| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Systematics and Phytogeography |
2Department of Botany and the Genetics Institute, University of Florida, Gainesville, Florida 32611-5826 USA; 3Florida Museum of Natural History and the Genetics Institute, University of Florida, Gainesville, Florida 32611-7800 USA; 4School of Biological Sciences, Washington State University, Pullman, Washington 99164-4236 USA
Received for publication January 22, 2002. Accepted for publication May 7, 2002.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Key Words: Bayesian inference • hypothesis tests • phylogeny • seed plants
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
; Sanderson et al., 2000
), uncertainty remains. Analyses of morphological data generally concur in suggesting that angiosperms and Gnetales are sister groups (the "anthophyte" hypothesis), with extant gymnosperms paraphyletic (Crane, 1985
; Doyle and Donoghue, 1986
, 1992
; Loconte and Stevenson, 1990
; Rothwell and Serbet, 1994
; Doyle, 1996
). However, the sister-group relationship of Gnetales and angiosperms has not been supported by most molecular analyses. In fact, early molecular studies based on single genes provided conflicting results regarding the relationship of Gnetales and angiosperms (see Doyle, 1998
, for review). For example, some analyses of rbcL alone and analyses of partial 18S and 26S rRNA sequences placed Gnetales as sister to all other seed plants; angiosperms were sister to a clade of cycads, Ginkgo, and conifers (e.g., Hamby and Zimmer, 1992
; Albert et al., 1994
). In contrast, a maximum likelihood (ML) analysis of rbcL placed angiosperms as the sister to a gymnosperm clade (Hasebe et al., 1992
).
Single-gene, multigene, and multigenome analyses provide conflicting reconstructions of seed plant phylogeny. Analyses of 18S rDNA alone suggested that the gymnosperms constitute a clade (Chaw et al., 1997
; Soltis et al., 1999
), as have plastid (Goremykin et al., 1996
) and mitochondrial (Malek et al., 1996
) sequences, although measures of support are generally weak, where provided. Recent multigene analyses have found strong support for the monophyly of extant gymnosperms (80% bootstrap value in Chaw et al., 2000
; >90% bootstrap values in Bowe, Coat, and dePamphilis, 2000
). In addition, both Bowe, Coat, and dePamphilis (2000)
and Chaw et al. (2000)
obtained strong support for cycads followed by Ginkgo as successive sisters to a conifer clade that includes Gnetales.
Analyses of both a four-gene data set (rbcL, 18S rDNA, and two slowly evolving mitochondrial genes, atpA and cox1; Bowe, Coat, and dePamphilis, 2000
) and a three-gene data set (mt SSU rDNA, 18S rDNA, and rbcL; Chaw et al., 2000
) provided strong support for the sister-group relationship of Pinaceae and Gnetales (the "gne-pine" hypothesis of Chaw et al., 2000
). The possible derivation of Gnetales from within conifers is a result supported largely by mitochondrial genes. However, the first and second codon positions of rbcL (Chaw et al., 2000
) and the chloroplast photosystem genes psaA and psbB (Sanderson et al., 2000
) further support the Gnetales plus Pinaceae relationship, although analyses of all positions of rbcL (Chaw et al., 2000
) and third positions of psaA and psbB (Sanderson et al., 2000
) placed Gnetales as sister to all other seed plants. Thus, different data partitions support alternative topologies, and relationships among extant seed plants remain unresolved.
