| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Systematics |
2Department of Biology, Faculty of Sciences, Kyushu University, 4-2-1 Ropponmatsu, Chuo-ku, Fukuoka 810-8560, Japan; 3Faculty of Agriculture, Ehime University, 3-5-7 Tarumi, Matsuyama, Ehime 790-8566, Japan; 4Department of Organismic and Evolutionary Biology, Harvard University, Cambridge, Massachusetts 02138 USA
Received for publication May 22, 2004. Accepted for publication January 24, 2005.
ABSTRACT
Dipterocarpaceae, trees that dominate tropical rain forests in Southeast Asia consist of many economically and ecologically important species. We determined partial sequences of the PgiC gene from species of Shorea, Hopea, Neobalanocarpus, and Parashorea to elucidate phylogenetic relationships among the species of these genera, which have been regarded as interrelated. The sequences generated a gene tree with better resolution than previous cpDNA trees. The PgiC tree is essentially consistent with cpDNA trees, except for the placement of Neobalanocarpus. The PgiC tree shows that Neobalanocarpus is nested within White Meranti of Shorea, whereas this genus forms a clade with Hopea in cpDNA trees. This conflict suggests that Neobalanocarpus is derived via hybridization between White Meranti of Shorea and Hopea. Species belonging to each of three timber groups (Yellow Meranti, Balau, and Red Meranti) within Shorea are monophyletic. Together they form a monophyletic clade distinct from White Meranti. Botanical sections within Red Meranti appear not to be monophyletic. An extensive number of shared polymorphisms among species and consequential lack of monophyly of intraspecific haplotypes are found in Red Meranti. Potential causes of this phenomenon, including persistence of ancestral polymorphisms and gene flow via interspecific hybridization, are discussed.
Key Words: Dipterocarpaceae • Hopea • interspecific hybridization • molecular phylogeny • Neobalanocarpus • PgiC • Shorea
The tropical rainforests in Southeast Asia are characterized by a high species diversity of trees (Whitmore, 1984
). The extreme floristic richness is largely due to co-occurrence of a great number of species within the same community (Whitmore, 1998
). In particular, Borneo has one of the highest species diversity of trees among the world's tropical rainforests. In lowland Southeast Asian tropical forests, dipterocarp species dominate the forest canopy (Ashton, 1988
). Because of their economic and ecological significance, the dipterocarp trees have served as representative species in a number of tropical biological studies. However, human impact on tropical rainforests has been increasing in the past half-century, and many primary forests have been degraded by logging and shifting cultivation (Richards, 1996
; Whitmore, 1998
).
Dipterocarpaceae consist of more than 500 species and are divided into three subfamilies, Dipterocarpoideae, Monotoideae, and Pakaraimoideae (Ashton, 1982
). Although the phylogenetic placement of Dipterocarpaceae within angiosperms has been problematic, a recent molecular phylogenetic analysis suggests that this family should be assigned to the order Malvales and that Sarcolaenaceae is the closest relative of the Dipterocarpaceae (Dayanandan et al., 1999
).
The Asian subfamily Dipterocarpoideae includes 13 genera and 470 species (Ashton, 1982
). Detailed taxonomic study by Ashton, initially focused on Borneo, substantially relied on androecium and bark characters for classification (Ashton, 1962
, 1963
, 1964
, 1967
). These subsequently led to a regional monograph (Ashton, 1982
). Ashton has retained most of the previous classification of Symington (1943)
, but some of the groups were reclassified at lower taxonomic rank. For Shorea, which we study here, Symington (1943)
divided this genus into four groups that are equivalent to the timber groups classified by timber characters (i.e., Balau, White Meranti, Yellow Meranti, and Red Meranti) and treated Pentacme as an independent genus being closely related to Shorea. Ashton (1982)
, primarily on the basis of shared fruit calyx characters and differences in androecium and bark morphology, reduced Doona Thw. and Pentacme A. DC. as sections within Shorea and divided the genus into 11 sections (see Fig. 1). Although Ashton treated all 11 sections as having equivalent status (section), Maury (1978
, 1979
; summarized in Maury-Lechon and Curtet, 1998
) argued, primarily on the basis of embryo and leaf epidermal characters, that some sections of Shorea have unequal hierarchic ranks. She concluded that Ashton's sections Doona, Pentacme, and Anthoshorea, Shorea and Richetioides, which correspond to Symington's informal groups White Meranti, Balau and Yellow Meranti, respectively, should have higher ranks than sections such as Ovalis and Rubella, which are members of the Red Meranti. Thus, a molecular phylogenetic approach would be helpful to clarify controversial relationships in the family.
|
; Kajita et al., 1998
; Kamiya et al., 1998
; Dayanandan et al., 1999
). The phylogenetic analyses based on the cpDNA data revealed two distinct clades within Dipterocarpoideae, corresponding to the two tribes, Dipterocarpeae and Shoreae. Those studies demonstrated that cpDNA phylogenies are largely consistent with morphological classification at higher taxonomic levels, but several generic circumscriptions conflict with each other. Kajita et al. (1998)
and Dayanandan et al. (1999)
showed that genus Shorea is not monophyletic; the White Meranti of Shorea together with Doona form a clade that is sister to the clade of Hopea and Neobalanocarpus, while other members of Shorea form a cluster with Parashorea. In other phylogenetic studies with large numbers of species of Shorea, Tsumura et al. (1996)
and Kamiya et al. (1998)
also supported the paraphyletic relationships of Shorea, but these studies could not resolve most of the intrageneric relationships due to an insufficient number of informative characters in the cpDNA.
