Am. J. Bot. Join the BSA
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (33)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Wang, X.-R.
Right arrow Articles by Szmidt, A. E.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Wang, X.-R.
Right arrow Articles by Szmidt, A. E.
Agricola
Right arrow Articles by Wang, X.-R.
Right arrow Articles by Szmidt, A. E.
(American Journal of Botany. 1999;86:1742-1753.)
© 1999 Botanical Society of America, Inc.

Phylogenetic relationships of Eurasian pines (Pinus, Pinaceae) based on chloroplast rbcL, MATK, RPL20-RPS18 spacer, and TRNV intron sequences1

Xiao-Ru Wang1, Yoshihiko Tsumura2, Hiroshi Yoshimaru2, Kazutoshi Nagasaka2 and Alfred E. Szmidt2,,1

1 Department of Forest Genetics and Plant Physiology, Swedish University of Agricultural Sciences, S-901 83 Umeå, Sweden; and 2 Bio-resource Technology Division, Forestry and Forest Products Research Institute, Ibaraki 305, Japan

Received for publication December 8, 1998. Accepted for publication March 25, 1999.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 LITERATURE CITED
 
The sequence divergence of chloroplast rbcL, matK, trnV intron, and rpl20-rps18 spacer regions was analyzed among 32 Pinus species and representatives of six other genera in Pinaceae. The total aligned sequence length is 3570 bp. Of the four sequences examined, matK evolved much faster than rbcL in Pinus and in other Pinaceae genera. The two noncoding regions did not show more divergence than the two coding regions, especially within each Pinus subgenus. Phylogenetic analyses based on these four sequences gave consistent results and strongly supported the monophyly hypothesis for the genus Pinus and its two recognized subgenera. Pinus krempfii, the two-flat-needle pine endemic to Vietnam, was placed in subgen. Strobus and showed closer affinity to subsect. Gerardianae. The ancient character of sect. Parrya is further confirmed. However, monophyly of the sect. Parrya is not supported by our data. Among the Eurasian pines of subgen. Pinus, Mediterranean pines formed one clade and the Asian members of subsect. Sylvestres formed another. The Himalayan P. roxburghii showed considerable divergence from all the other hard pines from both regions. Pinus merkusii was distinctly separated from all the Asian members of subsect. Sylvestres. The implications of our results for Pinus classification are discussed.

Key Words: matK • phylogeny • Pinaceae • Pinus • rbcL • rpl20-rps18 • sequence divergence • trnV intron


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 LITERATURE CITED
 
The genus Pinus is one of the most widely distributed genera of conifer trees in the Northern Hemisphere. The genus is usually divided into two subgenera Strobus (= Haploxylon, soft pines) and Pinus (= Diploxylon, hard pines), which are further divided into sections and subsections (Little and Critchfield, 1969 ). Classification of the genus differs among authors. In this paper, the classification scheme of Little and Critchfield (1969) is followed. Recently, research has become very active, in an attempt to achieve a better understanding of the evolution of the genus by various approaches (e.g., Strauss and Doerksen, 1990 ; Govindaraju, Lewis, and Cullis, 1992 ; Wang and Szmidt, 1993 ; Perez de la Rosa, Harris, and Farjon, 1995 ; Farjon, 1996 ; Krupkin, Liston, and Strauss, 1996 ; Wang, Szmidt, and Nguyen, 1999 ). The difficulties in genetic delineation are evident in the case of several species occurring in Asia and the Mediterranean part of Europe. The positions of several rare endemic species such as P. krempfii, P. merkusii, P. heldreichii, and P. roxburghii, as well as the relationships among and between Asian and Mediterranean pines are still not well settled (Schirone et al., 1991 ; Frankis, 1993 ; Krupkin, Liston, and Strauss, 1996 ; Liston et al., 1999 ). In most phylogenetic investigations, these species are seldom included. The study by Liston et al. (1999) , based on nuclear ribosomal DNA internal transcribed spacer (ITS) sequences, involved a broad sampling of the Pinus subsections and covered a wide range of geographic regions. However, the topologies of the recovered phylogenetic trees gave weak support for many of the clades, possibly because of the rapidly evolving nature of the ITS sequence.

Mediterranean pines present an interesting group in the evolution of the genus, linking different geographic regions as well as different evolutionary lineages (Mirov, 1967 ; Klaus, 1989 ). According to Klaus (1989) , Mediterranean pines represent an extremely heterogeneous assembly and consist mainly of relic pines from the Cretaceous–Tertiary period. Morphological, biochemical, and molecular data all indicate that Mediterranean hard pines are less uniform than the Asian taxa (Klaus, 1989 ; Schirone et al., 1991 ; Krupkin, Liston, and Strauss, 1996 ). Some Asian pines have been suggested to have close relationships with Mediterranean pines. The Himalayan P. roxburghii and P. wallichiana have been considered as close relatives of P. canariensis of the Canary Islands and P. peuce of the Balkan Peninsula, respectively, for instance (Mirov, 1967 ; Klaus, 1989 ). However, recent analyses of chloroplast (cp) DNA restriction site data and ITS sequences have suggested high levels of divergence among them (Wang and Szmidt, 1993 ; Liston et al., 1999 ). Therefore, more evidence is needed to clarify the relationships among this group of pines.

CpDNA sequences, especially the rbcL gene, have been used extensively to infer plant phylogenies, including those of a number of gymnosperms (e.g., Bousquet et al., 1992 ; Chase et al., 1993 ; Gadek and Quinn, 1993 ; Brunsfeld et al., 1994 ). However, some studies have shown that this coding sequence alone is sometimes too conserved to clarify relationships between closely related taxa (Doebley et al., 1990 ; Plunkett, Soltis, and Soltis, 1997 ; Xiang, Soltis, and Soltis, 1998 ). Following the use of rbcL, the matK gene has become another sequence candidate for phylogenetic analysis. Recent studies have demonstrated the utility of matK for resolving lower level relationships in angiosperms (Johnson and Soltis, 1994, 1995 ; Steele and Vilgalys, 1994 ; Liang and Hilu, 1996 ; Xiang, Soltis, and Soltis, 1998 ). However, matK sequence divergence and its phylogenetic application in Pinus and other conifers have not been previously investigated.

