Content uploaded by Brian Dixon
Author content
All content in this area was uploaded by Brian Dixon
Content may be subject to copyright.
Coral Reefs (2009) 28:277–288
DOI 10.1007/s00338-008-0423-2
123
REPORT
Genetic structure of juvenile cohorts of bicolor damselWsh
(Stegastes partitus) along the Mesoamerican barrier reef: chaos
through time
R. I. Hepburn · P. F. Sale · B. Dixon · Daniel D. Heath
Received: 11 June 2008 / Accepted: 29 August 2008 / Published online: 18 September 2008
© Springer-Verlag 2008
Abstract Dispersal in marine systems is a critical compo-
nent of the ecology, evolution, and conservation of such
systems; however, estimating dispersal is logistically diY-
cult, especially in coral reef Wsh. Juvenile bicolor damsel-
Wsh (Stegastes partitus) were sampled at 13 sites along the
Mesoamerican Barrier Reef System (MBRS), the barrier
reefs on the east coast of Central America extending from
the Yucatan, Mexico to Honduras, to evaluate genetic
structure among recently settled cohorts. Using genotype
data at eight microsatellite loci genetic structure was esti-
mated at large and small spatial scales using exact tests for
allele frequency diVerences and hierarchical analysis of
molecular variance (AMOVA). Isolation-by-distance mod-
els of divergence were assessed at both spatial scales.
Results showed genetic homogeneity of recently settled
S. partitus at large geographic scales with subtle, but sig-
niWcant, genetic structure at smaller geographic scales.
Genetic temporal stability was tested for using archived
juvenile S. partitus collected earlier in the same year (nine
sites), and in the previous year (six sites). The temporal
analyses indicated that allele frequency diVerences among
sites were not generally conserved over time, nor were
pairwise genetic distances correlated through time, indica-
tive of temporal instability. These results indicate that
S. partitus larvae undergo high levels of dispersal along the
MBRS, and that the structure detected at smaller spatial
scales is likely driven by stochastic eVects on dispersal cou-
pled with microgeographic eVects. Temporal variation in
juvenile cohort genetic signature may be a fundamental
characteristic of connectivity patterns in coral reef Wshes,
with various species and populations diVering only in the
magnitude of that instability. Such a scenario provides a
basis for the reconciliation of conXicting views regarding
levels of genetic structuring in S. partitus and possibly
other coral reef Wsh species.
Keywords Dispersal · Coral reef Wsh · Microsatellite ·
Genetic structure · Temporal · Spatial
Introduction
Like many marine organisms, most coral reef Wsh possess
two distinct life history phases: a dispersive pelagic larval
stage and a relatively sedentary demersal adult stage (Leis
1991). There are varied hypotheses concerning the evolu-
tionary and ecological signiWcance of the pelagic larval dis-
persal strategy. Johannes (1978) proposed that larval
dispersal was a mechanism to send untended young to a ref-
uge, ‘safe’ from predators. Others have postulated that
deep-water dispersal is an advantage in patchy environ-
ments, such as coral reefs, as (i) a means to overcome habi-
tats that are unstable over evolutionary time (Barlow 1981)
or (ii) a risk-spreading strategy (Doherty et al. 1985).
Larval dispersal is thought to be the exclusive mechanism
for linking populations across patchy reef habitats (Ehrlich
1975). Regardless of the reasons for the evolution of
Communicated by Dr Mark McCormick
R. I. Hepburn · P. F. Sale · D. D. Heath (&)
Great Lakes Institute for Environmental Research,
University of Windsor, 401 Sunset Avenue, Windsor,
ON, Canada N9B 3P4
e-mail: dheath@uwindsor.ca
R. I. Hepburn · P. F. Sale · D. D. Heath
Department of Biological Sciences, University of Windsor,
401 Sunset Avenue, Windsor, ON, Canada N9B 3P4
B. Dixon
Department of Biology, University of Waterloo,
200 University Avenue West, Waterloo, ON, Canada N2L 3G1
278 Coral Reefs (2009) 28:277–288
123
pelagic larval dispersal, it provides these organisms the
potential for large-scale transport from their natal origins.
Directly measuring larval dispersal in the open ocean is
technically diYcult and thus indirect estimates of coral reef
Wsh dispersal provide the majority of dispersal estimates
(but see Jones et al. 2005). Plankton tows (Leis 1991,
1993), growth increment and microchemical signatures of
the otoliths (Thresher et al. 1989; Sponaugle and Cowen
1994; Swearer et al. 1999), and genetic analyses (Shulman
and Bermingham 1995; HoVman et al. 2005; Jones et al.
2005; Purcell et al. 2006) all have been used. Key Wndings
using plankton sampling are that reef Wsh larvae have dis-
persal ranges of 10–100 km (rarely >500 km) from natal
reefs, while larvae from species with a pelagic spawning
life history are typically found farther from shore than the
larvae of demersally spawning species (Leis 1991, 1993).
Studies utilizing the daily growth increments in otoliths of
tropical Wsh have found that, in general, there is a positive
correlation between larval duration and species distribution
range (Doherty et al. 1995). However, larval duration has
been found to be of limited use as a predictor of dispersal
(Thresher et al. 1989; Sponaugle and Cowen 1994; Shul-
man and Bermingham 1995), primarily due to diVerences in
both the maximum age of larvae and plasticity in the larval
duration (Thresher et al. 1989; Sponaugle and Cowen 1994;
see Bay et al. 2006). Determining the origin of larvae and
inferring their dispersal patterns via analysis of trace ele-
ments found in the daily growth rings of otoliths has been
successful in determining larval retention to natal or proximate
reefs (Robertson 1992; Swearer et al. 1999) and for identifying
among-natal reef patterns (Patterson and Kingsford 2005).
Although natural microchemical markers found in otoliths
have some technical limitations, artiWcial microchemical
marking has proven very eVective in directly estimating
dispersal (Jones et al. 2005; Almany et al. 2007).
Published estimates of connectivity and dispersal in
coral reef Wsh vary widely and likely reXect biotic and abi-
otic variation in the factors determining dispersal (e.g., life
history, larval development, ocean currents, meteorological
events, etc.). Ecological larval recruitment surveys have
found that many species of coral reef Wsh in the PaciWc
have high temporal variability in recruitment (Sale et al.
2005). While several Caribbean species display consistent
spatial and temporal recruitment patterns, these vary annu-
ally in intensity (Robertson et al. 1988; Robertson 1992)
indicating temporal variation in coral reef Wsh dispersal.
Genetic methods are being used to estimate temporal varia-
tion in population structure, and, hence connectivity,
among reefs (Lenfant and Planes 2002; Jones and Barber
2005; Purcell et al. 2006).
