Lien,ySosaku Ikeda,* and Ko Harada1,* *Faculty of Agriculture, Ehime University, 3-5-7 Tarumi, Matsuyama 790-8566, Japan; and y Mangrove Ecosystem Research Center, Center for Natural Res
Trang 1GENETIC VARIATION OF TWO MANGROVE SPECIES IN KANDELIA
(RHIZOPHORACEAE) IN VIETNAM AND SURROUNDING AREA
REVEALED BY MICROSATELLITE MARKERS
Le H Giang,* Gretel L Geada,* Phan N Hong,yMai S Tuan,yNguyen T H Lien,ySosaku Ikeda,* and Ko Harada1,*
*Faculty of Agriculture, Ehime University, 3-5-7 Tarumi, Matsuyama 790-8566, Japan; and y Mangrove Ecosystem Research Center, Center for Natural Resources and Environmental Studies, Vietnam National University, 22, Ngo Luong Su B, Quoc Tu Giam Street, Hanoi, Vietnam
Genetic variation of the mangrove genus Kandelia (Rhizophoraceae) in the South China Sea region, in four
populations in Vietnam and in one population each in Iriomote, Japan, and Bako, Borneo, was evaluated using
microsatellite markers A total of 54 alleles in the six populations were detected by using four microsatellite
loci The two northern Vietnamese populations (Don Rui and Xuan Thuy) showed a high allelic diversity (40
alleles in total) and a high level of gene diversity (HE¼ 0:773 on average) In contrast, the two southern
Vietnamese populations (Can Gio and Ngoc Hien) showed low allelic diversity (11 alleles in total) and a low
level of gene diversity (HE¼ 0:244 on average) There was only one allele common to the two regions The
Iriomote population was genetically related to the northern Vietnamese populations, while the Bako
pop-ulation was related to the southern poppop-ulations The findings and the morphological observations indicate that
these two genetically differentiated vicariant lineages represent two different species groups, Kandelia obovata
Sheue, Liu, and Yong for northern Vietnam and Japan and Kandelia candel (L.) Druce for southern Vietnam
and Borneo The difference in the amount of genetic variation shows that these two species experienced
a different adaptive process during the past glacial ages
Keywords: genetic variation, mangrove, microsatellite, Kandelia candel, Kandelia obovata, Rhizophoraceae,
South China Sea
Introduction
Mangroves are constituent plants of tropical and
subtropi-cal intertidal forest communities, typisubtropi-cal for the estuaries of
large rivers that run over a shallow continental shelf They
occupy two separate hemispheric regions but are more
abun-dant in the Old World than in the New World tropics
(Tom-linson 1986) Vietnam is rich in mangrove species, especially
along the coast of the Gulf of Tonkin in the north and along
the coast of Mekong Delta in the south Thirty-four true
man-grove species have been identified in the area (Hong and San
1993)
In Vietnam, mangrove forests have been used for supplying
diverse commercial and traditional products However, the
mangrove forests, especially in the southern part of Vietnam,
were seriously damaged by chemical warfare from 1962 to
1971 (Stellman et al 2003) Although the forests have since
largely regenerated, conversion of mangrove forests to
agri-cultural and aquaculture farms and clearance for settlements
have caused rapid degradation and decline of the remaining
forests in the area Since mangrove forests are important not
only economically but also environmentally, reforestation
and rehabilitation projects have been carried out in many
countries, including Vietnam (Kogo and Kogo 1997) In these
projects, however, genetic variation and local adaptation of mangrove plants have seldom been considered One reason for this is that appropriate methods for measuring genetic variation have been developed only recently Analysis of ge-netic variation should reveal its extent and distribution within
a species, indicating the range of local adaptation This infor-mation could be used for choosing germplasm for regenera-tion and planning sustainable management of mangrove forests
Kandelia Wight and Arnold 1834, a genus in the mangrove family Rhizophoraceae, was known as a monotypic genus with the single species, Kandelia candel (L.) Druce Kandelia candel was said to be distributed from the Ganges Delta, Myanmar, through Southeast Asia to south China, the Ryu-kyu Islands, and south Japan (Tomlinson 1986) The eastern limit is northern Borneo (Tomlinson 1986) Recent studies on leaf anatomy (Naskar and Mandal 1999), physiological ad-aptation (Nakagoshi and Nehira 1986; Maxwell 1995), chro-mosome numbers (Yoshioka et al 1984; Das et al 1995), and molecular phylogeography (Huang and Chen 2000; Chiang et al 2001) have shown that the species can be sepa-rated into two well-distinguished geographical groups, with the border at the northern end of the South China Sea Tak-ing these results and the detailed morphological comparisons
of vegetative and reproductive characters into account, Sheue
et al (2003) concluded that K candel could be split into two vicariant lineages representing the two geographically differentiated species groups Because K candel was first named for the species in India (Nicolson et al 1988), the
1 Author for correspondence; phone 946-9870; fax
81-89-946-9868; e-mail kharada@agr.ehime-u.ac.jp.