In this study we selected four cpDNA regions for sequencing: rbcL, matK, the trnV intron, and the spacer between the rpl20 and rps18 genes. Considering the close relationships among pines within each subgenus, we selected matK to complement the rbcL information. Noncoding sequences tend to evolve faster than coding sequences and, thus, may provide more informative characters for phylogeny reconstruction. The trnV intron and the rpl20-rps18 spacer were selected for this reason, in the expectation that they might provide more variable characters for better phylogenetic tree resolution at the tips. We included all the Mediterranean pines, most of the Asian, and four American pines in this study. In addition, six taxa representing six different genera of Pinaceae were selected as outgroups to Pinus. Our main objectives in the study presented here were: (1) to compare sequence divergence of coding and noncoding regions in Pinus and Pinaceae; (2) to evaluate the relative utility of the different sequences for phylogenetic inferences in Pinus; (3) to provide additional information for the assessment of relationships among and between the Asian and Mediterranean pines; and (4) to reexamine the classification of several uncertain taxa in the light of our new sequence data.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 LITERATURE CITED
 
Species sampled
Species sampled in this analysis are listed in Table 1. Our sampling mainly focused on Eurasian pines, including 17 species of subgen. Pinus and 15 species of subgen. Strobus (including P. krempfii), four of which are American pines. Six species, Picea abies, Cathaya argyrophylla, Larix decidua, Pseudolarix amabilis, Keteleeria davidiana, and Abies numidica, representing six other genera in Pinaceae were selected as outgroup species (Table 1). In total, 38 taxa were included in the present study. All the samples for each species were collected either from documented individuals grown by different institutions or from natural stands (Table 1).


View this table:
[in this window]
[in a new window]
  		Table 1. List of species sampled. Pinus classification according to Little and Critchfield (1969). Plant material origin: 1 Lac Duong, Lamdong, Vietnam; 2 Umeå, Sweden; 3 Forestry and Forest Product Research Institute, Japan; 4 Tree Breeding Center, Japan; 5 Hørsholm Aboretum, Denmark; 6 Chiang Mai, Thailand; 7 Royal Botanical Garden Godawari, Nepal; 8 Nanjing Forest University, China; 9 Institute of Forest Genetics at Placerville, USA; 10 DNA samples from G.G. Vendramin, 11 Escola Superior Agraria de Castelo Branco, Portugal; 12 Sichuan, China; 13 Wakasugi et al. (1994) published sequence

 
DNA isolation, PCR amplification, and sequencing
Genomic DNA was isolated from needles of individual trees using the cetyltrimethylammonium bromide (CTAB) method (Doyle and Doyle, 1987 ). Four regions (rbcL, matK, trnV intron, and rpl20-rps18) on the cp genome were selected for polymerase chain reaction (PCR) amplification. The primers used to amplify these regions were designed on appropriate sequences from the whole sequenced cp genome of P. thunbergii (Wakasugi et al., 1994 ). The primer sequences and their positions on the P. thunbergii cp genome are presented in Table 2. The PCR reaction mix contained 50–100 ng DNA template, 200 µmol/L of each deoxyribonucleotide (dNTP, GibcoBRL, Life Technologies, USA), 0.5 µmol/L of each of the primer pair, and 1.5 units of Taq DNA polymerase (GibcoBRL) in a total volume of 50 µL. PCR amplification was carried out at 94°C, 3 min for initial denaturation, followed by 30 cycles of denaturation at 94°C for 45 sec, primer annealing at 58°C for 50 sec, extension at 72°C for 80 sec, and termination by 5 min at 72°C.


View this table:
[in this window]
[in a new window]
  		Table 2. Primer sequences for template PCR amplification and sequencing. The primer positions are relative to the Pinus thunbergii chloroplast genome

 
The PCR products were purified by passage through SUPRECTM-02 filter columns (TaKaRa, Japan) to remove the nonincorporated primers and nucleotides. Sequencing reactions were carried out using the BigDye Terminator Cycle Sequencing Kit (Perkin Elmer) according to the manufacturer's instruction on GeneAmp PCR System 9600 (Perkin Elmer). The sequencing reaction products were purified through CENTRI-SEP columns (Princeton Separations Inc., USA) and then applied to ABI 377 automatic sequencer (Perkin Elmer). All the four selected regions were sequenced in both directions for all the operational taxonomic units (OTUs), except for P. thunbergii, the sequence for which was retrieved from the EMBL database. Each sequencing run from each of the primer pairs allowed complete overlap of forward and reverse sequences, to ensure high accuracy of data scoring. Automated sequencing output was further checked visually for correction of the automated base calling. The primers used in sequencing are listed in Table 2.

Sequence alignment
The sequences of each species were aligned using Clustal V software, as implemented in Sequence Navigator (ABI, Perkin Elmer, USA) and further modified manually. In most cases the placement of gaps was straightforward. Insertion/deletions (indels) in the aligned sequences were coded as 1/0 binary characters in the data matrix. Gaps of more than 1 bp in length and shared by two or more taxa were treated as a single event. Overlapping gaps were treated as multiple-event length mutations and positioned to minimize the number of required mutational events for creation of the indel. All gaps were weighted equally. Separate alignments were carried out for the subgen. Pinus and Strobus, for all the 32 Pinus taxa, and for all the 38 taxa, including the outgroups.

Phylogenetic analysis
Four kinds of phylogenetic analyses differing in the treatment of gaps were carried out. In the first analysis, gaps were treated as missing data, sequences across the gaps were included, and indels were coded as binary characters. In the second analysis, sequences across the alignment gaps were excluded, but each indel was coded as a binary character. In the third analysis, both sequences across the alignment gaps and the coded indels were excluded. In the fourth analysis, indels were excluded and only the point substitutions were included. Parsimonious analysis of the four data sets produced nearly identical topology. Thus, only the results from scheme 1 are presented in this paper. Maximum parsimony analysis was performed using the PAUP v. 3.1.1 program (Swofford, 1993 ). Heuristic searches were performed with random sequence addition with 100 replicates, MULPARS, tree-bisection-reconnection (TBR) branch swapping, and ACCTRAN branch length optimization. All character states, including indels, were specified as unordered and equally weighted. To evaluate relative robustness of the clades found in the most parsimonious trees, bootstrap (Felsenstein, 1985 ), consistency index (CI) (Kluge and Farris, 1969 ), retention index (RI) (Farris, 1989 ), and decay index (Bremer, 1988 ; Donoghue et al., 1992 ) were calculated. Decay indices were calculated using the AutoDecay program v. 4.0 (T. Eriksson, Department of Botany, Stockholm University, Sweden). The bootstrap analysis was conducted with simple sequence addition, 1000 replicates, and nearest-neighbor interchanges (NNI) branch swapping. Sequence divergence in different regions was computed as the average number of nucleotide differences per site between two sequences according to Nei (1987 ; Eqs. 10.5 or 10.6, uncorrected p distance), and Jukes and Cantor (1969) , using the DnaSP 3.0 program (Rozas and Rozas, 1999 ). The distance matrices for all pairwise sequence combinations were analyzed with the neighbor-joining (NJ) method of phylogenetic tree construction (Saitou and Nei, 1987 ) with 1000 bootstrap replications, using the program Clustal X (Thompson, Higgins, and Gibson, 1994 )


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 LITERATURE CITED
 
Sequence characterization
Our rbcL sequence included 1331 nucleotides for all the 38 OTUs. The rbcL gene in P. thunbergii is 1428 bp long and is located between positions 43046 and 44473 (Wakasugi et al., 1994 ). Our 1331-bp sequence starts at position 43122 and ends at 44452, covering 93.2% of the gene. There is no insertion/deletion in this region, and all the 38 aligned sequences have the same length (Table 3).