Genetic markers, such as allozymes, mitochondrial DNA
sequence, and microsatellite DNA, can provide insight into
coral reef Wsh connectivity through examination of the
genetic signals of settling larval cohorts over time. It has
been found that larval cohorts have a clear genetic signal,
readily discernable from existing genetic signatures (juve-
nile and adult) on the reef (Planes et al. 2002; Jones and
Barber 2005). However, in terms of connectivity (i.e., gene
Xow), the interpretation of this genetic signal will depend
on how it is created. Stochastic dispersal or sweepstakes
eVects will result in genetic patterns that change unpredict-
ably over time, while stable dispersal (or no dispersal, i.e.,
self-recruitment) will tend to result in consistent temporal
patterns. Thus, determining the genetic patterns of reef Wsh
juvenile recruits over space and time can provide important
insight into the nature and magnitude of the dispersal pro-
cess, and, hence, the connectivity among populations.
The bicolor damselWsh (Stegastes partitus: Pomacentri-
dae) is a species abundant throughout most of the tropical
western Atlantic (DeLoach 1999). Spawning and settlement
follow a unimodal lunar cycle from May to October (Rob-
ertson et al. 1988; DeLoach 1999). During a week-long
reproductive cycle, female S. partitus demersally spawn
every 2 days, laying up to 5,000 eggs into nests defended
by males (Knapp 1993). S. partitus has a planktonic dura-
tion of 27–31 days, and the age of settlers ranges between
31 and 35 days (Robertson et al. 1988). As a demersal
spawner with a relatively lengthy pelagic larval duration,
the bicolor damselWsh would be expected to have high dis-
persal capabilities. As high dispersal rates are expected to
homogenize gene pools, and, therefore, eliminate local
genetic diVerences, it would follow that only very low lev-
els of genetic diVerentiation should be detectable among
settling cohorts of S. partitus.
This paper describes an analysis of temporal and spatial
genetic variation in recently settled bicolor damselWsh
across the Mesoamerican Barrier Reef System (MBRS).
Eight polymorphic microsatellite markers were used to
characterize genetic relationships at two spatial scales
(within and among atolls), and at two temporal scales,
annual (2002 vs. 2003), and seasonal (early summer versus
late summer, 2003) to test the hypothesis that newly
recruited juveniles exhibit no genetic structure. The com-
bined analysis of spatial and temporal genetic structure in
newly recruited juvenile S. partitus cohorts provides an
opportunity to estimate the relative contribution of predict-
able, temporally stable dispersal versus stochastic dispersal
processes along the MBRS.
Materials and methods
Sampling
The MBRS extends from Isla Contoy on the northern
Yucatan Peninsula to the Bay Islands of Honduras (Fig. 1).
Coral Reefs (2009) 28:277–288 279
123
Sampling took place from 12 to 16 June 2002 oV TurneVe
Atoll, Belize (three sites), and oV Banco Chinchorro, Mex-
ico (three sites; Fig. 1). In 2003, early-season sampling
took place from 16 to 19 June oV TurneVe Atoll (four sites),
Banco Chinchorro (two sites), and Roatan Island, Honduras
(three sites; Fig. 1). Also in 2003, late-season sampling
took place from 13 to 24 August oV TurneVe Atoll (six
sites), Banco Chinchorro (four sites), and Roatan Island
(three sites) (Fig. 1). The late-season sampling in 2003 was
the most intensive and was designed to test for spatial
genetic structure in recently settled juveniles at the within-
atoll (small scale) and among-atoll (large scale) geographi-
cal scales. The early 2003 and the 2002 samples were origi-
nally collected for other purposes and are used here to test
for temporal genetic variation. All collections were carried
out targeting newly settled S. partitus <25 mm total length
(estimated to be <2 weeks post settlement): juvenile Wsh
were collected by divers using hand nets, after anesthetiz-
ing and immobilizing the individuals with clove oil. A Wn
clip, or the whole Wsh, was preserved in 95% ethanol for
later DNA extraction.
DNA extraction and microsatellite analysis
DNA was extracted from caudal Wn clips using the Wiz-
ard® Genomic PuriWcation Kit (Promega, Madison, WI)
following the manufacturer’s protocol for DNA extraction
from animal tissue. Microsatellite markers developed for
S. partitus (Williams et al. 2003) were screened for reli-
able ampliWcation and suitable size ranges. Eight markers
were chosen, but six primer sequences were modiWed to
adjust amplicon molecular size to facilitate running com-
binations of loci on an automated sequencer (Table 1).
Polymerase chain reactions (PCR) were carried out in
reactions comprising of: 2.5 l 10£ PCR BuVer (100 mM
Tris–HCl (pH-8.4) 500 mM KCl), 2.5 mM MgCl2,
200 M each dNTP, 0.05 g of each primer, 0.5 units
DNA Taq polymerase, and 50–100 ng of genomic tem-
plate DNA with ddH2O added to bring the Wnal reaction
volume to 25 l. PCR was performed on an MJ Research
Tetrad DNA Engine model PTC-0225 (MJ Research,
Waltham, MA) with the following reaction proWle: 2 min
initial denaturation (94°C); 35 cycles of 1 min denatur-
ation (94°C), 1 min annealing (variable temperatures, see
Table 1), 1 min extension (72°C); 3 min concluding
extension cycle (72°C). AmpliWcations were analyzed for
fragment size using a CEQ 8000 automated DNA
sequencer with appropriate size standards (Beckman-
Coulter, Fullerton, CA). Approximately 5% of all PCR
reactions were replicated to verify repeatability.
Genetic analyses
An exact test for goodness-of-Wt to Hardy–Weinberg equi-
librium was performed using the conventional Monte Carlo
method for each locus within each site (10 batches, 2,000
permutations per batch; Raymond and Rousset 1995) using
TOOLS FOR POPULATION GENETIC ANALYSES (MP Miller,
TFPGA 1.3: A Windows program for the analysis of allo-
zyme and molecular population genetic data). To account
for multiple, simultaneous tests, Hardy–Weinberg results
were adjusted for signiWcance using the sequential Bonfer-
roni correction procedure (Rice 1989). All further analyses
were performed using ARLEQUIN version 2.0 (Schneider
et al. 2000), unless otherwise stated.
Spatial structure (Late 2003)
Genetic structure was evaluated at two spatial scales: (1)
large scale (among atolls) and (2) small scale (within
atolls). Since departure from Hardy–Weinberg equilibrium
at some marker loci can bias FST estimates, pairwise FST
was calculated twice, once using all loci and once after
excluding two loci that exhibited marked departure from
equilibrium. An hierarchical analysis of molecular variance
(AMOVA) was carried out to partition observed variance
into among-atolls, among-sites within atolls, and among-
individuals components, as described in ExcoYer et al.
(1992).
Fig. 1 A map of the Mesoamerican Barrier Reef System (MBRS),
depicting the sampling locations on the three atolls: Banco Chinchorro,
Mexico; TurneVe Atoll, Belize; and Roatan Island, Honduras. Inset
shows study area relative to Central America (shaded box)
Mexico
Pacific Ocean
Atlantic
Ocean
Banco
Chinchorro
Belize
Guatemala
Honduras
100 km
Turneffe
Atoll
Roatan Island
89o10’ W 88o20’ W 87o30’ W 86o40’ W 85o50’ W
15o00’ N
15o50’ N
16o40’ N
17o30’ N
18o20’ N
19o10’ N
280 Coral Reefs (2009) 28:277–288
123
Pairwise exact tests for diVerences in allele frequency
distributions among atolls (sites combined) and among all
sites were performed (1,000 dememorization steps, 10
batches, 2,000 permutations per batch; Raymond and Rous-
set 1995) using TFPGA. Populations Version 1.2.14 (O.