Manuscript received November 2004; revised manuscript received October
2005.
Ó 2006 by The University of Chicago All rights reserved.
1058-5893/2006/16702-0012$15.00
Trang 2populations north of the South China Sea were named
Kan-delia obovata Sheue, Liu, and Yong (Sheue et al 2003)
Specimens collected at the Gulf of Tonkin were identified as
K obovata (Sheue et al 2003), but no specimens have been
collected from other areas of Vietnam
Kandelia is one of the main mangroves used for
afforesta-tion in coastal areas of Vietnam, and propagules from the
northern part of Vietnam have often been used as a seed
source for the southern part of Vietnam Genetic information
on the geographical structure of Kandelia populations is
needed for an appropriate choice of propagules We collected
specimens of Kandelia from four Vietnamese populations,
two in the northern region and two in the southern region,
and additionally from one population each in the Ryukyu
Islands, Japan, and in Borneo, and we examined genetic
varia-tion, using microsatellite markers Microsatellites are
codom-inant markers and are known to be rich in genetic variation
from high mutation rates (Levinson and Gutman 1987; Wolff
et al 1991) These characteristics make the markers useful
for genome mapping, paternity testing, and individual
identi-fication (Jeffreys et al 1985; Nakamura et al 1987; Silver
1992) The usefulness of the markers for population genetics
studies on mangrove species has also been demonstrated in
Avicennia marina (Maguire et al 2000; Giang et al 2003)
The aim of this study was to investigate how much genetic
variation is maintained in populations of Kandelia in
Viet-nam, the Ryukyu Islands, and Borneo, and how they are
re-lated each other These findings should give us clues for
understanding the adaptive process of the two species in
dif-ferent climatic conditions
Material and Methods Sample Collection and DNA Isolation
Leaf samples from natural populations of Kandelia were
collected from two populations in the northern part of
Viet-nam (Dong Rui and Xuan Thuy), two populations in the
southern part of Vietnam (Can Gio, Ngoc Hien), one
popula-tion in Iriomote Island (Urauchi River) in the Ryukyu
Is-lands, and one population in Bako, Borneo (table 1; fig 1)
Sampling was planned so as to collect ca 30 adult
individu-als at each location, separated from each other by at least
30 m in order to avoid collecting half-sibs and to maximize
the probability of selecting diverse genotypes The leaf
sam-ples were individually sealed in airtight plastic bags and placed on ice in the field before freezing at 20°C About 0.5 g of leaf tissue was ground to powder in liquid nitrogen Total genomic DNA was extracted by the modified CTAB method (Doyle and Doyle 1990) Genomic DNA was further purified using a DNA binding matrix of the FastDNA kit (Qbiogene, Montreal)
Microsatellite Analysis
Primer sequences specific for four microsatellite loci de-scribed by Sugaya et al (2002) were used (table 2) The PCR amplification reaction mix was reaction buffer (20 mM Tris-HCl, pH 8.3, 50 mM KCl, 2 mM MgCl2, 0.001% gelatin),
2 mM dNTP, 0.5 mM forward and reverse primers, 0.3 units Taq DNA polymerase (Sigma-Aldrich, St Louis) and 10 ng ge-nomic DNA in a total volume of 10 mL The forward primers were fluorescent-labeled with FAM, VIC, and NED (Applied Biosystems, Foster City, CA) Microsatellite amplification was carried out using the following cycling parameters: preheating for 3 min at 94°C, followed by 35 cycles of denaturing at 90°C for 30 s, annealing at primer-specific temperatures of 55°–60°C for 30 s, and extension for 1 min at 72°C Reactions were com-pleted by incubation at 72°C for 5 min and holding at 4°C The PCR products were denatured for 3 min at 95°C and then sepa-rated by capillary electrophoresis on a 310 Genetic Analyzer (Applied Biosystems) GeneScan software (Applied Biosystems) was used for microsatellite analysis
Because these loci were cloned from Kandelia from a north-ern population (Amami in the Ryukyu Islands), we subse-quently subcloned the four microsatellite loci from individuals sampled from two southern populations, Can Gio and Bako, and determined the sequence in order to assess whether these loci are conserved across the populations One individual from each population was sampled The microsatellite regions were amplified using COD-Plus DNA polymerase (Toyobo, Osaka, Japan) and subcloned using Zero Blunt TOPO Cloning Kit (Invitrogen, Carlsbad, CA) Plasmid DNA was extracted and sequenced on a 310 Genetic Analyzer (Applied Biosystems) using Big Dye Terminator, version 1.1 (Applied Biosystems)
Data Analysis
Population genetics parameters were computed using the GDA program (ver 1.1) (Lewis and Zaykin 1999) The fol-lowing diversity statistics in each population were calculated: average number of alleles, unbiased expected heterozygosity