Accordingly, to clarify the relationships at lower taxonomic levels, nuclear gene sequences are required to obtain additional sources of characters (Kamiya et al., 1998
). In addition, a comparison between cpDNA and nuclear DNA phylogenetic analyses sometimes provides a strongly conflicting signal due to hybridogenous genomic constitution (Shi et al., 2001
; Oh and Potter, 2003
; Mummenhoff et al., 2004
). Therefore, we use the nuclear gene PgiC, which encodes cytosolic phosphoglucose isomerase, an essential enzyme of glycolysis and gluconeogenesis, to clarify the phylogenetic relationships between Shorea and its allied genera, and among taxonomic groups within Shorea. In this study, we examined partial sequences of the PgiC gene from species of Shorea and its allied genera Hopea, Neobalanocarpus, and Parashorea. The aims of our study are to (1) elucidate phylogenetic relationships among the species of Shorea using nuclear PgiC sequences, (2) compare the nuclear PgiC phylogeny with cpDNA based phylogenies, and (3) investigate the utility of the PgiC sequences for phylogenetic reconstruction at lower taxonomic levels.
MATERIALS AND METHODS
Sample collection
Some of the samples were collected from the two permanent plots, the Canopy Biology Plot (8 ha) and the Long Term Ecological Research Plot (52 ha) in Lambir Hills National Park, Sarawak, Malaysia (Lee et al., 2002
). All individuals in the plots at Lambir have been tagged, mapped, and identified to species by Sarawak Forest Department staff in collaboration with the Smithsonian's Center for Tropical Forest Science, the Center for Ecological Research in Kyoto University and Osaka City University. DNA samples that we used in this study include those used in the cpDNA analysis of Kamiya et al. (1998)
. Additional leaf samples were collected from the Dipterocarp Arboretum in FRIM (Forest Research Institute Malaysia), and some DNA samples were kindly provided by S. L. Lee. A total of 78 accessions throughout 48 species of Shorea, representing all recognized sections of Ashton (1982)
except for the Sri Lanka endemic section Doona and also Pentacme, six species representing all the sections and subsections of Hopea, the monotypic genus Neobalanocarpus, and one species from Parashorea were collected. Dipterocarpus palembanicus subsp. bornensis (tribe Dipterocarpeae) was used as an outgroup taxon. Dipterocarpus is known as the closest sister genus of tribe Shoreae based on previous molecular phylogenies (Tsumura et al., 1996
; Kajita et al., 1998
; Kamiya et al., 1998
; Dayanandan et al., 1999
). Total DNA was extracted using the CTAB procedure of Doyle and Doyle (1990)
or DNeasy Plant Mini kit (Qiagen, Valencia, CA). Information on the specimens is shown in Table 1.
|
Data analyses
The nucleotide sequences were aligned with ClustalX (Thompson et al., 1997
) and then manually edited using SeqPup 0.6 (available at website http://iubio.bio.indiana.edu/soft/molbio/seqpup). The aligned sequence data were analyzed by maximum-parsimony and neighbor-joining (Saitou and Nei, 1987
) methods implemented in PAUP* 4b10 (Swofford, 2002
). A heuristic search was conducted with random sequence addition with 100 replicates, tree-bisection-reconnection (TBR) branch swapping, and the number of rearrangement was limited to 100 000 for each replication. All characters were equally weighted, and gaps were treated as missing data. Relative robustness for clades was examined using a bootstrap analysis (Felsenstein, 1985
) with 5000 replications of fast bootstrapping. Mort (2000)
indicated that fast bootstrapping (without branch swapping) generates a smaller value than the standard bootstrap analysis (with branch swapping), and this method is suitable for analyzing large data sets because of easier and faster computation. The neighbor-joining tree was constructed based on Kimura's two-parameter distance (1980)
, and 1000 replicates of bootstrap were performed to obtain cluster supports.
RESULTS
Characteristics of the PgiC sequences
PCR amplifications using the primer set produced mostly single clear bands in the agarose gel electrophoresis. Furthermore, sequences from multiple clones from one individual were homologous to the PgiC sequence of Arabidopsis in all cases, indicating nonspecific products were not generally amplified. Therefore, each PCR product must be a result of amplification from a single specific region coding for PGIC. The cloned sequences from some of the accessions contained polymorphic sites where the direct sequencing has double peaks, and thus such individuals are heterozygous at the locus. Length variations were rare within individuals, allowing us to determine heterozygous sites easily. As expected for diploid outcrossers, 63 of 84 (except for four accessions of polyploid species) had two different haplotypes per individual. Each individual of Shorea ovalis subsp. sericea, Hopea odorata, and H. subalata, which are known to be polyploid species (Ashton, 1982
), had three or four haplotypes, as expected with polyploidy. In total, we identified 161 PgiC haplotypes throughout 88 accessions. Most variable sites were found in introns or at synonymous sites. Newly determined DNA sequences were deposited in the DNA Data Bank of Japan database under accession numbers AB189478 to AB189638.
The PgiC sequences that we determined corresponded to exons 13 to 19 of the PgiC of Arabidopsis thaliana (e.g., Kawabe and Miyashita, 2000
). The coding sequences of the dipterocarp PgiC were about 85% identical to that of Arabidopsis, although the introns of dipterocarps and Arabidopsis could not be aligned. Numbers of exons and their lengths in Arabidopsis and dipterocarp were the same. Lengths of the sequences vary from 1233 to 1449 bp due to insertions and deletions (indels) found in the introns. Alignment of the sequences was unambiguous in both exons and introns, and the alignment matrix used for phylogenetic analyses contained 1597 characters.