The monophyly of extant gymnosperms and the close relationship of Gnetales and conifers are supported by other molecular data, although the taxon sampling in these studies was sparse (e.g., Hansen et al., 1999
; Winter et al., 1999
; Frohlich and Parker, 2000
). For example, Hansen et al. (1999)
analyzed sequence data for a 9.5-kilobase (kb) portion of the chloroplast genome, but included only Pinus, Gnetum, and three angiosperms and used Marchantia as the outgroup. Winter et al. (1999)
analyzed MADS box genes and found similar results for five genes, but only Gnetum was used to represent Gnetales, and cycads and Ginkgo were not included. Analyses of Floricaula/LEAFY (LFY) sequences also suggested gymnosperms are monophyletic and that Gnetales are sister to pines (Frohlich and Parker, 2000
), but only five gymnosperms were analyzed. Additional analyses, aimed at inferring relationships within tracheophytes as a whole (e.g., Pryer et al., 2001
) or within angiosperms (e.g., Qiu et al., 1999
, 2000
; Soltis, Soltis, and Chase, 1999
; Graham and Olmstead, 2000
), also concur in reconstructing the monophyly of the gymnosperms and/or the sister-group relationship of Gnetales and Pinaceae. Because of small sample sizes (at least for gymnosperms) and low internal support, many of these analyses should be considered equivocal regarding seed plant relationships in general and the phylogenetic position of Gnetales in particular (see also Doyle, 1998
; Donoghue and Doyle, 2000
).
To examine the problem of seed plant relationships further, we constructed a data set containing sequences from four cpDNA genes (rbcL, atpB, psaA, and psbB), three mtDNA genes (mtSSU, coxI, and atpA), and one nuclear gene (18S rDNA) for 16 seed plants and three outgroups. All major groups of seed plants are represented, and the total aligned sequence per taxon is 15 772 base pairs (bp). We addressed the following questions: (1) Do the genome partitions (i.e., plastid, mitochondrial, and nuclear) yield consistent phylogenetic results? (2) For protein-coding genes, do separate partitions of first and second vs. third codon positions contain the same phylogenetic information? (3) Does method of analysis affect the relationships reconstructed for seed plants? Although some of these questions have been addressed previously (e.g., Bowe, Coat, and dePamphilis, 2000
; Chaw et al., 2000
; Sanderson et al., 2000
), none of these studies addressed all three of these questions using a data set with more than two genes from each of the organellar genomes.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
, Soltis et al. (1997)
, and Soltis et al. (1999)
, and the alignment follows that of Soltis et al. (1999)
. As reported previously, a few small regions were difficult to align and were excluded from phylogenetic analysis. Three genes from the mitochondrial genome were used: mitochondrial small subunit (mtSSU) sequences, atpA (= atp1), and coxI. The mtSSU sequences and alignment are from Parkinson, Adams, and Palmer (1999)
and Chaw et al. (2000)
. All but one of the atpA (= atp1) sequences are from Bowe, Coat, and dePamphilis (2000)
; the atpA sequence for Welwitschia is from Qiu et al. (1999)
. The coxI sequences are from Bowe, Coat, and dePamphilis (2000)
. The alignments for atpA and coxI are those of Bowe, Coat, and dePamphilis (2000)
. Known or presumed RNA editing sites were excluded (Bowe, Coat, and dePamphilis, 2000
; C. dePamphilis, Penn State University, personal communication). Sequences of four chloroplast genes were used: rbcL, psaA, psbB, and atpB. Most of the rbcL sequences are from Chase et al. (1993)
and Qiu et al. (1993)
; others are from Hasebe et al. (1992
, 1995
). The psaA and psbB sequences are from Sanderson et al. (2000)
. The atpB sequences are from Yoshinaga et al. (1992)
, Wolf (1997)
, Wolf et al. (1998)
, Qiu et al. (1999)
, Savolainen et al. (2000)
, and Pryer et al. (2001)
(Table 1). Alignment of sequences was straightforward and accomplished by eye (see Chase et al., 1993
; Sanderson et al., 2000
; Savolainen et al., 2000
; Pryer et al., 2001
).
|
To construct this data set, different congeneric species or closely related genera were occasionally used for different genes as placeholders for a lineage (Table 1). For example, Cryptomeria represents the conifer family Taxodiaceae for three genes (18S rDNA, rbcL, and mtSSU), whereas Metasequoia was used for atpA, coxI, and atpB and Sequoia was used for psaA and psbB. For Pinaceae, Pseudotsuga was used for 18S rDNA, rbcL, and mtSSU, and Pinus was used for atpA, coxI, psaA, psbB, and atpB (Table 1). In trees derived from multigene analyses, the name of the taxon included for 18S rDNA is provided.