View this table:
[in this window]
[in a new window]
  		Table 3. Summary of sequence variation among the analyzed species and phylogenetic tree statistics, excluding uninformative characters

 
The matK gene in P. thunbergii is 1548 bp long and is located in the intron of trnK between positions 1715 and 3262 (Wakasugi et al., 1994 ). Our primers for the matK region cover about half (863 bp) of the matK gene and 176 bp of the 3'-flanking region within the trnK intron. Relative to the P. thunbergii cp genome, our matK sequence lies between positions 1539 and 2577. Length variation was found in this region among the 38 OTUs. Within subgen. Pinus, P. canariensis has the longest sequence (1052 bp) and P. nigra the shortest (1033 bp). All the Asian members of subgen. Pinus, except for P. merkusii and P. roxburghii, have the same length as P. thunbergii (1039 bp). Pinus merkusii, P. roxburghii, and the other members of Mediterranean pines, as well as all the species of subgen. Strobus have a length of 1046 bp. Among the six outgroups, P. abies, L. decidua, and P. amabilis, have sequences 1046 bp long. The K. davidiana and C. argyrophylla sequences are 1058 and 1051 bp long, respectively. Abies numidica has the longest sequence (1060 bp).

The aligned sequence length for the matK region is 1076 bp, and it contains nine indels of different lengths (1–12 bp). Most of the indels (five out of nine) were introduced by the outgroups, and they are mainly located in the matK 3'-end and the 3'-flanking region. When only the 32 Pinus species are included in the alignment, the aligned sequence length is 1052 bp (Table 3). Two deletions in the matK 3'-flanking region, one of 6 bp and another of 1 bp, were found in the Asian members of subgen. Pinus, but not in P. merkusii and P. roxburghii (Table 4). In addition, an insertion of 6 bp and a deletion of 6 bp were found in the coding region of matK in P. canariensis and P. nigra, respectively. No alignment gaps were found among the taxa of subgen. Strobus. Among the nine indel characters, only three are phylogenetically informative (Table 4).


View this table:
[in this window]
[in a new window]
  		Table 4. Placement of phylogenetic informative indels in the aligned sequences from representative taxa. Positions correspond to the 38 OTUs alignment matrix for each sequence region. Dashes represent alignment gaps; dots indicate the same nucleotide as the reference P. thunbergii.

 
The trnV intron in P. thunbergii is 543 bp long, lying between positions 47471 and 48013 (Wakasugi et al., 1994 ). Our sequence for this region includes the whole intron and six nucleotides from the 3'-end of the trnV exon1. The length variation in this region is very minor (Table 3). Five gaps were found in the aligned matrix among the 38 OTUs, four of 1 bp and one of 5 bp in length. Four of the five gaps were introduced by the outgroup taxa, and only two of the five indel characters are phylogenetically informative (Table 4). The aligned sequence length for this region is 555 bp. Subgenus Pinus and subgen. Strobus differed by 1 bp in length.

The spacer between rpl20 and rps18 proved to be the most length-variable region among the four analyzed in this study. Our sequence for this region covers half (186 bp) of the rpl20 gene (360 bp), the spacer in between (256 bp), and half (134 bp) of the rps18 gene (303 bp), between positions 31383 and 31958 on the P. thunbergii cp genome. The sequence length varied between 555 bp in subgen. Strobus to 590 bp in A. numidica. The aligned sequence length is 608 bp (Table 3). Nineteen gaps of 1–20 bp were found in the aligned sequences, but only six are informative (Table 4). All the gaps, except for a 4-bp insertion in A. numidica, a 6-bp insertion in C. argyrophylla, and a 20-bp deletion in subgen. Strobus within rpl20, were found in the spacer region between the rpl20 and rps18 genes. Subgenus Strobus differed from subgen. Pinus by having a 20-bp deletion at the 3'-end of rpl20 and a 1-bp deletion in the spacer region (Table 4). No length variation was found within subgen. Strobus. Within subgen. Pinus, five gaps of 1–5 bp were introduced into the spacer by including P. merkusii, P. pinea, and P. canariensis. The other gaps were introduced by the addition of the outgroup species.

Sequence divergence
For P. krempfii, P. peuce, and P. wallichiana each region was sequenced for two individuals. The two samples of each species gave identical sequences on all the four regions analyzed. The other taxa were each sequenced using one individual. All the 148 sequences (37 OTUs and four sequences each) reported in this paper will appear in the DDBJ/EMBL/GenBank nucleotide sequence databases with accession numbers from AB019794 to AB019941 (Table 1).

Sequence variation in each region is summarized in Table 3. When all the 32 Pinus species were compared, 57 variable sites were found in the rbcL region, 91 in the matK region, 16 in the trnV intron, and 32 in the rpl20-rps18 region. The inclusion of the six outgroups introduced much additional variation to all the four regions. The variable sequence characters among all the 38 OTUs numbered 128 for rbcL, 208 for matK, 44 for trnV intron, and 102 for rpl20-rps18 (Table 3). When the four regions were combined, the total data matrix for the 38 OTUs consisted of 3570 sequence characters and 33 binary 1/0 indel characters. There were 482 variable sequence sites, of which 243 were phylogenetically informative. The 33 coded indels contributed an additional 11 informative characters (Table 3). The positions of these informative indels are presented in Table 4.

The average number of nucleotide substitutions for the four sequences analyzed in this study is presented in Table 5. In general, the sequence divergence is low across DNA regions and clades. The uncorrected distance and Jukes and Cantor (1969) distance gave very similar results, thus only the Jukes and Cantor (1969) measures are cited below. Comparison of nucleotide substitution rates among the four sequences between the two subgenera revealed similarly low divergence within each subgenus, except for matK, the sequence divergence for which in subgen. Pinus (0.0109) was 1.8 times higher than in subgen. Strobus (0.0061). The two noncoding regions did not show higher divergence than the two coding regions within each subgenus. Within subgen. Strobus, sequence divergences for rbcL (0.0063) and matK (0.0061) were similar, and both were higher than that for trnV intron and rpl20-rps18. However, within subgen. Pinus, sequence divergence for matK (0.0109) was much higher (2.7 times) than for rbcL (0.0041). When the two subgenera were combined, sequence divergence increased noticeably for all the four regions. The nucleotide divergence in genus Pinus for the trnV intron (0.0109) was lower than the divergence observed for the coding rbcL (0.0116) and matK (0.0220) sequences (Table 5). Sequence divergence in rpl20-rps18 (0.0197) was higher than in rbcL but lower than in matK. The matK sequence appears to have evolved 1.9 times faster than the rbcL sequence in Pinus. A similar pattern of sequence divergence was found among the six outgroups. Although the substitution rate was higher for rpl20-rps18 (0.0607) than for matK (0.0530), the matK region, as in Pinus, evolved two times faster than rbcL (0.0264) among the six outgroup genera (Table 5).