Table 1
Collection Sites of the Mangrove Genus Kandelia in Northern and
Southern Regions of the South China Sea
Population Location
Latitude (N)
Longitude (W)
Forest area (ha)
Sample size Northern region:
Iriomote Ryukyu 24°199 123°459 Ca 100 25 Dong Rui Quang Ninh Province 21°139 107°229 Ca 3500 22 Xuan Thuy Nam Dinh Province 20°149 106°339 Ca 1600 23 Southern region:
Can Gio Ho Chi Minh City 10°229 106°469 Ca 4700 28 Ngoc Hien Ca Mau Province 8°509 105°099 Ca 51,500 25 Bako Borneo 2°369 110°199 Ca 16,000 30
Trang 3(HE), and observed proportion of heterozygotes (HO)
Unbi-ased estimates of HEand F statistics (FIS, FIT, and FST) were
estimated for each locus The 95% confidence interval (CI)
of F statistics was determined by bootstrap analysis using
1000 replicates The outcrossing rate was calculated as
ð1 FÞ=ð1 þ FÞ (Weir 1996) Hardy-Weinberg equilibrium
(HWE) and linkage disequilibrium were evaluated, and the
significant levels were determined using the Markov chain
method with GENEPOP, version 3.4 (Raymond and Roussett
1995) The default parameters in GENEPOP with 100
batches of 1000 iterations each were used for the test
Results Conservation of the Microsatellite Loci
Four microsatellite loci were subcloned from two
individu-als, one each from the Can Gio and Bako populations, and
the sequences were determined (accession nos AB236960– AB236963 for Can Gio and AB236964–AB236967 for Bako)
In all four loci, primer regions were completely matched with those reported by Sugaya et al (2002) A total of four nucle-otide substitutions (one in Kc09, one in Kc11, and two in Kc34) and four indels (three with a length of 10 bp, 12 bp, and a single nucleotide in Kc11 and one with a length of
4 bp in Kc34) were detected in a total length of 922 bp of franking regions Franking sequences were identical in Can Gio and Bako in all four loci Interestingly, three indels (of
10, 12, and 4 bp) were found on one or both sides of the di-nucleotide repeats Large differences were found in the num-ber of dinucleotide repeats For Kc04, the TC unit repeated
35 times in Sugaya’s sequence, whereas it repeated 11 and 12 times in the clones from Can Gio and Bako, respectively Sim-ilarly, for Kc11, the TC unit repeated 34 times in Sugaya’s sequence, whereas it repeated nine and eight times in Can Gio and Bako, respectively For Kc09 and Kc34, the GA and AG units repeated 19 and 26 times, respectively, in Sugaya’s se-quence, whereas they repeated 14 and 10 times, respectively,
in both of Can Gio and Bako In all of the loci, the repeat number was much smaller in southern populations These findings indicate that the microsatellite loci are conserved across the northern and southern populations, although they are considerably differentiated
Allelic Diversity
Four microsatellite loci were consistently resolved in the six populations of Kandelia A total of 54 alleles, 17 for Kc04, 12 for Kc34, 7 for Kc09, and 18 for Kc11, were detected Among them, a total of 50 alleles were found in the Vietnamese popu-lations Allele frequency distribution for each locus is graphi-cally shown in figure 2 Many of the same alleles were found
in the two northern Vietnamese populations A total of 40 al-leles were found in this region Some of these alal-leles were also found in the Iriomote population A total of 11 alleles were found in the two southern Vietnamese populations, and seven
of them were shared between the populations There is only one common allele between the southern and northern Viet-namese populations, Kc09-283 A total of five alleles were de-tected in Bako, and all of them were also found in the southern populations in Vietnam These results show that pop-ulations in the northern and southern parts of Vietnam are clearly differentiated The northern populations are genetically related to the Iriomote population, while the southern popula-tions are related to the Bako population
Fig 1 Collection sites of the mangrove genus Kandelia in the
South China Sea region The line in the bottom represents the equator.
Table 2
Genetic Variation of Microsatellite Loci in Six Populations of the Mangrove Genus Kandelia Northern Vietnam Southern Vietnam Ryukyu Borneo Dong Rui Xuan Thuy Can Gio Ngoc Hien Iriomote Bako Locus Repeat unit A H E H O A H E H O A H E H O A H E H O A H E H O A H E H O
Kc04 (CT) 35 9 0.850 0.591 10 0.866 0.652 2 0.457 0.321 3 0.275 0.130 4 0.676 0.370 1 0.000 0.000 Kc34 (AG) 26 9 0.747 0.636 7 0.704 0.348 2 0.382 0.071 2 0.040 0.040 3 0.584 0.458 2 0.444 0.357 Kc09 (GA) 20 6 0.634 0.667 3 0.623 0.545 1 0.000 0.000 2 0.042 0.042 2 0.085 0.000 1 0.000 0.000 Kc11 (CT) 34 14 0.911 0.818 9 0.850 0.652 2 0.249 0.286 3 0.509 0.040 3 0.523 0.148 1 0.000 0.000 Note A ¼ total number of alleles; H E ¼ expected heterozygosity; H O ¼ observed heterozygosity.