Sequences of 18S rDNA, rbcL, and mtSSU were available for all 19 taxa. However, sequences of the remaining genes were not available for all taxa. Angiopteris lacks coxI, and Araucaria is missing atpB. The largest numbers of missing sequences are for psbB (Podocarpus, Nymphaea, Illicium, Gnetum, Cryptomeria), atpA (Angiopteris, Asplenium, Taxus, Ephedra), and psaA (Podocarpus, Nymphaea, Illicium, Gnetum). To examine the impact of missing data, a second eight-gene data set was constructed for 13 taxa (11 seed plants, two outgroups), with only a single taxon missing any gene. For this smaller data set, phylogenetic analyses of the 18S rDNA, combined mtDNA, combined cpDNA, and combined eight genes were conducted. Although the results of these analyses are not shown, the topologies and conclusions are identical to those obtained using the 19-taxon data set and will not be discussed further.
Phylogenetic analyses: maximum parsimony
Maximum parsimony (MP) analyses were conducted on the genomic partitions separately and on the entire data set. The cpDNA genes were also analyzed by codon position. Heuristic searches were conducted using PAUP* 4.02 (Swofford, 1999
), with tree-bisection-reconnection (TBR) branch swapping and 1000 random taxon addition replicates, saving all most parsimonious trees. For analyses of individual plastid genes, searches that did not complete were terminated after a minimum of 20 000 shortest trees were obtained. Internal support for relationships was assessed using fast bootstrap analyses with 1000 replicates. Fast bootstrapping provides equivalent values to bootstrapping with branch swapping when support is high and slightly lower values when support is less (Mort et al., 2000
).
Phylogenetic analyses: maximum likelihood
Maximum likelihood (ML) analyses were conducted using PAUP 4.0b2 (Swofford, 1999
) and an HKY85 + 
model of nucleotide substitution (Hasegawa, Kishino, and Yano, 1985
; Yang, 1994
). The MP tree (if more than one) with the greatest –ln score was used to estimate the model parameters and base frequencies. We used one of the most parsimonious trees and ten equiprobable random trees as starting trees and swapped on all trees, saving all optimal trees. The cpDNA data were analyzed in three ways: (1) all sites; (2) first and second codon positions; and (3) third codon postions.
Phylogenetic analyses: Bayesian approach
The Bayesian phylogenetic analyses were conducted using MrBayes version 1.10 (Huelsenbeck, 2000
). We used uniform prior probabilities, the general time-reversible + model of molecular evolution, and a random starting tree. The Markov chain was run for 41 000 generations and sampled every 100 generations. We ran four chains of the Markov Chain Monte Carlo, sampling every 100 generations for 41 000 generations, starting with a random tree. Stationarity was reached at approximately generation 5500; thus, the first 55 trees were the "burn in" of the chain, and phylogenetic inferences are based on those trees sampled after generation 5500.
Likelihood ratio and Shimodaira-Hasegawa tests of alternative topologies
We used two methods to investigate the degree to which alternative phylogenetic hypotheses were supported by the cpDNA data, the likelihood ratio (LR) test using the parametric bootstrap (e.g., Swofford et al., 1996
; Goldman, Anderson, and Rodrigo, 2000
) and the Shimodaira-Hasegawa test (SH test; Shimodaira and Hasegawa, 1999
).