View this table:
[in this window]
[in a new window]
  		Table5. Average number of nucleotide substitutions for different clades, based on uncorrected p distance (Nei, 1987) and Jukes and Cantor (1969) distance (in parentheses), averaged for all pairwise comparisons, gap sites were removed

 
Phylogenetic reconstruction
The overall topology of phylogenetic trees based on individual sequences all strongly supported the monophyly of the genus Pinus and the two subgenera, Strobus and Pinus (data not shown). However, as suggested by sequence variation analysis, the trees based on trnV intron and rpl20-rps18 were poorly resolved at the section level. The rbcL gene tree gave weak support for most of the branches. The matK tree is nearly identical to the topology of the combined data sets presented in Fig. 1, but with weaker bootstrap values for many branches. Although the overall topologies of individual sequence trees were similar, they differed in the placement of a few unstable taxa. For example, on the rbcL tree, P. krempfii and P. peuce formed one group sister to P. bungeana and P. gerardiana. However, on the matK tree, P. krempfii grouped with P. gerardiana, being sister to P. bungeana. On the rpl20-rps18 tree, P. merkusii grouped together with P. pinea and P. canariensis rather than with the Asian members of subsect. Sylvestres. The combined matK-rbcL tree (data not shown) had good resolution, gave strong support for most of the clades, and agreed well with the phylogenetic tree based on the combined four sequences. When all the four regions were combined, maximum parsimonious heuristic search of this data set produced 12 equally parsimonious trees requiring 450 steps (CI = 0.6378; RI = 0.9046). Compared to the individual sequences, the combined data set gave stronger support for internal clades, and more clades received good bootstrap support (>75%) in the analysis of all four regions than in analyses of the separate data sets. The strict consensus tree of the 12 equally parsimonious trees based on the combined data set is shown in Fig. 1.



View larger version (31K):
[in this window]
[in a new window]
  		Fig. 1. Strict consensus tree of the 12 most parsimonious trees based on combined sequences from 32 Pinus species and six outgroups. Bootstrap percentages (above) of 1000 replicates and decay values (below, preceded by "d") are mapped along each branch

 
On the strict consensus tree (Fig. 1) the 32 Pinus species were split into two distinct groups corresponding to the subgenera Pinus and Strobus. Within subgen. Strobus, P. aristata and P. balfouriana from the sect. Parrya, subsection Balfourianae, formed a well-supported (99%) basal group. Pinus bungeana and P. gerardiana from subsect. Gerardianae, also in the sect. Parrya, formed another separate group. Pinus krempfii was sister species to the sect. Strobus clade. Within the sect. Strobus clade, P. peuce was sister species to the rest of the section. The two American taxa of subsect. Strobi, P. strobus and P. monticola, were separated from the Asian members of the subsection. The remaining Eurasian species of subsects. Strobi and Cembrae formed one unresolved cluster. The topology of the NJ tree (Fig. 2) is essentially the same as the strict consensus tree. However, on the NJ tree P. krempfii was grouped together with P. bungeana and P. gerardiana, with 69% bootstrap support.



View larger version (25K):
[in this window]
[in a new window]
  		Fig. 2. The neighbor-joining tree based on overall pairwise substitution rates (p distance) of the combined sequences. Branch lengths are proportional to the scale of substitution rate given above. Bootstrap percentages of 1000 replicates are noted for each branch. Only bootstrap values higher than 50% are shown.

 
In the subgen. Pinus clade, the Himalayan P. roxburghii was a sister species to all the remaining taxa of the subgenus. Its divergent character was shown consistently on all the individual sequence trees (data not shown). The remaining 16 species were split into two distinct clades. One of these two clades included pines occurring in the Mediterranean. In this clade, the rare and endangered P. heldreichii appeared as sister species to the rest of the cluster. Pinus halepensis and P. brutia formed a strongly supported (98%) group. The remaining species, P. pinea, P. canariensis, and P. pinaster, formed one group with weak (<50%) bootstrap support on both the NJ and strict consensus trees. The second clade of the subgen. Pinus, consisted of species from subsect. Sylvestres, including all the Asian members and the European P. nigra. In this clade P. merkusii was placed on a separate long branch, sister to the rest of the Asian members (Fig. 2). Among the remaining species, P. yunnanensis, P. kesiya, P. hwangshanensis, P. tabuliformis (previous spelling P. tabulaeformis), and P. thunbergii formed one unresolved group, and P. sylvestris and P. densiflora formed a separate strongly supported group on both the NJ and strict consensus trees.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
LITERATURE CITED
 
Sequence divergence in Pinus
Sequence divergence varied considerably among the four cpDNA regions analyzed, as well as among clades for a given sequence. Previous studies have suggested that the rbcL gene is conservative within lineages of seed plants (Bousquet et al., 1992 ; Chase et al., 1993 ). Our results give further confirmation of the conservative nature of the rbcL sequence within Pinus and among genera of Pinaceae (Table 5). In this region all the 38 OTUs had the same length, and only 57 variable sites were found among the 32 pines in a sequence 1331 bp long. Based on rbcL sequence alone, the topology of the recovered tree was not well resolved, and most branches had weak support.

On the other hand, the matK gene has been often found to be more variable than other coding cpDNA sequences tested (Johnson and Soltis, 1994, 1995 ; Soltis et al., 1996 ; Plunkett, Soltis, and Soltis, 1997 ; Xiang, Soltis, and Soltis, 1998 ). Previous studies in angiosperms have shown that matK evolves much faster (2–3 times) than rbcL (Johnson and Soltis, 1994 ; Soltis et al., 1996 ; Plunkett, Soltis, and Soltis, 1997 ; Xiang, Soltis, and Soltis, 1998 ). Thus far, there has been no report on matK variation in gymnosperms. Our results provide the first information on this subject. The matK sequences analyzed in our study suggested there was a distinctly higher rate of evolution in this region than in the rbcL sequence, both within subgen. Pinus, and among different genera of Pinaceae. In addition, the variation of matK in Pinus was even higher than that of the noncoding regions. Surprisingly, however, within our samples of subgen. Strobus, unlike in subgen. Pinus, the matK diverged at a rate very similar to rbcL. One scenario that could explain this observation is uneven rate of divergence over time among lineages for different sequences. Another possible explanation is homogenizing sequence evolution within different lineages caused by differing types of recurrent mutations. The unequal evolution rate of different cpDNA sequences within and among lineages found in this study and the interspecific rate heterogeneity reported by Bousquet et al. (1992) indicate that care must be taken when selecting sequence candidates for estimating branching dates.