Langella, http://bioinformatics.org/project/?group_id=84)
was used to estimate pairwise Cavalli-Sforza and Edwards’
(1967) chord distance (DC) between all sites. These values
were used to evaluate the isolation-by-distance model of
genetic divergence among the S. partitus populations
within the study area. DC was used in the analysis of isola-
tion-by-distance since it is less aVected by departures from
Hardy–Weinberg equilibrium, and since many pairwise FST
values were negative. SpeciWc tests of the correlation
between geographic distance (measured as the shortest
water route) and DC were performed at large and small geo-
graphic scales (i.e., all data and only within atoll compari-
sons). At the small scale, not all pairwise comparisons were
included (between-atoll comparisons excluded) therefore
Mantel tests were not performed, but rather linear regres-
sion and the Spearman’s Rank correlation were used to test
for a relationship between DC and geographic distance.
Mantel tests were used to evaluate correlations between
distance matrices wherever complete matrices were avail-
able.
Temporal genetic analysis
To test a single population’s genetic stability over time, an
exact test for allele frequency diVerentiation was performed
between temporal samples for each population (as above)
using TFPGA. Cavalli-Sforza and Edwards’ (1967) DC
were calculated for each between-population comparison
within each time period (see above). The resulting among-
population DC values were compared in a pairwise manner
at each temporal scale. The pairwise DC comparisons were
plotted against one another to determine if they were pre-
dictive (for example, do early summer DC values correlate
to late summer DC values between the same populations?).
Mantel tests were used to test for signiWcance of the
expected relationship, and linear regression was used to
estimate the variance explained (r2).
Results
Genetic variation
All loci were judged to show acceptable levels of repeat-
ability based on the replicate genotyping (>95% agree-
ment). The eight microsatellite loci showed considerable
allelic variation (3–28 alleles per population; Table 2). Of
the 224 tests performed, 40 showed signiWcant deviation
from Hardy–Weinberg equilibrium after sequential Bonfer-
roni correction, all due to heterozygote deWciencies
(Table 2). Two loci, SpAAT9-2 and SpAAT39, were
responsible for 36 of the 40 signiWcant departures from
Hardy–Weinberg equilibrium (Table 2). Departures from
Hardy–Weinberg equilibrium can lead to biases in some
genetic analyses, and those two loci were removed from
analyses that are subject to such biases (noted above).
Spatial structure (Late 2003)
Pairwise exact tests for diVerences in allele frequency dis-
tribution among atolls resulted in no signiWcant diVerences.
Table 1 Primer sequences,
annealing temperatures, and
base pair size ranges of ampli-
Wed fragments for molecular
markers used in microsatellite
analyses of juvenile bicolor
damselWsh from the MBRS
Locus name Primer sequence 5⬘–3⬘Ta (°C) Size range (bp)
SpAAT39 TGCCAAGTTAAACGTAGACAC 59 140–220
CTCCCTTCAGTGTATTTCAGAA
SpGATA40 TTGCCTGCTGATAATTAACG 60 140–280
ATGCCTCACAATGATGTATATTT
SpAAT9-2aAGCCTCAAGGAACTTGTTGG 60 210–290
GATCTTGTATGACTCTCAATGCTAAT
SpAAT40-1aTGTTTCACCTGACATCCAAGA 57 250–310
AGCCTCCCACTGAACACACT
SpAAC44-1aTGCTGTAAACCACCAGGAGA 60 100–165
GCAAACAGAAGGAGCAGTGG
SpAAC33-1aTCACACCTGCTGAGTTCCTG 59 100–173
CATGTACCTCCAATACAGGAAAAA
SpAAC42-1aTGTTGAAGGGCAGGAAGC 54 100–160
TCTCAACAAAATGTCCCATCAG
SpAAC41-1aAGTCTGTGGTTTTGCCAACAT 60 310–410
TGGTGCAGTTATTGCTTAGA
All sequences are from Williams
et al. (2003); aprimers were
modiWed from the published
primer sequence to facilitate
combined locus sequence
analyses
Coral Reefs (2009) 28:277–288 281
123
Table 2 Sample sizes (N), number of alleles (A), and observed and
expected heterozygosity (HO and HE) at eight microsatellite loci for
j
uvenile Stagastes partitus collected at 13 sites in the MBRS (see
Fig. 1) within three geographic locations: TurneVe Atoll, Belize (six
sites), Banco Chinchorro, Mexico (four sites), and Roatan Island,
Honduras (three sites)
Locus Sample date TurneVe Atoll (TA) Banco Chinchorro (BC) Roatan (RI)
1234567 8 9 10 111213
SpAAC42-1 Early 2002 N20 25 23 –––– 15 25 26 –––
A15 19 16 –––– 13 21 19 –––
HO0.85 0.88 0.83 –––– 1.00 0.88 0.92 –––
HE0.91 0.91 0.91 –––– 0.91 0.94 0.93 –––