Level of variation in PgiC gene sequences
As expected, sequences of the nuclear PgiC had a higher proportion of informative characters (21.5% with 141 sequences) than noncoding regions of cpDNA (3.8% with 30 sequences based on data obtained from Kamiya et al., 1998
) for the Shorea species. Sequence variation at the PgiC locus (2.4%) is about 2.5 times as high as that in the chloroplast noncoding region (0.9%), due to a lower substitution rate in chloroplast than nuclear genomes (Wolfe et al., 1987
).
Mean pairwise sequence divergences among species within each clade and among clades are shown in Table 2. Mean divergences between species range from 0.0089 to 0.0422 within clades. The divergences range from 0.0247 to 0.0585 between clades within the ingroup, and 0.0623 to 0.0711 between the ingroup and outgroup.
|
Phylogenetic relationships inferred from the PgiC sequences
One of the 6376 most parsimonious trees (length [L] = 827; CI = 0.748; RI = 0.921) is shown in Fig. 2 and the neighbor-joining (NJ) tree is shown in Fig. 3. Most parsimonious (MP) and NJ trees are basically congruent with each other, with some differences between poorly supported nodes. Figure 2 shows several polytomies where supports for the corresponding nodes in the NJ tree are low. The analyses using Dipterocarpus as an outgroup identify six clades: I, White Meranti of Shorea + Neobalanocarpus; II, Hopea; III, Parashorea; IV, Yellow Meranti of Shorea; V, Balau of Shorea; and VI, Red Meranti of Shorea, with diversification of the groups estimated to have occurred in this order. All of these clades are supported by >70% of bootstrap values (BS), except for clade I (BS = 62% in the MP tree and 84% in the NJ tree), and clade II (BS = 50% in the MP and 63% in the NJ trees). In White Meranti/Neobalanocarpus clade (I), three subclades, consisting of Shorea roxburghii, Neobalanocarpus heimii, and White Meranti of Shorea, are recognized. Shorea roxburghii, a widely distributed species throughout Indian subcontinent to Malaysia, is a typical member of White Meranti (Ashton, 1982
), but our result shows that this species is more distantly related to the other Malaysian species of White Meranti than Neobalanocarpus. Clade II consists of two subclades, corresponding to sections Dryobalanoides and Hopea of Ashton (1982)
. In Balau (clade V), four distinct subclades are identified: (1) two species of section Shorea subsection Barbata, (2) S. isoptera (section Neohopea), (3) S. laevis (subsection Barbata), and (4) nine species of subsection Shorea. Red Meranti (clade VI) involves five distinct sections according to the classifications of Ashton (1982)
, but these taxonomical circumscriptions are only partially resolved in the gene tree. Although several well-supported subclades are identified, and species within Ashton's sections are largely concentrated within them, the species in some sections are placed in more than one separate subclade while the species of section Brachypterae and the subspecies of Shorea ovalis (sole member of section Ovalis) in particular, and some others to less extent, are spread over several subclades.
|
|
DISCUSSION
Nuclear- vs. chloroplast-based topologies
The topology of the PgiC gene tree is essentially consistent with the consensus of previous cpDNA trees (Tsumura et al., 1996
; Kajita et al., 1998
; Kamiya et al., 1998
; Dayanandan et al., 1999
), except for the placement of Neobalanocarpus (Fig. 4). The PgiC tree indicates that Neobalanocarpus is nested in White Meranti of Shorea, although the previous cpDNA analyses show that this genus forms a clade with Hopea. The phylogenetic placements of the Neobalanocarpus are supported well by bootstrap values in both phylogenetic trees of the cpDNA (BS > 99% in Kajita et al., 1998
) and the PgiC (BS > 84% in this study). Overall, the PgiC gene tree shows better resolution than the cpDNA trees; our analysis clarifies the relationship of three timber groups as (Yellow Meranti—(Balau + Red Meranti)) where the cpDNA trees could not resolve the relationship. Another incongruence is identified among the PgiC and cpDNA-based topologies; the PgiC tree shows that White Meranti of Shorea is at the basal position of the remaining groups, but the cpDNA topology exhibits a monophyletic clade containing White Meranti of Shorea, Neobalanocarpus and Hopea; this clade is sister to the clade of Parashorea and the remaining groups of Shorea. Because the basal position of White Meranti of Shorea receives lower bootstrap values (<56%) in the PgiC tree, we suppose that this incongruence is not the "hard incongruence" of Johnson and Soltis (1998)
.
|
; Ashton, 1982
). This species produces timber of good quality and is locally known as "chengal." On the basis of leaf morphology and wood anatomy, Neobalanocarpus seems to be rather close to Hopea (Symington, 1943
). Yet, Neobalanocarpus has unique characters such as a linear anther in the flower, and sub-equal short woody fruit sepals as many species of Hopea and many groups of Shorea. These obscure both its generic affinity, which is based on the number of wing-like fruit sepals in those species that have them, and its sectional affinity, which is based on anther characters (Ashton, 1982
). Therefore, it has been difficult to determine whether the genus is more closely related to Hopea or Shorea. An interesting feature of this species is irregular segregation of chromosomes during meiosis (Jong and Lethbridge, 1967
). These authors hypothesized that the irregular meiotic behavior may imply a hybrid origin of this taxon, and the equivocal placement of Neobalanocarpus based on morphology supports this hypothesis. The hybridization hypothesis could explain the incongruence between cpDNA and PgiC trees. Our data indicates that this genus could be derived via hybridization between the ancestral lineage leading to White Meranti of Shorea and that leading to Hopea. Our data further implies the Hopea lineage as the maternal progenitor and Shorea White Meranti lineage as the paternal progenitor of Neobalanocarpus. In this scenario, the maternal haplotype of the nuclear PgiC seems to have been eliminated through random genetic drift or excesses of gene flow from the paternal lineage.