Trang 4Eleven private alleles were detected in the two northern
Vietnamese populations: nine (Kc34-260, Kc34-274,
Kc09-283, Kc09-297, Kc11-190, Kc11-192, Kc11-216, Kc11-240,
and Kc11-248) in Dong Rui and three (Kc04-324, Kc11-218,
and Kc11-234) in Xuan Thuy Four private alleles were
de-tected in the two southern populations: one (Kc04-306) in
Can Gio and three (Kc04-300, Kc04-304, and Kc11-180) in
Ngoc Hien A total of 12 alleles were found in the Iriomote
population, and four (Kc04-344, Kc04-360, Kc04-362, and
Kc34-250) of them were private alleles
Genetic Diversity
The observed and expected heterozygosities were
deter-mined for each locus (table 2) The loci with longer
dinucleo-tide repeats (Kc04 and Kc11) commonly revealed greater
values of heterozygosity than did the two other loci (Kc34
and Kc09) The expected heterozygosity was greater than the
observed heterozygosity across all of the loci and all of the populations except Kc09 in Don Rui and Kc11 in Can Gio The average and the effective numbers of alleles (na and
ne, respectively), together with the expected and observed heterozygosities over all loci (HEand HO, respectively), were calculated for each population (table 3) The number of al-leles was much larger in the northern Vietnamese populations than in the southern populations The Dong Rui population showed the largest na, and the Ngoc Hien population showed the smallest na The average number of alleles in the north was 8.38, while that in the south was 2.25 The greatly re-duced number of effective alleles in the north (4.78) indicates that many rare alleles are involved
The HOwas high in the two northern Vietnamese popula-tions, 0.549 in Xuan Thuy and 0.678 in Dong Rui, and the value was much lower in the southern populations, 0.063 in Ngoc Hien and 0.170 in Can Gio It is notable that both
Fig 2 Allele frequencies at four microsatellite loci, Kc04, Kc34, Kc09, and Kc11 Numbers in the graphs refer to the nucleotide lengths of amplified fragments.
Trang 5allelic diversity and gene diversity were lowest in Bako,
where three of the four microsatellite loci were
monomor-phic Gene diversity in Iriomote was lower than that in the
two northern Vietnamese populations Observed heterozygosity
was lower than expected heterozygosity in all of the
popula-tions, leading to positive inbreeding coefficients Highly
sig-nificant departure from HWE was detected in Xuan Thuy,
Can Gio, Ngoc Hien, and Iriomote populations The
inbreed-ing coefficient (F) was low in the two northern populations,
the average value being 0.210, while the values were higher
in the two southern populations, varying from 0.380 (Can
Gio) to 0.713 (Ngoc Hien) Consequently, the outcrossing
rate was high in the north (0.560 in Xuan Thuy and 0.754 in
Dong Rui) and low in the south (0.168 in Ngoc Hien and
0.449 in Can Gio) (table 3)
Genetic Differentiation among the Populations
The fixation index, FST, was calculated between the two
northern populations and the two southern populations in
Vietnam (table 4) The value was 0.014, with 95% CIs of
0.000 and 0.050 between the two northern populations,
indi-cating no genetic differentiation and suggesting the
popula-tions to be panmictic, while the value was 0.134 with 95%
CIs of 0.086 and 0.175 between the two southern popula-tions, indicating moderate genetic differentiation The FST was also calculated among the populations in the northern South China Sea region (Dong Rui, Xuan Thuy, and Irio-mote) and the southern South China Sea region (Can Gio, Ngoc Hien, and Bako) (table 4) Relatively large genetic dif-ferentiation was found among the populations in the north-ern region (FST¼ 0:207), while differentiation was larger in the southern region (FST¼ 0:30) A UPGMA tree was con-structed on the basis of Nei’s genetic distance (Nei 1972) It clearly divided the populations into two clusters, one consist-ing of the two northern Vietnamese populations and Iriomote and the other of the two southern Vietnamese populations and Bako (fig 3) These groupings were supported by high bootstrap values
Linkage Disequilibrium
Linkage disequilibrium was tested for six pairwise combi-nations of the four microsatellite loci in each population using Fisher’s exact test (table 5) Highly significant disequi-librium was detected in two combinations in Iriomote (Kc04-Kc11 and Kc34-(Kc04-Kc11) Slightly significant or insignificant disequilibrium was detected for the rest of the populations
Table 3
Genetic Variation of Six Populations of Mangrove Genus Kandelia over All Microsatellite Loci Population Sample size n a n e H E H O F Outcrossing rate Northern Vietnam:
Dong Rui 22 9.50 5.22 0.785 0.678 0.140* 0.754 Xuan Thuy 23 7.25 4.57 0.761 0.549 0.282 *** 0.560 Mean 8.38 4.78 0.773 0.614 0.210 0.657 Southern Vietnam:
Can Gio 28 2.25 1.43 0.272 0.170 0.380 *** 0.449 Ngoc Hien 25 2.25 1.40 0.216 0.063 0.713 *** 0.168 Mean 2.25 1.43 0.244 0.116 0.528 0.309 Outside Vietnam:
Iriomote 25 3.00 2.18 0.467 0.244 0.482 *** 0.345 Bako 30 1.25 1.19 0.111 0.089 0.199 0.593 Note n a ¼ average number of alleles per locus; n e ¼ effective number of alleles per locus;
H E ¼ expected heterozygosity; H O ¼ observed heterozygosity; F ¼ inbreeding coefficient.