Early 2003 N44 13 26 –22 –22 –25 –23 16 25
A19 15 22 –17 –16 –17 –15 17 19
HO0.86 0.85 0.85 –0.91 –0.86 –0.84 –0.91 0.81 0.80
HE0.92 0.90 0.93 –0.92 –0.92 –0.91 –0.89 0.93 0.91
Late 2003 N39 25 25 25 20 20 29 17 23 25 34 24 26
A24 23 20 16 15 17 19 12 19 19 21 18 16
HO0.87 0.92 0.92 0.88 0.80 0.95 0.90 1.00 0.91 0.76 0.88 0.88 0.92
HE0.94 0.94 0.92 0.92 0.91 0.92 0.92 0.90 0.93 0.91 0.92 0.91 0.89
SpAAT40-1 Early 2002 N20 24 22 –––– 14 24 23 –––
A12 13 13 –––– 10 13 12 –––
HO0.95 0.83 0.82 –––– 0.86 0.92 0.91 –––
HE0.90 0.87 0.90 –––– 0.88 0.86 0.87 –––
Early 2003 N37 12 25 –22 –22 –24 –21 17 23
A12 10 13 –14 –12 –14 –11 11 12
HO0.78 0.83 0.92 –0.86 –0.77 –0.83 –0.76 0.88 0.78
HE0.88 0.88 0.82 –0.90 –0.84 –0.87 –0.78 0.85 0.90
Late 2003 N39 25 25 24 20 19 27 17 22 24 32 23 24
A14 15 13 14 10 13 14 11 13 13 13 10 14
HO0.82 0.84 0.88 0.96 0.75 0.79 0.93 0.82 0.77 0.88 0.78 0.83 0.88
HE0.85 0.83 0.90 0.88 0.84 0.88 0.89 0.88 0.85 0.89 0.86 0.87 0.87
SpAAC44-1 Early 2002 N20 25 23 ––– 15 25 26 –––
A4711––– 498–––
HO0.10 0.24 0.65 ––– 0.27 0.40 0.35 –––
HE0.19 0.22 0.63 ––– 0.24 0.41 0.40 –––
Early 2003 N44 13 27 –22 –26 –25 –23 17 25
A939–10 –8–6–768
HO0.41 0.31 0.44 –0.45 –0.42 –0.36 –0.43 0.53 0.56
HE0.36 0.27 0.44 –0.46 –0.37 –0.32 –0.41 0.44 0.49
Late 2003 N39 25 26 25 20 19 29 17 22 25 35 24 26
A10866768 11 7 7 968
HO0.38 0.36 0.15 0.28 0.25 0.42 0.48 0.76 0.32 0.44 0.40 0.25 0.31
HE0.36 0.38 0.28 0.26 0.39 0.40 0.44 0.60 0.39 0.44 0.37 0.30 0.37
SpGATA40 Early 2002 N20 25 23 –––– 15 25 26 –––
A19 21 18 –––– 15 23 23 –––
HO1.00 0.88 0.87 –––– 0.80 0.84 0.88 –––
HE0.93 0.94 0.92 –––– 0.91 0.94 0.95 –––
Early 2003 N44 13 27 –21 –26 –24 –23 17 25
A28 15 24 –23 –25 –21 –23 18 22
HO0.89 0.92 0.96 –0.90 –0.85 –0.88 –0.83 0.88 0.80
HE0.95 0.91 0.94 –0.94 –0.94 –0.93 –0.95 0.93 0.94
282 Coral Reefs (2009) 28:277–288
123
Table 2 continued
Locus Sample date TurneVe Atoll (TA) Banco Chinchorro (BC) Roatan (RI)
1234567 8 9 10 111213
Late 2003 N38 24 26 25 20 19 29 17 22 25 35 23 25
A25 21 23 22 19 21 25 15 19 20 25 23 18
HO0.84 0.92 0.88 0.92 0.85 0.84 0.69 0.59 0.77 0.84 0.91 0.74 0.80
HE0.93 0.93 0.94 0.93 0.93 0.93 0.94 0.90 0.93 0.94 0.95 0.94 0.92
SpAAC33-1 Early 2002 N20 25 22 –––– 15 25 26 –––
A11 11 12 –––– 81313–––
HO0.90 0.76 0.73 –––– 0.87 0.84 0.81 –––
HE0.83 0.86 0.83 –––– 0.85 0.85 0.87 –––
Early 2003 N44 13 27 –22 –26 –25 –23 17 24
A13 7 14 –11 –10 –13 –13 9 10
HO0.91 0.92 0.74 –0.91 –0.88 –1.00 –1.00 0.82 0.83
HE0.88 0.80 0.89 –0.86 –0.83 –0.86 –0.86 0.84 0.85
Late 2003 N39 25 26 25 19 20 29 17 22 25 35 24 25
A1412141391313131012141310
HO0.67 0.84 0.85 0.76 0.74 0.95 0.76 0.88 0.86 0.80 0.83 0.83 0.80
HE0.85 0.86 0.87 0.84 0.82 0.87 0.87 0.87 0.87 0.86 0.87 0.86 0.88
SpAAC41-1 Early 2002 N20 25 23 –––– 15 25 26 –––
A18 22 21 –––– 17 20 21 –––
HO0.70 0.80 1.00 –––– 1.00 0.80 0.88 –––
HE0.91 0.95 0.93 –––– 0.91 0.92 0.94 –––
Early 2003 N44 12 25 –21 –25 –24 –23 17 25
A26 15 19 –22 –20 –20 –23 20 25
HO0.84 0.92 0.84 –0.90 –0.72–0.83 –0.91 0.94 0.96
HE0.94 0.92 0.92 –0.94 –0.94 –0.93 –0.93 0.93 0.95
Late 2003 N37 24 24 22 19 17 26 14 22 25 33 22 24
A24 23 23 23 21 21 21 17 21 18 25 21 23
HO0.81 0.96 0.88 0.86 0.79 0.88 0.92 1.00 0.77 0.80 0.88 0.95 1.00
HE0.93 0.94 0.94 0.94 0.94 0.93 0.94 0.93 0.92 0.92 0.94 0.93 0.95
SpAAT9-2 Early 2002 N20 25 23 –––– 15 25 23 –––
A18 21 20 –––– 17 24 24 –––
HO0.70 0.76 0.70 –––– 0.600.32 0.39 –––
HE0.93 0.94 0.93 –––– 0.92 0.94 0.94 –––
Early 2003 N44 13 26 –22 –26 –25 –22 17 25
A30 16 21 –25 –25 –20 –19 20 23
HO0.70 0.69 0.81 –0.55–0.69–0.32–0.73 0.82 0.68
HE0.95 0.90 0.94 –0.95 –0.94 –0.94 –0.92 0.93 0.94
Late 2003 N39 25 26 25 20 20 29 17 21 25 35 24 26
A25 23 22 21 21 22 24 15 23 24 29 19 25
HO0.67 0.48 0.65 0.64 0.55 0.65 0.59 0.53 0.71 0.72 0.74 0.38 0.65
HE0.95 0.93 0.93 0.93 0.93 0.94 0.94 0.90 0.94 0.94 0.95 0.93 0.94
SpAAT39 Early 2002 N19 25 21 –––– 15 25 26 –––
A14 15 20 –––– 15 18 21 –––
HO0.74 0.48 0.81 –––– 0.470.56 0.62 –––
HE0.91 0.90 0.94 –––– 0.91 0.92 0.94 –––
Early 2003 N44 13 27 –22 –23 –25 –23 17 25
A23 14 20 –18 –16 –16 –18 18 21
Coral Reefs (2009) 28:277–288 283
123
Pairwise exact tests among sites within atolls showed very
limited evidence for genetic diVerentiation after Bonferroni
correction, although three signiWcant comparisons were
found: two within TurneVe Atoll and one between TurneVe
Atoll and Banco Chinchorro sites (Table 3; Late 2003).
