Journal of Heredity 2013:104(2):248–262
doi:10.1093/jhered/ess088
Advance Access publication November 1, 2012
© The American Genetic Association. 2012. All rights reserved.
For permissions, please email: [email protected].
Genetic Variation and Structure
within 3 Endangered Calanthe Species
(Orchidaceae) from Korea: Inference of
Population-Establishment History and
Implications for Conservation
Mi Yoon Chung, Jordi López-Pujol, Masayuki Maki, Myung-Ok Moon, Jin Oh Hyun, and
Myong Gi Chung
From the Department of Biology and the Research Institute of Natural Science, Gyeongsang National University, Jinju
660-701, Republic of Korea (M. Y. Chung and M. G. Chung); the BioC, GReB, Laboratori de Botànica, Facultat de Farmàcia,
Universitat de Barcelona, Barcelona 08028, Spain (López-Pujol); the Division of Ecology and Evolutionary Biology,
Graduate School of Life Sciences, Tohoku University, Aoba, Sendai 980-8578, Japan (Maki); Northeastern Asia Plant
Institute, Seoul 135-943, Republic of Korea (Hyun); the Department of Biology, Jeju National University, Jeju 690-756,
Republic of Korea (Moon).
Address correspondence to Myong Gi Chung at the address above, or e-mail: [email protected].
Abstract
It is thought that the warm-temperate vegetation of Korea, which at present is limited to southern coastal areas, was
founded by individuals from glacial refugia putatively located in southern Japan and/or southern China. Two scenarios of
postglacial recolonization can be hypothesized: 1) extant Korean populations are derived from multiple source populations
or, 2) they originated from a single source. To test which of these scenarios is more likely, we surveyed patterns of genetic
diversity in 3 congeneric terrestrial orchids that are typical of Korean warm-temperate vegetation, Calanthe discolor, C. sieboldii, and C. reflexa. In total, we studied 14 populations with 17 allozyme loci. To complement the study, we also tested the 2
scenarios with previous population-genetics data reported for other warm-temperate plant species native to Korea. Levels
of genetic variation in the 3 Calanthe species were substantially higher than those typical of allozyme-based studies in other
terrestrial orchid species, which clearly suggest a pattern of postglacial recolonization fitting the first scenario. However, previous studies with native Korean species showed equivocal patterns and suggest conflicting scenarios. The Calanthe species
studied here and the previous genetic surveys conducted on other species indicate that the establishment of contemporary
warm-temperate vegetation in Korea has been diverse since deglaciation. From a conservation perspective, C. reflexa should
be regarded as endangered in Korea, needing special attention to preserve its current genetic diversity and to prevent further
decreases in population sizes.
Key words: allozymes, founder effect, glacial refugia, Korea, population history, warm-temperate vegetation
Contemporary levels of genetic variation within and among
populations of plant species are shaped by a complex
suite of species’ life-history traits, as well as ecological
and historical factors (e.g., past climate oscillations)
(reviewed in Loveless and Hamrick 1984; Hamrick and
Godt 1989; Hewitt 1999, 2000). Species with high rates
of outcrossing, high fecundities, wind pollination, windmediated seed dispersal, and long generation time usually
maintain high intrapopulation genetic variation and low
248
interpopulation differentiation (reviewed in Loveless
and Hamrick 1984; Hamrick and Godt 1989). However,
marginal or peripheral populations within a species’ range
tend to harbor lower levels of within-population genetic
variation than geographically central populations, whereas
among-population differentiation shows a reverse trend
(reviewed in Eckert et al. 2008). The glacial–interglacial
oscillations during the Quaternary period have also played
an important role in shaping the current distribution of
Chung et al. • Genetic Variation and Structure within 3 Endangered Calanthe Species
plant species and thus, their levels and partitioning of
genetic diversity within and among populations (Hewitt
1999, 2000).
On the Korean Peninsula, broad-leaved evergreen
shrubs and trees currently occur in a narrow zone of warmtemperate and subtropical climate along the coast (Yi 2011).
Before the Quaternary period, these plants were also present in lowlands at higher latitudes. Fossil records indicate
the occurrence of some broad-leaved evergreen tree species
(Castanopsis, Cinnamomum, Cyclobalanopsis, Ilex, and Myrica) in
eastern coastal areas during the Miocene (Kong and Watts
1993; Kong 1996). During glacial advances, climatic cooling caused southward shifts in the geographic distribution
of broad-leaved evergreen forests. However, paleovegetation in the southeastern and southern coastal areas of
the Korean Peninsula during the Last Glacial Maximum
(LGM, ca., 21 000 years ago) is not well understood,
largely due to the lack of adequate pollen records (Kong
and Watts 1993; Kong W-S, personal communication). The
few available pollen records suggest, nevertheless, that the
evergreen broad-leaved forest vegetation was absent from
Korea during the LGM (e.g., Choi 1998; Chung et al. 2006,
2010; Chung 2007; Yi and Kim 2010). The lack of vegetation characteristic of a warm-temperate or subtropical
climate during the LGM on the Korean Peninsula is also
indicated by vegetation reconstructions (e.g., Adams and
Faure 1997; Harrison et al. 2001; Hope et al. 2004; Prentice
et al. 2011). The first appearance of broad-leaved evergreen vegetation on the Korean Peninsula after the LGM
was approximately 8500 years ago (Chung 2011), and during the middle Holocene optimum (ca., 6000 years ago) it
reached latitudes up to 37°N (Choi 1998; Yi 2011). During
the late Holocene under dry, cool-temperature climatic
conditions, warm-temperate and subtropical plant species
retreated toward their current distribution in southwestern
and southern coastal areas (i.e., latitudes well below 36°N;
Yi 2011). Although little is known about the past distribution of many orchids in southwestern and southern Korea,
it is likely that they would have experienced similar range
shifts as broad-leaved evergreen plants because they currently coexist in a narrow warm-temperate and subtropical
climatic zone along the Korean coast, the more extensive
warm-temperate regions of Japan (Ohwi 1965), as well
as subtropical China (Hou 1983; Wu et al. 2009). Thus,
Korean populations of orchid species are on the margin of
their distribution ranges.
Currently, populations of many orchids worldwide are
small and isolated, and have increasingly been a major subject of conservation concern (Swarts and Dixon 2009a,
2009b; Seaton et al. 2010; Vereecken et al. 2010; Phillips et al.
2011). The same situation is true for orchids on the Korean
Peninsula (Lee and Choi 2006). Although many orchids
were once abundant and continuously distributed in southwestern and southern Korea, currently they are subjected to
human disturbance (e.g., mass collection) and the increasing
disappearance of their habitats (Lee and Choi 2006; Chung
and Nason 2007). Since information on demographic
and genetic structure of surviving plant populations and
species is essential for designing comprehensive plans for
their conservation (Falk and Holsinger 1991; Hamrick and
Godt 1996), some conservation-genetics studies on marginally distributed orchid populations have been conducted
in southern Korea. For example, marginal populations of
the terrestrial orchid Cymbidium goeringii, which are characterized by large sizes and continuous distribution, maintain
high levels of genetic variation and low among-population
differentiation (Chung MY and Chung MG 1999). In contrast, marginal populations of the 2 lithophytic orchids
Bulbophyllum drymoglossum and Sarcanthus scolopendrifolius,
characterized by their small size and high isolation, exhibit
low levels of genetic variation and high among-population
differentiation (Chung et al. 2007). In the last example, rare
and highly isolated populations of the terrestrial orchids
Oreorchis coreana and Tipularia japonica have no allozyme variation within- and among populations (Chung 2009; Chung
et al. 2012).
Based on these prior studies, together with the information available on the paleoecology of the Korean Peninsula,
we hypothesize 2 alternative scenarios for the origin of
warm-temperate orchid populations native to southern
Korea. If extant populations are recently derived from
multiple source populations (i.e., multiple glacial refugia),
presumably from southern Japan and/or southern China,
we would expect high levels of genetic variation within
populations and low or high degree of genetic divergence
among the extant populations depending on ecological factors (Hamrick and Nason 1996). Large populations that are
continuously distributed should exhibit low interpopulation
variation probably due to high recurrent gene flow between
adjacent populations. In contrast, high genetic divergence
would be expected among historically small disjunct populations because of low rates of gene flow between isolated
populations. Alternatively, if current Korean populations
of warm-temperate orchid species are derived from a single
source (i.e., a single glacial refugium), then within-population genetic variation will be low and genetic differentiation among populations would be low or high depending on
rates of contemporary gene flow among the Korean populations. To date, these hypotheses have never been empirically tested for the warm-temperate and subtropical orchids
native to the Korean Peninsula.
In this study, we selected 3 congeneric terrestrial orchids
that occur in Korea, Calanthe discolor Lindley, C. reflexa
Maximowicz, and C. sieboldii Decaisne ex Regel. These
species are part of the warm-temperate flora of southern
Korea, central and southern subtropical mountain regions
in China, and southern Japan. We surveyed the levels
and distribution of genetic diversity in these 3 species to
test which of the 2 postglacial immigration hypotheses
is most likely. Related to this study, we further test the 2
proposed scenarios using the published data for several
warm-temperate plant species native to southern Korea,
which also occur in East Asia. Finally, genetic data obtained
from this study will provide guidelines for the recovery and
management of these currently rare orchids in the Korean
Peninsula.
249
Journal of Heredity 2013:104(2)
Materials and Methods
Study Species
Calanthe discolor occurs in broad-leaved evergreen forests
in subtropical and warm-temperate regions of central and
southern China (ca., 800–1500 m above sea level [a.s.l.]), lowlands of central and southern Japan (including southwestern
Hokkaido), and lowlands of southwestern and southern
Korea, including several islands (Chen et al. 2009; Lee 2011).