An unexpected result from this study was the low sequence divergence of the two noncoding regions, especially within each Pinus subgenus. This was particularly manifest in the case of the trnV intron, which appeared to evolve slower than the rbcL sequence and contained only one 1-bp indel in the alignment matrix of the 32 Pinus species. The basis for the apparently slow evolution of this intron cannot be elucidated with our data.

Although the sequence divergence across the four DNA regions was generally low within each subgenus, a sharp increase was noticed when the two subgenera were combined. This can result from the differences in mutation sites between the two groups. Although both have similar mutation rates, individual mutations can occur in a different genome region in each group. As a consequence, a sharp increase in divergence would occur when we combine them, because one group contributes changes that do not occur in the other. In fact, this is the case for many of the mutations we observed in our data set, which further stresses the distinct split between the two subgenera.

Phylogenetic implications—Subgenus Strobus
One of the most morphologically unique species in Pinus is P. krempfii, which is endemic to Vietnam. Morphologically it differs from all the other pines by having two flat leaf-like needles rather than typical pine needles (Lecomte, 1921, 1924 ). Several specific morphological and wood anatomy features giving unusual combinations of characters have made classification of this taxon difficult (Chevalier, 1944 ; De Ferré, 1948, 1953 ; Buchholz, 1951 ; Erdtman, Kimland, and Norin, 1966 ). It has been suggested that the taxon may represent a link between the genus Pinus and other genera of the family Pinaceae such as Keteleeria and Pseudolarix (De Ferré, 1948, 1953; Mirov, 1967 ). However, both previous cpDNA RFLP analysis (Wang, Szmidt, and Nguyen, 1999 ) and the present sequence data do not support this hypothesis. The relationship between P. krempfii and Keteleeria and Pseudolarix is clearly remote. Chevalier (1944) elevated this taxon to an independent monospecific genus in Pinaceae and named it Ducampopinus krempfii. Other authors, however, created a separate subgenus, Ducampopinus, in the genus Pinus to accommodate this taxon (De Ferré, 1953 ; Gaussen, 1960 ; Little and Critchfield, 1969 ). In Pilger's (1926) classification, the species was placed in the same subsection, Balfourianae, as P. aristata and P. balfouriana. Farjon (1984) following the subdivision of Van der Burgh (1973) placed P. krempfii in sect. Parrya, monospecific subsect. Krempfianae. Our previous analysis of cpDNA restriction site variation could only place this taxon in subgen. Strobus (Wang, Szmidt, and Nguyen, 1999 ). In the present study, however, P. krempfii was found outside sect. Strobus and could not be placed in the subsect. Balfourianae; rather it seems to have a closer affinity to subsect. Gerardianae, as indicated by the matK and rbcL trees and the combined NJ tree. Taking into account its unique morphology, our results tend to support the placement of P. krempfii in the sect. Parrya, monotypic subsect. Krempfianae. Although by now the available molecular data clearly suggest the placement of P. krempfii in genus Pinus, subgen. Strobus, the evolution of its unique needle morphology remains to be explained.

Species representing subsects. Balfourianae and Gerardianae of the sect. Parrya were placed in two separate, strongly supported groups. Our results, similar to the conclusions of Perez de la Rosa, Harris, and Farjon (1995) and Liston et al. (1999) , also showed that this section is not monophyletic. The distinct character of these subsections has been recognized in most other phylogenetic studies (Strauss and Doerksen, 1990 ; Wang and Szmidt, 1993 ). It has been suggested that sect. Parrya represents the most ancient pines (Farjon, 1984, 1996 ; Strauss and Doerksen, 1990 ). Strauss and Doerksen (1990) suggested that the ancestral species in Parrya are perhaps North American and gave rise to the Asian group subsect. Gerardianae, which then gave rise to the section Strobus. The position of P. krempfii revealed in the present study suggests the taxon might represent a link between sect. Parrya and sect. Strobus. According to Millar (1998 , and references therein) pine originated in the early-middle Mesozoic in middle latitudes. At the beginning of the Mesozoic, there was one landmass. By the early Jurassic, a northern super-continent, Laurasia, separated and began to drift from a southern continent. During the Cretaceous, the genus was already differentiated into the two subgenera, and pines were widely distributed throughout the Northern Hemisphere, indicating that wherever within middle latitudes they originated, their main route of migration was east and west. Cretaceous fossil records of sect. Parrya, especially Balfourianae and Gerardianae, are very poor, making it difficult to track the path of these pines. Nevertheless, our results support the ancient character of sect. Parrya, and the divergence between subsects. Balfourianae and Gerardianae seems very advanced.

The close relationship of subsects. Strobi and Cembrae has been revealed by several previous analyses (Strauss and Doerksen, 1990 ; Govindaraju, Lewis, and Cullis, 1992 ; Wang and Szmidt, 1993 ). However, as in other phylogenetic analyses, further divisions among pines from these two subsections were not resolved in the present study. The two American species, P. strobus and P. monticola, from the subsect. Strobi were separated from Eurasian members of subsects. Strobi and Cembrae, indicating advanced divergence between Old World and New World soft pines as well as relatively recent diversification among the Eurasian taxa. Since only four North American pines were sampled in this study, patterns of divergence between North American and Eurasian pines cannot be generalized. A wider sampling of North American pines from different subsections would be necessary for such a comparison. In general, patterns of divergence among species within and between continents would largely depend on their origin and development history. Many of the extant pines were developed from scattered refugia throughout the Tertiary (Millar, 1998 ). In the sect. Strobus clade, P. peuce was clearly separated from the remaining members, which suggests the species is distinctly different from others in this section. According to Klaus (1989) , P. peuce is a small-cone relative of the Himalayan P. wallichiana and represents an Eurasian relic pine that has been isolated from other pines of subgen. Strobus since Tertiary times (Mirov, 1967 ; Klaus, 1989 ). However, from our sequence data P. wallichiana was clearly associated with other Asian soft pines, which together formed an unresolved group. Thus, the relationship between P. peuce and P. wallichiana requires further investigation.

