico) sampled by Shulzitski et al. (2009) were very near 
two of the localities sampled in this study, it appears 
that mutton snapper from the Leeward (northeastern) 
Islands in the Lesser Antilles to the Central American 
coast to the eastern Gulf of Mexico may be homogeneous 
in frequencies of alleles at microsatellite markers. 
Shulzitski et al. (2009) suggested that genetic homo- 
geneity among mutton snapper in the region stemmed 
from long-distance larval dispersal or adult migration 
to spawning aggregations. Estimates of long-term mi- 
gration rates (m) in our study between geographically 
proximal localities, some separated by less than 100 
km, ranged from 0.33% to 0.54%. These estimates of m, 
however, should be viewed as heuristic, in part because 
Migrate tends to underestimate m and because confi- 
dence intervals for m are generally unreliable (Abdo et 
al. 2004), and in part because of potential bias intro- 
duced by the necessity of running subsets of data owing 
to the computational demands of Migrate (Palstra et al., 
2007). On the other hand, even if our estimates of m 
were 20 times higher, there still could be independent 
response of local populations to environmental or other 
(e.g., fishing) perturbations (Hastings, 1993, Hauser 
and Carvalho, 2008). Because the genetic markers 
used here and in Shulzitski et al. (2009) are presumed 
to be selectively neutral, genetic homogeneity in this 
case could be decoupled from genetic factors that affect 
adaptability and sustainability of local populations. 
That is, patterns of variation in genes affecting traits 
influenced by natural selection do not necessarily follow 
the same patterns as selectively neutral genes (or ge- 
netic markers) and geographic differences in adaptively 
useful genes (or alleles) can be maintained even in the 
face of substantial gene flow (Conover et al., 2005). 
Our estimates of average long-term migration also are 
consistent with the argument of Roberts (1997) that 
regional currents in the Caribbean Sea are insufficient 
for larval dispersal across the region. 
The estimates of average long-term effective size 
varied three-fold among the localities sampled and 
the lowest and highest effective size was found in the 
samples from St. Croix (AI e =341) and the Florida Keys 
(Af,=1066), respectively. Briefly, N e is the number of 
breeding individuals in an ideal population that expe- 
rience the same amount of genetic drift and show the 
same dispersion of allele frequencies or inbreeding as 
the population under consideration (Wright, 1931) and 
is of importance as a measure of a population’s response 
to evolutionary and ecological forces (Waples, 2010). 
For the conservation and management of exploited bio- 
logical resources, effective size reflects fixation of del- 
eterious alleles, loss of adaptive genetic variance, and 
the capacity to respond to either natural selection or 
Table 3 
Estimates of average long-term mutation-scaled migra- 
tion ( M ) and rate of migration ( m , proportion of migrant 
individuals/generation), and of average long-term effec- 
tive size ( N e ) for mutton snapper ( Lutjanus analis). Esti- 
mates ofM and m are presented for pair-wise comparison 
of geographically adjacent sample localities; distance (in 
km) between pairs of localities is approximate. Estimates 
of N e and 95% confidence intervals (Cl) are presented for 
each of five sample sites. PR = Puerto Rico. 
Comparison 
M 
m 
Distance 
St. Croix and 
21.46 
0.0054 
60 
St. Thomas 
St. Croix and 
15.16 
0.0038 
90 
PR-east 
St. Thomas 
35.11 
0.0053 
90 
and PR-east 
PR-east and 
13.27 
0.0033 
200 
PR-west 
PR-west and 
15.07 
0.0038 
>1,600 
Florida Keys 
Lower 
Upper 
Site 
K 
95% Cl 
95% Cl 
St. Croix 
341 
314 
372 
St. Thomas 
922 
847 
1007 
PR-east 
828 
766 
896 
PR-west 
646 
607 
689 
Florida Keys 
1066 
987 
1155 
environmental perturbation (Franklin, 1980; Anderson, 
2005). Long-term estimates of N e represent a harmonic 
mean of N e over approximately 4 N e generations (Hare 
et al., 2011), meaning that 1) smaller values (that may 
have occurred either in the past or recently) will have a 
greater weight on average values, and 2) the time over 
which long-term N e in mutton snapper was estimated 
ranged between -1500 and >4000 generations. Because 
of the time period usually involved, estimates of long- 
term N e are not necessarily reliable indicators of con- 
temporary N e but do provide a baseline for evaluating 
management planning (Hare et al, 2011). Differences in 
long-term N e , however, do indicate possible differences 
in long-term demographic dynamics that potentially af- 
fect the number of individuals over time that produce 
surviving offspring (and hence population sustainabil- 
ity). Demographic factors that generate differences in 
effective size are difficult to assess empirically and can 
stem from varying numbers of breeding individuals 
across generations or from variance in reproductive 
success of either or both sexes (Charlesworth, 2009). In 
both cases, a number of factors including food availabil- 
ity, habitat quality, predation, or mortality are likely 
involved (Saillant and Gold, 2006). 
The low effective size observed for the sample of 
mutton snapper taken off the southwest coast of St. 