Its rhizomes are short and thick, and from April to May,
adults produce scapes (30–50 cm tall), each bearing 10–20
flowers with dark brown to greenish sepals (13–20 mm long)
and white or whitish-rose, deeply three-lobed lips (15–20 mm
long). Calanthe discolor is pollinated by Eucera bees (Ishikawa
and Suzuki 1992; Sugiura N, personal observation). Similar
to C. discolor, C. reflexa is widely distributed in subtropical and
warm-temperate regions of central and southern China under
broad-leaved evergreen forests (altitude 600–2500 m a.s.l.)
and lowlands (Chen et al. 2009) and also in southern Japan
(Honshu, Shikoku, and Kyushu; Figure 1). In Korea, however,
Figure 1. Map of East Asia and locations of sampled populations of C. discolor (CD-1 to CD-9), C. sieboldii (CS-1 to CS-3),
and C. reflexa (CR-1 and CR-2) in the Korean Peninsula. Dotted line in the upper figure indicates exposed coastal lines during the
LGM (Shi 2002; Xu et al. 2010).
250
Chung et al. • Genetic Variation and Structure within 3 Endangered Calanthe Species
it only occurs in a few localities in the lowlands on Jeju Island
off the coast of southern Korea (Lee 2011). Rhizomes of
C. reflexa are short, and flowering scapes (20–40 cm tall) are
produced during May to June in China and July to August
both in Korea and Japan, each bearing 10–20 pale purple
flowers (sepals are 15–20 mm long; petals are 12–15 mm
long and the apex of the middle segment of the lip is cuspidate). Flowers of C. reflexa are pollinated by Bombus workers (Sugiura N, unpublished data). Compared with C. discolor
and C. reflexa, C. sieboldii has a narrower distribution, occurring
in subtropical regions of a few mountainous areas (altitude
1200–1500 m a.s.l.) in southern China (Hunan and Taiwan),
in lowlands of southern Japan (Shikoku and Kyushu), including the Ryukyu Islands, and in lowlands of a few islands of
southern Korea (Ulleung, Heuksan, Gageo, Hong, and Jeju
Islands; Figure 1) (Chen et al. 2009; Lee 2011). The 3 Calanthe
species often co-occur on Jeju Island. Like the 2 other Calanthe
species, rhizomes of C. sieboldii are short and thick. From
February to March in China and April to June both in Korea
and Japan, adults produce scapes (20–50 cm tall), each bearing
5–20 yellow flowers (sepals and petals are 13–20 mm long and
the apex of the middle segment of the lip is slightly lobed).
Flowers of C. sieboldii are pollinated by Xylocopa bees (Sugiura
N, unpublished data). Without observing the flowers, it is very
difficult to distinguish it from C. discolor. As found in many
orchids, fruits (capsules; 1.5–2 cm long) of the 3 species contain large numbers of tiny, dust-like seeds.
In Korea, the 3 species were once common with large
population sizes along the coastal areas in southwestern
and southern Korea but the numbers of individuals and
populations have dramatically decreased since early 1970s
primarily due to over-collection and habitat destruction (Lee
and Choi 2006). We do not have information on the generation times for the 3 Calanthe species, although we can
assume to have long generation times (10–20 years) as occurs
in many other orchids (e.g., Curtis 1943; Nicolè et al. 2005;
COSEWIC 2010). According to population genetics theory,
little alteration in the levels of genetic diversity is expected
because there have not been sufficient generations after overcollection for original genetic diversity to be substantially
eroded by random genetic drift (Chung and Nason 2007).
Thus, human alterations should not significantly mask the
natural genetic patterns of the studied orchids.
Study Populations and Sampling
We collected 535 leaf samples of C. discolor (NT, a mean of
59 samples per population, ranging from 14 to 134; Table 1)
from 9 coastal and insular populations in Korea (CD-1 to
CD-9; Figure 1). For C. sieboldii, we sampled a total of 101
leaf samples (NT, a mean of 34 samples per population,
ranging from 25 to 49; Table 1) from 1 population located on
Gageo Island (CS-1; Figure 1) and 2 populations from Jeju
Island (CS-2 and CS-3; Figure 1). For C. reflexa, we sampled
46 leaf samples (NT) from 2 populations CR-1 (22) and CR-2
(24) on Jeju Island (Table 1; Figure 1). We collected from
every individual but did not collect from closely (ca., less than
20 cm) spaced shoots that appeared to be members of the
same clone. Since C. sieboldii and C. reflexa were extremely rare
on Gageo and Jeju Islands, we collected leaf samples from
Table 1 Summary of parameters of clonal diversity and distribution observed in populations of the 3 Calanthe species
Species/population
Calanthe discolor
CD-1
CD-2
CD-3
CD-4
CD-5
CD-6
CD-7
CD-8
CD-9
Average
Calanthe sieboldii
CS-1
CS-2
CS-3
Average
Calanthe reflexa
CR-1
CR-2
Average
NT
NG
Pgen FIS
Psex FIS
R
D
ED
β
(95% CI) of b
82
41
80
14
77
31
38
134
38
59
55
22
23
10
22
11
16
59
23
27
1.99–4
4.60–5
1.28–3
4.12–3
2.98–4
1.97–4
3.89–3
1.05–3
5.74–5
1.23–3
0.011
0.001
0.033
0.096
0.008
0.008
0.081
0.052
0.002
0.031
0.67
0.51
0.28
0.69
0.28
0.33
0.41
0.44
0.59
0.47
0.99
0.96
0.93
0.92
0.87
0.87
0.94
0.98
0.97
0.93
0.95
0.89
0.92
0.42
0.79
0.81
0.92
0.95
0.89
0.84
1.860
1.033
0.589
0.904
0.351
0.553
0.807
0.989
1.000
0.898
(–5.25, 1.53)
(–1.58, –0.48)
(–0.86, –0.31)
(–3.24, 1.43)
(–0.47, –0.24)
(–0.89, –0.21)
(–1.26, –0.35)
(–1.37, –0.61)
(–1.62, –0.38)
49
25
27
34
38
15
15
23
7.05–5
8.46–5
1.87–4
5.79–5
0.000
0.002
0.006
0.003
0.76
0.58
0.53
0.62
0.98
0.96
0.94
0.96
0.75
0.92
0.85
0.84
1.191
1.094
1.881
1.055
(–1.20, –0.38)
(–6.54, 4.36)
(–1.52, –0.24)
22
24
23
13
15
14
5.30–3
2.46–3
3.88–3
0.111
0.039
0.075
0.57
0.61
0.59
0.94
0.95
0.95
0.86
0.85
0.86
1.050
1.114
1.082
(–1.76, –0.34)
(–1.72, –0.51)
NT, the number of individuals sampled (including clonal ramets); NG, the number of individuals excluding clonal ramets; Pgen FIS, probability of the identical
MLG occurring by chance due to sexual reproduction by taking into account departures from H-W equilibrium; Psex FIS, probability of the second encounter
of identical MLG via sexual reproduction by taking into account departures from H-W equilibrium; R, genotypic richness; D, Simpson diversity index of
clonal heterogeneity; ED, Simpson evenness index; β, parameter for Pareto distribution (–1 × regression slope, b).; and 95% CI.
251
Journal of Heredity 2013:104(2)
Table 2 Summary of genetic diversity measures and mean fixation (FIS) values observed in populations of the 3 Calanthe species
Species/population
Calanthe discolor
CD-1
CD-2
CD-3
CD-4
CD-5
CD-6
CD-7
CD-8
CD-9
Average
Total samples
Calanthe sieboldii
CS-1
CS-2
CS-3
Average
Total samples
Calanthe reflexa
CR-1
CR-2
Average
Total samples
Alt (m) Area (m)
%P
AR
A
Ho (SE)
He (SE)
FIS (95 CI)
43
114
63
174
141
282
260
146
219
160
65 × 38
50 × 50
25 × 25
70 × 120
90 × 200
15 × 15
16 × 12
30 × 30
60 × 60
76.5
70.6
64.7
70.6
64.7
64.7
64.7
70.6
70.6
68.6
88.2
1.91
1.98
1.95
1.82
1.87
1.81
1.86
1.83
2.07
1.90
2.07
2.18
2.06
2.18
1.82
1.94
1.82
1.94
2.00
2.18
2.01
2.59
0.183 (0.044)
0.254 (0.065)
0.151 (0.041)
0.194 (0.053)
0.198 (0.048)
0.187 (0.051)
0.154 (0.036)
0.184 (0.049)
0.217 (0.047)
0.191(0.010)
0.190 (0.038)
0.242 (0.052)
0.283 (0.059)
0.210 (0.052)
0.194 (0.049)
0.230 (0.050)
0.216 (0.052)
0.179 (0.046)
0.211 (0.055)
0.282 (0.051)
0.227 (0.012)
0.244 (0.047)
0.244* (0.047, 0.441)
0.101ns (–0.149, 0.351)
0.285* (0.088, 0.482)
0.003ns (–0.190, 0.196)
0.141* (–0.108, –0.389)
0.135ns (–0.070, 0.340)
0.138ns (–0.118, 0.393)
0.126ns (–0.086, 0.338)
0.229* (–0.006, 0.464)
0.176a
120
238
219
192
40 × 100
90 × 100
100 × 100
76.5
58.8
64.7
66.7
76.5
2.03
1.76
1.88
1.89
2.03
2.24
1.76
1.88
1.96
2.35
0.214 (0.052)
0.191 (0.048)
0.169 (0.039)
0.191 (0.013)
0.199 (0.048)
0.288 (0.056)
0.281 (0.060)
0.270 (0.056)
0.280 (0.003)
0.293 (0.053)
0.258* (0.016, 0.500)
0.345* (0.178, 0.512)
0.380* (0.187, 0.573)
0. 296a
185
168
177
100 × 100
100 × 100
47.1
47.1
47.1
47.1
1.47
1.52
1.5
1.5
1.47
1.52
1.50
1.53
0.109 (0.036)
0.149 (0.050)
0.129 (0.020)
0.130 (0.042)
0.173 (0.050)
0.197 (0.055)
0.185 (0.012)
0.180 (0.051)
0.376* (0.102, 0.650)
0.246* (–0.082, 0.574)
0. 299a
Alt (m), altitude above sea (a.s.l.) in meters; %P, percentage of polymorphic loci; AR, mean allelic richness (adjusted for a sample size of 10 plants in C. discolor, 15 plants in C. sieboldii, and 13 plants in C. reflexa); A, mean number of alleles per locus; Ho, observed heterozygosity; He, H-W expected heterozygosity
or genetic diversity; and FIS, fixation index within populations. SE and 95% CI denote standard error and 95% confidence intervals, respectively.
aSignificant through a permutation test (1000 randomizations of alleles among individuals within samples) over populations.
nsNot significant.