Gene duplication is also a potential cause of phylogenetic incongruence (Wendel and Doyle, 1998
). Gene duplication at the PgiC locus is found in Clarkia (Gottlieb and Ford, 1996
) and in Arabidopsis halleri subsp. gemmifera (Kawabe and Miyashita, 2002
). Assuming that the cpDNA tree is identical to the true phylogeny, we must hypothesize that gene duplication occurred in the stem lineage of clades I to VI. If this is true, at least four independent gene losses at specific positions are necessary to explain the PgiC tree topology. The explanation does not seem parsimonious and thus occurrence of gene duplication appears unrealistic although further analyses using other nuclear loci are necessary to confirm it. We favor the hypothesis of the hybrid origin of Neobalanocarpus, because of cytological evidence for it (Jong and Lethbridge, 1967
) and the confusion about its affinity to Hopea and Shorea (Symington, 1943
; Ashton, 1982
).
Intrageneric relationships within genus Shorea
The PgiC tree from representative Shorea species shows two distinct clades, White Meranti and a clade consisting of Yellow Meranti, Balau and Red Meranti. The paraphyly of the genus Shorea and the sister relationship of Parashorea to this second group of Shorea in the PgiC tree correspond to the previous phylogeny of Kajita et al. (1998)
. One striking difference between the phylogenetic analyses based on PgiC sequences and the previously studied cpDNA is the levels of resolution below the genus level. Whereas the cpDNA phylogenies based on PCR-RFLP (Tsumura et al., 1996
) and sequences of the noncoding regions (Kamiya et al., 1998
) could not resolve many of the relationships within Shorea, the PgiC tree could identify three distinct lineages that are concordant with the wood anatomical characters.
All sampled species of the Yellow Meranti (the botanical section Richetioides) form a strongly supported group (BS = 100%), and this group is sister to a clade of Balau (the botanical section Shorea) and Red Meranti (section Rubroshorea of Meijer) (BS > 70%). The Yellow Meranti group is distributed in the Malay Peninsula, Borneo, Sumatra, and the Philippines (Ashton, 1982
). The flowers and bark of members of this group are uniform. The species with short subequal fruit sepals were formerly referred to Balanocarpus, but Symington (1938
, 1943
) first recognized the group as a natural botanical entity, the ‘Richetia’ group. Later, Ashton (1963)
formalized Symington's group as section Richetioides Heim. The PgiC gene tree confirms that species such as S. maxima, S. patoiensis, and S. xanthophylla, all with wingless fruits, are not separated from the species having wings (S. faguetiana). The monophyly of this group was supported from the cpDNA analysis, and the interspecific divergences were fewer than those in other groups (Kamiya et al., 1998
). Mean divergence among species in this group is the least compared with the other groups at the nuclear PgiC locus (0.0089 in Table 2). These results suggest that the species of Yellow Meranti have experienced recent adaptive radiation.
The Balau group, known to foresters in Borneo as Selangan Batu, produces rather hard and glistening textured timbers; it is widely distributed from India to Malesia, except for East of Wallace's Line (Symington, 1943
; Ashton, 1982
). This group was formerly divided into three subgroups, named Ciliata, Isoptera, and Barbata, based on the morphology of flower and fruiting calyx, although there are no reliable diagnostic characters for wood anatomy. Symington (1943)
suggested that the two subgroups Ciliata and Isoptera, based on differences in the equality of the fruit calyx lobes, are not synapomorphic, and Ashton (1982)
further confirmed this when newly described Bornean materials were included. Consequently, Ashton reduced these two groups within section Shorea. Furthermore, if new materials from Borneo are added, the Balau contains two sections, Shorea and Neohopea (Ashton, 1963
, 1982
). The section Shorea is further divided into two subsections Shorea and Barbata, and the latter corresponds to the Barbata subgroup of Symington (Ashton, 1982
). In the PgiC tree as well as the previous cpDNA tree (Kamiya et al., 1998
), the Balau timber group as a whole is monophyletic (BS > 76%). Within this group, the PgiC tree recognizes a distinct group, consisting of the species of section Shorea subsection Shorea (BS > 90%). The two species of Symington's Isoptera subgroup (S. seminis and S. sumatrana) are not of distinct lineage from the members of Ciliata group within subsection Shorea. The members of subsection Barbata form a well-supported group (BS = 98%), except for S. laevis, and this species clusters with section Neohopea (S. isoptera) in the NJ tree (BS = 70%). This suggests that section Shorea as well as subsection Barbata is not a monophyletic group.