Significant departure from Hardy-Weinberg equilibrium (HWE): P < 0:05.
Significant departure from HWE: P < 0:0001.
Table 4
F Statistics of Four Microsatellite Loci in the Mangrove Genus Kandelia for Each Region
Northern Vietnam: Dong
Rui and Xuan Thuy
Southern Vietnam: Can Gio and Ngoc Hien
Northern South China Sea:
Dong Rui, Xuan Thuy, and Iriomote
Southern South China Sea: Can Gio, Ngoc Hien, and Bako Locus F IS F IT F ST F IS F IT F ST F IS F IT F ST F IS F IT F ST
Kc04 0.279 0.273 0.000 0.377 0.475 0.158 0.336 0.446 0.164 0.376 0.515 0.222 Kc34 0.331 0.320 0.000 0.747 0.790 0.169 0.297 0.447 0.241 0.466 0.688 0.415 Kc09 0.038 0.109 0.073 0.002 0.002 0.003 0.104 0.349 0.273 0.003 0.001 0.004 Kc11 0.169 0.183 0.016 0.548 0.589 0.090 0.316 0.435 0.174 0.547 0.636 0.196 Overall 0.216 0.223 0.014 0.517 0.582 0.134 0.282 0.431 0.207 0.456 0.620 0.301 Upper CI a 0.303 0.294 0.050 0.756 0.772 0.175 0.331 0.459 0.262 0.523 0.683 0.412 Lower CI a 0.114 0.151 0.000 0.359 0.456 0.086 0.170 0.382 0.167 0.357 0.495 0.188
a Confidence interval of the 95% probability calculated for overall F statistics by the bootstrapping method.
Trang 6Discussion Genetic Differentiation of Kandelia in Vietnam
The genetic analysis using microsatellite markers displayed
a large divergence between the northern populations and the
southern populations These loci are conserved across the
populations in the north and the south A total of 50 alleles
were detected in the four Vietnamese populations by using
four microsatellite loci Forty of those 50 alleles were
distrib-uted in the northern populations, and 11 were in the
south-ern populations Of these, one allele (Kc09-283) was found
in both regions This allele was distributed restrictively in
Dong Rui and Ngoc Hien, which are more than 1700 km
apart, and had a very low frequency (0.02 in each
popula-tion), indicating homoplasious occurrence of the allele The
two northern Vietnamese populations shared a large number
of frequent alleles (70% of common alleles), while the two
southern Vietnamese populations shared 54.5% of the
al-leles Frequent alleles in Iriomote were disseminated in both
of the northern populations, with some alleles occurring at
high frequencies Five alleles found in Bako were all found
in the southern Vietnamese populations Moreover, the
UPGMA tree indicated that the two populations in northern
Vietnam and Iriomote can be grouped together and
differen-tiated from the other group in southern Vietnam and Bako
(fig 3)
Correspondingly, Kado et al (2004) examined 1526 bp of
the matK gene region in the chloroplast DNA (cpDNA) in 15
individuals from Don Rui and in 15 and 10 individuals from
Can Gio and Ngoc Hien, respectively, and they found clear
genetic differentiation between the two regions Cang Gio and
Ngoc Hien populations were fixed in the same haplotype,
while four haplotypes appeared in Don Rui, none of which were found in Can Gio and Ngoc Hien Furthermore, there are 12 nucleotide substitutions and three indels between hap-lotypes in the northern and southern regions Interestingly, the haplotype in cpDNA on Okinawa Island (Ryukyus) was identical to the most frequent haplotype of Dong Rui (Kado
et al 2004) The nucleotide diversity between the two re-gions was 0.0046 (Kado et al 2004), and this is comparable
to that found between species in the matK gene in Fagopy-rum (Ohsako and Ohnishi 2001) and Clintonia (Hayashi
et al 2001)
Accordingly, Chiang et al (2001) also found that there were two lineages corresponding to the geographically iso-lated populations, using chloroplast DNA and mitochondrial DNA of Kandelia They showed that a northern Vietnamese population (Quang Ninh) was clustered with a population in the Ryukyus, while the population in Bako, Borneo, was clustered with a population in Ranong on the west side of the Malay Peninsula Moreover, the most pronounced differ-ence was seen in the minimum spanning network constructed from the trnL-trnF spacer (Chiang et al 2001) Eight clades
of the network were divided into two geographical groups with 34 mutational changes; the first group included popula-tions in the northern South China Sea region, and the other consisted of the Sarawak (Bako) and Ranong populations Within the network, closely related chlorotypes were mostly linked by single mutations
In Vietnam, Kandelia is distributed in the north and the south but is interrupted in the vast central area (Hong and San 1993) The central coast ranges from Lach Truong to Vung Tau Cape The seacoast is parallel to the Truong Son Range Most of the rivers here, unlike those in the coastal re-gions of the north and the south, originate in the Truong Son Range and are generally precipitous The geographical struc-ture in the central coast prevents mangroves from developing large populations along the seashore, so that they often exist
as small, patchy, discontinuous populations along the narrow strips of brackish water (Hong and San 1993) Although Kandelia vegetation was reported in this area by Hong and San (1993), we could not find Kandelia in our survey at some places in the central region (Canh Duong and Lang Co, near Hue)
One of us (P N H.) found that Kandelia in the north varies in height, from small trees branching at ca 1 m to
a maximum of ca 8 m Flowering time lasts from early June
Fig 3 UPGMA tree based on Nei’s genetic distance showing the
relationships among Kandelia populations in Vietnam, Iriomote, and
Borneo.
