ro
jan
RO
SS
Box
ment
?Bishop Museum, 1525 Bernice Street, Honolulu, HI 96817 USA, ?
9100, Box 0948, DPO AA, 34002-0998, USA
Received 1 September 2009; revision received 8 November 2009; accepted 24 November 2009
Islands in the central Pacific. This natural range does Islands (Oda & Parrish 1982; Randall 1987). The intro-
lines and transferred to floating pens. When sufficient
fish had been captured, the fish were transferred to the
Correspondence: Michelle Gaither, Fax: +1 808 236 7442;
E-mail: gaither@hawaii.edu
Molecular Ecology (2010) 19, 1107?1121 doi: 10.1111/j.1365-294X.2010.04535.xnot include the Hawaiian Islands, which has only a sub-
set of the Indo-Pacific flora and fauna, lacking many
duction of L. kasmira was conducted in two discrete
events. In preparation for introduction, juvenile fish
(approximately 100?120 g) were caught using handIntroduction
The Bluestriped Snapper, Lutjanus kasmira (Forsska?l,
1775), is a widely distributed coral reef fish with a natu-
ral range from South Africa to the Marquesas and Line
taxa such as most shallow water snappers and groupers
(Randall 2007). In an effort to fill a perceived empty
ecological niche, and to enhance local fisheries, the
Hawaii Division of Fish and Game (HDFG) introduced
L. kasmira, among other reef fishes, to the Hawaiian 2010 BlackAbstract
A half century ago the State of Hawaii began a remarkable, if unintentional, experiment
on the population genetics of introduced species, by releasing 2431 Bluestriped Snappers
(Lutjanus kasmira) from the Marquesas Islands in 1958 and 728 conspecifics from the
Society Islands in 1961. By 1992 L. kasmira had spread across the entire archipelago,
including locations 2000 km from the release site. Genetic surveys of the source
populations reveal diagnostic differences in the mtDNA control region (d = 3.8%;
/ST = 0.734, P < 0.001) and significant allele frequency differences at nuclear DNA loci
(FST = 0.49; P < 0.001). These findings, which indicate that source populations have been
isolated for approximately half a million years, set the stage for a survey of the Hawaiian
Archipelago (N = 385) to determine the success of these introductions in terms of genetic
diversity and breeding behaviour. Both Marquesas and Society mtDNA lineages were
detected at each survey site across the Hawaiian Archipelago, at about the same
proportion or slightly less than the original 3.4:1 introduction ratio. Nuclear allele
frequencies and parentage tests demonstrate that the two source populations are freely
interbreeding. The introduction of 2431 Marquesan founders produced only a slight
reduction in mtDNA diversity (17%), while the 728 Society founders produced a greater
reduction in haplotype diversity (41%). We find no evidence of genetic bottlenecks
between islands of the Hawaiian Archipelago, as expected under a stepping-stone model
of colonization, from the initial introduction site. This species rapidly colonized across
2000 km without loss of genetic diversity, illustrating the consequences of introducing
highly dispersive marine species.
Keywords: alien species, bottleneck, introductions, invasion biology, mtDNA, Marquesas,
nuclear DNA, Papahanaumokuakea, Society IslandsGenetic consequences of int
of Bluestriped Snapper (Lut
MICHELLE R. GAITHER,* BRIAN W. BOWEN,*
VANESSA MESSMER,? JOHN EARLE? and D. RO
*Hawaii Institute of Marine Biology, University of Hawaii, PO
Centre de Recherches Insulaires et Observatoire de l?Environnewell Publishing Ltdducing allopatric lineages
us kasmira) to Hawaii
BERT J . TOONEN,* SERGE PLANES,?
ROBERTSON?
1346, Kaneohe, HI 96744 USA, ?UMS 2978 CNRS-EPHE,
(CRIOBE), BP 1013, 98729 Moorea, French Polynesia,
Smithsonian Tropical Research Institute (Panama?) Unit
variable sea temperatures (Randall 2001; Gaither et al.
1108 M. R. GAITHER ET AL .bait wells of the transport vessel and brought to Hawaii
for transplant. In 1958, 2431 fish from Nuku Hiva in the
Marquesas Islands, and, in 1961, 728 fish from Moorea
in the Society Islands were released on Oahu (Fig. 1;
3.4: 1 ratio; HDFG records). L. kasmira quickly spread
through the archipelago at a rate of about 60 km per
year (Oda & Parrish, 1982; Randall 2007). In 1992, just
34 years after the initial introduction, L. kasmira was
recorded at the far reaches of the archipelago at Mid-
way Atoll (Randall et al. 1993) over 2000 km from the
release site. This successful introduction provides a
number of research opportunities relating to under-
standing founder ? colonization processes.
The Marquesas and Society populations of L. kasmira
are phylogenetically distinct with diagnostic differences
Hawaii
Society
Marquesas
2431
728
Fig. 1 Map of the Pacific Ocean. Lutjanus kasmira were intro-
duced to the Hawaiian Island of Oahu from two source loca-
tions: Nuku Hiva in the Marquesas Islands and Moorea in the
Society Islands. The number of fish introduced from each loca-
tion is shown. (Photo credit: Keoki Stender.)in the mitochondrial genome (average sequence diver-
gence for cytochrome b = 0.53%; Gaither et al. 2010).
The introduction of L. kasmira to the Hawaiian Islands
from two genetically divergent populations, resulted in
the sympatric distribution of lineages that have been
separated for about a quater to a half a million years
(Gaither et al. 2010). The genetic divergence between
these two populations is a result of the phylogenetic
distinction of the Marquesas population relative to
other Indo-Pacific populations (Gaither et al. 2010). The
Marquesas have the third highest level of endemism
among shorefishes (11.6%) in Oceania (Randall 2001)
and high levels of genetic differentiation in three of five
species of non-endemic reef fishes examined to date
(Planes & Fauvelot 2002; Craig et al. 2007; Schultz et al.
2007; Gaither et al. 2010). The distinction of the Marque-
sas shorefish fauna has been attributed to a combination
of geographic isolation (enhanced by the westerly
Southern Equatorial Current) and adaptation to unusually2010). Because geographic isolation and ecological
divergence may both promote speciation in fishes
(Rogers & Bernatchez 2006; Rocha & Bowen 2008), we
ask whether L. kasmira populations separated for half a
million years can freely interbreed in sympatry. The
genetic distinctiveness of the two source populations
provides an opportunity to identify the descendants of
the Hawaiian introductions, to assess their relative suc-
cess in the archipelago, and to determine if the two
genetic lineages are mixing.
Colonization events, including human-mediated
introductions, often involve a severe reduction in popu-
lation size and isolation from the larger parental popu-
lation. The dramatic decrease in effective population
size that accompanies such founder events is expected
to lead to decreased genetic diversity (Nei et al. 1975).
However, the accumulation of data indicates that
genetic bottlenecks in introduced populations are not
an invariable outcome (Bossdorf et al. 2005; Wares et al.
2005; Roman & Darling 2007). Interpreting patterns of
genetic diversity in introduced populations is con-
founded by the fact that in most cases the source popu-
lation and the number of founding individuals are
unknown. Under these conditions, researchers must
reconstruct the history of introductions by combining
molecular and geographic data to identify source popu-
lations (Wares et al. 2005). In some cases high genetic
diversity in the introduced range can be attributed to
admixture of genetically divergent populations (Kolbe
et al. 2004; Genton et al. 2005; Carmeron et al. 2008; Ro-
senthal et al. 2008). In cases where only a single source
population can be identified, high genetic diversity in
the introduced range is often attributed to either a large
number of colonizers, rapid population expansion fol-
lowing the founder event, or both (Hassan et al. 2003;
Stepien et al. 2005). Discussions concerning the effect of
founder events on genetic diversity would be greatly
informed if more empirical data concerning the effects
of founder population size on genetic diversity were
available. Intentional and well documented introduc-
tions, where the source population and founder popula-
tion size are confidently known, offer powerful test
cases.
The intentional introduction of L. kasmira to the
Hawaiian Islands provides a rare opportunity to
directly evaluate the effects of founder population size
on genetic diversity in recently established populations.
Here we capitalize on two unique aspects of the intro-
duction of L. kasmira to Hawaii: (i) the introduction
occurred in two well documented events with known
numbers of founders and source populations; (ii) the
source populations (Nuka Hiva in the Marquesas
Islands and Moorea in the Society Islands) are geneti- 2010 Blackwell Publishing Ltd
cally distinct, allowing us to identify their descendents.
divergent populations interbreeding in the Hawaiian
that are sufficient to retain the genetic diversity of par-
NOAA R ? V Hi?ialakai, as part of an initiative by the
was inferred by direct sequencing.
THE I NTRO DUCT ION OF LUTJANUS KASMIRA TO HAW AII 1109ent populations.
Materials and methods
Study species
Lutjanus kasmira has broad habitat preferences, occupy-
ing hard substrata from shallow waters to at least
265 m (Randall 1987) and has a generalized predatory
diet that includes fish, crustaceans, and cephalopods
(Randall & Brock 1960; Oda & Parrish 1982; Schuma-
cher & Parrish 2005). This species reaches sexual matu-
rity at 1?2 years (Rangarajan 1971; Morales-Nin &
Ralston 1990) and engages in mass spawning (Suzuki &
Hioki 1979). Long-distance movement between isolated
patches of adult habitat (reefs) occurs during a highly
dispersive pelagic larval phase that, in other species of
Lutjanus, lasts 20?44 days (Zapata & Herron 2002; Denit
& Sponaugle 2004).