The Red Meranti group, like the Yellow Meranti, is confined to the biogeographic region of western Malesia but with a single species also in the Moluccas; most species are found in Borneo (Symington, 1943
). Although the members of Red Meranti are, with some exceptions, easily defined by reddish inner bark and wood in the field, they are botanically heterogeneous (Symington, 1943
). Symington suggested, in his monograph on the basis of specimens from the Malay Peninsula, that this group could be divided into three distinct subgroups, Pauciflora, Ovalis, and Parvifolia. Ashton (1963
, 1982
) defined two new sections, Pachycarpae and Rubella, neither of which occur on the Malay Peninsula. Finally, Ashton (1982)
proposed a total of five sections in Red Meranti based primarily on flower and bark characters: Brachypterae Heim referring to Symington's Pauciflora subgroup, Ovalis Ashton to his Ovalis subgroup and Mutica Brandis to his Parvifolia subgroup, and two new sections, Pachycarpae and Rubella, which include species confined to Borneo or the Philippines. The previous cpDNA study suggested that Red Meranti is not monophyletic, probably due to an insufficient number of informative molecular characters (Kamiya et al., 1998
). In this study, the Red Meranti is a monophyletic group (BS > 72%), while neither Symington's (1943)
three subgroups nor Ashton's (1982)
five sections within this group are consistently recognized in the gene tree. Although most species in each of these sections cluster together in the subclades and the two subsections of the large section Mutica are mostly well defined, S. ovalis and several species in section Brachypterae, notably, are widely scattered within the Red Meranti clade. This must imply either that Ashton's sections are not monophyletic, or that they have arisen too recently to be resolved by PgiC as is suggested by the low level of interspecific variation that we found among some species (discussed later).
Maury (1978
, 1979
; summarized in Maury-Lechon and Curtet, 1998
) was the first to reassess phylogenetic relationships within Ashton's Shoreae. She divided Ashton's tribe Shoreae into tribes Hopeae and Shoreae, with Shoreae further divided into three subtribes: Anthoshorinae (to include Anthoshorea, Doona, and Pentacme), Shorinae (Shorea, Richetia, and Rubroshorea), and Parashorinae (the genus Parashorea, Fig. 1). She therefore concluded that Ashton's 11 sections of Shorea have unequal hierarchic levels and that sections such as Anthoshorea, Shorea, and Richetioides (also Doona and Pentacme), should be raised to generic rank. When we compare our phylogenetic study with current arguments of classification, the results of the topology of PgiC gene tree and the level of sequence divergences among the clades support the argument of Maury. Our results support Maury's three genera in Shoreae: Richetia (clade IV), Shorea (clade V), and Rubroshorea (clade VI). Moreover, Parashorea has an equal level of divergence as those groups of Shorea when we consider levels of sequence divergence among the clades (Table 2).
These results urge us to reconsider the taxonomic ranks of Shorea in relation to closely related genera such as Parashorea. Until now, however, the morphological and anatomical characters by which Maury recognized her supraspecific taxa within Shorea and Parashorea have been examined in only 20–30% (according to the character) of the c. 200 species in the genus Shorea, though these taxa were representative of all of Ashton's taxa. Furthermore, the species examined by Maury were consistently assignable to her proposed genera solely on the position of the hypocotyl and, in the case of some, their stomatal anatomy. Most challenging, no group of characters have yet been identified that consistently identifies species of Rubroshorea from other taxa of proposed equivalent rank. It would be imprudent to assign new generic names to a majority of species in such an economically and ecologically important genus as Shorea, whose species are so easily recognized in the field, until most species have been examined for the characters proposed for its division and more readily observable characters found.
Inter- and intraspecific variation within the Red Meranti of Shorea
We can also discuss polymorphism and divergence across closely related species in Red Meranti because we collected multiple samples from some of the species. It is noteworthy that some species have large intraspecific variation relative to interspecific divergence (Tables 1 and 2). Accordingly, for such species, no monophyletic clustering of intraspecific haplotypes is revealed (Figs. 2 and 3). This is due to an extensive number of variable sites that are shared among species (shared polymorphisms). Within Red Meranti, the observed 241 variable sites can be divided into three categories: (1) one or more species have a derived nucleotide, which is fixed in the species, but others have the ancestral nucleotide (fixed difference), (2) one species is polymorphic at the site, but others have the ancestral nucleotide (polymorphism exclusive to one species), and (3) more than one species is polymorphic at the site (shared polymorphism). When we classify observed variable sites into these categories, the number of sites in categories 1, 2, and 3 are 6, 160, and 75, respectively. Note that the number of shared polymorphisms may increase as the sample size increases, because the sites currently categorized as fixed may become polymorphic if the sample size becomes larger. Shared polymorphism can be explained by the persistence of polymorphism from the ancestral population, gene flow via interspecific hybridization, or reverse/parallel mutations (Machado et al., 2002
). Because the species of Red Meranti have low sequence divergences among species (Table 2), reverse or parallel mutations may explain only a small fraction of shared polymorphisms between species. Consequently, our concern is whether the shared polymorphisms are better explained by persistence of polymorphisms from the common ancestor or by gene flow via interspecific hybridization.
Polymorphisms from a common ancestor persist even at neutral loci if the number of generations after the speciation is insufficient for the ancestral polymorphic alleles to be fixed (Clark, 1997
). Because fixations of neutral alleles depend on the number of generations since speciation and effective populations size, the longer generation time of tree species (>50 yr) certainly will extend their persistence in terms of absolute time. Although the time when the most recent common ancestor of Red Meranti existed is not known, the low levels of divergences in the nuclear PgiC and cpDNA (Kamiya et al., 1998
) suggest relatively recent diversification of this group. Morley (2000)
has shown that the extensive lowlands under an aseasonal wet climate to which the mixed dipterocarp forests of western Malesia, and the Yellow and Red Meranti groups of Shorea are confined, has its origin in the early Miocene c. 20 mya. This can therefore be regarded as the earliest date for the origin of these groups. A reasonable average age to flowering of Shorea in primary forests would be 50 years, predicting 400 000 generations in 20 million years. Therefore, retention of ancestral polymorphisms currently seems to be a likely explanation for the shared polymorphisms among several recent diverged species. Alternatively, balancing selection may have been promoting maintenance of polymorphism and thus has increased coalescence time, resulting in slower lineage sorting (Broughton and Harrison, 2003
). It is difficult to discriminate these two possibilities—recent speciation or balancing selection—solely with the present data. However, because demographic factors affect all loci similarly while natural selection acts on specific loci, a comparative DNA approach surveying more loci will enable us to discriminate these two possibilities.