Subgenus Pinus
Within subgenus Pinus, the 16 species excluding P. roxburghii were split into two strongly supported clades, one containing all Asian members of subsect. Sylvestres and P. nigra, and the other comprising all the Mediterranean pines of subsects. Sylvestres, Pineae, and Canarienses. The Himalayan P. roxburghii was found as a strongly divergent taxon from all the remaining hard pines. Based on analysis of the ITS region, Liston et al. (1999) found P. roxburghii had a sister relationship to the American and Mexican pines of subsects. Ponderosae, Leiophyllae, Contortae, Oocarpae, Australes, and Attenuatae and that it was paraphyletic to the Asian and Mediterranean hard pines. The strong morphological resemblance of P. roxburghii to P. canariensis has promoted the classification of the two taxa into the same subsection, Canarienses (Little and Critchfield, 1969 ; Farjon, 1984 ; Klaus, 1989 ). Klaus (1989) suggested that P. roxburghii originated from Mediterranean ancestors of P. canariensis that followed the Tethys coast to the east and reached the Himalayan region in the Upper Cretaceous–Lower Tertiary and led to the rise of P. roxburghii. On the other hand, Mirov (1967) suggested an eastern Asia origin for P. roxburghii, from where it purportedly migrated to the Himalayas via the mountain ranges that once extended from eastern Asia to the Caucasus and farther west. By this route, Mirov (1967) suggested, the closely related P. canariensis reached the Canary Islands. The highly divergent character of P. roxburghii revealed by our present and other (Liston et al., 1999 ) molecular evidence does not clearly support a Mediterranean descent for P. roxburghii, rather it suggests a very early split of the P. roxburghii lineage from the Mediterranean pines. Alternatively, P. roxburghii might represent an ancestral stock to the Eurasian hard pines.

In the clade comprising the Mediterranean pines, P. heldreichii was a sister species to the remaining members. Pinus heldreichii is an endemic species that grows in southern Italy and the Balkan Peninsula (Mirov, 1967 ). By some authors this species is called P. leucodermis (Farjon, 1984 ; Schirone et al., 1991 ; Boscherini et al., 1994 ). The taxonomic position of P. heldreichii remains uncertain, and it has seldom been studied at the molecular level (Schirone et al., 1991 ; Boscherini et al., 1994 ). In general, P. heldreichii is regarded as more closely related to P. nigra, P. sylvestris, and other Asian hard pines than to the true Mediterranean pines (Klaus, 1989 ). Shaw (1914) even considered it as a variety of P. nigra. However, chemical analysis revealed that P. heldreichii has a different terpene composition than P. nigra (Mirov, 1967 ). Seed protein analysis revealed a "divider" position for P. heldreichii between Mediterranean pines and other members of subsect. Sylvestres, though it is more closely related to the Mediterranean taxa (Schirone et al., 1991 ). Our present results clearly support a distinct taxonomic status for this rare and endangered pine and its close affinity to the "true" Mediterranean pines.

Within the Mediterranean pine clade, P. halepensis and P. brutia formed a highly supported (98% on both NJ tree and consensus tree) group. The clear resemblance in their seed protein profiles (Schirone et al., 1991 ), and allozyme patterns (Conkle, Schiller, and Grunwald, 1988 ) and their ability to hybridize in nature (Panetsos et al., 1997 ) all indicate a close relationship between the two. Pinus brutia is even described as a variety of P. halepensis by some authors (Farjon, 1984 ). Allozyme (Conkle, Schiller, and Grunwald, 1988 ) and morphology (Frankis, 1993 ) studies have suggested that P. halepensis is derived from a P. brutia-like ancestor and that P. brutia has retained greater ancestral variation, showing affinities not only to P. halepensis but also to other Mediterranean pines, e.g., P. pinaster and P. canariensis (Frankis, 1993 ). Our present results support this suggestion.

Pinus pinaster, P. pinea, and P. canariensis formed one group, albeit with weak (<50%) bootstrap support. Pinus pinea is considered by many authors as an enigmatic and isolated species (Mirov, 1967 ; Klaus, 1989 ). Traditionally, P. pinea is placed in the monotypic subsect. Pineae (Little and Critchfield, 1969 ; Farjon, 1984 ). However, our present results do not reveal such a distinct separation of P. pinea from other Mediterranean pines. Klaus (1989) noted that P. pinea, P. pinaster, and P. canariensis share many cone and vegetative characters. Frankis (1993) combined P. pinaster, P. canariensis, P. halepensis, and P. brutia into one subsection, Pinaster, but both authors still placed P. pinea in a separate subsection. Our present results lend additional support to the grouping of these species into one subsection, Pinaster, suggested by Frankis (1993) , but indicate that P. pinea may also belong to this subsection.

The Asian members of the subsect. Sylvestres formed a strongly (94% on the NJ tree and 87% on the consensus tree) supported monophyletic group that is clearly separated from the Mediterranean clade. In this clade, P. merkusii appeared as strongly diverged from all the other members (Fig. 2). Morphological, chemical, and population studies have revealed that P. merkusii is very different from other neighboring Asian hard pines (Cooling, 1968 ; Weissmann and Lange, 1987 ; Szmidt, Wang, and Changtragoon, 1996 ). Its distinct separation from the rest of the Asian members of subsect. Sylvestres at the molecular level was first reported in a study based on cpDNA restriction site data (Wang and Szmidt, 1993 ) and was further confirmed by a recent analysis of the nuclear ITS region (Liston et al., 1999 ). It appears that the distinct character of P. merkusii is a result of an early separation and prolonged isolation of this species from other Asian members. During the Jurassic and Cretaceous periods, tropical pines were present in southeastern Asia (Mirov, 1967 ). It is possible that P. merkusii has continued to develop in this region ever since, while the other extant pines migrated to and developed in southeastern Asia not earlier than the Tertiary (Mirov, 1967 ). Considering all these lines of evidence, it appears that P. merkusii could be excluded from subsect. Sylvestres. A similar suggestion was made by Frankis (1993) . Based on the similarity of cones of P. merkusii and P. brutia, Frankis (1993) placed the former species in subsect. Pinaster together with other Mediterranean pines. On our rbcL and trnV intron trees the position of P. merkusii appeared uncertain. On the rpl20-rps18 tree this taxon was grouped together with P. pinea and P. canariensis. However, on the matK tree and the combined sequence tree, P. merkusii was placed in the same clade as other Asian members of subsect. Sylvestres. Thus, the classification scheme proposed by Frankis (1993) is not fully supported by our combined sequence topology. We feel reluctant to express a strong opinion about its placement in subsect. Sylvestres or Pinaster. Taking into account inconsistent characters of the available morphological and molecular evidence, we believe that additional studies are necessary for its placement in subgen. Pinus.