*Denotes significance (P < 0.05) based on permutation (999 replicates) under the null hypothesis of FIS = 0.
relatively larger areas (0.4–1 ha; Table 2). As we collected
most of the individuals occurring within each population,
we could consider “sample sizes” per population as roughly
equivalent to “population sizes.” To minimize the damage to
these endangered orchids, we cut only 3 cm from the tip of
leaf per shoot. Voucher specimens were deposited at the herbarium of Gyeongsang National University (GNUC).
Allozyme Electrophoresis
Leaf samples were wrapped in damp paper towels, stored
in plastic bags, returned to the laboratory, and stored at
4 °C until protein extraction. For extraction, leaf samples
were crushed using chilled mortars and pestles by adding
a phosphate–polyvinylpyrrolidone buffer (Mitton et al.
1979), and enzyme extracts were absorbed onto 4 × 6 mm
paper wicks (Whatman 3MM chromatography paper). We
conducted electrophoresis on 13% starch gels, with 2 buffer
systems. We used a modification (Haufler 1985) of the system
6 of Soltis et al. (1983) to resolve alcohol dehydrogenase
(Adh), cathodal peroxidase (Per), diaphorase (Dia-1 and Dia2), fluorescent esterase (Fe), leucine aminopeptidase (Lap),
phosphoglucomutase (Pgm), and triosephosphate isomerase
(Tpi-1, Tpi-2, and Tpi-3). We also used the morpholine-citrate
buffer system (pH 6.1) of Clayton and Tretiak (1972) to resolve
fructose-1,6-diphosphatase (F1,6), isocitrate dehydrogenase
252
(Idh-1 and Idh-2), malate dehydrogenase (Mdh-1 and Mdh-2),
6-phosphogluconate dehydrogenase (6Pgd), and shikimic
acid dehydrogenase (Skdh). We followed stain recipes from
Soltis et al. (1983) except for diaphorase (Cheliak and Pitel
1984). We designated putative loci sequentially, with the most
anodally migrating isozyme designated as 1, the next 2, and
so on. We also designated different alleles within each locus
sequentially by a, the next b, and so on. The observed enzyme
banding patterns were consistent with their typical subunit
structure and subcellular compartmentalization in diploid
plants (Weeden and Wendel 1989).
Data Analysis on Clonal Structure, Genetic Diversity,
and Genetic Structure
As multiple ramets (NT) representing allozyme-based
identical multilocus genotypes (MLGs) could result
either from distinct sexual reproduction events or clonal
propagation, it is essential to discriminate these cases. To
do this, we calculated Pgen FIS, the probability of the MLG
occurring by chance due to sexual reproduction by taking
into account departures from Hardy–Weinberg (H-W)
equilibrium (Parks and Werth 1993; Arnaud-Haond et al.
2007). We averaged Pgen FIS estimates generated from one
such value for each MLG in each population. We further
estimated Psex FIS, the probability of the second encounter
Chung et al. • Genetic Variation and Structure within 3 Endangered Calanthe Species
of identical MLGs via sexual reproduction by taking into
account departures from H-W equilibrium (Parks and Werth
1993; Arnaud-Haond et al. 2007). We also averaged Psex
FIS from one such value for each ramet per MLG in each
population. Psex FIS is considered the most conservative
estimate of clonal identity (Arnaud-Haond et al. 2007). We
used a P < 0.05 cut off for the discrimination of ramets
versus genets.
Arnaud-Haond et al. (2007) and Becheler et al. (2010)
recommend the use of 4 parameters to describe clonal
diversity and distribution. We calculated genotypic richness
(R; Dorken and Eckert 2001), Simpson diversity index
(Pielou 1969) of clonal heterogeneity (D, the probability of
encountering distinct MLGs when randomly taking 2 units
in a population) and its equitability (ED, Simpson evenness;
Hurlbert 1971), and the parameter of the Pareto distribution
(β = [–1 × regression slope, b], that is, the negative slope
of the power law describing the frequency distribution of
clonal membership for populations (Arnaud-Haond et al.
2007). Simulations revealed that β increases exponentially
with increasing evenness of the distribution of sampling
units into MLGs (Arnaud-Haond et al. 2007). Both R and
ED influence the shape of the Pareto distribution and its
associated β value. High R and ED (i.e., clonal ramets all
having approximately equal sizes) will result in a high β
value (a steep slope), whereas the result of low R and ED
(i.e., a skewed clonal distribution with very few, large clonal
lineages and many small ones) will be a shallow slope (low
β value) (Arnaud-Haond et al. 2007). Two other cases could
be expected by comparing populations with similar R and
distinct ED or, conversely, with similar ED and distinct R.
For all these calculations, we used the program GenClone
2.0 (Arnaud-Haond and Belkhir 2007). Finally, to determine
whether β was significantly different among populations,
we estimated the 95% confidence intervals (CIs) around b
using classical least-squares regression theory (Zar 1999;
S. Arnaud-Haond, personal communication).
To calculate genetic diversity and structure parameters, we
prepared a second data set (NG) by excluding clonal ramets
in which each distinct MLG was only represented once. We
considered a locus as polymorphic within a population if the
frequency of the most common allele does not exceed 0.99
(Young et al. 1996). We estimated the following allele and
genetic diversity parameters using the programs POPGENE
(Yeh et al. 1999) and FSTAT (Goudet 2002): percentage of
polymorphic loci (%P ), mean number of alleles per locus
(A), allelic richness (AR) using a “rarefaction method” that
compensates for uneven sample sizes (Hurlbert 1971; El
Mousadik and Petit 1996), observed heterozygosity (Ho), and
Nei’s (1978) unbiased gene diversity (He ).
To gain insight into the role of bottlenecks in relation
to genetic diversity within populations, we conducted correlation analyses between log-transformed sample size (NG)
and AR, %P, A, and He. Among these parameters, allelic
richness (AR) is known to be highly dependent on effective
population size and is, therefore, more useful for identifying
historical processes such as bottlenecks (Luikart et al. 1998a;
Widmer and Lexer 2001). The correlation analyses were
conducted only for C. discolor because of small population
numbers of C. sieboldii (3 populations) and C. reflexa (2 populations). To test for recent decreases in effective population
size (bottlenecks), we evaluated individual loci for difference
between the H-W He and the equilibrium heterozygosity
(Heq) expected from the number of alleles assuming mutation-drift equilibrium. These differences were evaluated
using a sign test and a Wilcoxon sign-rank test conducted
across loci under an infinite allele model using the program
BOTTLENECK (Cornuet and Luikart 1996). Since allelic
diversity is generally lost more rapidly than He (Nei et al.
1975), recently bottlenecked populations will exhibit an
excess of H-W He relative to that Heq (Cornuet and Luikart
1996; Luikart et al. 1998b).
To measure deviations from H-W equilibrium, we calculated Wright’s (1965) FIS and FST over loci following Weir
and Cockerham (1984). These fixation indices measure the
average deviation from H-W equilibrium of individuals
relative to their local populations (FIS, a measure of local
inbreeding) and local populations relative to the total population (FST, also a measure of differentiation among local
populations). The significance of multi-population FIS and
FST estimates was determined by a permutation test (1000
randomizations of alleles among individuals within samples and 1000 randomizations of genotypes among populations, respectively). These calculations were performed
using FSTAT (Goudet 2002). Further, we used the program
SPAGeDi (Hardy and Vekemans 2002) to calculate population-level FIS (inbreeding) and its significance level by 1000
permutations under the null hypothesis of FIS = 0. To test
for differences between populations, we approximated 95%
CI around FIS as 1.96 times the jackknifed standard error
under the assumption of normality (Loiselle et al. 1995;
Chung et al. 2011; Chung et al. 2012).
To determine whether populations of C. discolor are at
regional equilibrium between gene flow and drift (i.e., isolation by distance), we conducted a correlation analysis
(Hutchison and Templeton 1999) between pairwise FST and
linear geographic distance (kilometers) to determine the relative importance of gene flow and genetic drift. Hutchison
and Templeton (1999) formulated 3 cases of nonequilibrium
between gene flow and drift based on degrees of scatter in
plotted points between pairwise FST and linear geographic
distance (kilometers): 1) gene flow relatively more important
than genetic drift, 2) gene flow relatively less important than
genetic drift, and 3) gene flow more important at shorter
distances and drift more important at greater distances.
Using the FSTAT, we tested a linear regression model using
the Mantel test (Mantel 1967) by making 999 replicates. To
gain insight into the patterns of gene flow between individual populations, we estimated recent migration rates using
the program BayesAss v. 1.3 (Wilson and Rannala 2003).
Following Yamada and Maki (2012), we ran 3 × 106 Markov
chain Monte Carlo iterations, with a burn-in of 999 999 iterations and a sampling frequency of 2000 by setting delta to be
0.15 (the default value).
253
Journal of Heredity 2013:104(2)
Finally, using the POPGENE, we calculated Nei’s (1978)
unbiased genetic identity (I) and distance (D) between pairs
of populations of each species and between the 3 species to
determine the degree of genetic divergence among populations of the 3 Calanthe species. In addition, a neighbor-joining tree was drawn from genetic distance (Nei et al. 1983)
matrix with branch support produced by 1000 bootstrapping
over loci, using Populations v. 1.2.30 (Langella 1999) and
TREEVIEW v. 1.6 (Page 1996).