No signiWcant pairwise FST values were found after cor-
recting for multiple simultaneous comparisons (using either
all data or after excluding the two non-equilibrium loci);
the mean FST was 0.0031 (range = 0.0–0.0091) with nega-
tive values set to zero. Isolation by distance over the MBRS
was not apparent when geographic distance was plotted
against all pairwise FST or Cavalli-Sforza and Edwards’
(1967) DC, nor was there a signiWcant relationship based on
a Mantel test (Fig. 2a). In fact, the pairwise comparisons
Table 2 continued
SigniWcant deviations from Hardy–Weinberg equilibrium after sequential Bonferroni corre ction are underlined in boldface. D ashes indicate no data
Locus Sample date TurneVe Atoll (TA) Banco Chinchorro (BC) Roatan (RI)
1234567 8 9 10 111213
HO0.61 0.77 0.70 –0.50–0.57 –0.48–0.74 0.71 0.80
HE0.94 0.91 0.93 –0.92 –0.92 –0.91 –0.92 0.92 0.93
Late 2003 N39 25 26 25 20 20 29 16 23 25 35 23 26
A20 21 20 21 19 16 19 16 19 20 21 15 19
HO0.44 0.76 0.62 0.72 0.75 0.45 0.66 0.56 0.70 0.60 0.46 0.43 0.62
HE0.93 0.92 0.93 0.92 0.92 0.92 0.92 0.92 0.93 0.92 0.93 0.91 0.93
Table 3 SigniWcant pairwise
comparisons for exact tests of al-
lele frequency diVerences
among juvenile bicolor damsel-
Wsh populations for three tempo-
ral replicates on the
Mesoamerican Barrier Reef
(TurneVe Atoll, Belize (TA);
Banco Chinchorro, Mexico
(BC); Roatan Island, Honduras
(RI))
Sites compared Early 2002 Early 2003 Late 2003
1 (TA) 2 (TA) 0.004*0.19 0.02
1 (TA) 3 (TA) 0.0009*0.07 0.005*
1 (TA) 5 (TA) –0.35 0.003*
1 (TA) 8 (BC) 0.002*–0.32
1 (TA) 9 (BC) 0.007*0.0007*0.08
1 (TA) 10 (BC) 0.06 –0.03
1 (TA) 13 (RI) –0.60 0.006
2 (TA) 3 (TA) 0.01 0.52 0.25
2 (TA) 7 (BC) –0.26 0.04
2 (TA) 8 (BC) 0.29 –0.04
2 (TA) 9 (BC) 0.07 0.02 0.79
2 (TA) 10 (BC) 0.04 – 0.03
2 (TA) 11 (RI) –0.020.47
3 (TA) 7 (BC) –0.59 0.02
3 (TA) 8 (BC) 0.69 –0.01
3 (TA) 9 (BC) 0.29 0.008 0.16
3 (TA) 10 (BC) 0.01 – 0.26
3 (TA) 11 (RI) –0.005*0.22
5 (TA) 8 (BC) – – 0.004*
5 (TA) 9 (BC) –0.10 0.67
5 (TA) 13 (RI) –0.80 0.03
7 (BC) 9 (BC) –0.030.17
8 (BC) 9 (BC) 0.72 –0.02
8 (BC) 10 (BC) 0.34 –0.01
8 (BC) 11 (RI) – – 0.01
9 (BC) 11 (RI) – 0.0001*0.73
9 (BC) 13 (RI) –0.040.06
10 (BC) 13 (RI) – – 0.02
Dashes indicate no sample tak-
en; signiWcant (P<0.05) results
before and after Bonferroni cor-
rection are indicated by boldface
and asterisks, respectively. Site
abbreviations are as described in
Fig. 1
284 Coral Reefs (2009) 28:277–288
123
formed three distinct clusters that reXect the sampling
pattern (within-atolls (0–50 km), among close atolls
(100–200 km) and between Banco Chinchorro and Roatan
Island (250–300 km)); however, no variation among those
spatial scales in genetic distance is evident (Fig. 2a). The
signiWcant negative relationship between geographic and
genetic distance at the 100–200 km scale is an artifact
driven by the consistently higher geographic distances cou-
pled with lower genetic distances for the Roatan Island—
TurneVe Atoll comparisons (circled points; Fig. 2a). Ezer
et al. (2005) noted an eddy isolating the Mexican sampling
sites from the more southerly sampled sites, potentially
explaining this anomalous pattern of genetic divergence.
Isolation by distance analysis at the within-atoll scale
produced a signiWcant positive relationship (Spearman’s
Rank = 0.41; n= 24; P=0.039; r2= 0.29; Fig. 2b); how-
ever, such a correlation analysis is subject to biases that
may inXate the signiWcance of the relationship since the
pairwise measures of geographical and genetic distance are
not independent. Molecular variance among atolls, among
sites within atolls and among individuals was inXuenced
exclusively by variation among individuals (100%), with
negligible or no variance explained by among-atoll com-
parisons (0.06%) or by sites within atolls (0.0%).
Temporal genetic analysis
Exact tests for changes in allele frequency distribution for
individual populations over time detected two signiWcant
changes after Bonferroni correction, both on an annual
scale (one on TurneVe Atoll and one on Banco Chincharro;
Table 4), although an additional six marginally signiWcant
(signiWcant prior Bonferroni correction) diVerences were
found (Table 4). Temporal genetic stability implies a corre-
lation between genetic distances across sampling years.
However, Mantel tests showed no signiWcant correlation in
either the annual (2002 vs. Late 2003) or seasonal (Early
2003 vs. Late 2003) pairwise DC comparisons (P>0.35;
Fig. 3). The variance explained by the linear regression was
low (annual r2= 0.04; seasonal r2= 0.09) and the slopes
were negative.
The lack of genetic stability is also apparent when com-
parisons are made among the results of the exact tests of
allele frequency distribution within the three temporal repli-
cates (Table 3). Overall nine signiWcant allele frequency
diVerences among sampled populations across all sampling
times were found; however, only two site comparisons show
repeated signiWcant diVerences over time (Table 3). Finally,
no signiWcant isolation-by-distance correlations were identi-
Wed in any time period except Late 2003 (see above) based
on Spearman Rank or Mantel correlation analyses, although
this may be a function of greatly reduced number of pair-
wise comparisons in the earlier temporal samples.
Discussion
The spatial analysis of genetic structure in juvenile bicolor
damselWsh along the MBRS did not identify structuring at
large geographic scales (among atolls) but did at small geo-
graphic scales (within atolls). Although a number of pair-
wise comparisons of allele frequency distribution were
found to be signiWcant, there was no obvious geographic
pattern to those diVerences, and they tended to be found
more often at the within-atoll spatial scale. Such stochastic
genetic divergence in recently settled juvenile Wsh implies
equally chaotic adult population structure, since adult
populations are an integration of multiple juvenile cohorts.
Limited or no genetic structure in coral reef Wsh at various
geographic scales has been reported in previous studies uti-
lizing a variety of genetic markers (Shaklee 1984; Lacson
and Morizot 1991; Planes et al. 1993; Doherty et al. 1995;
Planes and Doherty 1997; Bernardi et al. 2001; McCartney
Fig. 2 Geographic versus genetic distance plots for bicolor damselWsh
(S. partitus) sample sites in the Mesoamerican Barrier Reef System.