Table 5
Linkage Disequilibrium between Microsatellite Loci Northern Vietnam Southern Vietnam Ryukyu Borneo Locus pair Dong Rui Xuan Thuy Can Gio Ngoc Hien Iriomote Bako Kc04-Kc34 1.000 0.843 0.049* 0.220 0.054 1.000 Kc04-Kc09 0.490 0.033* 1.000 0.090 0.467 1.000 Kc04-Kc11 1.000 0.488 0.224 0.239 0.003 ** 1.000 Kc34-Kc09 0.841 0.112 1.000 1.000 0.646 1.000 Kc34-Kc11 0.175 0.935 0.821 0.392 0.009 ** 1.000 Kc09-Kc11 1.000 0.629 1.000 0.082 1.000 1.000
Significant departure from linkage equilibrium: P < 0:05.
Significant departure from linkage equilibrium: P < 0:01.
Trang 7to July, propagules appear on the trees from July on, and
mature fruits remain until the following May Kandelia in
the south, however, is taller, with trees branching from ca
5 to 12 m Flowering time lasts from September to late
November Propagules appear from November on, and fruits
remain until the following September He also found that
the leaves of Kandelia in the north are elliptic-obovate in
shape and small in size, whereas the leaves in the south are
usually elliptic-oblong and larger Other obvious
morphologi-cal and anatomimorphologi-cal differences in Kandelia between the two
regions were also found in flowers, ovules, cotyledons, and
seeds
The new species, Kandelia obovata, is distributed in the
northern South China Sea and East China Sea regions, from
the Gulf of Tonkin to the Ryukyu Islands and southern Japan,
while Kandelia candel is distributed from western India,
Burma, and Thailand to the Malay Peninsula and northern
Borneo (Sheue et al 2003) Sheue et al (2003) also reported
that specimens collected in northern Vietnam at the Gulf of
Tonkin were K obovata Considering the finding of Sheue
et al (2003) together with the genetic and morphological
dif-ferentiation between Kandelia in the northern and southern
parts of Vietnam, it is obvious that there are two species of
Kandelia in Vietnam, K obovata in the north and K candel
in the south
Genetic Diversity
Expected heterozygosity (HE) or gene diversity (Nei 1973)
over all loci (table 3) in the two northern Vietnamese
popula-tions was high (HE¼ 0:785 in Dong Rui and 0.761 in Xuan
Thuy, with an average of 0.773), whereas gene diversity was
much lower in the two southern populations (HE¼ 0:272 in
Can Gio and 0.216 in Ngoc Hien, with an average of 0.244)
This tendency was consistent with the allelic diversity
ob-served in each population (table 3) Gene diversities in the
northern populations (K obovata) were higher than those
reported for Avicennia marina worldwide (HE¼ 0:494;
Maguire et al 2000) and in Vietnam (HE¼ 0:322; Giang
et al 2003) Coastline currents in the northern South China
Sea and East China Sea are in a northerly direction in
sum-mer (Wang et al 1995) Corresponding to the maturation of
propagules of K obovata at the Gulf of Tonkin in June to
July, those populations with high genetic diversity are
thought to be potential sources of genetic variation
north-ward
In contrast, the average value of genetic variation in the
two southern populations was much lower than that in the
two northern populations It is surprising that an even smaller
level of genetic variation was found in Bako in Borneo,
where there is a large estuary area with diverse mangrove
species The low level of genetic variation in K candel in
Vietnam and Borneo suggests that these populations
experi-enced a severe bottleneck or were founded very recently In
addition, great genetic differentiation occurred among the
pop-ulations in southern Vietnam and Bako This indicates either
that the effective population size is small or that gene flow is
limited Since the most frequent alleles are common among
these populations in all four microsatellite loci, it is thought
that they are derived from a common ancestral population
with limited genetic variation
Results of paleoceanographical studies have shown that re-gions in Southeast Asia were connected, resulting in the for-mation of Sundaland, when the sea level was lowered during glacial ages (Tjia 1980; Wang et al 1995) In the last glacial maximum (LGM), the sea level was lowered by 100–120 m, and the South China Sea closed and changed into a semi-enclosed basin, with Bashi Strait as its only water passage to the Pacific Ocean (Wang et al 1995) The relatively large ge-netic variation retained in K obovata indicates that this spe-cies remained in coastal areas of the South China Sea region