Results
Identification of Clones
For C. discolor, 15 out of the 17 putative resolved loci were
polymorphic whereas 2 (Per and Skdh) were monomorphic across the 9 populations. For C. sieboldii, 13 loci were
polymorphic across the 3 studied populations, and 4 (Adh,
F1,6, Per, and Skdh) were monomorphic. Compared with the
other 2 species, relatively low polymorphism was detected in
the 2 populations of C. reflexa: only 8 loci were polymorphic, whereas the remaining 9 were monomorphic (Adh, Fe,
Lap, Mdh-2, 6Pgd, Pgm, Tpi-1, Tpi-2, and Tpi-3). Most of the
MLGs at 14 populations of the 3 species had Pgen FIS and
Psex FIS values less than 0.05 (Table 1). Thus, we considered
that ramets sharing identical MLGs were repeated samples
from the same clone and identified a total of 337 MLGs
(NG) out of 682 total samples (NT) (Table 1). For C. discolor,
estimates of genotypic richness (R) were relatively low and
ranged from 0.28 (CD-3 and CD-5) to 0.69 (CD-4), with a
mean of 0.47. Slightly higher estimates were found for C. sieboldii (mean = 0.62) and C. reflexa (mean = 0.59; Table 1).
All Simpson diversity indices (D) were more than 0.87, and
showed similar values across populations and among species
(with species’ means between 0.93 and 0.96; Table 1). Except
for CD-4 of C. discolor, all values of the Simpson evenness
index (ED) were over 0.75, with a mean of 0.84 for both
C. discolor and C. sieboldii and 0.86 for C. reflexa (Table 1). For
C. discolor, slopes of Pareto distribution (β, –1 × regression
slope, b) were highly variable, ranging from 0.351 (CD-5) to
1.860 (CD-1) with a mean of 0.898 (Table 1). Since population CD-1 had high R and ED estimates (i.e., clonal ramets
all having approximately equal sizes), it showed the highest
β value (i.e., a steep slope). Contrary to CD-1, CD-5 showed
the lowest R and ED values (i.e., a skewed clonal distribution
with very few, large clonal lineages and many small ones) and
the lowest βvalue (i.e., a shallow slope). Similar and moderate β values were found in C. sieboldii and C. reflexa (Table 1).
Except for two comparisons (CD-5 vs. CD-8 and CD-5 vs.
CR-2), the 95% CIs around b using classical least-squares
regression theory overlapped among populations (Table 1).
Table 3 Results of statistical tests for evidence of recent
population bottlenecks in the 3 Calanthe species
Species/population
Calanthe discolor
CD-1
CD-2
CD-3
CD-4
CD-5
CD-6
CD-7
CD-8
CD-9
Calanthe sieboldii
CS-1
CS-2
CS-3
Calanthe reflexa
CR-1
CR-2
Sign test
Wilcoxon sign-rank test
0.412
0.361
0.505
0.019
0.364
0.324
0.084
0.484
0.609
0.271
0.117
0.880
0.961
0.500
0.880
0.994
0.235
0.285
0.474
0.021
0.400
0.122
0.003
0.246
0.246
0.121
0.125
0.037
Numbers reported are P-values (probability for heterozygosity excess)
of sign and Wilcoxon sign-rank tests conducted using the program
BOTTLENECK.
alleles/locus (A) were 88.2 and 2.59, respectively, and the
observed (Ho) and expected heterozygosity (He) were 0.190
and 0.244, respectively (Table 2). Similar or slightly lower
values for these genetic diversity measures were obtained at
the population level: %P = 68.6, A = 2.01, Ho = 0.191, and
He = 0.227; Table 2). Populations CD-6 and CD-9 harbored
the lowest and highest allelic richness (AR = 1.81 and 2.07,
respectively), and populations CD-7 and CD-2 showed the
lowest and the highest expected heterozygosity (He = 0.179
and 0.283, respectively; Table 2). Similarly, high levels of genetic variation within populations and total samples as a whole
(i.e., at the species level) were found in C. sieboldii (Table 2).
Compared with the other two species, the two studied
populations (CR-1 and CR-2) of C. reflexa harbored much
lower levels of genetic variation with means of %P = 47.1,
AR = 1.50, A = 1.50, Ho = 0.129, and He = 0.185; Table 2).
For C. discolor, there was a significant correlation between
population size (NG) and mean A (Spearman’s rank correlation coefficient, rs = 0.789, P = 0.012). However, we found
no significant correlations for %P (rs = 0.433, P = 0.248),
AR (rs = 0.479, P = 0.192), and He (rs = 0.353, P = 0.352).
Consistent with these findings, we detected only 4 cases of
significant indications of recent bottlenecks in 3 populations
based on both sign (2 populations) and Wilcoxon sign-rank
tests (2 populations) conducted using BOTTLENECK
(Table 3).
Levels of Genetic Diversity in the 3 Calanthe Species
Population Genetic Structure and Inbreeding in the 3
Calanthe Species
We found high levels of genetic variation within populations
of C. discolor and C. sieboldii (Table 2). For C. discolor, when
we pooled samples over all populations (NG = 241), the percentage of polymorphic loci (%P) and the mean number of
For C. discolor, population-level FIS estimates calculated over
polymorphic loci did not differ significantly among populations (broadly overlapping 95% CIs; Table 2). All FIS values
were positive, ranging from 0.003 (CD-4) to 0.285 (CD-3),
254
Chung et al. • Genetic Variation and Structure within 3 Endangered Calanthe Species
Figure 2. Differentiation among C. discolor populations
in the Korean Peninsula. Multilocus estimates of pairwise
differentiation (FST) are plotted against linear geographic
distances (in kilometer) according to Hutchison and Templeton
(1999).
but 4 populations (CD-1, CD-3, CD-5, and CD-9) showed
significant FIS values at 0.05 level (Table 2). These results,
as well as a significant multipopulation-level FIS (FIS = 0.176,
P = 0.001; Table 2), indicate a deficit of heterozygotes within
populations that is consistent with local inbreeding. A significant excess of homozygotes within all the populations of
C. sieboldii and C. reflexa was detected (multipopulation-level
FIS estimates were 0.296 and 0.299, respectively; Table 2).
Differences in allele frequencies calculated over polymorphic
loci among populations of both C. discolor and C. sieboldii were
low but significant (FST = 0.068 and 0.072, both P = 0.001).
However, we found no significant differences between two
adjacent populations of C. reflexa on Jeju Island (FST = 0.006,
P = 0.614).
For C. discolor, we found a very weak and nonsignificant
linear relationship between pairwise genetic distance and
linear geographic distance (r = –0.184, P = 0.156), suggesting that populations are not at regional equilibrium of gene
flow and drift (Figure 2). Consistent with this, the neighbor-joining tree showed little association between populations of C. discolor in relation to their geographic location
(Figure 3). Results of BayesAss analysis suggested recent
gene flow from CD-3 to CD-4 (the linear distance between
them is 77.3 km; Figure 1) and from CD-1 to CD-9 (with
a linear distance of 282.2 km, the longest distance among
population pairs; Figure 1) (Table 4). If the patterns of gene
flow follow an isolation-by-distance model, we would expect
a significant negative relationship between pairwise recent
migration rates and linear geographic distances among populations. This prediction is not the case in C. discolor: a Mantel
test showed a positive and significant correlation between
pairwise migration rates below the diagonal of Table 4 (i.e.,
from CD-1 to CD-2, etc.) and linear distances (r = 0.338,
P = 0.040) and a negative but not significant correlation
between pairwise migration rates above the diagonal (i.e.,
from CD-2 to CD-1, etc.) and linear distances (r = –0.173,
P = 0.311).
Figure 3. Neighbor-joining tree of the 3 Calanthe species
based on Nei et al.’s (1983) genetic distances between
populations of C. discolor (CD-1 to CD-9), C. sieboldii (CS-1 to
CS-3), and C. reflexa (CR-1and CR-2). Numbers above branches
represent bootstrap support for 1000 replicates, and only
values greater than 50% are shown.
Pairwise Nei’s (1978) mean I values for conspecific populations of the 3 species were high (0.956–0.995; Table 5), which
falls well within the range of values for most conspecific
orchid populations (average I = 0.955 ± 0.051, N = 84; Chung
MY and Chung MG 2012). Mean values for congeneric pairs
ranged from 0.504 (C. sieboldii vs. C. reflexa) to 0.844 (C. discolor vs. C. sieboldii) (Table 5), which are higher than the average value reported for congeneric orchid species pairs (average
I = 0.453 ± 0.274, N = 190; Chung MY and Chung MG 2012).
Genetic differences between species are primarily due to the
occurrence of specific (i.e., unique) alleles: C. reflexa had 8
species-specific alleles (allele frequencies from 0.161 at Pera to
1.000 at Lapa; Table 6), C. discolor had 7 unique alleles with 5
of them at low frequencies (<0.05), and C. sieboldii also had
7 alleles with 3 of them at low frequencies (<0.05) (Table 6).
Discussion
Genetic Diversity and Structure in the 3 Calanthe Species
Calanthe discolor and C. sieboldii should be regarded as variable species, since their levels of within-population genetic
255
Journal of Heredity 2013:104(2)
Table 4 Mean value of the posterior distribution of the recent migration rate of each Calanthe discolor population estimated from
allozyme data using the BayesAss program
From
To
CD-1
CD-2
CD-3
CD-4
CD-5
CD-6
CD-7
CD-8
CD-9
CD-1
CD-2
CD-3
CD-4
CD-5
CD-6
CD-7
CD-8
CD-9
0.8850
0.0094
0.0094
0.0295
0.0123
0.0744
0.0112
0.0038
0.1934
0.0225
0.8964
0.0229
0.0138
0.0385
0.0241
0.0057
0.0017
0.0279
0.0079
0.0110
0.9254
0.1159
0.0118
0.0149
0.0073
0.0038
0.0150
0.0021
0.0049
0.0035
0.6949
0.0049
0.0109
0.0056
0.0016
0.0066
0.0033
0.0296
0.0059
0.0253
0.7809
0.0400
0.0602
0.0037
0.0145
0.0034
0.0066
0.0098
0.0133
0.0059
0.7091
0.0083
0.0019
0.0267
0.0032
0.0247
0.0065
0.0255
0.0965
0.0631
0.8560
0.0029
0.0093
0.0696
0.0121
0.0064
0.0688
0.0144
0.0529
0.0390
0.9534
0.0210
0.0029
0.0053
0.0103
0.0130
0.0348
0.0106
0.0068
0.0271
0.6855
Values on the diagonal in bold underlined are the proportions of individuals derived from source populations. Values higher than 0.1 are represented in bold.