(a) All pairwise genetic (Cavalli-Sforza and Edwards (1967) chord dis-
tance—DC) versus geographic distance (measured as shortest water
distance) comparisons. The three clusters correspond to within atoll,
between adjacent atolls and between furthest atolls (i.e., Chinchorro—
Roatan). Circled points in the center cluster are those for TurneVe—
Roatan comparisons (un-circled points are TurneVe-Chinchorro com-
parisons). (b) Pairwise genetic and geographic distance plotted for the
within-atoll comparisons, excluding all between-atoll comparisons
(solid line is signiWcant regression line)
0 100 200 300
0102030405
0
0.30
0.35
0.40
0.45
0.50
0.35
0.40
0.45
Geographical Distance (km)
Genetic Distance (DC)
a
b
Coral Reefs (2009) 28:277–288 285
123
et al. 2003; Rhodes et al. 2003; van Herwerden et al. 2003;
Purcell et al. 2006; Gerlach et al. 2007; Haney et al. 2007),
while other studies have documented remarkable levels of
population diVerentiation (Taylor and Hellberg 2003; HoV-
man et al. 2005; Gerlach et al. 2007). The variation in the
magnitude of genetic structure may depend critically on the
life history of the study species (Gerlach et al. 2007). In
theory, predictable recruitment and connectivity patterns
are expected to drive stable genetic structure among popu-
lations or study sites (Purcell et al. 2006), while panmixia
resulting from random and widespread larval dispersal
should not result in any form of consistent genetic substruc-
ture (Selkoe et al. 2006). However, if the observed genetic
structure is present, but weak, then discriminating between
those two possibilities can prove problematic.
The low genetic divergence among S. partitus juvenile
cohorts across the MBRS is not an unexpected result and,
indeed, is in agreement with most past population-level
genetic studies of Stegastes species in the Hawaiian archipel-
ago and Florida Keys (Shaklee 1984; Lacson and Morizot
1991), although those studies did not examine a wide geo-
graphic range. Shulman and Bermingham’s (1995) study,
however, detected ‘very strong’ population structuring in
adult S. leucostictus in the Caribbean using mitochondrial
restriction fragment length polymorphisms. Lacson et al.’s
(1989) allozyme study of S. partitus detected large-scale
structure in the Florida Keys which they concluded was
most likely the result of a population bottleneck caused by a
storm event earlier in the season. Generally, damselWsh
species show considerable variation in genetic structure in
the Indo-PaciWc and Atlantic regions (Doherty et al. 1995;
Planes and Doherty 1997; Planes et al. 1998) although Das-
cyllus aruanus (Planes et al. 1993) and the three-spot das-
cyllus, D. trimaculatus (Bernardi et al. 2001) of French
Polynesia showed genetic homogeneity at all spatial scales
until West PaciWc populations (>1,000 km) were included
in the comparisons. Some recent microsatellite studies also
report low or no genetic structuring among reef Wsh popula-
tions up to very large spatial scales (McCartney et al. 2003;
van Herwerden et al. 2003; Purcell et al. 2006; Haney et al.
2007), while other studies report strong large-scale diVeren-
tiation patterns (Rhodes et al. 2003). Although no general
relationship between spatial scale and the likelihood of
population-level genetic divergence emerges based on pub-
lished coral reef Wsh genetics, it is evident that a variety of
Fig. 3 Scatterplots of pairwise genetic distances (Cavalli-Sforza and
Edwards (1967) chord distance—DC) comparisons between temporal
replicates of MBRS juvenile S. partitus. (a) Seasonal comparison
(Early 2003 vs. Late 2003) of newly settled juveniles. (b) Annual com-
parison (2002 vs. Late 2003) of newly settled juveniles. The dashed
line represents the one-to-one relationship expected if genetic distance
values remain static over time (i.e., high predictability)
0.30 0.35 0.40 0.45 0.50
Early 2003 pairwise DC
0.30
0.35
0.40
0.40 0.45 0.50
2002 pairwise DC
0.35
0.40
0.45
Late 2003 pairwise DC
a
b
Table 4 Exact test probabilities of allele frequency diVerences (eight
microsatellite loci) between temporal replicate samples of juvenile bi-
color damselWsh sampled at 13 sites on the MBRS (TurneVe Atoll, Be-
lize (TA); Banco Chinchorro, Mexico (BC); Roatan Island, Honduras
(RI))
All signiWcant results are in boldface; those signiWcant after Bonferroni
correction are indicated by asterisks. Site abbreviations are as de-
scribed in Fig. 1
Temporal comparison Population P (exact test)
Early 2002 vs. Early 2003 1 (TA) 0.0001*
2 (TA) 0.031
3 (TA) 0.453
9 (BC) 0.007*
Early 2002 vs. Late 2003 1 (TA) 0.041
2 (TA) 0.115
3 (TA) 0.042
8 (BC) 0.464
9 (BC) 0.188
10 (BC) 0.143
Early 2003 vs. Late 2003 1 (TA) 0.234
2 (TA) 0.022
3 (TA) 0.012
5 (TA) 0.356
7 (BC) 0.262
9 (BC) 0.013
11 (RI) 0.396
12 (RI) 0.174
13 (RI) 0.494
286 Coral Reefs (2009) 28:277–288
123
oceanographic as well as biological (e.g., behavioral, repro-
ductive, and early life history) factors likely play a role in
determining population genetic structure at medium to
large spatial scales (Cowen et al. 2006; Purcell et al. 2006;
Selkoe et al. 2006; Gerlach et al. 2007). The oceanographic
Xow regime of the MBRS is seasonally variable both in
velocity and direction of currents, and further variation
is introduced in conjunction with eddies and gyres asso-
ciated with the Caribbean Current (Ezer et al. 2005;
Tang et al. 2006). Furthermore, oceanographic events,
such as hurricanes, can dramatically alter patterns of
larval movement in the sampled region, facilitating gene
Xow between previously divergent populations and
resulting in increased genetic homogeneity at larger
spatial scales. However, small-scale genetic diVerences
may be generated by local oceanographic eVects (Cowen
et al. 2006), and such diVerences are likely to be tempo-
rally transient.
In this study, genetic structuring among recently settled
MBRS bicolor damselWsh was detected at small spatial
scales. The nature of that genetic structure was curious;
while a few signiWcant diVerences among sites were identi-
Wed, they did not constitute a pattern. A linear relationship
between genetic and geographic distances at the within-atoll
scale was found, a pattern usually indicative of spatial limi-
tation of gene Xow (Purcell et al. 2006). However, here
juvenile cohorts were used, and hence the relationship
between genetic and geographic distance is not likely medi-
ated by simple variation in gene Xow. Nevertheless, varia-
tion in the genetic structure of settling cohorts is expected to
translate into adult population genetic structure. Similar to
the large-scale genetic studies described above, small-scale
genetic diVerentiation is usually lacking (Planes et al. 1993;
Bernardi et al. 2001; McCartney et al. 2003; Purcell et al.
2006; Gerlach et al. 2007), although some studies have
reported remarkable levels of genetic divergence on rela-
tively small geographic scales that were attributed to behav-
ioral and life history eVects (Lacson et al. 1989; HoVman
et al. 2005; Gerlach et al. 2007). In general, small spatial
scale processes may play an important role in the dispersal
of larvae, and hence adult population connectivity and
genetic structure (Selkoe et al. 2006). Variation in local Xow
conditions brought about by deep-water and coastal current
interactions can result in high levels of local retention
(Swearer et al. 1999; Cowen et al. 2006; Selkoe et al. 2006).