Gene flow via interspecific hybridization is another potential source of generating shared polymorphisms. Many tropical forest trees are thought to be highly outcrossed, but low hybrid fitness and rarity of fertile interspecific hybrid populations has been suggested in Dipterocarpaceae of the aseasonal mixed dipterocarp forest zone (Ashton, 1969
; Murawski et al., 1994
). However, a few examples of hybridization have been reported among dipterocarps of seasonal Indo-Burma (Ashton, 1982
; Murawski et al., 1994
; Bawa, 1998
). The interspecific hybrids between S. curtisii and S. leprosula have been reported from several localities in the Malay Peninsula and from Singapore (Ashton, 1982
). A recent study of sequence variation at the nuclear GapC locus of four Shorea species (S. acuminata, S. curtisii, S. leprosula, and S. parvifolia) found a part of a sequence in one species resembling that in another species (chimeric haplotypes), probably resulting from recombination between two divergent haplotypes (Ishiyama et al., 2003
). From this result, coupled with successful intercrossing between closely related species of Shorea (Chan, 1981
), Ishiyama and colleagues have suggested that introgressive hybridization is likely to have occurred in natural populations of these Shorea species. Such a chimeric sequence is not found at the PgiC locus, and no identical haplotypes are found between different species. Although we did not collect enough population samples, the fact that some haplotypes from one species are unexpectedly clustered with haplotypes found in other species suggests the possibility of introgressive hybridization occurring currently and/or in the past. With continuing interspecific gene flows, we would expect identical haplotypes from two different co-occurring species. However, the lack of identical haplotypes from co-occurring species suggests at least that interspecific hybrids are not abundant at the present time, even though it could have occurred to some extent in the past.
In Quaternary glacial periods, most of the regions in Southeast Asia have been covered with savanna and deciduous forests. Consequently, tropical rain forests were confined to a few refugia in northern Borneo, northern Sumatra, and the Mentawai islands (Gathorne-Hardy et al., 2002
). This indicates that the tropical rain forests expanded their ranges after the last glacial period. Paleogeographical evidence also suggested that during the glacial periods coinciding with the low sea level, Borneo was connected to Southeast Asian mainland, Java, and Sumatra (Morley, 2000
). The forest fragmentation and recolonization caused by such historical climate fluctuations may have influenced the patterns and levels of intraspecific variation, species differentiation, and interspecific hybridization that we observed in Shorea.
At this time, we cannot say whether ancestral polymorphism, balancing selection, or introgressive hybridization is the more important to explain the shared polymorphisms among the Shorea species. Currently, we do not have enough data to accept either one of those hypotheses. Although shared polymorphisms lead to ambiguous results in phylogenetic analyses, studies of polymorphisms and divergence between closely related species are important to understand mechanisms of how a great variety of species evolved and now coexist in the tropical ecosystem.
Conclusion
The study presented here assessed phylogenetic relationships among species of Shorea, Hopea, Neobalanocarpus, and Parashorea within Dipterocarpaceae, a dominant tree family in Asian tropical rainforests, based on the partial sequences of the nuclear PgiC gene. The PgiC gene tree is essentially compatible with the previous cpDNA trees, with the exception of the placement of Neobalanocarpus. This conflict suggests that Neobalanocarpus is derived via hybridization between Shorea and Hopea (Hopea is presumed to be the female parent considering the maternal inheritance of cpDNA usual in angiosperms.).
The nuclear PgiC gene sequences have more potentially informative sites than cpDNA as expected and provide a better-resolved phylogenetic tree. Our results with respect to Shorea support the phylogeny and proposed classification of Maury (1978)
. Three timber groups within Shorea are shown to constitute monophyletic groups, and the phylogenetic relationship among those groups is clearly elucidated from the PgiC gene tree. However, the PgiC tree does not yield a consistent grouping of the five sections of Ashton within Red Meranti. Our result suggests that the Ashton's botanical sections such as Anthoshorea, Richetioides, and Shorea, each of which is recognized as a monophyletic group in our analysis, have higher ranks than the other sections of Red Meranti, which are less distinct from each other.
Extensive numbers of shared polymorphisms are found among species within Red Meranti, and this results in lack of monophyly of intraspecific haplotypes. Ancestral polymorphisms, natural selection, and introgressive hybridization could be considered as potential causes of shared polymorphism among these species. Multilocus approaches of polymorphisms and divergences across the closely related taxa are needed to infer accurately phylogenetic relationships and to reveal historical, demographic, and selective factors that have contributed to a high rate of speciation and a great variety of dipterocarp species. In addition, more extensive sampling of individuals and populations are required to assess these aspects.
For the present, though, most species must be examined, and more readily observable key characters found, before it would be wise to divide the well known and easily recognized groups of genus Shorea into several separately named entities.
|
|
1 The authors thank Prof. K. Ogino (The University of Shiga Prefecture) and Dr. H. S. Lee (Forest Department Sarawak, Malaysia) for their guidance and encouragement of our study; Dr. S. L. Lee (Forest Research Institute Malaysia) for kindly providing some DNA materials and useful comments; Dr. T. Kado, H. Ishiyama, and Dr. A. E. Szmidt (Kyushu University) for providing primer sequences of the PgiC; Prof. A. Latiff, A. Zainudin, and M. Sani (Universiti Kebangsaan Malaysia) for support of fieldwork; Prof. M. M. Clyde, Dr. R. Wickneswari, Dr. C. Y. Choong (Universiti Kebangsaan Malaysia), and two anonymous reviewers for helpful comments on the manuscript. This study was partly supported by a JSPS Research Fellowship for Young Scientists to K. Kamiya and by a Grant-in-Aid for Scientific Research on Priority Areas "Molecular mechanism of speciation" No. 16057201 to H. Tachida. 