Table 5 Average Nei’s (1978) identity values and ranges (in
parenthesis) for pairwise comparisons of the 3 Calanthe species.
Mean intraspecific identities are given on the diagonal
C. discolor
C. sieboldii
C. reflexa
C. discolor
C. sieboldii
C. reflexa
0.956
(0.898–0.991)
0.844
(0.780–0.886)
0.965
(0.954–0.978)
0.559
(0.522–0.601)
0.504
(0.463–0.535)
0.995
diversity are substantially higher than those typical of the terrestrial orchids examined by allozymes (means %P = 46.2,
A = 1.83, He = 0.119; Case 2002). Compared with C. discolor
and C. sieboldii, populations of C. reflexa harbor lower levels
of within-population variation, but they still exhibit similar
levels to the abovementioned reference values.
The significant deficit of heterozygotes within 4 populations of C. discolor and in all populations of C. sieboldii and
C. reflexa could be caused by nonrandom mating with relatives (from limited seed dispersal at finer spatial scale and
short pollinator movements) and/or geitonogamous pollination (selfing) because the 3 species are self-compatible
(Chung MY and Chung MG, unpublished data). Our ongoing
study on the fine-scale genetic structure (FSGS) conducted in
CD-1 revealed a significant genetic structuring; bF, the regression slope of Fij (kinship coefficients; Loiselle et al. 1995),
estimated between individuals i and j on the natural logarithm
of distance interval, was significantly strong (bF = –0.037 and
a strength of FSGS, Sp = 0.0507; Chung MY and Chung
MG, unpublished data). If inbreeding was solely the result of
mating among neighboring plants, then we may expect FIS to
be approximately equal to Fij at the smallest distance interval (2 m) (Chung et al. 2011). We found higher FIS (0.244)
than Fij(2m) (0.203), suggesting that partial selfing could also
be a cause of the excess of homozygotes in the population
(Vekemans and Hardy 2004). In addition, since we collected
some samples from relatively large populations (from 0.4 to 1
ha; Table 2), sampling may have occurred over substructured
256
populations, which could also increase observed homozygosities in several populations (i.e., Wahlund effect). Finally, a
low degree of among-population genetic differentiation was
found in the 3 Calanthe species (FST = 0.006–0.072), estimates which are substantially below the average for orchids
(mean = 0.187 from FST or from analogous statistics, N = 76;
Forrest et al. 2004).
The observed high within-population variation and low
degree of genetic differentiation among populations indicate
that extant populations of the 3 Calanthe species were probably founded by multiple, genetically diverse individuals, and
that most of this initial diversity would have been maintained
in stable habitats in lowlands dominated by broad-leaved
evergreen shrubs and trees in southwestern and southern
Korea from the mid-Holocene optimum to the present (i.e.,
during the past 6000 years; Choi 1998; Yi 2011). Before the
early 1970s, populations of C. discolor were large and relatively continuously distributed in southern Korea, including
many islands (Cheon CP, personal communication). The
same seems to be true for C. sieboldii and C. reflexa on Jeju
Island (Kim CS, personal communication). The low FST estimates found in the 3 species suggest that historical gene flow
among neighboring populations of these terrestrial orchids
has been high on the Korean Peninsula. If this is true, we
should have found a significant correlation between genetic
distance and geographic distance among populations of
C. discolor (i.e., regional equilibrium between gene flow and
drift). However, the graphic representation of the relationship between genetic and geographic distances found for
this species resembles Hutchison and Templeton’s (1999)
theoretical “model III” (i.e., “lack of regional equilibrium,
with genetic drift much more influential than gene flow”).
Moreover, results from BayesAss revealed that the patterns
of recent gene flow do not strictly follow an isolation-bydistance model in C. discolor. Unlike C. discolor, we found a
significant positive linear relationship between pairwise FST
and linear geographic distance among relatively undisturbed
and continuously distributed populations of the terrestrial
orchid Oreorchis patens across montane areas in South Korea
(r = 0.541, P = 0.000; Chung et al. 2012).
Chung et al. • Genetic Variation and Structure within 3 Endangered Calanthe Species
Table 6 Average allele frequencies for 11 polymorphic loci harboring unique alleles (boldface) for each of the 3 Calanthe species
Locus
Allele
CD
Dia
a
b
c
d
e
0.237
a
b
c
d
e
0.025
0.823
0.152
a
b
c
0.789
0.211
Fe
Idh-1
Idh-2
Lap
a
b
c
d
e
a
b
c
0.763
CS
CR
0.463
0.573
0.666
0.004
0.271
0.059
0.712
0.288
0.965
0.035
0.626
0.270
0.104
0.912
0.088
0.691
0.309
0.122
0.878
Locus
Allele
CD
CS
CR
Mdh-1
a
b
c
d
0.127
0.683
0.164
0.026
0.342
0.612
0.046
0.017
0.840
0.143
6Pgd
a
b
c
d
0.880
0.076
0.039
0.005
0.908
0.092
1.000
Per
a
b
c
1.000
1.000
a
b
c
d
e
0.078
0.830
0.092
0.591
0.374
0.009
0.004
0.022
Skdh
a
b
1.000
1.000
Tpi-1
a
b
c
d
0.063
0.897
0.040
0.027
0.947
1.000
0.644
0.357
Pgm
0.401
0.599
1.000
0.161
0.840
1.000
0.318
0.682
1.000
0.026
Abbreviations: CD, Calanthe discolor; CS, C. sieboldii; CR, C. reflexa
Inference of Population-Establishment History in the 3
Calanthe Species
Levels of genetic diversity found in the 3 Calanthe species
(relatively high within-population genetic variation and low
interpopulation differentiation) are consistent with the first
scenario: extant populations were founded by colonizers that
arrived from multiple source populations (i.e., from multiple putative glacial refugia). Furthermore, the species have
likely maintained relatively large populations during postglacial times and high historical rates of gene flow among
populations. Most populations did not experience recent bottlenecks although in general larger populations harbor more
alleles suggesting some gene drift is occurring. However,
many of these populations were once large, which retards
the loss of alleles through random genetic drift (Sun 1996),
enabling the species to retain high levels of genetic variation.
Our data do not allow inference regarding the locations
of original source populations of the Korean populations of
the 3 Calanthe species, although southern Japan is one of the
most likely sources. Refugia for much of warm-temperate
and subtropical flora during the LGM have been proposed
to occur in Kyushu and in southern Honshu (Tsukada 1984;
Hattori 1985; Matsuoka and Miyoshi 1998; Aoki et al. 2004;
Gotanda and Yasuda 2008). Orchids produce a large number
of dust-like seeds that may be transported long distance on
air currents (Dressler 1981). Thus, the mode of migration for
the orchid species is most likely through long-distance dispersal of their minute seeds by wind. Therefore, migrations
via seeds from southern China, Taiwan, or even from some
locations offshore in the southern part of the East China
Sea (as this sea was largely exposed due to the marine glacial
regressions until at least 10 000 years ago; Xu et al. 2010)
cannot be ruled out (see Harrison et al. 2001; Hope et al.
2004; Prentice et al. 2011). Finally, we might also consider
the remote possibility of recolonization from small, scattered hospitable pockets (“microrefugia”; Rull 2009) that
would have been located within Korea; however, as explained
above, it is unlikely that warm-temperate vegetation would
had remained within the Korean Peninsula during the LGM.
Test of the 2 Proposed Scenarios in Other
Warm-Temperate Plants in Southern Korea
To gain insight into the origin of the warm-temperate elements of the current Korean vegetation, it is of great interest
to check which of the 2 scenarios fits better with previously
studied species from southern Korea (Table 7). First, as seen
in Table 7, the terrestrial orchid Cymbidium goeringii and the 3
broad-leaved evergreen trees Camellia japonica, Eurya japonica,
and E. emarginata would fit the first scenario along with the
results of the 3 Calanthe species. All these species share some
life history and ecological characteristics, such as occurring
in historically large and continuously distributed populations,
and having a high potential for gene flow; C. goeringii, like most
orchids, has a high potential for long-distance seed dispersal
(Chung MY and Chung MG 1999), in C. japonica the pollen is
carried by birds (Wendel and Parks 1985), and in Eurya japonica
257
Journal of Heredity 2013:104(2)
Table 7 Summary of mean allozyme diversity statistics for plant species in subtropical and warm-temperate regions in the Korean
Peninsula
Species
Distributiona Family
NP
NL
%PP
Ardisia japonica
Bulbophyllum
drymoglossum
Calanthe discolor
Calanthe sieboldii
Calanthe reflexa
Camellia japonica
Cymbidium goeringii
Eurya japonica
Eurya emarginata
Hemerocallis
hongdoensis
Lycoris sanguinea
Oreorchis coreana
Sarcanthus
scolopendrifolius
Tipularia japonica
Average
C, J, K
J, K
Myrsinaceae
Orchidaceae
11
2
18
21
18.7
2.4
C, J, K
C, J, K
C, J, K
J, K
C, J, K
C, I, J, K, M
C, J, K
J, K
Orchidaceae
Orchidaceae
Orchidaceae
Theaceae
Orchidaceae
Theaceae
Theaceae
Liliaceae
9
3
2
17
16
20
6
3
17
17
17
16
14
12
18
15
J, K
K
J, K
Amaryllidaceae 7
Orchidaceae
4
Orchidaceae
3
J, K
Orchidaceae
8
7.9
AP
AS
HeP
HeS
GST or FST Ref.b
50.0
4.8
1.25
1.03
2.11
1.05
0.061
0.011
0.194
0.016
0.533
0.253
1
2
68.6
66.7
47.1
68.4
63.0
94.1
73.2
55.6
88.2
76.5
47.1
83.0
71.0
100.0
78.0
75.0
2.01
1.96
1.50
2.63
2.08
3.79
2.22
2.11
2.59
2.35
1.53
4.29
2.71
6.17
3.17
3.25
0.227
0.280
0.185
0.265
0.240
0.462
0.296
0.165
0.244
0.293
0.186
0.322
0.251
0.496
0.321
0.177
0.068
0.072
0.006
0.129
0.098
0.069
0.079
0.058
3
3
3
4
5
6
7
8
19
17
21
15.3
0.0
1.6
21.1
0.0
4.8
1.17
1.00
1.02
1.26
1.00
1.10
0.052
0.000
0.002
0.058
0.000
0.015
0.092
na
0.899
9
10
2
24
18
0.0
41.1
0.0
50.0
1.00
1.77
1.00
2.40
0.000
0.160
0.000
0.184
na
0.196
11
%PS
Except Hemerocallis hongdoensis, Lycoris sanguinea, and Oreorchis coreana, most plant species listed are widely distributed in subtropical and warm-temperate
regions, largely in southern China, central and southern Japan, and southwestern and southern Korea. Subscripts P and S represent population and total
sample levels, respectively.