Likelihood ratio test
Our MP analysis of first and second codon positions of cpDNA genes place Gnetales within the conifers sister to Pseudotsuga, in agreement with several other analyses (see Results). We tested the hypothesis that for first and second codon positions there is no significant difference between the placement of Gnetales as sister to all conifers (conifers monophyletic) and a topology in which Gnetales are sister to Pseudotsuga (conifers nonmonophyletic). We first approximated the null distribution using parametric bootstrapping (Huelsenbeck and Hillis, 1996
). To conduct the parametric bootstrap we constructed a constraint tree with Gnetales as sister to a monophyletic conifer clade. We then performed MP analysis, enforcing the constraints imposed by the placement of Gnetales sister to a conifer clade. The constrained analysis found a single most parsimonious tree, from which we estimated, using maximum likelihood, parameter values for the HKY85 + model of molecular evolution and base frequencies (Hasegawa, Kishino, and Yano, 1985
; Yang, 1994
). We then used the most parsimonious tree found in the constrained analysis (Gnetales as sister to monophyletic conifers) to simulate 500 data sets using seq-gen (Rambaut and Grassly, 1997
), with the size of the data set identical to the 19-taxon data set (i.e., 19 taxa, 15 772 bp), for each of the two hypotheses under consideration: Gnetales sister to Pseudotsuga (conifers nonmonophyletic) vs. Gnetales sister to all conifers (conifers monophyletic).
Each of the simulated data sets was analyzed via MP using methods similar to those described above: TBR branch swapping, 100 random taxon addition replicates, and saving all most parsimonious trees. We enforced a topological constraint of Gnetales sister to Pseudotsuga and saved all tree scores and trees. These analyses of the simulated data allowed us to calculate the null distribution of the likelihood-ratio test statistic, delta = (log L1 – log L0) where log L1 is the –ln tree score of the unconstrained tree using data simulated under the assumption that Gnetales are sister to a monophyletic conifer clade and log L0 is the –ln tree score of the tree(s) obtained from the constrained analysis, with Gnetales sister to Pseudostuga and the data simulated assuming that Gnetales are sister to monophyletic conifers. We then tested whether delta calculated from analysis of the real 19-taxon data set fell within this null distribution of delta.
Shimodaira-Hasegawa test
Using the Shimodaira-Hasegawa test (SH test; Shimodaira and Hasegawa, 1999
) as implemented in PAUP 4.02 we tested three topologies using both the full cpDNA data set (all codon positions) and the partition containing the first and second codon positions of the cpDNA data set. We compared the optimal tree obtained (Gnetales sister to Pseudotsuga; conifers nonmonophyletic) to: (1) Gnetales sister to all other seed plants (the topology obtained in MP analyses of third codon positions—see Results), (2) Gnetales sister to a conifer clade (conifers monophyletic), and (3) Gnetales sister to angiosperms. In ML analyses of first and second codon positions, as well as in ML analyses of all codon positions, we found Gnetales sister to Pseudotsuga (see Results). For the other hypotheses of relationships, we performed ML analyses in which we constrained Gnetales to be sister to: (1) all other seed plants, (2) monophyletic conifers, and (3) angiosperms. For the SH test, we obtained the test distribution using the re-estimated log likelihoods (RELL) approximation with 1000 nonparametric bootstrap replicates.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
, 2000
; Soltis et al., 1999
). The difference between our MP tree and the topologies from these earlier studies likely reflects the smaller taxon sampling we used; the sequences included here are taken from the previous studies.
|
; Chaw et al., 2000
). Furthermore, also as in earlier analyses, bootstrap support is very high. The mtDNA tree provides strong support (93%) for the monophyly of the gymnosperms, which appear as the sister group to a well-supported angiosperm clade (100%). Within the gymnosperms, there is also strong support for cycads (95%) followed by Ginkgo (96%) as successive sisters to the remaining gymnosperms and for a conifer clade that includes Gnetales (96%). The sister-group relationship of Gnetales and Pseudotsuga of the Pinaceae (the "gne-pine" hypothesis of Chaw et al., 2000
) is also strongly supported.
|
|
Analyses of psaA and psbB sequences provided nearly identical topologies in all analyses, except one. The MP analyses of all three codon positions for psaA, as well as psbB, did not resolve the position of Gnetales. However, the MP analyses of first and second positions of psaA, and of psbB, indicated that Gnetales are sister to Pseudotsuga, whereas MP analyses of third codon positions for psaA, as well as psbB, suggested that Gnetales are sister to all other seed plants. Using ML and both psaA and psbB, analyses of all codon positions and first and second codon positions found Gnetales sister to Pseudotsuga. In ML analysis of third codon positions for psaA, Gnetales again appeared as sister to Pseudotsuga. However, ML analysis of third codon positions of psbB placed Gnetales sister to all other seed plants.