as fairly large panmictic populations The winter sea surface temperature in the LGM was estimated to be not more than 23°C in the South China Sea region (Wang et al 1995) Since the sea surface temperature is critical for the survival of man-groves, K candel, which can adapt to a high temperature (Maxwell 1995), retreated to the south, and the remainder of the species was probably extinguished in the closed South China Sea during the LGM, while the main body of K can-del was retained in the Indian Ocean rim (the Ranong popu-lation probably being one of them) This is plausible because the current sea surface temperature in winter in the southern part of Vietnam (the northern limit of K candel) is around 25°C (Wang et al 1995) Kandelia candel could have ex-tended to Southeast Asian regions after the LGM, when re-gions were disconnected by the rise in sea level mediated by the ocean current through the Strait of Malacca to the South China Sea Greatly reduced genetic variation in K candel in Vietnam and Borneo could be explained by a founder effect
Reproductive Systems
In this study, all four populations of Kandelia in Vietnam showed significant departure from HWE because of heterozy-gosity deficits It is notable that the two northern populations showed higher outcrossing rates and lower levels of inbreed-ing (F ¼ 0:140 in Dong Rui, 0.282 in Xuan Thuy) These values are parallel to the value reported in K obovata popu-lations in Hong Kong (Sun et al 1998) In contrast, the two southern populations, especially the southernmost population
in Ngoc Hien, exhibited much higher levels of inbreeding (F ¼ 0:713) Actually, in Can Gio and Ngoc Hien, the habi-tats are restricted to small patches and separated from each other High levels of inbreeding may be due to the Wahlund principle (Hartl and Clark 1997) and/or geitonogamous pol-lination, as reported by Chen (2000) in Fuijan, China A high level of inbreeding also occurred in Iriomote A geomor-phological survey on Iriomote Island showed that seismic up-lift occurred at intervals of 1000–2000 yr, with huge earthquakes inducing fluctuation between destruction and re-generation of mangrove forests (Miyagi and Fujimoto 2003) The last seismic uplift occurred ca 1000 yr ago (Miyagi and Fujimoto 2003) This indicates that the population at Urau-chi River was established very recently by colonization by
a limited number of individuals from the surrounding area This explains the high inbreeding rate and the significant linkage disequilibrium detected among the microsatellite loci
Conclusions
We have demonstrated that two obviously different spe-cies, Kandelia obovata and Kandelia candel, exist in the
Trang 8north and the south in Vietnam, respectively With our
lim-ited sampling, it is not known whether these two species have
a clear boundary or overlap in the central region, although the
paleogeographical discussion indicates the latter Much smaller
genetic variation has been indicated in the populations of K
candel in the South China Sea region This indicates the
vul-nerability of this species to climate changes Comparative
stud-ies of genetic variation in K candel of this region with that of
the Indian Ocean region are required to clarify the route of
migration and to understand the adaptive process of this spe-cies in the South China Sea region after the last glacial age
Acknowledgments
We thank J J Kendawang and J Augustine for collecting samples of Kandelia in Bako, Sarawak, Malaysia This work was partly supported by grant-in-aid 13575001 from the Japan Society for Promotion of Science (JSPS)
Literature Cited
Chen XY 2000 Effects of plant density and age on the mating system
of Kandelia candel Druce (Rhizophoraceae), a viviparous mangrove
species Hydrobiologia 432:189–193.
Chiang TY, YC Chiang, YJ Chen, CH Chou, S Havanond, PN Hong,
S Huang 2001 Phylogeography of Kandelia candel in East Asiatic
mangroves based on nucleotide variation of chloroplast and
mitochondrial DNAs Mol Ecol 10:2697–2710.
Das AB, UC Basak, P Das 1995 Karyotype diversity and genomic
variability in some Indian tree mangroves Caryologia 48:319–328.
Doyle JJ, JL Doyle 1990 Isolation of plant DNA from fresh tissue.
Focus 12:13–15.
Giang LH, K Harada, PN Hong, MS Tuan 2003 Genetic variation of
Avicennia marina (Forsk.) Vierh (Avicenniaceae) in Vietnam
revealed by microsatellite and AFLP markers Gene Genet Syst
78:339–407.
Hartl DL, AG Clark 1997 Principles of population genetics 3rd ed.
Sinauer, Sunderland, MA 543 pp.
Hayashi K, S Yoshida, FH Utech, S Kawano 2001 Molecular
systematics in the genus Clintonia and related taxa based on rbcL
and matK gene sequence data Plant Species Biol 16:119–137.
Hong PN, HT San 1993 Mangroves of Vietnam IUCN, Bangkok.
173 pp.