NP, number of populations sampled; NL, number of loci analyzed; GST, the average amount of variation distributed among populations (Nei 1973, 1977).
aDistribution: C, China; I, India; J, Japan; K, Korea; M, Malaysia.
bSource references: 1, Cheon et al. (2000); 2, Chung et al. (2007); 3, this study; 4, Chung and Kang (1996); 5, Chung and Chung (1999); 6, Chung and Kang
(1994); 7, Chung and Kang (1995); 8, Kang and Chung (2000); 9, Chung (1999); 10, Chung et al. (2012); 11, Chung (2009).
and E. emarginata seed dispersal is via frugivorous birds (Manabe
and Yamamoto 1997; Kominami et al. 2003). In addition, the
liliaceous species Hemerocallis hongdoensis, which is restricted to
several islands (including the Korean Hong, Heuksan, Gageo,
and Jeju Islands, and Tsushima Island of Japan; Figure 1), also
harbors high levels of within-population variation and low
among-population genetic divergence (Table 7), but populations of this species are highly isolated on islands and are characterized by gravity seed dispersal and insect pollination.
Second, low levels of genetic variation within populations
and moderate or high genetic divergence among populations
were observed in Ardisia japonica, a very small shrub (height
approximately 10 cm; Cheon et al. 2000) and the 2 lithophytic
orchids Bulbophyllum drymoglossum and Sarcanthus scolopendrifolius (Table 7), which are characterized by small, highly isolated populations (Chung et al. 2007). Populations of these
species appear to have originated from fewer sources (perhaps from a single refugium), with low rates of recurrent
gene flow among the highly isolated Korean populations.
Lycoris sanguinea, whose seeds are dispersed by frugivorous
birds (Chung 1999), would also fit the second scenario in
that, after the colonization event, most (but not all) variation
would have been lost in the founding event and through multiple generations of small population sizes. For this species,
genetic differentiation among populations is low (Table 7)
because of its high potential for long-distance seed dispersal.
Additionally, recolonization from a single source may also
lead to no variation within- and among populations. This scenario (i.e., a severe founder effect) appears to explain the lack
258
of genetic diversity found in 2 terrestrial orchids, which are
characterized by small, highly isolated populations (Tipularia
japonica; Chung 2009), and rarity and endemism on Jeju Island
(Oreorchis coreana; Chung et al. 2012).
Implications for Conservation
Although C. discolor and C. sieboldii harbor high levels of
genetic variation and low interpopulation divergence, they
warrant implementation of conservation measures (both in
situ and ex situ) because of their rarity from a demographic
perspective (more human impacts on natural populations can
be anticipated). In particular, C. reflexa populations on Jeju
Island need special attention to preserve their current genetic
diversity, which is lower compared with that of the other 2
Calanthe species. Since C. reflexa is restricted to a single island
in Korea (Jeju Island, which has an area of less than 2000
km2) and is extremely rare there; it is essential that the current
threat according to the IUCN criteria (IUCN 2001) be evaluated. Because the “extent of occurrence” (“the area contained within the shortest continuous imaginary boundary”)
estimated for C. reflexa is less than 200 km2 and the “area of
occupancy” (“the area within its extent of occurrence which
is occupied by a taxon”) is very small (<20 km2), the species
can be categorized as “Endangered” (EN) in Korea following the criteria B1b(iii)c(iv). From a long-term conservationgenetics perspective, the extreme rarity and small population
sizes characteristic of C. reflexa, coupled with its relatively low
genetic diversity and that more human disturbance is likely
Chung et al. • Genetic Variation and Structure within 3 Endangered Calanthe Species
to occur (Jeju Island has become one of the most popular
tourist spots in South Korea), place it at a high risk of extinction. Thus, in situ and ex situ conservation efforts should be a
high priority for this species. First of all, we recommend all
known populations to be protected in situ by law to prevent
further decreases in population size (e.g., Ramsay and Stewart
1998). More importantly, detailed ecological studies including
germination ecology, seedling establishment, identification
of associated mycorrhizal fungi and demographic dynamics, should be immediately initiated to achieve an effective
conservation of this rare orchid (Swarts et al. 2009a, 2009b).
Conclusions and Perspectives
We found high (or relatively high) within-population genetic
variation and low interpopulation differentiation in the 3
Calanthe species, indicating that the postglacial recolonization
pattern of these 3 species fits the first scenario; extant populations are probably derived from multiple glacial refugia located
outside of the Korean Peninsula. Previous studies on genetic
diversity of other warm-temperate plant species in southern
Korea show mixed results in terms of population-establishment history, with some species falling into the first scenario
while others fell into the second scenario. The large variability
of genetic diversity values reported in these studies are in agreement with the very complex patterns of East Asia vegetation
Quaternary migrations revealed by paleoecological data (e.g.,
Takahara et al. 2000; Gotanda and Yasuda 2008; Ni et al. 2010;
Prentice et al. 2011) and by the few (but increasingly more) phylogeographic regional studies (reviewed in Qiu et al. 2011).
Unfortunately, there are no detailed molecular phylogeographic studies of warm-temperate plant species native
to Korea. To unravel the putative glacial refugia of these
taxa and their recolonization routes, future studies should
use cpDNA sequences or other adequate phylogeographic
markers (e.g. AFLP); of course a thorough sampling of the
populations occurring within the presumable refuges (southern Japan, southern and eastern continental China, Taiwan
Island, or even the Ryukyu Islands) is needed. As many
authors suggest (e.g., Médail and Diadema 2009; Verboom
et al. 2009; López-Pujol et al. 2011a, 2011b), an accurate
identification of glacial refugia is a high priority for developing regional conservation strategies because these are centers
of genetic diversity and key areas for the species persistence.
More importantly, this information would be crucial for the
design of future intercountry (i.e, between China, Japan,
and Korea) cooperative conservation strategies (e.g., reintroduction of rare and endangered species into and out of
each country). Such transboundary conservation schemes
have shown great success in some parts of the world, such
as Europe, southern Africa, and North America (King and
Wilcox 2008; Vasilijević and Pezold 2011).
Funding
Basic Science Program through the National Research
Foundation of Korea (KRF) funded by the Ministry of
Education, Science and Technology (MOEHRD) (KRF2007-521-C00280) to M.G.C.
Acknowledgments
The authors thank Jae Min Chung, Jin Seok Kim, Chung Pyo
Chun, and Chan Soo Kim for helping us in locating populations of the 3 Calanthe species; Naoto Sugiura for providing
his unpublished data on pollination biology of the 3 Calanthe
species; Beom Jin Shim and Myeong Soon Park for laboratory assistance; Ann C. Chung for checking the English on
an earlier version of the article; John D. Nason for discussing
the scenarios for colonization histories of warm-temperate
plant species in the Korean Peninsula; and Jim Hamrick
and Dorset Trapnell for providing comments and editing a
­current version of the article.
References
Adams JM, Faure H. 1997. Review and atlas of palaeovegetation: preliminary land ecosystem maps of the world since the Last Glacial Maximum. Oak
Ridge National Laboratory: Oak Ridge (TN); [cited 2012 Mar 24]. Available
from: http://www.esd.ornl.gov/projects/qen/adams1.html.
Aoki K, Suzuki T, Hsu TW, Murakami N. 2004. Phylogeography of the component species of broad-leaved evergreen forests in Japan, based on chloroplast DNA variation. J Plant Res. 117:77–94.
Arnaud-Haond S, Belkhir K. 2007. GENCLONE: a computer program to
analyze genotypic data, test for clonality and describe spatial clonal organization. Mol Ecol Notes. 7:15–17.
Arnaud-Haond S, Duarte CM, Alberto F, Serrão EA. 2007. Standardizing
methods to address clonality in population studies. Mol Ecol. 16:5115–5139.
Becheler R, Diekmann O, Hily C, Moalic Y, Arnaud-Haond S. 2010. The
concept of population in clonal organisms: mosaics of temporally colonized
patches are forming highly diverse meadows of Zostera marina in Brittany.
Mol Ecol. 19:2394–2407.
Case MA. 2002. Evolutionary patterns in Cypripedium: inferences from
allozyme analysis. In: Clark J, Elliott WM, Tingley G, Biro J, editors.
Proceedings of the 16th world orchid conference; 1999 April 28–May 2,
Vancouver: Richmond (BC): Vancouver Orchid Society. p. 192–202.