View this table:
[in this window]
[in a new window]
  		Table 4. Continued

 

   FOOTNOTES
 
1 The authors thank Dr. G. G. Vendramin for the DNA samples of Mediterranean pines, M. M. Ribeiro for supplying samples of P. pinaster and P. canariensis, Dr. J. W. Leverenz for samples of P. aristata and P. balfouriana, Dr. L. M. Xiong for the sample of Cathaya, Dr. N. Tani for the collection of P. wallichiana and P. roxburghii from Nepal, Forest Tree Breeding Center, Japan, for the access to their collection, and Dr. N. Minaka for valuable discussions on phylogeny reconstruction. This study was conducted during XRW's stay in Japan as an STA fellow, sponsored by JISTEC/JRDC, Japan. Support was also provided by the Grant-in-Aid Baikal Drilling Project from the Science and Technology Agency of Japan. AES acknowledges support from SJFR and from the Carl Tryggers Foundation, Sweden. Back

4 Author for correspondence. Back


   LITERATURE CITED
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 LITERATURE CITED
 
Boscherini, G., M. Morgante, P. Rossi, and G. Vendramin. 1994 Allozyme and chloroplast DNA variation in Italian and Greek populations of Pinus leucodermis. Heredity 73: 284–290.

Bousquet, J., S. H. Strauss, A. H. Doerksen, and R. A. Price. 1992 Extensive variation in evolutionary rate of rbcL gene sequences among seed plants. Proceedings of the National Academy of Sciences, USA 89: 7844–7848. [Abstract/Free Full Text]

Bremer, K. 1988 The limits of amino acid sequence data in angiosperm phylogenetic reconstruction. Evolution 42: 795–803. [CrossRef][ISI]

Brunsfeld, S. J., P. S. Soltis, D. E. Soltis, P. A. Gadek, C. J. Quinn, D. D. Strenge, and T. A. Ranker. 1994 Phylogenetic relationships among the genera of Taxodiaceae and Cupressaceae: evidence from rbcL sequences. Systematic Botany 19: 253–262. [CrossRef][ISI]

Buchholz, J. T. 1951 A flat-leaved pine from Annam, Indo-China. American Journal of Botany 38: 245–252. [CrossRef][ISI]

Chase, M. W., et al. 1993 Phylogenetics of seed plants—an analysis of nucleotide sequences from the plastid gene rbcL. Annals of the Missouri Botanical Garden 80: 528–580. [CrossRef][ISI]

Chevalier, A. 1944 Notes sur les conifères de l'Indochine. Revue de Botanique Appliquée et d'Agriculture Tropicale 24: 7–34.

Conkle, M. T., G. Schiller, and C. Grunwald. 1988 Electrophoretic analysis of diversity and phylogeny of Pinus brutia and closely related taxa. Systematic Botany 13: 411–424. [CrossRef][ISI]

Cooling, E. N. G. 1968 Pinus merkusii. Commonwealth Forestry Institute, Department of Forestry, Oxford University, Oxford.

De Ferré, Y. 1948 Quelques particularités anatomiques d'un pin Indochinois: Pinus krempfii. Bulletin de la Société d'Histoire Naturelle de Toulouse 83: 1–6.

———. 1953 Division du genre Pinus en quatre sous-genres. Academie des Sciences Compte Rendu 236: 226–228.

Doebley, J., M. Durbin, E. M. Golenberg, M. T. Clegg, and D. P. Ma. 1990 Evolutionary analysis of the large subunit of carboxylase (rbcL) nucleotide sequence among grasses (Gramineae). Evolution 44: 1097–1108. [CrossRef][ISI]

Donoghue, M. J., R. G. Olmstead, J. F. Smith, and J. D. Palmer. 1992 Phylogenetic relationships of Dipsacales based on rbcL sequences. Annals of the Missouri Botanical Garden 79: 333–345.

Doyle, J. J., and J. L. Doyle. 1987 A rapid DNA isolation procedure for small quantities of fresh leaf tissue. Phytochemical Bulletin 19: 11–15.

Erdtman, H., B. Kimland, and T. Norin. 1966 Wood constituents of Ducampopinus krempfii (Lecomte) Chevalier (Pinus krempfii Lecomte). Phytochemistry 5: 927–931. [CrossRef][ISI]

Farjon, A. 1984 Pines: drawings and descriptions of the genus. E. J. Brill/DR. W. Backhuys, Leiden.

———. 1996 Biodiversity of Pinus (Pinaceae) in Mexico: speciation and palaeo-endemism. Botanical Journal of the Linnean Society 121: 365–384. [CrossRef]

Farris, J. S. 1989 The retention index and the rescaled consistency index. Cladistics 5: 417–419. [ISI]

Felsenstein, J. 1985 Confidence limits on phylogenies: an approach using the bootstrap. Evolution 39: 783–791. [CrossRef][ISI]

Frankis, M. P. 1993 Morphology and affinities of Pinus brutia. In International Symposium on Pinus brutia Ten., 11–18. Ministry of Forestry, Ankara.

Gadek, P. A., and C. J. Quinn. 1993 An analysis of relationships within the Cupressaceae sensu stricto based on rbcL sequences. Annals of the Missouri Botanical Garden 80: 581–586. [CrossRef][ISI]

Gaussen, H. 1960 Les gymnospermes actuelles et fossiles. Fassicule VI. Les conifèrales. Chap. 11. Généralités, Genre Pinus. Travaux du Toulous Université Laboratoire Forestier Tome 2, Sect. 1, vol. 1, 1–272.

Govindaraju, D., P. Lewis, and C. Cullis. 1992 Phylogenetic analysis of pines using ribosomal DNA restriction fragment length polymorphisms. Plant Systematics and Evolution 179: 141–153.

Johnson, L. A., and D. E. Soltis. 1994 matK DNA sequences and phylogenetic reconstruction in Saxifragaceae s. str. Systematic Botany 19: 143–156.

———, and ———. 1995 Phylogenetic inference in Saxifragaceae sensu stricto and Gilia (Polemoniaceae) using matK sequences. Annals of the Missouri Botanical Garden 82: 149–175. [CrossRef][ISI]

Jukes, T. H., and C. R. Cantor. 1969 Evolution of protein molecules. In H. N. Munro [ed.], Mammalian protein metabolism, 21–132. Academic Press, New York, NY.

Klaus, W. 1989 Mediterranean pines and their history. Plant Systematics and Evolution 162: 133–163. [CrossRef][ISI]

Kluge, A. G., and J. S. Farris. 1969 Quantitative phyletics and the evolution of Anurans. Systematic Zoology 18: 1–32.

Krupkin, A. B., A. Liston, and S. H. Strauss. 1996 Phylogenetic analysis of the hard pines (Pinus subgenus Pinus, Pinaceae) from chloroplast DNA restriction site analysis. American Journal of Botany 83: 489–498. [CrossRef][ISI]

Lecomte, H. 1921 Un pin remarquable de l'Annam. Bulletin du Museum National d'Histoire Naturelle Paris 25: 191–192.

———. 1924 Additions au sujet de Pinus krempfii H. Lec. Bulletin du Museum National d'Histoire Naturelle Paris 30: 321–325.