Collections
A total of 385 specimens of Lutjanus kasmira were col-
lected from 10 locations across the Hawaiian Archipel-
ago by scuba divers using polespears (Table 1, Fig. 2).
Specimens from the uninhabited NW Hawaiian Islands
were obtained during research expeditions on theIslands? (v) Is there evidence of genetic bottlenecks at
the introduction site or as the fish spread throughout
the archipelago? The circumstances of this study offer
unprecedented opportunities to study species introduc-
tions and invasions, pertinent to management of marine
resources including the Papahanaumokuakea Marine
National Monument (PMNM) that traverses 2000 km of
the north-western (NW) Hawaiian Islands. At least 350
alien marine species occupy the inhabited Main Hawai-
ian Islands (Eldredge & Smith 2001) and few studies
have addressed the threat these aliens pose to the unin-
habited (and nearly pristine) ecosystems of the NW
Hawaiian Islands. Hence an ongoing concern is the
level of connectivity between the Main Hawaiian
Islands and the NW Hawaiian Islands (see Eble et al.
2009). Here we document an extreme scenario of rapid
colonization into the NW Hawaiian Islands, in numbersWe employ both mitochondrial and nuclear sequence
data to ask the following questions (i) Did fish from
both source populations become established in the
Hawaiian Islands? If so, (ii) how are their descendents
distributed in the archipelago? (iii) Were fish from both
source populations equally successful at reproducing
and colonizing the islands? (iv) Are these genetically 2010 Blackwell Publishing LtdSequences for each locus were aligned and edited
using SEQUENCHER 4.8 (Gene Codes, Ann Arbor, MI,
USA) and trimmed to a common length. The mtDNA
control region contained multiple indels which varied
from 1 to 3 bp in length. Alignment of the mtDNA
sequences was confirmed using default parameters in
CLUSTAL W 1.81 (Thompson et al. 1994). Unique mtDNA
haplotypes and nuclear alleles were identified with the
merge taxa option in MACCLADE 4.05 (Maddison &Papahanaumokuakea Marine National Monument
(http://hawaiireef.noaa.gov/) to monitor and character-
ize this vast protected area. Tissue samples (fin clips or
gill filaments) were preserved in either 95% ethanol
(EtOH) or saturated NaCl solution (Seutin et al. 1991),
and stored at room temperature. Fifty L. kasmira sam-
ples from each of the Marquesas and Society source
populations, previously analysed in Gaither et al. (2010),
were also used in this study.
DNA extraction, PCR amplifications, and sequencing
All DNA extraction, PCR cycling, cloning, and sequenc-
ing protocols used here are identical to those in Gaither
et al. (2010). The growth hormone (GH) and adenine
nucleotide transporter translocase (ANT) intron
sequences obtained from each of the Marquesas and
Society populations in Gaither et al. (2010) were used in
this study [GenBank accession numbers FJ754178?
FJ754184 (GH intron), FJ754157?FJ754177 (ANT intron)].
All 385 specimens of L. kasmira collected from the Hawai-
ian Islands were sequenced at these two loci. Addition-
ally, approximately 215 bp of the third intron in the
gonadotropin-releasing hormone 3 (GnRH3-3) were
amplified using the primers GnRH3F (5?-GCCCAAACC-
CAAGAGAGACTTAGACC-3?) and GnRH3R (5?-
TTCGGTCAAAATGACTGGAATCATC-3?) (Hassan et al.
2002) and approximately 575 bp of the mitochondrial
control region were amplified using the primers Lutjf1
(5?-GCACTCTGAAATGTCAAGTGAAAGG-3?) and
CRA (5?-TTCCACCTCTAACTCCCAAAGCTAG-3?) (Lee
et al. 1995) in all 484 samples (Hawaii = 385, Marque-
sas = 50, and Society = 49). PCR protocols and cycling
conditions for both the GnRH3-3 intron and the mtDNA
control region were carried out as described in Gaither
et al. (2010) using an annealing temperature of 60 C.
Due to the presence of multiple indels at the GnRH3-3
locus, that would require extensive cloning to phase
alleles, analysis of this locus was restricted to the pres-
ence or absence of a 10 bp indel near the reverse priming
site. The presence of the indel was confirmed by cloning
ten individuals and comparing alleles to direct
sequences. The allelic state of the remaining individuals
Table 1 Molecular diversity indices for the mitochondrial control region sequences for the two source populations of Lutjanus kas-
of sp
rted
mito
Frig
esas
Nh
47
(0.012) (0.009)
?
Introduced range
33
26
28
1110 M. R. GAITHER ET AL .Oahu 50 40 30 0.992
(0.006)
0.033
(0.017)
40
Kona 50 41 35 0.989
(0.007)
0.038
(0.019)
28
Hilo 51 38 30 0.985 0.037 33mira and ten populations across the introduced range. Number
(Ns), haplotype diversity (h), and nucleotide diversity (p) as repo
deviations. Values for the entire data set (All Data) and for each
listed. See Fig. 2 for locations of Hawaiian Islands (FFS = French
All data Marqu
N Nh Ns h p N
Source populations
Marquesas 50 47 45 0.997
(0.005)
0.019
(0.010)
50
Society 49 31 23 0.970 0.017 ?Maddison 2002). All control region haplotypes and
nuclear alleles unique to Hawaii were deposited in
GenBank [accession numbers: GU123931?GU124148
(control region), GU192444?GU192447 ANT intron)]
Data analysis
Mitochondrial control region. Summary statistics includ-
ing mtDNA haplotype diversity (h), and nucleotide
diversity (p) were estimated using algorithms in Nei
(1987) as implemented in ARLEQUIN 3.11 (Excoffier et al.
2005). A statistical parsimony network was constructed
using the program TCS 1.21 (Clement et al. 2000).
The resulting network was simplified using standard
tie-breaking rules. In keeping with the cytochrome b
data in Gaither et al. (2010) the control region sequences
in the Marquesas and Society samples fell into two dis-
(0.008) (0.018)
Maui Nui 39 32 28 0.985
(0.011)
0.034
(0.017)
29 24
Kauai 36 30 25 0.989
(0.010)
0.035
(0.018)
25 23
Necker 49 38 31 0.986
(0.008)
0.035
(0.018)
34 29
Maro 21 18 16 0.981
(0.023)
0.036
(0.018)
15 15
FFS 40 38 36 0.997
(0.006)
0.033
(0.017)
31 30
Midway 40 33 28 0.989
(0.009)
0.046
(0.023)
28 26
Kure 9 9 9 1.000
(0.052)
0.034
(0.019)
7 7
All Hawaii
specimens
385 172 92 0.990
(0.001)
0.037
(0.018)
270 142
All
specimens
484 218 123 0.991
(0.001)
0.037
(0.018)
320 170ecimens (N), number of haplotypes (Nh), number of singletons
by ARLEQUIN 3.11 are listed. Numbers in parenthesis are standard
chondrial lineage (Marquesas Lineage and Society Lineage) are
ate Shoals)
lineage Society lineage
Ns h p N Nh Ns h p
45 0.997
(0.005)
0.019
(0.010)
? ? ? ? ?
? ? ? 49 31 23 0.970
(0.012)
0.017
(0.009)
26 0.991
(0.008)
0.023
(0.012)
10 7 4 0.933
(0.062)
0.018
(0.010)
25 0.992
(0.013)
0.019
(0.010)
22 15 10 0.957
(0.028)
0.018
(0.009)
24 0.989 0.020 18 10 6 0.915 0.019tinct lineages. Average percent difference between pop-
ulations was calculated by dividing the average number
of nucleotides (corrected; Tamura & Nei 1993) that dif-
fer between the two source populations (as calculated
in ARLEQUIN) by the total number of base pairs. The aver-
age percent difference between populations is reported
here as sequence divergence (d).
The number of individuals from the Hawaiian Islands
that grouped with either the Marquesas or the Society
mtDNA lineage was calculated, and deviations from
the initial introduction ratio of 3.4 Marquesas:1.0 Society
were tested using Fisher?s exact test (Sokal & Rohlf
1995). To test for the loss of genetic diversity in the
introduced range, while controlling for unequal sample
sizes (Leberg 2002), we estimated haplotype richness
using rarefaction analysis. For this method we deter-
mined the haplotype frequency distribution for the
(0.011) (0.010) (0.041) (0.010)
21 0.980
(0.017)
0.018
(0.010)
10 8 7 0.933
(0.077)
0.019
(0.011)
21 0.993
(0.013)
0.020
(0.011)
11 7 4 0.909
(0.066)
0.016
(0.009)
25 0.989
(0.010)
0.018
(0.010)
15 9 6 0.905
(0.054)
0.016
(0.009)
15 1.000
(0.024)
0.019
(0.011)
6 3 1 0.733
(0.155)
0.016
(0.010)
29 0.998
(0.009)
0.021
(0.011)
9 8 7 0.972
(0.064)
0.013
(0.008)
24 0.995
(0.011)
0.029
(0.015)
12 7 4 0.894
(0.063)
0.034
(0.018)
7 1.000
(0.076)
0.023
(0.013)
2 2 2 1.000
(0.500)
0.008
(0.009)
80 0.993
(0.001)
0.021
(0.011)
115 30 12 0.930
(0.011)
0.019
(0.010)
99 0.993
(0.001)
0.021
(0.010)
164 48 24 0.946
(0.007)
0.018
(0.009)
 2010 Blackwell Publishing Ltd
50
Fig. 2 Map of the Hawaiian archipel-
THE I NTRO DUCT ION OF LUTJANUS KASMIRA TO HAW AII 1111largest sample in the comparison. From this larger
sample we randomly subsampled haplotypes (size of
subsample = N of smaller sample) with replacement
10 000 times to estimate the number of haplotypes that
would occur in the smaller sample. We compared the
distribution of the subsamples with the number of
haplotypes found in the smaller sample. P-values were
calculated based on the number of times in 10 000 subs-
amples that as many or more haplotypes were found in
the larger sample as found in the smallest. Rarefaction
curves plotting the number of individuals sampled
against the expected number of mitochondrial haplo-
types were constructed using ANALYTIC RAREFACTATION 1.4
(UGA Stratigraphy Lab website; http://www.uga.edu/
~strata/software/).