5 E-mail: kamiyarcb{at}mbox.nc.kyushu-u.ac.jp
LITERATURE CITED
Ashton P. S. 1962 Some new Dipterocarpaceae from Borneo. The Gardens' Bulletin, Singapore 19: 253-319
Ashton P. S. 1963 Taxonomic notes on Bornean Dipterocarpaceae. The Gardens' Bulletin, Singapore 20: 229-284
Ashton P. S. 1964 A manual of the dipterocarp trees of Brunei State. Oxford University Press, Oxford, UK
Ashton P. S. 1967 Taxonomic notes on Bornean Dipterocarpaceae III. The Gardens' Bulletin, Singapore 22: 259-352
Ashton P. S. 1969 Speciation among tropical forest trees: some deductions in the light of recent evidence. Biological Journal of Linnean Society 1: 155-196
Ashton P. S. 1982 Dipterocarpaceae. Flora Malesiana 9: 237-552
Ashton P. S. 1988 Dipterocarp biology as a window to the understanding of tropical forest structure. Annual Review of Ecology and Systematics 19: 347-370
Bawa K. S. 1998 Conservation of genetic resources in the Dipterocarpaceae. In S. Appanah and J. M. Turnbull [eds.], A review of dipterocarps taxonomy, ecology and silviculture, 45–55. Center for International Forestry Research, Bogor, Indonesia
Broughton R. E. R. G. Harrison 2003 Nuclear gene genealogies reveal historical, demographic and selective factors associated with speciation in field crickets. Genetics 163: 1389-1401
Chan H. T. 1981 Reproductive biology of same Malaysian dipterocarps. III. Breeding systems. The Malaysian Forester 44: 28-35
Clark A. G. 1997 Neutral behavior of shared polymorphism. Proceedings of the National Academic of Sciences, USA 94: 7730-7734[CrossRef]
Dayanandan S. P. S. Ashton S. M. Williams R. B. Primack 1999 Phylogeny of the tropical tree family Dipterocarpaceae based on nucleotide sequences of the chloroplast rbcL gene. American journal of Botany 86: 1182-1190
Doyle J. L. Doyle 1990 Isolation of plant DNA from fresh tissues. Focus 12: 13-15
Felsenstein J. 1985 Confidence limits on phylogenies: an approach using the bootstrap. Evolution 39: 783-791[CrossRef][ISI]
Gathorne-Hardy F. J. Syaukani R. G. Davies P. Eggleton D. T. Jones 2002 Quaternary rainforest refugia in south-east Asia: using termites (Isoptera) as indicators. Biological Journal of the Linnean Society 75: 453-466[CrossRef]
Gottlieb L. D. V. S. Ford 1996 A recently silenced, duplicate PgiC locus in Clarkia. Molecular Biology and Evolution 14: 125-132[ISI]
Ishiyama H. T. Kado M. Iwasaki N. Ab Shukor A. E. Szmidt T. Yamazaki 2003 Nucleotide variation in the GapC region of four species of Shorea and its putative hybrids. Tropics 13: 89-99
Johnson L. D. E. Soltis 1998 Assessing congruence: empirical examples from molecular data. In P. S. Soltis, D. E. Soltis, and J. J. Doyle [eds.], Molecular systematics of plants II: DNA sequencing, 297–348. Kluwer Academic Publishers, Dordrecht, Netherlands
Jong K. A. Lethbridge 1967 Cytological studies in the Dipterocarpaceae. I. Chromosome numbers of certain Malaysian genera. Notes from the Royal Garden Edinburgh 27: 175-184
Kajita T. K. Kamiya K. Nakamura H. Tachida R. Wickneswari Y. Tsumura H. Yoshimaru T. Yamazaki 1998 Molecular phylogeny of Dipterocarpaceae in Southeast Asia based on nucleotide sequences of matK, trnL intron, and trnL-trnF intergenic spacer region in chloroplast DNA. Molecular Phylogenetics and Evolution 10: 202-209[CrossRef][ISI][Medline]
Kamiya K. K. Harada T. Kajita T. Yamazaki H. S. Lee K. Ogino P. S. Ashton 1998 Molecular phylogeny of dipterocarp species using nucleotide sequences of two noncoding regions in chloroplast DNA. Tropics 7: 195-207
Kawabe A. N. T. Miyashita 2000 DNA polymorphism at the cytosolic phosphoglucose isomerase (PgiC) locus of the wild plant Arabidopsis thaliana. Genetics 156: 1339-1347
Kawabe A. N. T. Miyashita 2002 Characterization of duplicated two cytosolic phosphoglucose isomerase (PgiC) loci in Arabidopsis halleri ssp. gemmifera. Genes & Genetic Systems 77: 159-165
Kimura M. 1980 A simple method for estimating evolutionary rates of base substitutions through comparative studies of nucleotide sequences. Journal Molecular Evolution 16: 111-120[CrossRef]
Lee H. S. S. J. Davies J. V. LaFrankie S. Tan T. Yamakura A. Itoh T. Ohkubo P. S. Ashton 2002 Floristic and structural diversity of mixed dipterocarp forest in Lambir Hills National Park, Sarawak, Malaysia. Journal of Tropical Forest Sciences 14: 379-400
Machado C. A. R. M. Kliman J. A. Markert J. Hey 2002 Inferring the history of speciation from multilocus DNA sequence data: the case of Drosophila pseudoobscura and close relatives. Molecular Biology and Evolution 19: 472-488