Liang, H. P., and K. W. Hilu. 1996 Application of the matK gene sequences to grass systematics. Canadian Journal of Botany 74: 125–134. [CrossRef]

Liston, A., W. A. Robinson, D. Pinero, and E. R. Alvarez-Buylla. 1999 Phylogenetics of Pinus (Pinaceae) based on nuclear ribosomal DNA internal transcribed spacer region sequences. Molecular Phylogenetics and Evolution 11: 95–109. [CrossRef][ISI][Medline]

Little, E. L., Jr., and W. B. Critchfield. 1969 Subdivisions of the genus Pinus (pines). USDA Forest Service, Washington DC, Miscellaneous Publication Number 1144.

Millar, C. I. 1998 Early evolution of pines. In D. M. Richardson [ed.], Ecology and biogeography of Pinus, 69–91. Cambridge University Press, Cambridge.

Mirov, N. T. 1967 The genus Pinus. Ronald Press Company, New York, NY.

Nei, M. 1987 Molecular evolutionary genetics. Columbia University Press, New York, NY.

Panetsos, K., A. Scaltsoyiannes, F. A. Aravanopoulos, K. Dounavi, and A. Demetrakopoulos. 1997 Identification of Pinus brutia Ten., P. halepensis Mill. and their putative hybrids. Silvae Genetica 46: 253–257. [ISI]

Perez de la Rosa, J., S. A. Harris, and A. Farjon. 1995 Noncoding chloroplast DNA variation in Mexican pines. Theoretical and Applied Genetics 91: 1101–1106. [ISI]

Pilger, R. 1926 Coniferae. In A. Engler and K. Prantl [eds.], Die Naturlichen Pflanzenfamilien, vol. 13 (ed.2) Gymnospermae, 121–403. Wilhelm Engelmann, Leipzig.

Plunkett, G. M., D. E. Soltis, and P. S. Soltis. 1997 Clarification of the relationship between Apiaceae and Araliaceae based on matK and rbcL sequence data. American Journal of Botany 84: 565–580. [Abstract]

Rozas, J., and R. Rozas. 1999 DnaSP version 3: an integrated program for molecular population genetics and molecular evolution analysis. Bioinformatics 15: 174–175. [Abstract/Free Full Text]

Saitou, N., and M. Nei. 1987 The neighbor-joining method: a new method for reconstructing phylogenetic trees. Molecular Biology and Evolution 4: 406–425. [Abstract]

Schirone, B., G. Piovesan, R. Bellarosa, and C. Pelosi. 1991 A taxonomic analysis of seed proteins in Pinus spp. (Pinaceae). Plant Systematics and Evolution 178: 43–53. [CrossRef][ISI]

Shaw, G. R. 1914 The genus Pinus. Publication of Arnold Arboretum Number 5. Riverside Press, Cambridge, MA.

Soltis, D. E., R. K. Kuzoff, E. Conti, R. Gornall, and K. Ferguson. 1996 matK and rbcL gene sequence data indicate that Saxifraga (Saxifragaceae) is polyphyletic. American Journal of Botany 83: 371–382. [CrossRef][ISI]

Steele, K. P., and R. Vilgalys. 1994 Phylogenetic analyses of Polemoniaceae using nucleotide sequences of the plastid gene matK. Systematic Botany 19: 126–142.

Strauss, S. H., and A. H. Doerksen. 1990 Restriction fragment analysis of pine phylogeny. Evolution 44: 1081–1096. [CrossRef][ISI]

Swofford, D. L. 1993 PAUP: Phylogenetic analysis using parsimony, version 3.1.1. Computer program distributed by the Illinois Natural History Survey, Champaign, IL.

Szmidt, A. E., X.-R. Wang, and S. Changtragoon. 1996 Contrasting patterns of genetic diversity in two tropical pines: Pinus kesiya (Royle ex Gordon) and P. merkusii (Jungh et De Vriese). Theoretical and Applied Genetics 92: 436–441. [CrossRef][ISI]

Thompson, J. D., D. G. Higgins, and T. J. Gibson. 1994 CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Research 22: 4673–4680. [Abstract/Free Full Text]

Van der Burgh, J. 1973 Hölzer der niederrheinischen Braunkohlenformation 2. Hölzer der Braunkohlengruben "Maria Theresia" zu Herzogenrath, "Zukunft West" zu Eschweiler und "Victor" (Zülpich Mitte) zu Zülpich. Nebst einer systematisch-anatomischen Bearbeitung der Gattung Pinus L. Review of Palaeobotany and Palynology 15: 73–275.

Wakasugi, T., S. I. Tsudzuki, M. Shibata, and M. Sugiura. 1994 A physical map and clone bank of the black pine (Pinus thunbergii) chloroplast genome. Plant Molecular Biology Reporter 12: 227–241.

Wang, X.-R., and A. E. Szmidt. 1993 Chloroplast DNA-based phylogeny of Asian Pinus species (Pinaceae). Plant Systematics and Evolution 188: 197–211. [CrossRef][ISI]

———, ———, and H. N. Nguyen. 1999 The phylogenetic position of the endemic flat-needle pine Pinus krempfii (Pinaceae) from Vietnam, based on PCR-RFLP analysis of chloroplast DNA. Plant Systematics and Evolution (in press).

Weissmann, V. G., and W. Lange. 1987 Zusammensetzung der Neutralteile des Balsamkolophoniums von Pinus massoniana Lamb., Pinus merkusii Jungh. und Pinus luchuensis Mayr. Holzforschung 41: 147–154.

Xiang, Q. Y., D. E. Soltis, and P. S. Soltis. 1998 Phylogenetic relationships of Cornaceae and close relatives inferred from matK and rbcL sequences. American Journal of Botany 85: 285–297. [Abstract]




This article has been cited by other articles:


Home page
 ANN BOT (LOND)Home page
C.-Y. Qiao, J.-H. Ran, Y. Li, and X.-Q. Wang
Phylogeny and Biogeography of Cedrus (Pinaceae) Inferred from Sequences of Seven Paternal Chloroplast and Maternal Mitochondrial DNA Regions
Ann. Bot., September 1, 2007; 100(3): 573 - 580.
[Abstract] [Full Text] [PDF]

Home page
 ANN BOT (LOND)Home page
Q. CAI, D. ZHANG, Z.-L. LIU, and X.-R. WANG
Chromosomal Localization of 5S and 18S rDNA in Five Species of Subgenus Strobus and their Implications for Genome Evolution of Pinus
Ann. Bot., May 1, 2006; 97(5): 715 - 722.
[Abstract] [Full Text] [PDF]

Home page
 Am. J. Bot.Home page
D. M Erwin and H. E Schorn
Pinus baileyi (section Pinus, Pinaceae) from the Paleogene of Idaho, USA
Am. J. Botany, February 1, 2006; 93(2): 197 - 205.
[Abstract] [Full Text] [PDF]