Generally, behavioral and life history factors would tend to
produce predictable, temporally stable patterns, while small-
scale oceanographic factors would tend to produce chaotic
and unpredictable patterns of genetic divergence (Selkoe
et al. 2006). Quantitative estimates of the patterns of genetic
divergence among juvenile cohorts may provide data valu-
able for evaluating connectivity mechanisms independent of
the level of adult population genetic structure.
At present, studies examining temporal genetic stability
using suitably variable genetic markers (e.g., microsatellite
DNA) in coral reef Wshes are uncommon, but a pattern of
limited temporal stability is emerging (see Selkoe et al.
2006). Rhodes et al. (2003) employed three microsatellite
markers to examine the population genetic structure of the
camouXage grouper (Epinephelus polyphekadion) over a
2-year period; of the three regional groups identiWed, one
was genetically unstable over time. Lacson and Morizot
(1991) used allozymes to examine the temporal stability of
population genetic structure in S. partitus in the Upper
Florida Keys, and showed that previously diVerentiated
populations (Lacson et al. 1989) were almost completely
genetically homogenous 3 years after the Wrst study. How-
ever, other studies have reported no detectable temporal het-
erogeneity in population genetic structure among coral reef
Wsh populations (Bernardi et al. 2001; Purcell et al. 2006;
Gerlach et al. 2007); however, many of those studies did not
perform speciWc tests for temporal stability or changes in
genetic structure over time, and hence are diYcult to evaluate.
The apparently chaotic and temporally unstable genetic
structure of the recently settled MBRS S. partitus reported
here may reXect genetic “patchiness” in the larval pool
resulting from high reproductive variance and oceano-
graphic factors (Selkoe et al. 2006). Organisms with high
fecundity and high levels of early mortality are likely to
have elevated reproductive variance due to random matches
of larval cohorts with the optimal environmental condi-
tions—the driving mechanism behind the “sweepstakes-
chance matching” hypothesis (Hedgecock 1994). As the
bicolor damselWsh has relatively high fecundity (Knapp
1993) and since coral reef Wsh larvae typically have high
early mortality, high reproductive variance is expected, and
thus perhaps sweepstakes-chance matching recruitment
patterns. The correlation between geographic and genetic
distance at the within-atoll scale reported here may in fact
provide indirect evidence for sweepstakes-chance matching
in the MBRS S. partitus. Since no evidence for a correla-
tion between genetic and geographic distance at larger geo-
graphic scales was found, it is unlikely that the small-scale
correlation is due to a migration-drift equilibrium normally
thought to drive isolation by distance models. An alterna-
tive explanation is that related cohorts of settling larvae
(resulting from sweepstakes-chance matching) are spatially
limited, but span more than one of the small-scale sampling
sites in this study. Thus, the geographic-genetic distance
correlation at small scales in this study may simply reXect
spatial autocorrelation in the settling pattern of related
cohorts such that physically close sampling sites are more
likely to be related than distant sampling sites. Such a
hypothesis would explain the unexpected pattern of signiW-
cance of genetic-geographic distance measures seen in this
study, and perhaps other studies. Analyses of small-scale
Coral Reefs (2009) 28:277–288 287
123
genetic-geographic distance correlations in recently settled
cohorts could provide additional empirical evidence for
spatial autocorrelation in larval relatedness.
The genetic signatures of newly recruited bicolor dam-
selWsh were variable over time in this study, providing a
possible explanation for the diYculty in predicting, and the
inconsistency in estimating, connectivity and recruitment in
reef Wsh populations. Unpredictable changes in the genetic
structure of newly settled juvenile cohorts highlight a
potential confounding factor in the interpretation of genetic
structure studies in coral reef Wshes. It is possible that tem-
poral variation in the recruitment mechanism is the norm
for reef Wsh populations, and diVerent species and popula-
tions may simply vary in the degree that the biotic and abi-
otic factors drive temporal instability. Such a possibility
would lead to unpredictable variation aVecting all methods
of assessing connectivity, including otoliths tags, and thus
temporal variation in cohort composition may also explain
disagreement among diVerent methods of quantifying con-
nectivity. Cowen et al.’s (2006) model showed that unpre-
dictable dispersal patterns arose from combinations of
larval behavior and oceanographic currents, while in this
MBRS study area, current patterns and estimated retention
times vary dramatically (Tang et al. 2006). More temporal
genetic data for recruiting cohorts will be required before a
robust evaluation of the generality of variability over time
in reef Wsh connectivity can be made; however, fundamen-
tally chaotic connectivity would have important implica-
tions for future management and research on coral reef Wsh.
Acknowledgments We thank S. Jamieson, M. Docker, C. Busch,
A. Ludusan, and Z. Mazman for laboratory and analysis contributions and
P. Chittaro, J. Kritzer, P. Usseglio, C. Mora, D. Hogan, W. Thompson,
C. Nolan, and the staV at the University of Belize, Institute of Marine
Studies, for Weld collection assistance. This work was supported by
Natural Science and Engineering Research Council (NSERC) grants to
DDH, BD, and PFS (CRO-227965) and by NSERC Canada Research
Chair funding to DDH.
References
Almany GR, Berumen ML, Thorrold SR, Planes S, Jones GP (2007)
Local replenishment of coral reef Wsh populations in a marine
reserve. Science 316:742–744
Barlow GW (1981) Patterns of parental investment, dispersal and size
among coral-reef Wshes. Environ Biol Fish 6:65–85
Bay LK, Crozier RH, Caley MJ (2006) The relationship between pop-
ulation genetic structure and pelagic larval duration in coral reef
Wshes on the Great Barrier Reef. Mar Biol 149:1247–1256
Bernardi G, Holbrook SJ, Schmitt RJ (2001) Gene Xow at three spatial
scales in a coral reef Wsh, the three-spot dascyllus, Dascyllus
trimaculatus. Mar Biol 138:457–465
Cavalli-Sforza LL, Edwards AW (1967) Phylogenetic analysis: mod-
els and estimation procedures. Am J Hum Genet 19:233–257
Cowen RK, Paris CB, Srinivasan A (2006) Scaling of connectivity in
marine populations. Science 311:522–527
DeLoach N (1999) Reef Wsh behavior: Florida, Caribbean, Bahamas.
New World Publications Inc., Jacksonville
Doherty PJ, Williams DM, Sale PF (1985) The adaptive signiWcance of
larval dispersal in coral reef Wshes. Environ Biol Fish 12:81–90
Doherty PJ, Planes S, Mather P (1995) Gene Xow and larval duration
in seven species of Wsh from the Great Barrier Reef. Ecology
76:2373–2391
Ezer T, Thattai DV, Kjerfve B, Heyman WD (2005) On the variability
of the Xow along the Meso-American barrier reef system: a
numerical model study of the inXuence of the Caribbean current
and eddies. Ocean Dynamics 55:458–475
Ehrlich PR (1975) The population biology of coral reef Wshes. Annu
Rev Ecol Syst 6:211–247
ExcoYer L, Smouse P, Quattro J (1992) Analysis of molecular vari-
ance inferred from metric distances among DNA haplotypes:
applications to human mitochondrial DNA restriction data.