Maximum parsimony searches of the combined data set of four cpDNA genes yielded two shortest trees (length = 6314; CI = 0.570; RI = 0.476). The two MP trees differed only in the placement of Araucaria, which appeared as sister to either Podocarpus or Taxus and Cryptomeria. The relationships in the MP trees are strongly supported but differ from those obtained with either 18S rDNA or mtDNA sequence data. Gnetales are sister to all other extant seed plants, the monophyly of which receives very strong support (99%). The angiosperms (100% support) are sister to a well-supported (100%) clade of cycads, Gingko, and conifers. Within this clade, the cycads are sister to Ginkgo and conifers, the latter clade receiving bootstrap support of 91% (Fig. 3). Thus, MP analyses of the combined cpDNA data appear to be in strong conflict with mtDNA data and 18S rDNA data, both of which suggest that the gymnosperms are monophyletic and that Gnetales are related to conifers (see also Chaw et al., 1997
, 2000
; Soltis et al., 1999
; Bowe, Coat, and dePamphilis, 2000
).
|
= 0.256 015. The ML tree (–ln = 36 519.524 72) is essentially identical to that obtained with mtDNA gene sequence data. The gymnosperms are monophyletic and sister to the angiosperms. Within the gymnosperms, cycads are sister to Ginkgo and the remaining gymnosperms. Gnetales are embedded within the conifers and appear as sister to Pseudotsuga (Fig. 4).
|
= 0.150 712. The ML tree was the same as the MP tree from which the parameters were estimated. Gnetales again appeared as sister to Pseudotsuga.
Maximum parsimony analysis of the cpDNA third positions produced a single most parsimonious tree (length = 4189; CI = 0.547; RI = 0.460) in which Gnetales appeared as sister to the rest of the seed plants. Maximum likelihood parameter values estimated from the single most parsimonious tree were: (1) base frequencies, A: 0.295 411, C: 0.171 523, G: 0.151 231, T: 0.381 834; (2) ti/tv: 4.915 934; and (3) gamma shape parameter = 0.810 807. In contrast to the MP topology, the ML tree (–ln 18 794.582 21) placed Gnetales as sister to Pseudotsuga and cycads as sister to the angiosperms.
The analysis of the combined cpDNA data set using a Bayesian approach yielded a tree (Fig. 5) comparable to that obtained with ML. Most clades, including the angiosperms (1.0), all gymnosperms (1.0), Gnetales and Pinaceae (0.987), and the successive sister-group relationships of cycads followed by Ginkgo were supported by posterior probabilities of 1.0 or nearly 1.0.
|
|
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
; Chaw et al., 2000
), phylogenetic studies of seed plants employing cpDNA genes have yielded varying topologies depending on the sampling of taxa, the method of analysis, and the codon positions analyzed (e.g., Hasebe et al., 1992
; Albert et al., 1994
; Goremykin et al., 1996; Chaw et al., 2000
; Sanderson et al., 2000
). The difference in signal we observed between first and second vs. third positions for cpDNA genes is a pattern observed for seed plants in other studies. Sanderson et al. (2000)
found strongly supported, but conflicting, topologies with psaA and psbB gene sequences when the data were partitioned into first and second vs. third codon positions. In our MP analyses of first and second positions, both genes supported the monophyly of the gymnosperms, with Gnetales related to Pinaceae; with third positions, Gnetales were sister to all other seed plants. The ML analyses of first and second as well as third positions of psaA yielded a topology with Gnetales sister to Pinaceae, but ML analyses of psbB third positions placed Gnetales as sister to all other seed plants.