0
Kauai
2.3:1
Necker
2.3:1
Oahu
4.0:1
FFS
3.4:1
Maro
2.5:1
Midway
2.3:1
Kure
3.5:1
3.4:1
Introduction RatioThe Akaike Information Criterion in MODELTEST 3.7
(Posada & Crandall 1998) was used to determine the
mutational model that best fit the control region data.
The best fit model is TVM+I+G with equal rates for all
sites and a Ti ? Tv ratio of 10.13. Because this model is
not implemented in ARLEQUIN (Excoffier et al. 2005), the
most similar model available (Tamura & Nei 1993) was
employed using a gamma value of 0.77, a transversion
weighting of 10.13 and a transition and deletion weight
of 1.0. To test for population genetic structure in Hawaii
an L. kasmira, an analysis of molecular variance (AMOVA)
was performed in ARLEQUIN using 20 000 permutations.
An analogue of Wright?s FST (/ST), which incorporates a
model of sequence evolution, was calculated for the
entire data set and for pairwise comparisons among all
locations. We maintained a = 0.05 among all pairwise
tests by controlling for the false discovery rate as rec-
ommended by Benjamini & Yekutieli (2001) and
reviewed by Narum (2006).
 2010 Blackwell Publishing LtdNuclear introns. Observed (HO) and expected (HE) het-
erozygosities were calculated for each locus and an
exact test of Hardy-Weinberg equilibrium (HWE) using
100 000 steps in a Markov chain was performed using
ARLEQUIN. Additionally, average HE was calculated for
the multi-locus data set. Linkage disequilibrium
between the three nuclear loci was assessed using the
likelihood ratio test with 20 000 permutations in ARLE-
QUIN. FST was calculated for the entire data set and for
pairwise comparisons between locations. The false dis-
covery rate among multiple comparisons was controlled
as described above.
Tests for loss of genetic diversity in the introduced
range were conducted with BOTTLENECK 1.2.02 (Piry et al.
1999) using the infinite alleles mutation (IAM) model
0 km
Hilo
1.8:1
Kona
1.3:1
Maui Nui
2.9:1
ago. Pie chart in bottom left corner
depicts the 3.4:1 introduction ratio of
Marquesas fish (black) to Society fish
(white). Pie charts for each sample loca-
tion in Hawaii show the ratio of Lutj-
anus kasmira in either the Marquesas or
Society lineage (see Fig. 3). Hilo and
Kona are locations on opposite sides of
Hawaii Island. The figure demonstrates
that fish from both source populations
are found at each sample location and
are in roughly the same ratio as the ori-
ginal introduction ratio of 3.4:1. Abbre-
viation: FFS = French Frigate Shoals.(Kimura & Crow 1964). The loss of rare alleles was
evaluated using the mode-shift test (Piry et al. 1999).
The Wilcoxon signed rank test, which assumes that
populations in mutation-drift equilibrium have an equal
probability of heterozygote excess or deficit, was used
to detect genetic bottlenecks (Cornuet & Luikart 1996).
Interbreeding. To determine if L. kasmira Hawaiian
descendents from the Marquesas and Society Islands
are interbreeding we employed the genealogical-fre-
quency and individual-assignment methods of Nason
et al. (2002). This method assigns individuals to one of
six genealogical categories using multi-locus diploid
data. Individuals are classified as either pure parental
(P1 and P2), crosses between pure parentals (F1), crosses
between F1 individuals (F2), or backcrosses (BP1 and
BP2). The program uses maximum-likelihood estimates
to assign each individual to one of the six genealogical
classes while providing estimates of statistical power
for correct classification. For comparison, the Bayesian
statistical model developed by Anderson & Thompson
(2000), which computes the posterior probability that
an individual belongs to each of the hybrid classes (P1,
P2, F1, F2, BP1, and BP2) was employed using the pro-
gram default settings. A third method of testing for
interbreeding utilized the chi-square (v2) goodness of fit
(Sokal & Rohlf 1995) to test whether the nuclear alleles
at each locus were randomly distributed among the two
mitochondrial lineages in the introduced range.
Results
Distribution of descendents of the two source
populations
Mitochondrial control region. We resolved a 521 bp seg-
ment of the mtDNA control region in 484 individuals
yielding 218 haplotypes, with a few common haplo-
types, 123 haplotypes observed in single individuals,
and 41 haplotypes observed in two individuals. The
number of specimens (N), the number of haplotypes
(Nh), the number of haplotypes observed in single indi-
viduals (N ), h, and p per location are listed in Table 1.
found at every sample site and exclusion of these hapl-
otypes from the parsimony network did not change the
overall structure (Fig. 3). Haplotypes observed in the
introduced range (Oahu, Kona, Hilo, Maui Nui, Kauai,
Necker, French Frigate Shoals (FFS), Maro, Midway,
and Kure) are grouped with either the Marquesas or
Society lineages (Fig. 3).
The overall ratio of the number of L. kasmira samples
from the Hawaiian Islands that fell into the Marquesas
lineage to those that fell into the Society lineage was
2.3:1. This value was significantly different than the
introduction ratio of 3.4:1 (Fisher?s exact text,
P = 0.027). Among the ten sample locations scattered
across the Hawaiian Archipelago only Kona on Hawaii
Island, differed significantly from the 3.4:1 introduction
ratio, with a ratio of 1.3:1 (Fisher?s exact text, P = 0.036)
(Fig. 2). Once these specimens were removed from the
analysis the overall ratio (2.6:1) was not significantly
different than the introduction ratio (Fisher?s exact text,
P = 0.131).
Nuclear introns. We resolved 148 bp of the GH intron in
482 specimens and 168 bp of the ANT intron in 471
specimens (Table 2). Three polymorphic sites yielded
1112 M. R. GAITHER ET AL .22 stepss
There were no shared haplotypes between the two
source populations (Marquesas and Society Islands).
A statistical parsimony network demonstrated that sam-
ples from the two source populations fell into distinct
lineages separated by 22 steps (Fig. 3; d = 3.8%
between source populations). Haplotypes observed in
one or two specimens (singletons and doublets) werefour alleles at the GH locus and 13 polymorphic sites
yielded 15 alleles at the ANT locus. The GnRH3-3
intron was scored for the presence or absence of a
10 bp indel in 480 specimens. Summary statistics are
listed in Table 2.
When all locations, from the native and introduced
ranges, were grouped together there was a significant
FFS
Marquesas
Society
Kauai
Necker
Maro
Midway
Oahu
Kona
Hilo
Maui Nui
Kure
Fig. 3 Statistical parsimony network
for 484 control region sequences of Lutj-
anus kasmira constructed using TCS 2.21
(Clement et al. 2000). Each circle repre-
sents one mitochondrial haplotype with
the area of each circle is proportional to
number of that particular haplotype in
the data set; dashes represent hypotheti-
cal haplotypes; colours represent collec-
tion location (see key). There were no
shared haplotypes between the two
source populations (Marquesas and
Society) which formed two distinct lin-
eages that are separated by 22 steps
(average percent sequence diver-
gence = 3.8% between source popula-
tions). For clarification, singletons and
doublets (164 haplotypes) were omitted.