Huang S, YC Chen 2000 Pattern of genetic variation of Kandelia
candel among populations around South China Sea Pages 59–64 in
TY Chian, TW Hsu, eds Wetland biodiversity Proceedings of
Symposium of Biodiversity in Wetlands Taiwan Endemic Species
Research Institute, Nantou.
Jeffreys AJ, V Wilson, SL Then 1985 Individual-specific
‘‘finger-prints’’ of human DNA Nature 316:76–79.
Kado T, A Fujimoto, LH Giang, MS Tuan, PN Hong, K Harada, H
Tachida 2004 Genetic structures of natural populations of three
mangrove species in Vietnam revealed by matK sequences of plastid
DNA Plant Species Biol 19:91–99.
Kogo K, M Kogo 1997 Mangrove reforestation in Vietnam by
supporting villagers’ activities: a trial to make a restoration model
on a global level Tropics 6:247–282.
Levinson G, GA Gutman 1987 Slipped-strand mispairing: a major
mechanism for DNA sequence evolution Mol Biol Evol 4:
203–221.
Lewis PO, D Zaykin 1999 Genetic data analysis: computer program
for the analysis of allelic data Version 1.0 http://hydrodictyon.
eeb.uconn.edu/people/plewis/software.php.
Maguire TL, P Saenger, P Baverstock, R Henry 2000 Microsatellite
analysis of genetic structure in the mangrove species Avicennia
marina (Forsk.) Vierh (Avicenniaceae) Mol Ecol 9:1853–1862.
Maxwell GS 1995 Ecogeographic variation in Kandelia candel from
Brunei, Hong Kong and Thailand Hydrobiologia 295:59–65.
Miyagi T, K Fujimoto 2003 Geoecological field investigation of
mangrove sedimentation process and carbon sequestration in the
Iriomote Island, Okinawa Pages 64–65 in Research Institute for
Subtropics, ed A summary on the mangrove study in Okinawa (FY 2000–2002) Research Institute for Subtropics, Naha.
Nakagoshi N, K Nehira 1986 Growth and mortality of mangrove seedlings transplanted to Hiroshima Hikobia 9:439–449 Nakamura Y, M Leppert, P O’Connell, R Wolff, T Holm, M Culver, C Martin, et al 1987 Variable number of tandem repeat (VNTR) markers for human gene mapping Science 235:1616–1622 Naskar K, R Mandal 1999 Ecology and biodeversity of Indian mangroves Daya, Delhi 754 pp.
Nei M 1972 Genetic distance between populations Am Nat 106: 283–282.
——— 1973 Analysis of gene diversity in subdivided populations Proc Natl Acad Sci USA 70:3321–3323.
Nicolson DH, CR Suresh, KS Manilal 1988 An interpretation of van Rheed’s Hortus Malabaricus Regnum Veg 119:1–378.
Ohsako T, O Ohnishi 2001 Nucleotide sequence variation of the chloroplast trnK/matK region in two wild Fagopyrum (Polygona-ceae) species, F leptopodum and F statice Gene Genet Syst 76: 39–46.
Raymond M, F Roussett 1995 GENEPOP (ver 3.1b): population genetics software and ecumenicism Heredity 86:248–249 Sheue CR, HY Liu HY, JWH Yong 2003 Kandelia obovata (Rhizophoraceae), a new mangrove species from Eastern Asia Taxon 5:287–294.
Silver LM 1992 Bouncing off microsatellites Nat Genet 2:8–9 Stellman JM, SD Stellman, R Christian, T Weber, C Tomasallo
2003 The extent and patterns of usage of Agent Orange and other herbicides in Vietnam Nature 422:681–686.
Sugaya T, T Takeuchi, H Yoshimaru, M Katsuta 2002 Development and polymorphism of simple sequence repeat DNA markers for Kandelia candel (L.) Druce Mol Ecol 2:65–66.
Sun M, KC Wong, JSE Lee 1998 Reproductive biology and population genetic structure of Kandelia candel (Rhizophoraceae),
a viviparous mangrove species Am J Bot 85:1631–1637.
Tjia HD 1980 The Sunda shelf, Southeast Asia Z Geomorphol 24: 405–427.
Tomlinson PB 1986 The botany of mangrove Cambridge University Press, Cambridge 419 pp.
Wang P, L Wang, Y Bian, Z Jian 1995 Late Quaternary paleoceanog-raphy of the South China Sea: surface circulation and carbonate cycles Mar Geol 127:145–165.
Weir BS 1996 Genetic data analysis II Sinauer, Sunderland, MA.
445 pp.
Wolff RK, R Plaeke, AJ Jeffreys, R White 1991 Unequal crossing over between homologous chromosomes is not the major mechanism involved in the generation of new alleles at VNTR loci Genomics 5: 382–384.
Yoshioka H, K Kondo, M Segawa, K Nehira, S Maeda 1984 Karyo-morphological studies in five species of mangrove genera in the Rhizophoraceae Kromosomo 35–36:1111–1116.