Cheliak WM, Pitel JP. 1984. Technique for starch gel electrophoresis of
enzyme from forest tree species. Information Report P1-X-42. Chalk River,
Ontario (Canada): Petawawa National Forestry Institute, Canadian Forestry
Service, Agriculture Canada.
Chen X, Cribb PJ, Gale SW. 2009. Calanthe R. Brown. In: Wu ZY, Raven PH,
Hong DY, editors. Flora of China. Vol. 25 (Orchidaceae). Beijing: Science
Press and St. Louis (MO): Missouri Botanical Garden Press. p. 292–309.
Cheon CP, Chung MY, Chung MG. 2000. Allozyme and clonal diversity in
Korean populations of Ardisia japonica and Ardisia crenata (Myrsinaceae). Isr
J Plant Sci. 48:239–245.
Choi K-R. 1998. The post-glacial vegetation history of the lowland in
Korean Peninsula. Korean J Ecol. 21:169–174.
Chung C-H. 2007. Vegetation response to climate change on Jeju Island,
South Korea, during the last deglaciation based on pollen record. Geosci J.
11:147–155.
Chung C-H. 2011. Hococene vegetation dynamics and its climatic implications inferred from pollen record in Boseong area, South Korea. Geosci J.
15:257–264.
Chung C-H, Lim HS, Yoon HI. 2006. Vegetation and climate changes during
the Late Pleistocene to Holocene inferred from pollen record in Jinju area,
South Korea. Geosci J. 10:423–431.
259
Journal of Heredity 2013:104(2)
Chung C-H, Lim HS, Lee HJ. 2010. Vegetation and climate history during the late Pleistocene and early Holocene inferred from pollen record in
Gwangju area, South Korea. Quatern Int. 227:61–67.
Chung MG. 1999. Allozyme diversity in the endangered herb Lycoris sanguinea
var. koreana (Amaryllidaceae). Ann Bot Fennici. 36:1–7.
Chung MG, Kang SS. 1994. Genetic variation and population structure in
Korean populations of Eurya japonica (Theaceae). Am J Bot. 81:1077–1082.
Chung MG, Kang SS. 1995. Allozyme diversity and genetic structure in
Korean populations of Eurya emarginata (Theaceae). Jpn J Genet. 70:387–398.
Chung MG, Kang SS. 1996. Genetic variation within and among populations
of Camellia japonica (Theaceae) in Korea. Can J For Res. 26:537–542.
Chung MY. 2009. Lack of allozyme diversity in populations of the rare,
endangered terrestrial orchids Tipularia japonica and Epipactis papillosa in
Korea. Plant Syst Evol. 278:203–209.
Chung MY, Chung MG. 1999. Allozyme diversity and population structure
in Korean populations of Cymbidium goeringii (Orchidaceae). J Plant Res.
112:139–144.
Chung MY, Chung MG. 2012. A review of the use of genetic markers
in orchid systematics with emphasis on allozymes. Biochem Syst Ecol.
41:62–73.
Chung MY, López-Pujol J, Maki M, Kim K-J, Chung JM, Sun B-Y, Chung
MG. 2012. Genetic diversity in the common terrestrial orchid Oreorchis patens
and its rare congener Oreorchis coreana: inference of species evolutionary history and implications for conservation. J Hered. 103:692–702
Chung MY, Nason JD. 2007. Spatial demographic and genetic consequences
of harvesting within populations of the terrestrial orchid Cymbidium goeringii.
Biol Conserv. 137:125–137.
Chung MY, Nason JD, Chung MG. 2011. Significant demographic and finescale genetic structure in expanding and senescing populations of the terrestrial orchid Cymbidium goeringii (Orchidaceae). Am J Bot. 98:2027–2039.
Chung MY, Park C-W, Chung MG. 2007. Extremely low levels of allozyme
variation in southern Korean populations of the two rare and endangered
lithophytic or epiphytic Bulbophyllum drymoglossum and Sarcanthus scolopendrifolius
(Orchidaceae): implications for conservation. Biodiv Conserv. 16:775–786.
Clayton JW, Tretiak DN. 1972. Amine citrate buffers for pH control in starch
gel electrophoresis. J Fisher Res Board Canad. 29:1169–1172.
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.
COSEWIC (Committee on the Status of Endangered Wildlife in Canada).
2010. COSEWIC assessment and status report on the purple twayblade
Liparis liliifolia in Canada. Ottawa (Canada): Committee on the Status of
Endangered Wildlife in Canada. Xii + 25 pp.
Curtis JT. 1943. Germination and seedling development in five species of
Cypripedium L. Am J Bot. 30:199–205.
Dorken ME, Eckert CG. 2001. Severely reduced sexual reproduction in northern populations of a clonal plant, Decodon verticillatus (Lythraceae). J Ecol.
89:339–350.
Dressler RL. 1981. The orchids: natural history and classification. Cambridge
(MA): Harvard University Press.
Eckert CG, Samis KE, Lougheed SC. 2008. Genetic variation across species’
geographical ranges: the central-marginal hypothesis and beyond. Mol Ecol.
17:1170–1188.
El Mousadik A, Petit RJ. 1996. High level of genetic differentiation for allelic
richness among populations of the argan tree [Argania spinosa (L.) Skeeels]
endemic to Morocco. Theor Appl Genet. 92:832–839.
Falk DA, Holsinger KE, editors. 1991. Genetics and conservation of rare
plants. Oxford: Oxford University Press.
Forrest AD, Hollingsworth ML, Hollingsworth PM, Sydes C, Bateman RM.
2004. Population genetic structure in European populations of Spiranthes
260
romanzoffiana set in the context of other genetic studies on orchids. Heredity.
92:218–227.
Gotanda K, Yasuda Y. 2008. Spatial biome changes in southwestern Japan
since the Last glacial Maximum. Quatern Int. 184:84–93.
Goudet J. 2002. FSTAT, a program to estimate and test gene diversities
and fixation indices (version 2.9.3.2). [cited 2011 Aug 18]. Available from:
http://www.unil.ch/izea/softwares/fstat.html.
Hamrick JL, Godt MJW. 1989. Allozyme diversity in plant species. In: Brown
AHD, Clegg MT, Kahler AL, Weir BS, editors. Plant population genetics,
breeding and genetic resources. Sunderland (MA): Sinauer Associates.
p. 43–63.
Hamrick JL, Godt MJW. 1996. Conservation genetics of endemic plant species. In: Avise JC, Hamrick JL, editors. Conservation genetics. New York:
Chapman & Hall. p. 281–304.
Hamrick JL, Nason JD. 1996. Consequences of dispersal in plants. In:
Rhodess OE Jr, Chesser RK, Smith MH, editors. Population dynamics
in ecological space and time. Chicago (IL): University of Chicago Press.
p. 203–236.
Hardy OJ, Vekemans X. 2002. SPAGeDi: a versatile computer program to
analyze spatial genetic structure at the individual or population levels. Mol
Ecol Notes. 2:618–620.
Harrison SP, Yu G, Takahara H, Prentice IC. 2001. Diversity of temperate
plants in east Asia. Nature. 413:129–130.
Hattori T. 1985. Synecological study on the lucidophyllous forest of
Castanopsis-Persea type in Japan proper. Bull Kobe Geobot Soc. 1:1–98.
(In Japanese with English abstract.)
Haufler CH. 1985. Enzyme variability and modes of evolution in Bommeria
(Pteridaceae). Syst Bot. 10:92–104.
Hewitt GM. 1999. Post-glacial re-colonization of European biota. Biol J
Linn Soc. 68:87–112.
Hewitt GM. 2000. The genetic legacy of the Quaternary ice ages. Nature.
405:907–913.
Hope G, Kershaw AP, van der Kaars S, Sun X, Liew P-M, Heusser LE,
Takahara H, McGlone M, Miyoshi N, Moss PT. 2004. History of vegetation and habitat change in the Austral-Asian region. Quatern Int.
118–119:103–126.
Hou X-Y. 1983. Vegetation of China with reference to its geographical distribution. Ann Missouri Bot Gard. 70:509–548.
Hurlbert SH. 1971. The nonconcept of species diversity: a critique and alternative parameters. Ecology. 52:577–586.
Hutchison DW, Templeton AR. 1999. Correlation of pairwise genetic and
geographic distance measures: inferring the relative influences of gene flow
and drift on the distribution of genetic variability. Evolution. 53:1898–1914.
Ishikawa R, Suzuki K. 1992. Tetralonia nipponensis. The Insectarium 29:126.
(In Japanese.)
IUCN. 2001. IUCN red list categories: version 3.1. Prepared by the IUCN
species survival commission. Gland, Switzerland and Cambridge (UK):
IUCN. [cited 2010 Oct 16]. Available from: http://www.iucn.org/themes/ssc
Kang SS, Chung MG. 2000. High levels of allozyme variation within populations and low allozyme divergence within and among species of Hemerocallis
(Liliaceae). Am J Bot. 87:1634–1646.
King B, Wilcox S. 2008. Peace Parks and jaguar trails: transboundary conservation in a globalizing world. GeoJournal. 71:221–231.
Kominami Y, Sato T, Takeshita K, Manabe T, Endo T, Endo A, Noma N.
2003. Classification of bird-dispersed plants by fruiting phenology, fruit size,
and growth form in a primary lucidophyllous forest: an analysis, with implications for the conservation of fruit-bird interactions. Ornithol Sci. 2:3–23.
Kong W-S. 1996. The distribution of dicotyledons in time and space in the
Korean peninsula. Korean J Quatern Res. 10:1–18. (In Korean with English
abstract.)
Chung et al. • Genetic Variation and Structure within 3 Endangered Calanthe Species
Kong W-S, Watts D. 1993. The plant geography of Korea with an emphasis on
the alpine zones. Dordrecht (The Netherlands): Kluwer Academic Publishers.
Langella O. 1999. Populations version 1.2.30. Available from: http://bioinformatics.org/~tryphon/populations/.
Lee J-S, Choi B-H. 2006. Distribution and red data of wild orchids in the
Korean Peninsula. Korean J Plant Taxon. 36: 335–360. (In Korean with
English abstract.)