Singletons and doublets were found at
every sample site and inclusion of these
haplotypes did not change the pattern
of the parsimony network. Abbrevia-
tion: FFS = French Frigate Shoals. 2010 Blackwell Publishing Ltd
Table 2 Number of specimens (N), number of alleles (Na), heterozygosity observed (H ), heterozygosity expected (H ), and the cor-
brium
ls)
tron
a H
6 0
5 0
8 0
9 0
4 0
6 0
5 0
7 0
5 0
6 0
7 0
3 0
2 0
5 0
THE I NTRO DUCT ION OF LUTJANUS KASMIRA TO HAW AII 1113deviation from HWE (Hardy?Weinberg equilibrium)
expectations at the GH and ANT loci (P = 0.02 for
each) (Table 2). In each case an excess of homozyg-
otes was detected. When samples were divided by
archipelago (Marquesas, Society, and Hawaiian
Islands) the Marquesas and Society populations devi-
ated from HWE expectations, with an excess of ho-
responding P-value for an exact test of Hardy?Weinberg equili
are listed for the multi-locus data set (FFS = French Frigate Shoa
GH intron ANT in
N Na HO HE P N N
Source populations
Marquesas 49 3 0.66 0.59 0.57 49
Society 50 3 0.10 0.10 1.00 47
Introduced range
Oahu 49 4 0.59 0.58 0.73 49
Kona 50 4 0.74 0.60 0.14 50
Hilo 52 3 0.50 0.61 0.09 47
Maui Nui 39 3 0.56 0.61 0.48 39
Kauai 35 3 0.60 0.58 0.57 36
Necker 50 3 0.60 0.59 0.75 48
Maro 20 3 0.60 0.61 0.90 20
FFS 40 3 0.50 0.56 0.43 38
Midway 39 3 0.62 0.58 0.57 40
Kure 9 3 0.22 0.54 0.03 9
All Hawaii specimens 383 4 0.58 0.59 0.68 375 1
All specimens 482 4 0.54 0.58 0.02 471 1mozygotes, at the ANT locus (P = 0.044 and P = 0.032
respectively). No evidence of linkage disequilibrium
between pairs of nuclear loci was detected (P > 0.05)
within populations from each of the three archipela-
gos.
The number of nuclear alleles at each locus was sim-
ilar for the two source populations (Tables 2, S1) how-
ever; there were strong shifts in allele frequencies
between the Marquesas and Society Islands (Table S1).
Populations in the introduced range had allele fre-
quencies intermediate between the two source popula-
tions (Table S1). Three putative private alleles are
found in each source population (Table S1; Marque-
sas = GH3, A1, A11; Society = GH4, A4, A5). The pres-
ence of these alleles at widely separated locations in
the introduced range (Table S1) provides additional
evidence that descendents of both source populations
spread throughout the archipelago. As expected for
introduced populations of mixed lineages, many of the
Hawaiian samples had a greater number of nuclear
alleles and higher heterozygosities (HO and HE at ANT
and GnRH3-3 loci) than either of the source popula-
tions (Table 2).
 2010 Blackwell Publishing LtdPopulation structure. Pairwise comparisons indicate sig-
nificant population structure between the two source
populations (Marquesas and Society Islands) with
mtDNA /ST = 0.734 (P < 0.001) and nDNA FST = 0.49
(P < 0.001) (Table 3). The Marquesas and Society popu-
lations were also significantly different than each of the
ten Hawaiian populations (Oahu, Kona, Hilo, Maui
O E
(HWE) are listed for each nuclear intron. Na and average HE
GnRH3-3 intron Multi-locus
O HE P N Na HO HE P Na HE
.29 0.37 0.05 50 2 0.16 0.18 0.39 17 0.38
.31 0.38 0.03 50 2 0.37 0.49 0.14 16 0.32
.59 0.57 0.95 50 2 0.56 0.48 0.37 23 0.54
.64 0.64 0.39 50 2 0.54 0.47 0.37 23 0.57
.62 0.56 0.94 51 2 0.33 0.43 0.11 24 0.54
.62 0.59 0.22 39 2 0.54 0.50 0.75 19 0.57
.57 0.51 1.00 36 2 0.39 0.43 0.69 19 0.50
.58 0.59 0.31 45 2 0.49 0.49 1.00 28 0.56
.35 0.36 0.35 21 2 0.52 0.49 1.00 14 0.49
.61 0.60 0.98 39 2 0.51 0.50 1.00 22 0.55
.75 0.69 0.72 40 2 0.50 0.49 1.00 22 0.58
.89 0.66 0.12 9 2 0.33 0.50 0.49 10 0.57
.61 0.59 0.85 380 2 0.48 0.48 0.83 59 0.53
.55 0.61 0.02 480 2 0.44 0.46 0.23 65 0.53Nui, Kauai, Necker, FFS, Maro, Midway, and Kure)
(Table 2). In the introduced range, there was no popu-
lation structure detected in the mtDNA (overall /ST =
0.001, P = 0.38) or nDNA (FST = 0.001, P = 0.30) data
sets (Table 3).
Interbreeding of the two populations in the Hawaiian
Islands. The likelihood model of Nason et al. (2002)
indicated that approximately 31% of the individuals
from the Hawaiian Islands were F1 X F1 crosses (F2
genealogical class) while the remainder (69%) were F2
X P1 backcrosses (BP1 genealogical class). The program
did not assign any individual from the introduced
range to either pure parental class (P1 or P2) or to the
P1 X P2 cross (F1 geneological class). This indicates that
all assayed specimens of L. kasmira in the Hawaiian
Islands are of mixed Marquesas and Society descent.
The Anderson & Thompson (2002) model indicated
similar results to the Nason model (data not shown).
The chi-square test corroborated the findings of the
Nason et al. (2002) model, demonstrating that nuclear
allele frequencies at the GH and ANT loci were not sig-
nificantly different than expected if the alleles were
Table 3 Pairwise F-statistics for the two source populations of Lutjanus kasmira and ten populations across the introduced range.
nd p
ise t
M
7
3
3 )
9 )
)0.002 )0.005 )0.002 0.007 0.016 )0.004 )0.000
8 ? 0.008 0.001 0.001 0.007 )0.006 0.010
1 0.000 ? 0.005 0.025 )0.001 0.011 0.025
2 )0.007 )0.006 ? 0.000 0.018 )0.006 )0.014
9 )0.004 0.003 )0.011 ? 0.047 )0.008 )0.020
9 )0.004 0.003 )0.019 )0.025 ? 0.028 0.066
5 )0.007 )0.005 )0.007 0.003 )0.006 ? )0.014
2 )0.036 )0.054 )0.037 )0.042 )0.041 )0.041 ?
ench Frigate Shoals.
Table 4 Results of rarefaction analyses. Number of specimens
(N), number of haplotypes (Nh), and mean number of haplo-
types (H) (?standard deviation) estimated from 10 000 random
subsamples (N = number of individuals sampled in the source
population) of the Hawaiian lineages are listed. The % lost is
the reduction in haplotypes when comparing the source popu-
lation to the corresponding Hawaiian lineage. P-values reflect
the number of times in 10 000 subsamples that as many (or
more) haplotypes, that were found in the source population,
were also found in the Hawaiian lineage. The difference in loss
of haplotypes (17% vs. 41%) was marginally significant (Fish-
er?s exact test, P = 0.058)
1114 M. R. GAITHER ET AL .randomly distributed between the Marquesas
(v2 = 0.227, P = 0.99; v2 = 0.157, P = 0.99 respectively)
and Society (v2 = 0.256, P = 0.96; v2 = 0.803, P = 0.79)
mitochondrial lineages in Hawaii. The Society mito-
chondrial lineage in Hawaii deviated significantly from
a random distribution of alleles at the GnRH3-3 locus
(v2 = 4.718, P = 0.03) while the Marquesas mitochon-
drial lineage did not (v2 = 1.729, P = 0.19).
Genetic consequence of founder event. The Marquesas and
Society populations had high mtDNA haplotype diver-
Pairwise /ST values for control region data are below diagonal a
diagonal. We maintained an alpha value of 0.05 among all pairw
by Benjamini & Yekutieli (2001) and reviewed by Narum (2006)
Sample Location
Source Populations
Marquesas Society Oahu Kona Hilo
Marquesas ? 0.490 0.089 0.089 0.06
Society 0.734 ? 0.295 0.248 0.29
Oahu 0.081 0.525 ? )0.005 )0.00
Kona 0.224 0.339 0.042 ? )0.00
Hilo 0.187 0.395 0.021 0.003 ?
Maui Nui 0.107 0.503 )0.006 0.022 0.00
Kauai 0.149 0.466 0.005 0.011 )0.01
Necker 0.130 0.455 )0.003 0.005 0.00
FFS 0.083 0.527 )0.011 0.033 0.01
Maro 0.117 0.532 )0.006 0.013 0.00
Midway 0.125 0.401 0.004 0.005 )0.00
Kure 0.119 0.575 )0.037 )0.004 )0.03
Values in bold are significant at the corrected a = 0.010. FFS = Frsity (h = 0.997 and 0.970 respectively). All 10 popula-
tions in the introduced range (Oahu, Kona, Hilo, Maui
Nui, Kauai, Necker, FFS, Maro, Midway, and Kure) had
similarly high h values (h = 0.985?1.000; ?All Data?
Table 1). Using the parsimony network in Fig. 3 we
divided the Hawaiian samples into either the Marque-
sas or Society mitochondrial lineage (Table 1). The
mtDNA haplotype diversity values in the Hawaiian
Islands ranged from 0.989 to 1.00 for the Marquesas
lineage and from 0.733 to 1.00 for the Society lineage.
Due to the lower sensitivity of heterozygosity to
losses of genetic diversity (Nei et al. 1975) we restricted
our statistical comparisons of diversity loss to haplotype
richness which we compare at the archipelago level. We
observed 47 haplotypes in the Marquesas (N = 50) and
31 haplotypes in the Society Islands (N = 49) (Table 4).