Genetics 131:479–491
Gerlach G, Atema J, Kingsford MJ, Black KP, Miller-Sims V (2007)
Smelling home can prevent dispersal of reef Wsh larvae. Proc Natl
Acad Sci USA 104:858–863
Haney RA, Silliman BR, Rand DM (2007) A multi-locus assessment
of connectivity and historical demography in the bluehead wrasse
(Thalassoma bifasciatum). Heredity 98:294–302
Hedgecock D (1994) Does variance in reproductive success limit eVec-
tive population sizes of marine organisms? In: Beaumont A (ed)
Genetics and evolution of aquatic organisms. Chapman and Hall,
London, pp 122–134
HoVman EA, Kolm N, Berglund A, Arguello R, Jones AG (2005)
Genetic structure in the coral-reef-associated Banggai cardinal-
Wsh, Pterapogon kauderni. Mol Ecol 14:1367–1375
Johannes RE (1978) Reproductive strategies of coastal marine Wshes in
the tropics. Environ Biol Fish 3:741–760
Jones ME, Barber PH (2005) Characterization of microsatellite loci for
the detection of temporal genetic shifts within a single cohort of
the brown demoiselle, Neopomacentrus Wlamentosus. Mol Ecol
Notes 5:834–836
Jones GP, Planes S, Thorrold SR (2005) Coral reef Wsh larvae settle
close to home. Curr Biol 15:1314–1318
Knapp RA (1993) The inXuence of egg survivorship on the subsequent
nest Wdelity of female bicolour damselWsh, Stegastes partitus.
Anim Behav 46:111–121
Lacson JM, Morizot DC (1991) Temporal genetic variation in subpop-
ulations of bicolor damselWsh (Stegastes partitus) inhabiting coral
reefs in the Florida Keys. Mar Biol 110:353–357
Lacson JM, Riccardi VM, Calhoun SW, Morizot DC (1989) Genetic
diVerentiation of bicolor damselWsh (Eupomacentrus partitus)
populations in the Florida Keys. Mar Biol 103:445–451
Leis JM (1991) The pelagic stage of reef Wshes: the larval biology of
coral reef Wshes. In: Sale PF (ed) The ecology of Wshes on coral
reefs. Academic Press, San Diego, pp 183–230
Leis JM (1993) Larval Wsh assemblages near Indo-PaciWc coral reefs.
Bull Mar Sci 53:362–392
Lenfant P, Planes S (2002) Temporal genetic changes between cohorts
in a natural population of a marine Wsh, Diplodus sargus. Biol J
Linn Soc 76:9–20
McCartney MA, Acevedo J, Heredia C, Rico C, Quenoville B,
Bermingham E (2003) Genetic mosaic in a marine species Xock.
Mol Ecol 12:2963–2973
Patterson HM, Kingsford MJ (2005) Elemental signatures of Acanth-
ochromis polycanthus otoliths from the Great Barrier Reef have
signiWcant temporal, spatial, and between-brood variation. Coral
Reefs 24:360–369
Planes S, Doherty PJ (1997) Genetic relationships of the colour morphs
of Acanthochromis polycanthus (Pomacentridae) on the northern
Great Barrier Reef. Mar Biol 130:109–117
288 Coral Reefs (2009) 28:277–288
123
Planes S, Bonhomme F, Galzin R (1993) Genetic structure of Dascyllus
aruanus populations in French Polynesia. Mar Biol 117:665–674
Planes S, Parroni M, Chauvet C (1998) Evidence of limited gene Xow
in three species of coral reef Wshes in the lagoon of New Caledo-
nia. Mar Biol 130:361–368
Planes S, Lecaillon G, Lenfant P, Meekan M (2002) Genetic and
demographic variation in new recruits of Naso unicornis. J Fish
Biol 61:1033–1049
Purcell JFH, Cowen RK, Hughes CR, Williams DA (2006) Weak
genetic structure indicates strong dispersal limits: a tale of two
coral reef Wsh. Proc Roy Soc Lond, B 273:1483–1490
Raymond M, Rousset F (1995) An exact test for population diVerenti-
ation. Evolution 49:1280–1283
Rhodes KL, Lewis RI, Chapman RW, Sadovy Y (2003) Genetic struc-
ture of camouXage grouper, Epinephelus polyphekadion (Pisces:
Serranidae), in the western central PaciWc. Mar Biol 142:771–776
Rice WR (1989) Analyzing tables of statistical tests. Evolution 43:
223–225
Robertson DR (1992) Patterns of lunar settlement and early recruit-
ment in Caribbean reef Wshes at Panama. Mar Biol 114:527–537
Robertson DR, Green DG, Victor BC (1988) Temporal coupling of
production and recruitment of larvae of a Caribbean reef Wsh.
Ecology 69:370–381
Sale PF, Danilowicz BS, Doherty P, Williams DM (2005) The relation
of microhabitat to variation in recruitment of young-of-year coral
reef Wshes. Bull Mar Sci 76:123–142
Schneider S, Roessli D, ExcoYer L (2000) Arlequin ver 2.000: a soft-
ware for population genetics data analysis. Genetics and Biome-
try Laboratory, University of Geneva, Geneva
Selkoe KA, Gaines SD, Caselle JE, Warner RR (2006) Current shifts
and kin aggregation explain genetic patchiness in Wsh recruits.
Ecology 87:3082–3094
Shaklee J (1984) Genetic variation and population structure in the
damselWsh, Stegastes fasciolatus, throughout the Hawaiian archi-
pelago. Copeia 3:629–640
Shulman MJ, Bermingham E (1995) Early life histories, ocean
currents, and the population genetics of Caribbean reef Wshes.
Evolution 49:897–910
Sponaugle S, Cowen RK (1994) Larval durations and recruitment pat-
terns of two Caribbean gobies (Gobiidae): contrasting early life
histories in demersal spawners. Mar Biol 120:133–143
Swearer SE, Caselle JE, Lea DW, Warner RR (1999) Larval retention
and recruitment in an island population of a coral-reef Wsh. Nature
402:799–802
Taylor MS, Hellberg ME (2003) Genetic evidence for local retention
of pelagic larvae in a Caribbean reef Wsh. Science 299:107–109
Tang L, Sheng J, Hatcher BG, Sale PF (2006) Numerical study of cir-
culation, dispersion, and hydrodynamic connectivity of surface
waters on the Belize shelf. J Geophys Res 111:1–18
Thresher RE, Colin PL, Bell LJ (1989) Planktonic duration, distribu-
tion and population structure of western and central PaciWc dam-
selWshes (Pomacentridae). Copeia 2:420–434
van Herwerden L, Benzie J, Davies C (2003) Microsatellite variation
and population genetic structure of the red throat emperor on the
Great Barrier Reef. J Fish Biol 62:987–999
Williams DA, Purcell J, Hughes CR, Cowen RK (2003) Polymorphic
microsatellite loci for population studies of the bicolor damselWsh,
Stegastes partitus (Pomacentridae). Mol Ecol Notes 3:547–549