Maury G. 1978 Diptérocarpacées du Fruit à la Plantule. Doctoral thesis, Université Paul Sabatier, Toulouse, France
Maury-Lechon G. 1979 Conséquences taxonomiques de l'étude des charactères des fruits-germinations, embryons et plantules des Diptérocarpacéees. Mémoires du Muséum National d'Histoire Naturelle Nouv. Ser. B 26: 139–144
Maury-Lechon G. L. Curtet 1998 Biogeography and evolutionary systematics of Dipterocarpaceae. In S. Appanah and J. M. Turnbull [eds.], A review of dipterocarps taxonomy, ecology and silviculture, 5– 44. Center for International Forestry Research, Bogor, Indonesia
Meijer W. G. H. S. Wood 1964 Dipterocarp of Sabah (North Borneo). Sabah Forest Record No. 5, Forest Department, Sandakan. Art Printing Works, Kuala Lumpur, Malaysia
Morley R. J. 2000 Origin and evolution of tropical rain forests. Wiley, Chichester, UK
Mort M. E. D. E. Soltis P. S. Soltis M. L. Mabry 2000 An empirical test of three methods for estimating internal support on phylogenetic trees. Systematic Biology 49: 160-171[CrossRef][ISI][Medline]
Mummenhoff K. P. Linder N. Friesen J. L. Bowman J. Y. Lee A. Franzke 2004 Molecular evidence for bicontinental hybridogenous genomic constitution in Lepidium sensu stricto (Brassicaceae) species from Australia and New Zealand. American Journal of Botany 91: 254-261
Murawski D. A. I. A. U. N. Gunatilleke K. S. Bawa 1994 The effect of selective logging on mating patterns of Shorea megistophylla (Dipterocarpaceae) from Sri Lanka. Conservation Biology 8: 997-1002[CrossRef][ISI]
Oh S. H. D. Potter 2003 Phylogenetic utility of the second intron of LEAFY in Neillia and Stephanandra (Rosaceae) and implications for the origin of Stephanandra. Molecular Phylogenetics and Evolution 29: 203-215[CrossRef][ISI][Medline]
Richards 1996 The tropical rain forest: an ecological study, 2nd ed. Cambridge University Press, Cambridge, UK
Saitou N. M. Nei 1987 The neighbor-joining method: a new method for reconstructing phylogenetic trees. Molecular Biology and Evolution 4: 406-425[Abstract]
Shi S. Y. Huang Y. Zhong Y. Du Q. Zhang H. Chang D. E. Boufford 2001 Phylogeny of the Altingiaceae based on cpDNA matK, PY-IGS and nrDNA ITS sequences. Plant Systematics and Evolution 230: 13-24[CrossRef][ISI]
Swofford D. L. 2002 PAUP*: phylogenetic analysis using parsimony (*and other methods), version 4. Sinauer Associates, Sunderland, Massachusetts, USA
Symington C. F. 1938 Notes on Malayan Dipterocarpaceae IV. Gardens' Bulletin Singapore 9: 319-354
Symington C. F. 1943 Foresters' manual of dipterocarps. Malaysian Forest Records No. 16. Penerbit Universiti Malaya, Kuala Lumpur, Malaysia
Thompson J. D. T. J. Gibson F. Plewniak F. Jeanmougin D. G. Higgins 1997 CLUSTAL X windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools. Nucleic Acids Research 24: 4876-4882
Tsumura Y. T. Kawahara R. Wickneswari K. Yoshimura 1996 Molecular phylogeny of Dipterocarpaceae in Southeast Asia using RFLP of PCR-amplified chloroplast genes. Theoretical and Applied Genetics 93: 22-29[CrossRef][ISI]
Wendel J. F. J. J. Doyle 1998 Phylogenetic incongruence: window into genome history and molecular evolution. In P. S. Soltis, D. E. Soltis, and J. J. Doyle [eds.], Molecular systematics of plants II: DNA sequencing, 265–296. Kluwer Academic Publishers, Dordrecht, Netherlands
Whitmore T. C. 1984 Tropical rain forests of the Far East, 2nd ed. Clarendon Press, Oxford, UK
Whitmore T. C. 1998 An introduction to tropical rain forests, 2nd ed. Oxford University Press, Oxford, UK
Wolfe K. H. W. H. Li P. M. Sharp 1987 Rates of nucleotide substitution vary greatly among plant mitochondrial, chloroplast, and nuclear DNAs. Proceedings of the National Academy of Sciences, USA 84: 9054-9058
This article has been cited by other articles:
|
|
 |
|
  P. J. Grubb, D. A. Coomes, and D. J. Metcalfe Comment on "A Brief History of Seed Size" Science, November 4, 2005; 310(5749): 783a - 783a. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
50%) resulting from 5000 replicates of "fast" bootstrap analysis. The branch labeled with an arrow collapses in the strict consensus tree. Numbers following species names represent accession numbers. Letters following the accession numbers refer to different haplotypes of the same accession. Six main clades corresponding to taxonomic circumscriptions are shown by gray brackets. Sections and subsections of Shorea abbreviated according to Ashton (1982)