By creating haplotype frequency distributions for the
larger Hawaiian sample sets and by randomly subsam-
pling (10 000 times with replacement) these populations
we found evidence of a small but significant decrease
(17%) in haplotypes from the Marquesas lineage inairwise FST values for the multi-locus nuclear data set are above
ests by controlling for the false discovery rate as recommended
Introduced Range
aui Nui Kauai Necker FFS Maro Midway Kure
0.102 0.039 0.104 0.144 0.060 0.114 0.163
0.303 0.356 0.266 0.233 0.434 0.250 0.184
0.003 0.001 )0.004 0.002 0.009 )0.002 0.007
0.007 0.002 )0.009 )0.004 0.020 )0.007 )0.014Hawaii which had a mean of 39.0 haplotypes per sub-
sample (N = 50, P-value <0.001). A greater decrease
(41%) was detected in the Society lineage in Hawaii
with a mean of 18.3 haplotypes (N = 49, P-value
<0.001). The difference in loss of haplotypes (17% vs.
41%) was marginally significant (Fisher?s exact test,
P = 0.058) (Table 4).
Rarefaction curves, that plotted the number of indi-
viduals sampled against the expected number of mito-
chondrial haplotypes, were constructed (Fig. 4).
Population N Nh H % lost P-value
Marquesas
Source 50 47
Hawaiian
lineage
270 142 39.0 ? 2.50 17.0% <0.001
Society
Source 49 31
Hawaiian
lineage
115 30 18.3 ? 2.06 41.0% <0.001
 2010 Blackwell Publishing Ltd
160
THE I NTRO DUCT ION OF LUTJANUS KASMIRA TO HAW AII 11150
20
40
60
80
100
120
140
0 20 40 60 80 100 120 140 160 180 200 220 240 260
0
5
10
15
20
25
30
35
0 10 20 30 40 50 60 70 80 90 100 110
Number of individuals sampled
N
um
be
r o
f h
ap
lo
ty
pe
s
A
BSamples from the introduced range were separated by
mitochondrial lineage and compared with their respec-
tive source population (Fig. 4). Due to the large confi-
dence intervals (95%) there was no significant
difference between expected number of mtDNA haplo-
types in the native and introduced ranges at low sam-
ple sizes. However, as sample size increased (N > 40)
the curves no longer over-lapped and a loss of mtDNA
haplotypes in the introduced range became evident in
the Society, but not the Marquesas lineage (Fig. 4).
Using the nuclear allele frequency data (Table S1)
there was no evidence of a genetic bottleneck in any of
the introduced populations using the Wilcoxon signed
rank test or the mode shift test implemented in BOTTLE-
NECK. However, it should be noted that three loci do not
provide high levels of power for these analyses and
due to the presence of shared alleles between the source
populations and admixture of lineages in the intro-
duced range we did not attempt to statistically compare
allelic richness values using the nuclear data set.
Fig. 4 Rarefaction curves plotting the number of individuals
sampled against the expected number of mitochondrial haplo-
types were calculated using the Analytic Rarefactation1.4 soft-
ware available at the UGA Stratigraphy Lab website (http://
www.uga.edu/~strata/software/). Samples belonging to the
Marquesas (A) and Society (B) lineages are plotted separately.
Grey lines represent data for source populations, black lines
represent data from the introduced range, and solid lines are
95% confidence intervals. The Society lineage in the introduced
range is significantly different than the source population indi-
cating a loss of rare haplotypes.
 2010 Blackwell Publishing LtdDiscussion
Establishment and spread of L. kasmira throughout
the Hawaiian Islands
The introduction of L. kasmira to the Hawaiian Islands
from two populations with diagnostic differences in the
mitochondrial genome (Fig. 3) and several private
nuclear alleles (Table S1) enables us to trace the fate of
their descendents in the introduced range. Individuals
from both source populations became established in the
Hawaiian Islands and both mtDNA lineages are found
on every island and atoll of the archipelago (Table S1,
Figs 2 and 3). The mtDNA data (Fig. 2) indicate that
the overall ratio of the two lineages across the Hawaiian
Islands is less than the 3.4:1 introduction ratio.
Although one Hawaii Island population had a ratio sig-
nificantly less than 3.4:1, the overall ratio among the
remaining populations (2.6:1) was not significant, albeit
with a tendency in the same direction.
Hawaii Division of Fish and Game (HDFG) records
show that the initial 2431 Marquesas fish began to
reproduce and spread very soon after the initial release.
When the 728 Society fish were introduced, 3 years after
the Marquesas fish, the former had already spread to
Hawaii Island about 500 km to the southeast of Oahu.
This, in combination with the fact that maturity occurs
at 1?2 years of age, indicates that the ratio at the time
of the introduction of Society fish likely was greater
than 3.4:1. After the 3 years head-start and rapid spread
of the Marquesas lineage, it might be predicted that its
descendents now would be proportionately more
numerous and widespread than those of the Society
lineage. This is not the case, however. Instead the data
indicate that numerically the Society lineage was able to
?catch up? with, and even surpass, the Marquesas line-
age, most notably on Hawaii Island. Given the esti-
mated time of divergence between these two source
populations (approximately half a million years) and
the differences in their native environments (Gaither
et al. 2010) it is possible that population specific adapta-
tions endowed Society-lineage fish with a higher fitness
in the Hawaiian environment. However, such an advan-
tage would be quickly lost by interbreeding. Another
likely advantage the Society lineage had over the Mar-
quesas lineage at the time of introduction was the pres-
ence of an established population when the former
were released. For the Society lineage this could have
alleviated many of the adverse consequences associated
with small population size, such as difficulty in finding
suitable mates.
The genetic data indicate that there is now a single
population of L. kasmira in the Hawaiian Islands. We
found no population structure in the mitochondrial data
set across the archipelago, and only one of 45 pairwise
or to prevent hybridization in trumpetfish species (Bo-
wen et al. 2001). Populations which diverge sufficiently
1116 M. R. GAITHER ET AL .comparisons of the nuclear data set was significant after
control for false discovery rate. The lack of genetic
structure coupled with the maintenance of genetic
diversity across the archipelago implies that there was
little or no loss of genetic lineages, as would be
expected under a stepping stone model of colonization,
as the fish spread through the islands. Instead our data
indicate that either L. kasmira colonized each island in
large enough numbers to capture most of the standing
genetic diversity, or gene flow between the islands is
sufficient to homogenize the geographic distribution of
the genetic diversity, or both.
The success of L. kasmira in Hawaii, as indicated by
HDFG catch records and corroborated here by our
genetic data, is especially notable because most other
introductions of reef fishes to the Hawaiian Islands
have failed. In the 1950s the Hawaiian Division of Fish
and Game introduced 11 non-native snappers and
groupers (Oda & Parrish 1982; Randall 1987). Six of the
eleven species were introduced in numbers greater than
1500 individuals (HDFG records) but 50 years later only
three are regularly recorded in Hawaiian waters.
Besides Lutjanus kasmira, these include the Blacktail
Snapper (Lutjanus fulvus) and the Peacock Grouper (Ce-
phalopholus argus). Notably, neither L. fulvus nor C. argus
has colonized north of French Frigate Shoals (FFS;
Fig. 2). While L. fulvus is not a common fish in the
lower Hawaiian Islands, C. argus is more common there
than in its natural range (Meyer 2008).
In the field of invasion biology, an intense debate
revolves around the factors that promote successful col-
onization of new habitat (Kolar & Lodge 2001; Allen-
dorf & Lundquist 2003), particularly the genetic factors
(Frankham 2005; Golani et al. 2007; Zayed et al. 2007).
Two of the primary factors that are thought to contrib-
ute to invasion success are large founder populations
and multiple introduction events (Lockwood et al. 2005;
Colautti et al. 2006). Certainly these conditions apply to
the introduction of L. kasmira in Hawaii, and it is possi-
ble that introductions of two genetically distinct popula-
tions have yielded a more robust fish than either
parental stock (hybrid vigour; Allendorf & Luikart
2007). Other traits that might apply specifically to the
invasion success of L. kasmira include mass spawning
(Suzuki & Hioki 1979), broad habitat preference (2?
265 m depth; Randall 1987), and a generalist diet (Ran-
dall & Brock 1960; Oda & Parrish 1982). Range-wide
mtDNA surveys also indicate that this species has a
much more dispersive larval stage than either Lutjanus
fulvus or Cephalopholus argus (Gaither et al. 2010; unpub-
lished data) which may explain why L. kasmira has
swiftly colonized the entire archipelago, while the other
two species have not.in allopatry might resume mating upon secondary con-
tact but resulting offspring could have lower fitness than
purebred offspring. Reinforcement theory predicts that
due to lower fitness of hybrids, natural selection will
favour the evolution of prezygotic isolating mechanisms
to maximize fitness, and therefore drive further diversifi-
cation (Coyne & Orr 2004). We see no evidence for rein-
forcement in the hybridization tests we performed here.
Notably, our study was conducted approximately thir-
teen generations after the initial introduction (see Mate-
rials and methods for references). If reproductive
barriers existed or preferential mating occurred during
initial contact of these two lineages, the genetic signature
has been lost. Furthermore, the mass spawning behav-
iour of this species (Suzuki & Hioki 1979) may have
reduced the potential for assortative mating by eliminat-
ing active mate choice.