Lee NS. 2011. Illustrated flora of Korean orchids. Seoul: Ewha Womans
University Press. (In Korean.)
Loiselle BA, Sork VL, Nason JD, Graham C. 1995. Spatial genetic structure
of a tropical understory shrub, Psychotria officinalis (Rubiaceae). Am J Bot.
82:1420–1425.
López-Pujol J, Zhang F-M, Sun H-Q, Ying T-S, Ge S. 2011a. Centres of
plant endemism in China: places for survival or for speciation? J Biogeogr.
38:1267–1280.
López-Pujol J, Zhang F-M, Sun H-Q, Ying T-S, Ge S. 2011b. Mountains
of southern China as ‘plant museums’ and ‘plant cradles’: evolutionary and
conservation insights. Mountain Res Develop. 31:261–269.
Loveless MD, Hamrick JL. 1984. Ecological determinants of genetic structure in plant populations. Annu Rev Ecol Syst. 15:65–95.
Luikart G, Allendorf FW, Cornuet JM, Sherwin WB. 1998a. Distortion of
allele frequency distributions provides a test for recent population bottlenecks. J Hered. 89:238–247.
Luikart G, Sherwin WB, Steele BM, Allendorf FW. 1998b. Usefulness of
molecular markers for detecting population bottlenecks via monitoring genetic change. Mol Ecol. 7:963–974.
Manabe T, Yamamoto S-I. 1997. Spatial distribution of Eurya japonica in an
old-growth evergreen broad-leaved forest, SW Japan. J Veg Sci. 8:761–772.
Mantel N. 1967. The detection of disease clustering and a generalized
regression approach. Cancer Res. 27:209–220.
Matsuoka K, Miyoshi N. 1998. Chapter III-4. In: Yasuda Y, Miyoshi N, editors. The illustrated vegetation history of the Japanese Archipelago. Tokyo
(Japan): Asakura-shoten. p. 224–236. (In Japanese.)
Médail F, Diadema K. 2009. Glacial refugia influence plant diversity patterns
in the Mediterranean Basin. J Biogeogr. 36:1333–1345.
Phillips RD, Brown AP, Dixon KW, Hopper SD. 2011. Orchid biogeography
and factors associated with rarity in a biodiversity hotspot, the Southwest
Australian floristic region. J Biogeogr. 38:487–501.
Pielou EC. 1969. An introduction to mathematical ecology. New York:
Wiley-Interscience.
Prentice IC, Harrison SP, Bartlein PJ. 2011. Global vegetation and terrestrial
carbon cycle changes after the last ice age. New Phytol. 189:988–998.
Qiu Y-X, Fu C-X, Comes HP. 2011. Plant molecular phylogeography in
China and adjacent regions: tracing the genetic imprints of Quaternary climate and environmental change in the world’s most diverse temperate flora.
Mol Phylogenet Evol. 59:225–244.
Ramsay MM, Stewart J. 1998. Re-establishment of the lady’s slipper orchid
(Cypripedium calceolus L.) in Britain. Bot J Linn Soc. 126:173–181.
Rull V. 2009. Microrefugia. J Biogeogr. 36:481–484.
Seaton PT, Hu H, Perner H, Pritchard HW. 2010. Ex situ conservation of
orchids in a warming world. Bot Rev. 76:193–203.
Shi Y. 2002. Characteristics of late Quaternary monsoonal glaciation on the
Tibetan Plateau and in East Asia. Quatern Int. 97–98:79–91.
Soltis DE, Haufler CH, Darrow DC, Gastony GJ. 1983. Starch gel electrophoresis of ferns: a compilation of grinding buffers, gel and electrode buffers, and staining schedules. Am Fern J. 73:9–27.
Sun M. 1996. Effects of population size, mating system, and evolutionary
origin on genetic diversity in Spiranthes sinensis and S. hongkongensis. Conserv
Biol. 10:785–795.
Swarts ND, Dixon KW. 2009a. Terrestrial orchid conservation in the age of
extinction. Ann Bot. 104:543–556.
Swarts ND, Dixon KW. 2009b. Perspectives on orchid conservation in
botanic gardens. Trends Plant Sci. 14:590–598.
Takahara H, Sugita S, Harrison SP, Miyoshi N, Morita Y, Uchiyama T. 2000.
Pollen-based reconstructions of Japanese biomes at 0, 6000 and 18,000 14C
yr BP. J Biogeogr. 27:665–683.
Tsukada M. 1984. A vegetation map in the Japanese Archipelago approximately 20,000 years B.P. Jpn J Ecol. 34:203–208. (In Japanese with English
abstract.)
Mitton JB, Linhart YB, Sturgeon KB, Hamrick JL. 1979. Allozyme polymorphisms detected in mature needle tissue of ponderosa pine. J Hered. 70:86–89.
Vasilijević M, Pezold T., editors. 2011. Crossing borders for nature. European
examples of transboundary conservation. Gland (Switzerland) and Belgrade
(Serbia): IUCN Programme Office for South-Eastern Europe.
Nei M. 1973. Analysis of gene diversity in subdivided populations. Proc Natl
Acad Sci USA. 70:3321–3323.
Vekemans X, Hardy OJ. 2004. New insights from fine-scale spatial genetic
structure analyses in plant populations. Mol Ecol. 13:921–935.
Nei M. 1977. F-statistics and analysis of gene diversity in subdivided populations. Ann Hum Genet. 41:225–233.
Verboom GA, Archibald JK, Bakker FT, Bellstedt DU, Conrad F, Dreyer LL,
Forest F, Galley C, Goldblatt P, Henning JF et al. 2009. Origin and diversification of the Greater Cape flora: ancient species repository, hot-bed of
recent radiation, or both? Mol Phylogenet Evol. 51:44–53.
Nei M. 1978. Estimation of average heterozygosity and genetic distance
from a small number of individuals. Genetics. 89:583–590.
Nei M, Maruyama T, Chakraborty R. 1975. The bottleneck effect and genetic variability in populations. Evolution. 29:1–10.
Nei M, Tajima F, Tateno Y. 1983. Accuracy of estimated phylogenetic trees
from molecular data. II. Gene frequency data. J Mol Evol. 19:153–170.
Ni J, Yu G, Harrison SP, Prentice IC. 2010. Paleovegetation in China during
the late Quaternary: Biome reconstructions based on a global scheme of
plant functional types. Palaeogeogr Palaeoclimatol Palaeoecol. 289:44–61.
Nicolè F, Brzosko E, Till-Bottraud I. 2005. Population viability analysis of
Cypripedium calceolus in a protected area: longevity, stability and persistence.
J Ecol. 93:716–726.
Ohwi J. 1965. Flora of Japan. Washington (DC): Smithsonian Institute.
Page RD. 1996. TreeView: an application to display phylogenetic trees on
personal computers. Comput Appl Biosci. 12:357–358.
Parks JC, Werth CR. 1993. A study of spatial features of clones in a population of bracken fern, Pteridium aquilinum (Dennstaedtiaceae). Am J Bot.
80:537–544.
Vereecken NJ, Dafni A, Cozzolino S. 2010. Pollination syndromes in mediterranean orchids—implications for speciation, taxonomy and conservation.
Bot Rev. 76:220–240.
Weeden NF, Wendel JF. 1989. Genetics of plant isozymes. In: Soltis DE,
Soltis PS, editors. Isozymes in plant biology. Portland (OR): Dioscorides
Press. p. 46–72.
Weir BS, Cockerham CC. 1984. Estimating F-statistics for the analysis of
population structure. Evolution. 38:1358–1370.
Wendel JF, Parks CR. 1985. Genetic diversity and population structure in
Camellia japonica L. (Theaceae). Am J Bot. 72:52–65.
Widmer A, Lexer C. 2001. Glacial refugia: sanctuaries for allelic richness, but
not for gene diversity. Trends Ecol Evol (Amst). 16:267–269.
Wilson GA, Rannala B. 2003. Bayesian inference of recent migration rates
using multilocus genotypes. Genetics. 163:1177–1191.
Wright S. 1965. The interpretation of population structure by F-statistics
with special regard to systems of mating. Evolution. 19:395–420.
261
Journal of Heredity 2013:104(2)
Wu ZY, Raven PH, Hong DY. editors. 2009. Flora of China. Vol. 25
(Orchidaceae). Beijing (China): Science Press and St Louis (MO): Missouri
Botanical Garden Press.
Xu D, Lu H, Wu N, Liu Z. 2010. 30 000-Year vegetation and climate change
around the East China Sea shelf inferred from a high-resolution pollen
record. Quatern. Int. 227:53–60.
Yamada T, Maki M. 2012. Impact of geographical isolation on genetic
differentiation in insular and mainland populations of Weigela coraeensis
(Caprifoliaceae) on Honshu and the Izu Islands. J. Biogeogr. 39:901–917.
Yeh FC, Yang RC, Boyle TBJ. 1999. POPGENE version 1.32, microsoft
window-based free ware for population genetic analysis. [cited 2011 May 14].
Available from: http://www.ualberta.ca/~fyeh/index.htm.
Yi S. 2011. Holocene vegetation responses to East Asian monsoonal
changes in South Korea. In: Blanco J, Kheradmand H, editors. Climate
262
change – geophysical foundations and ecological effects. Rijeka (Croatia):
InTech. p. 157–178.
Yi S, Kim S-J. 2010. Vegetation changes in western central region of
Korean Peninsula during the last glacial (ca. 21.1–26.1 cal kyr BP). Geosci J.
14:1–10.
Young A, Boyle T, Brown T. 1996. The population genetic consequences
of habitat fragmentation for plants. Trends Ecol Evol (Amst). 11:413–418.
Zar JH. 1999. Biostatistical analysis. Upper Saddle River (NJ): Prentice-Hall.
Received April 12, 2012; Revised September 7, 2012;
Accepted September 12, 2012
Corresponding Editor: Frank Andrew Jones