Genetic consequences of founding events
Contrary to expectations, alien species often retain high
levels of genetic diversity in their introduced range
(Bossdorf et al. 2005; Wares et al. 2005; Roman & Dar-
ling 2007). In cases where individuals from genetically
divergent source populations are introduced to the
same region (admixture), there may actually be an
increase in genetic diversity in the introduced range
(Kolbe et al. 2004; Genton et al. 2005; Roman & Darling
2007). This is the case for L. kasmira in the Hawaiian
Islands. At the archipelago level, L. kasmira in Hawaii
exhibit similar or slightly higher heterozygosities and a
similar or greater number of nuclear alleles than either
source population. The only possible exception to this
pattern (Kure) may simply be an artefact due to small
sample size (N = 9).Interbreeding and outbreeding
The Marquesas and Society source populations of L. kas-
mira demonstrate an average mitochondrial control
region sequence divergence of d = 3.8%. The control
region appears to evolve at 3?10% per million years in
shore fishes (Bowen et al. 2006; Lessios et al. 2008). Using
this range as a first order approximation we estimate
that these two populations have been separated for
about half a million years (380 000?1 300 000 years), a
value that overlaps the estimate from cytochrome b data
from the same samples (265 000?530 000 years; Gaither
et al. 2010). This time interval is sufficient to produce
gamete incompatibility in allopatric populations of sea
urchins (Lessios 1984; Palumbi & Metz 1991). However,
3?4 Myr is insufficient to prevent gamete compatibility
in geminate species of gobies (Rubinoff & Rubinoff 1971) 2010 Blackwell Publishing Ltd
Introductions that involve a large number of individ-
uals (high propagule pressure) are less likely to suffer
the loss of rare alleles and heterozygosity associated
with founder events (Lockwood et al. 2005). What is
unclear is how many individuals are required to pre-
vent such a loss of genetic diversity. The answer to this
question is dependent on both the genetic diversity of
the taxa involved and patterns of survivorship follow-
ing the founder event. Dlugosch & Parker (2008)
reviewed 80 surveys of molecular variation in intro-
duced species. These include 11 cases of intentional
introduction where the number of individuals is confi-
dently known and derived from a single source popula-
tion (Table 5). In these eleven cases, which cover a
variety of taxa, a loss of genetic diversity was detected
in all but one case involving the introduction of less
than 250 individuals. The introduction of 2385 Peacock
Groupers (Cephalopholis argus) to the Hawaiian Islands
(Planes & Lecaillon 1998) is the only example in this
review that involved a founder population of greater
than 250 individuals (Dlugosch & Parker 2008) and for
this species the authors found no loss of genetic diver-
sity in the introduced range. As with C. argus, we found
that 2431 L. kasmira were sufficient to prevent a major
loss of mtDNA haplotypes (17%) (Marquesas Lineage,
Table 1). In contrast, with a founder population size of
728 individuals (Society Lineage, Table 1), we detected
a larger decrease in haplotype richness (41%) indicat-
ing, that at least for L. kasmira introduced to Hawaii,
this founder size is at the level where we begin to
detect substantial losses of genetic diversity. This con-
clusion should be tempered with the recognition that
Table 5 Table is modified from Dlugosch & Parker (2008). Studies of molecular variation in eleven intentionally introduced species.
Only cases in which all individuals were derived from a single source population and the number of individuals released is confi-
dently known are listed. Locations indicate the regions that served as the source (S) and introduced (I) areas. Number of individuals
introduced (NI) and marker type (number of loci is in parentheses) are listed. Values for allelic richness (A) and expected heterozy-
gosity (HE) are averages per locus and population
Organism Location (S ? I) NI Marker A (S ? I) HE (S ? I) Reference
Birds
Common Myna India ? Australia 250 allozymes (21) 1.43 ? 1.30 0.06 ? 0.06 Baker &
Moeed 1987Acridotheres tristis
Eurasian Tree Sparrow Germany ? United States 20 allozymes (39) 1.50 ? 1.33 0.101 ? 0.078 St. Louis &
Barlow 1988Passer montanus
Reptiles
Jamaican Anole Jamaica ? Bermuda 71 allozymes (24) 1.75 ? 1.50 0.078 ? 0.064 Taylor &
Gorman 1975Anolis grahami
Mammals
Red-necked Wallaby Australia ? New Zealand 6?10 microsatellites (5) 8.4 ? 4.6 0.767 ? 0.586 Le Page et al. 2000
Macropus rufogriseus
Caribou Norway ? Iceland 35 allozymes (1) 8.0 ? 3.0 0.729 ? 0.332 Roed et al. 1985
Rangifer tarandus
Javan Rusa Deer New Caledonia ? Australia 7 microsatellites
(10,24)
7.60 ? 2.29 0.595 ? 0.467 Bonnet et al. 2002
Webley et al. 2004Cervus timorensis russa
Insects
all
m
all
all
m
n to e
from
as 5
urce
THE I NTRO DUCT ION OF LUTJANUS KASMIRA TO HAW AII 1117Mountain Butterfly E ? W Sudetans
(Czechia)
50*
Erebia epiphron silesiana
Amphibians
Marsh Frog Hungary ? Britain 12
Rana ridibunda
Crustaceans
Signal Crayfish Canada (Pitt Lake) ?
Sweden
200
Pacifastacus leniusculus
Fish
Peacock Grouper Society ? Hawaii 2385?
Cephalopholis argus
European Grayling NW Europe ?
Norway
?a small
number?Thymallus thymallus
*50 inseminated females were translocated. This species is know
would render that effective population size higher than expected
?The number of individuals released is reported in this reference
include the 1814 fish that were released in 1961 from the same so 2010 Blackwell Publishing Ltdozymes (17) 1.59 ? 1.47 0.100 ? 0.116 Schmitt et al. 2005
icrosatellites (5) 3.2 ? 2.2 0.522 ? 0.484 Zeisset &
Beebee 2003
ozymes (4) 1.50 ? 1.25 0.177 ? 0.079 Agerberg &
Jansson 1995
ozymes (9) 4.00 ? 3.78 0.046 ? 0.045 Planes &
Lecaillon 1998
icrosatellites (17) 3.75 ? 1.90 0.435 ? 0.170 Koskinen
et al. 2002a, b
ngage in multiple inseminations and to store sperm which
the census size.
71 (released in 1956). This number has been corrected here to
population (HDFG records).
even a small number of reproductive adults can retain
the case for marine fishes (Grant & Bowen 1998; 2006).
great numbers and with robust genetic diversity.
ported by research expeditions funded and staffed by the
Papahanaumokuakea Marine National Monument. We thank
1118 M. R. GAITHER ET AL .Acknowledgements
This study began with support from the National Geographic
Society (grant No. NGS 7269-02 to DRR), and was further sup-Conclusion
The introduction of L. kasmira to the Hawaiian Islands
is a remarkable case study for two reasons. First, the
introduction of this species occurred in two discrete
and well documented events with known chronology,
numbers, and source populations. Second, this species
was introduced from two genetically distinct popula-
tions at the Marquesas and Society Islands, allowing us
to trace the fate of their descendents in the introduced
range. Using mitochondrial and nuclear sequence data
we determined that individuals from both source popu-
lations became established in the archipelago, inter-
breed, and their descendents have colonized each
island and atoll surveyed (Figs 2 and 3). We found that
2431 L. kasmira were sufficient to prevent a substantial
loss of mtDNA diversity while 728 individuals resulted
in a 41% decrease in haplotype richness.
Previous reports document that L. kasmira colonized
from the inhabited Main Hawaiian Islands to the farthest
north-western (NW) Hawaiian Islands, a distance of over
2000 km in just 34 years (Oda & Parrish 1982; Randall
1987). More recently, a range-wide genetic survey of
L. kasmira demonstrated exceptional dispersal ability in
this species (Gaither et al. 2010). Here we conclude that
the rapid colonization across the NW Hawaiian Island
was accompanied by maintenance of high levels of
genetic diversity, indicating large numbers of colonists at
every island along the way. The NW Hawaiian Islands
now one of the largest marine protected areas in the
world (Papahanaumokuakea Marine National Monu-
ment), and subject to large-scale efforts to prevent and
eradicate alien introductions. In these circumstances,
managers need to know whether the 350+ marine exotics
in the inhabited Main Hawaiian Islands pose a threat to
the nearly pristine habitats of the NW Hawaiian Islands.
Our data indicate that highly dispersive species such as
L. kasmira may prove to be the most effective invaders,
and add a new layer to the findings of Oda & Parrish
(1982) and Randall (1987); not only can exotic species
jump to the NW Hawaiian Islands, they can do so ingenetic diversity over the short term (Spencer et al.
2000) and that even very large populations may not
retain genetic diversity if there is a high variance in
reproductive success (Hedgecock 1994), as is generallyReferences
Agerberg A, Jansson H (1995) Allozymic comparisons between
three subspecies of the freshwater crayfish Pacifastacus
leniusculus (Dana), and between populations introduced to
Sweden. Hereditas, 122, 33?39.
Allendorf FW, Luikart G (2007) Conservation and the Genetics of
Populations. Blackwell Publishing, New York.
Allendorf FW, Lundquist LL (2003) Introduction: population
biology, evolution, and control of invasive species.
Conservation Biology, 17, 24?30.
Anderson EC, Thompson EA (2002) A model-based method for
identifying species hybrids using multilocus genetic data.
Genetics, 160, 1217?1229.
Baker AJ, Moeed A (1987) Rapid genetic differentiation and
founder effect in colonizing populations of common mynas
(Acridotheres tristis). Evolution, 41, 525?538.
Benjamini Y, Yekutieli D (2001) The control of the false
discovery rate in multiple testing under dependency. Annals
of Statistics, 29, 1165?1188.
Bonnet A, Thevenon S, Maudet F, Maillard JC (2002) Efficiency
of semi-automated fluorescent multiplex PCRs with 11
microsatellite markers for genetic studies of deer
populations. Animal Genetics, 33, 343?350.
Bossdorf O, Auge H, Lafuma L, Rogers WE, Siemann E, Prati D
(2005) Phenotypic and genetic differentiation between native
and introduced plant populations. Oecologia, 144, 1?11.
Bowen BW, Bass AL, Rocha LA, Grant WS, Robertson DR
(2001) Phylogeography of the trumpetfish (Aulostomus spp.):Monument co-trustees NOAA Marine Sanctuaries, U.S. Fish
and Wildlife Service (USFWS), and the State of Hawaii. This
research was supported by NOAA National Marine Sanctuaries
Program MOA No. 2005-008 ? 66882 (BWB & RJT), Hawaii
Sea Grant Nos. NA05OAR4171048 and NA09OAR4170060
(BWB), National Science Foundation (NSF) grants No.
OIA0554657, OCE-0453167, OCE-0929031 (BWB), and OCE-
0623678 (RJT), and the Evolution, Ecology, and Conservation
Biology (EECB) specialization at the University of Hawaii. The
views expressed herein are those of the authors and do not nec-
essarily reflect the views of NSF, NOAA USFWS, the State of
Hawaii, or any of their sub-agencies. Greg Concepcion, Matt
Craig, Jonathan Dale, Toby Daly-Engel, Jeff Eble, Matt Iacchei,
Stephen Karl, Randall Kosaki, Carl Meyer, Yannis Papastama-
tiou, Luiz Rocha, Zoli Szabo, Jill Zamzow, Joshua Reece, and
the crew of the R.V. Hi?ialakai helped collect specimens. Aulani
Wilhelm, Jo-Ann Leong, Hoku Johnson, Danielle Carter, Daniel
Polhemus, Randall Kosaki, Ann Mooney, Elizabeth Keenen, and
Kelly Gleason provided crucial logistic assistance to this project.
We thank Marc Crepeau for valuable laboratory assistance and
protocol development, Sarah Daley, Rajesh Shrestha and Mindy
Mizobe of the HIMB EPSCoR core facility for their assistance
with DNA sequencing, and Shelley Jones and all the members
of the ToBo lab and staff at HIMB for their assistance, support
and feedback throughout this project. This is contribution
#1371 from the Hawaii Institute of Marine Biology, #7829 from
the School of Ocean and Earth Science and Technology and #JC-
06-30 from the University of Hawaii Sea Grant Program. 2010 Blackwell Publishing Ltd
ring species complex on a global scale. Evolution, 55, 1029? Grant WS, Bowen BW (2006) Living in a tilted world: climate
THE I NTRO DUCT ION OF LUTJANUS KASMIRA TO HAW AII 11191039.
Bowen BW, Muss A, Rocha LA, Grant WS (2006) Shallow
mtDNA coalescence in Atlantic pygmy angelfishes (genus
Centropyge) indicates a recent invasion from the Indian
Ocean. Journal of Heredity, 97, 1?12.
Cameron EK, Bayne EM, Coltman DW (2008) Genetic structure
of invasive earthworms Demdrpbaema octaedra in the boreal
forest of Alberta: insights into introduction mechanisms.
Molecular Ecology, 17, 1189?1197.
Clement M, Posada D, Crandall KA (2000) TCS: a computer
program to estimate gene genealogies. Molecular Ecology, 9,
1657?1659.
Colautti RI, Grigorovich IA, MacIsaac HJ (2006) Propagule
pressure: a null model for biological invasions. Biological
Invasions, 8, 1023?1037.
Coyne JA, Orr HA (2004) Speciation. Sinauer Associates, Inc.,
Sunderland, MA.
Cornuet JM, Luikart G (1996) Description and power analysis
of two tests for detecting recent population bottlenecks from
allele frequency data. Genetics, 144, 2001?2014.
Craig MT, Eble JA, Bowen BW, Robertson DR (2007) High
genetic connectivity across the Indian and Pacific Oceans in
the reef fish Myripristis berndti (Holocentridae). Marine
Ecology Progress Series, 334, 245?254.
Denit K, Sponaugle S (2004) Growth variation, settlement,
and spawning of gray snapper across a latitudinal
gradient. Transactions of the American Fisheries Society, 133,
1339?1355.
Dlugosch KM, Parker IM (2008) Founding events in species
invasions: genetic variation, adaptive evolution, and the role
of multiple introductions. Molecular Ecology, 17, 431?449.
Eble JA, Toonen RJ, Bowen BW (2009) Endemism and
dispersal: comparative phylogeography of three surgeonfish
species across the Hawaiian Archipelago. Marine Biology,
156, 689?698.
Eldredge LG, Smith CM (2001) A Guidebook of Introduced Marine
Species in Hawaii. Bishop Museum Technical Report 21.
Bishop Museum, Honolulu, HI.
Excoffier L, Laval G, Schneider S (2005) Arlequin ver. 3.0: an
integrated software package for population genetics data
analysis. Evolutionary Bioinformatics Online, 1, 47?50.
Frankham R (2005) Invasion biology: resolving the genetic
paradox in invasive species. Heredity, 94, 385?385.
Gaither MR, Toonen RJ, Robertson DR, Planes S, Bowen BW
(2010) Genetic evaluation of marine biogeographic barriers:
perspectives from two widespread Indo-Pacific snappers
(Lutjanus kasmira and Lutjanus fulvus). Journal of Biogeography,
37, 133?147.
Genton BJ, Shykoff A, Giraud T (2005) High genetic diversity
in French invasive populations of common ragweed,
Ambrosia artemisiifolia, as a result of multiple sources of
introduction. Molecular Ecology, 14, 4275?4285.
Golani D, Azzurro E, Corsini-Foka M, Falautano M, Andaloro
F, Bernardi G (2007) Genetic bottlenecks and successful
biological invasions: the case of a recent Lessepsian migrant.
Biology Letters, 3, 541?545.
Grant WS, Bowen BW (1998) Shallow population histories in
deep evolutionary lineages of marine fishes: insights from
the sardines and anchovies and lesions for conservation.
Journal of Heredity, 89, 415?426. 2010 Blackwell Publishing Ltdchange and geography limit speciation in Old World
anchovies (Engraulis; Engraulidae). Biological Journal of the
Linnean Society, 88, 673?689.
Hassan M, Harmelin-Vivien M, Bonhomme F (2003)
Lessepsian invasion without bottleneck: example of two
rabbitfish species (Siganus rivulatus and Siganus luridus).
Journal of Experimental Marine Biology and Ecology, 291, 219?
232.
Hassan M, Lemaire C, Fauvelot C, Bonhomme F (2002)
Seventeen new exon-primed intron-crossing polymerase
chain reaction amplifiable introns in fish. Molecular Ecology
Notes, 2, 334?340.
Hedgecock D (1994) Does variance in reproductive success
limit effective population sizes of marine organisms? In:
Genetics and Evolution of Aquatic Organisms (ed. Beaumont
AR), pp. 122?134. Chapman and Hall, London.
Kimura M, Crow JF (1964) The number of alleles that can be
maintained in a finite population. Genetics, 49, 725?738.
Kolar CS, Lodge DM (2001) Progress in invasion biology:
predicting invaders. Trends in Ecology & Evolution, 16, 199?204.
Kolbe JJ, Glor RE, Schettino LR, Lara AC, Larson A, Losos JB
(2004) Genetic variation increases during biological invasion
by a Cuban lizard. Nature, 431, 480?483.
Koskinen MT, Haugen TO, Primmer CR (2002)
Contemporary Fisherian life-history evolution in small
salmonid populations. Nature, 419, 826?830.
Koskinen MT, Nilsson J, Veselov AJ, Potutkin AG, Ranta E,
Primmer CR (2002) Microsatellite data resolve
phylogeographic patterns in European grayling, Thymallus
thymallus, Salmonidae. Heredity, 88, 391?401.
Leberg PL (2002) Estimating allelic richness: effects of sample
size and bottlenecks. Molecular Ecology, 11, 2445?2449.
Le Page SL, Livermore RA, Cooper DW, Taylor AC (2000)
Genetic analysis of a documented population bottleneck:
Introduced Bennett?s wallabies (Macropus rufogriseus
rufogriseus) in New Zealand. Molecular Ecology, 9, 753?763.
Lee W-J, Conroy J, Howell WH, Kocher TD (1995) Structure
and evolution of teleost mitochondrial control regions.
Journal of Molecular Evolution, 41, 54?66.
Lessios HA (1984) Possible prezygotic reproductive isolation in
sea urchins separated by the Isthmus of Panama. Evolution,
38, 1144?1148.
Lessios HA (2008) The great American schism: divergence of
marine organisms after the rise of the Central American
Isthmus. Annual Review of Ecology, Evolution, & Systematics,
39, 63?91.
Lockwood JL, Cassey P, Blackburn T (2005) The role of propagule
pressure in explaining species invasions. Trends in Ecology &
Evolution, 20, 223?228.
Maddison DR, Maddison WP (2002) MacClade 4: Analysis
of Phylogeny and Character Evolution. Sinauer Associates,
Sunderland.
Meyer A (2008) An ecological comparison of Cephalopholis argus
between native and introduced populations. PhD Thesis,
University of Hawaii at Manoa, Honolulu.
Morales-Nin B, Ralston S (1990) Age and growth of Lutjanus
kasmira (Forskal) in Hawaiian waters. Journal of Fish Biology,
36, 191?203.
Narum SR (2006) Beyond Bonferroni: less conservative analyses
for conservation genetics. Conservation Genetics, 7, 783?787.
Nason JD, Heard SB, Williams FR (2002) Host-associated Rubinoff RW, Rubinoff I (1971) Geographic and reproductive
1120 M. R. GAITHER ET AL .genetic differentiation in the goldenrod elliptical-gall moth,
Gnorimoschema gallaesolidaginis (Lepidoptera: Gelechiidae).
Evolution, 56, 1475?1488.
Nei M (1987) Molecular Evolutionary Genetics. Columbia Univer-
sity Press, New York.
Nei M, Maruyama T, Chakraborty R (1975) The bottleneck
effect and genetic variability in populations. Evolution, 29,
1?10.
Oda DK, Parrish JD (1982) Ecology of commercial snappers
and groupers introduced to Hawaiian reefs. Proceedings of the
Fourth International Coral Reef Symposium, Manila, 59?67.
Palumbi SR, Metz EC (1991) Strong reproductive isolation
between closely related tropical sea urchins Genus
Echinometra. Molecular Biology & Evolution, 8, 227?239.
Piry S, Luikart G, Cornuet JM (1999) Bottleneck: a computer
program for detecting recent reductions in the effective
population size using allele frequency data. Journal of
Heredity, 90, 502?503.
Planes S, Fauvelot C (2002) Isolation by distance and vicariance
drive genetic structure of a coral reef fish in the Pacific
Ocean. Evolution, 56, 378?399.
Planes S, Lecaillon G (1998) Consequences of the founder effect
in the genetic structure of introduced island coral reef fish
populations. Biological Journal of the Linnean Society, 63, 537?
552.
Posada D, Crandall KA (1998) MODELTEST: testing the model
of DNA substitution. Bioinformatics, 14, 817?818.
Randall JE (1987) Introductions of marine fishes to the
Hawaiian Islands. Bulletin of Marine Science, 47, 356?400.
Randall JE (2001) Four new cardinalfishes (Perciformes:
Apogonidae) from the Marquesas Islands. Pacific Science, 55,
47?64.
Randall JE (2007) Reef and Shore Fishes of the Hawaiian Islands.
Sea Grant College Program, University of Hawaii, Honolulu.
Randall JE, Brock VE (1960) Observation on the ecology of
epinepheline and lutjanid fishes of the Society Islands, with
emphasis on food habits. Transactions of the American Fisheries
Society, 89, 9?16.
Randall JE, Earle JL, Pyle RL, Parrish JD, Hayes T (1993)
Annotated checklist of the fishes of Midway Atoll,
Northwestern Hawaiian Islands. Pacific Science, 47, 356?400.
Rangarajan K (1971) Maturity and spawning of the snapper,
Lutjanus kasmira (Forskal) from the Andaman Sea. Indian
Journal of Fisheries, 18, 114?125.
Rocha LA, Bowen BW (2008) Speciation in coral-reef fishes.
Journal of Fish Biology, 72, 1101?1121.
Roed KH, Soldal AV, Thorisson S (1985) Transferring
variability and founder effect in Iceland reindeer, Rangifer
tarandus. Hereditas, 103, 161?164.
Rogers SM, Bernatchez L (2006) The genetic basis of intrinsic
and extrinsic postzygotic isolation jointly promoting
speciation in the lake whitefish species complex (Coregonus
clupeafornis). Journal of Evolutionary Biology, 19, 1979?1994.
Roman J, Darling JA (2007) Paradox lost: genetic diversity and
the success of aquatic invasions. Trends in Ecology &
Evolution, 22, 454?464.
Rosenthal DM, Ramakrishnan AP, Cruzan MB (2008) Evidence
for multiple sources of invasion and intraspecific
hybridization in Brachypodium sylvaticum (Hudson) Beauv. in
North America. Molecular Ecology, 17, 4657?4669.isolation in Atlantic and Pacific populations of Panamanian
Bathygobius. Evolution, 25, 88?97.
Schmitt T, Cizek O, Konvicka M (2005) Genetics of a butterfly
relocation: large, small and introduced populations of the
mountain endemic Erebia epiphron silesiana. Biological
Conservation, 123, 11?18.
Schultz JK, Pyle RL, DeMartini E, Bowen BW (2007) Genetic
connectivity among color morphs and Pacific archipelagos
for the flame anglefish, Centropyge loriculus. Marine Biology,
151, 167?175.
Schumacher BD, Parrish JD (2005) Spatial relationships
between an introduced snapper and native goatfishes on
Hawaiian reefs. Biological Invasions, 7, 925?933.
Seutin G, White BN, Boag PT (1991) Preservation of avian and
blood tissue samples for DNA analyses. Canadian Journal of
Zoology, 69, 82?92.
Sokal RR, Rohlf JF (1995) Biometry: The Principles and Practice of
Statistics in Biological Research. W.H. Freeman and Company,
New York.
Spencer CC, Neigel JE, Leberg PL (2000) Experimental
evaluation of the usefulness of microsatellite DNA for
detecting demographic bottlenecks. Molecular Ecology, 9,
1517?1528.
St Louis VL, Barlow JC (1988) Genetic differentiation among
ancestral and introduced populations of the Eurasian tree
sparrow (Passer montanus). Evolution, 42, 266?276.
Suzuki K, Hioki S (1979) Spawning behavior, eggs, and larvae
of the lutjanid fish, Lutjanus kasmira, in an aquarium. Japanese
Journal of Ichthyology, 26, 161?165.
Tamura K, Nei M (1993) Estimation of the number of nucleotide
substitutions in the control region of mitochondiral DNA in
humans and chimpanzees. Molecular Biology & Evolution, 10,
512?526.
Taylor CE, Gorman GC (1975) Population genetics of a
colonizing lizard: natural selection of allozyme morphs in
Anolis grahami. Heredity, 35, 241?247.
Thompson JD, Higgins DG, Gibson TJ (1994) CLUSTAL W:
improving the sensitivity of progressive multiple sequence
alignment through sequence weighting, position-specific gap
penalties and weight matrix choice. Nucleic Acids Research,
22, 4673?4680.
Wares JP, Hughes R, Grosberg K (2005) Mechanisms that
drive evolutionary change. In: Species Invasions: Insights
Into Ecology, Evolution, and Biogeography (eds. Sax DF,
Stachowicz JJ, Gaines SD), pp. 229?257. Sinauer Press,
Sunderland, MA.
Webley LS, Zenger KR, English AW, Cooper DW (2004) Low
levels of genetic variation within introduced Javan rusa deer
(Cervus timorensis russa) in Australia. European Journal of
Wildlife Research, 50, 137?140.
Zapata FA, Herron PA (2002) Pelagic larval duration and
geographic distribution of tropical eastern Pacific snappers
(Pisces: Lutjanidae). Marine Ecology Progress Series, 230, 295?
300.
Zayed A, Constantin S?A, Packer L (2007) Successful biological
invasion despite a severe genetic load. PLoS ONE 2, e868.
doi:10.1371/journal.pone.0000868.
Zeisset I, Beebee TJC (2003) Population genetics of a successful
invader: the marsh frog Rana ridibunda in Britain. Molecular
Ecology, 12, 639?646. 2010 Blackwell Publishing Ltd
M.R.G. is a PhD candidate at the University of Hawaii at
Manoa. She is interested in the dynamics of species invasions
and the phylogeography of marine organisms. B.W.B. studies
the phylogeography of marine vertebrates, with an emphasis
on how marine biodiversity is produced and maintained. R.J.T.
studies connectivity and phylogeography of coral reef species.
S.P. has focused his work on population genetics and evolu-
tionary biology of marine organisms, with special interest on
coral reef fish. V.M. is interested in patterns of genetic and spe-
cies diversity of coral reef fishes and the potential conse-
quences of biodiversity loss on coral reefs. J.E. is an avid scuba
diver, underwater photographer, and fish biologist. D.R.R. stu-
dies the biogeography of tropical reef fishes, and of neotropical
shorefishes and is working on the development of online infor-
mation systems for the shorefish faunas of three neotropical
biogeographic regions (the Tropical Eastern Pacific, the Greater
Caribbean, Brazil).
Supporting Information
Additional supporting information may be found in the online
version of this article.
Table S1 Allele frequencies at three nuclear introns in the two
source populations and across the introduced range.
Please note: Wiley-Blackwell are not responsible for the content
or functionality of any supporting information supplied by the
authors. Any queries (other than missing material) should be
directed to the corresponding author for the article.
THE I NTRO DUCT ION OF LUTJANUS KASMIRA TO HAW AII 1121 2010 Blackwell Publishing Ltd