A difference in phylogenetic signal between first and second vs. third positions is also apparent for rbcL. Chaw et al. (2000)
obtained different topologies for seed plants depending on the treatment of third positions. With all positions equally weighted, ML analyses of rbcL placed Gnetales as sister to the other seed plants, whereas MP analyses placed Gnetales as sister to the angiosperms. However, when third position transitions were excluded, both ML and MP placed Gnetales as sister to Pinaceae.
Our analyses of cpDNA gene sequences likewise indicate that the signal of third codon positions differs from that of first and second codon positions (see also Chaw et al., 2000
; Sanderson et al., 2000
). Sanderson et al. (2000)
concluded that a factor potentially contributing to bias and inconsistency in their analyses of psaA and psbB is the combination of extremely short branches at the base of the seed plants as well as extremely high rates of molecular evolution in Gnetales and the outgroups. Sanderson et al. (2000)
suggested that the difference in topology obtained with MP for first and second codon positions vs. third codon positions of these two cpDNA genes might be the result of saturation of third position sites at this deep phylogenetic level (i.e., all seed plants). Furthermore, analyses of the cpDNA genes rbcL, atpB, and rps4 across all tracheophytes indicate heterogeneous rates of evolution in first, second, and third codon positions (Soltis et al., 2002
) that could translate into conflicting phylogenetic signal.
Our results, coupled with those of Chaw et al. (2000)
and Sanderson et al. (2000)
, provide an intriguing paradox in that recent studies have pointed to the value of third codon positions of cpDNA genes in phylogeny reconstruction (e.g., Lewis, Mishler, and Vilgalys, 1997
; Källersjö et al., 1998
; Källersjö, Albert, and Farris, 1999
; Olmstead, Reeves, and Yen, 1998
). Even at very deep phylogenetic levels, such as all photosynthetic life, third positions in rbcL sequences provide most of the phylogenetic signal (Källersjö et al., 1998
; Källersjö, Albert, and Farris, 1999
). However, our results suggest that third codon positions of cpDNA genes may be misleading in phylogenetic analyses of seed plants.
The conflict between first and second vs. third codon positions in plastid genes that we observed has similarly been reported in other investigations of seed plant relationships (Sanderson et al., 2000
; Rydin, Källersjö, and Friis, 2002
). The nature of this conflict appears to be complex. Sanderson et al. (2000)
reported conflict between first and second vs. third codon positions in the plastid genes psaA and psbB. Rydin, Källersjö, and Friis (2002)
analyzed the plastid genes rbcL and atpB across seed plants and found that the conflict in codon position is more pronounced in rbcL than in atpB, but the conflict in these genes is less than that observed for psaA and psbB. Although third codon positions of plastid genes generally have most of the phylogenetic signal (e.g., Källersjö et al., 1998
; Olmstead, Reeves, and Yen, 1998
), it may be that third positions are saturated in some, but not all, instances. Chaw et al. (2000)
found that third positions of rbcL were saturated based on transition/transversion ratios across seed plants. In contrast, however, Sanderson et al. (2000)
found that for psaA the third codon positions were not sufficiently saturated to generate a misleading phylogenetic result. Adding to the complexity of the conflict between first, second, and third positions is the fact that transitions within each codon position conflict with transversions (Rydin, Källersjö, and Friis, 2002
).
The preponderance of evidence—gene trees based on (1) first and second positions of four cpDNA genes, as well as third positions with ML and all positions with ML and Bayesian approaches, (2) nuclear 18S rDNA, and (3) three mtDNA genes—supports a relationship between Gnetales and conifers, although 18S rDNA suggests that Gnetales are sister to all conifers rather than to Pinaceae (Soltis et al., 1999
; Chaw et al., 2000
).
General conclusions
An overall conclusion of our analyses is that some aspects of relationship among extant seed plants appear to be largely resolved. We conclude that extant gymnosperms are monophyletic and that cycads and Ginkgo are sisters to the remaining gymnosperms, with Gnetales sister to Pinaceae (if not embedded within Pinaceae) within the conifers. However, the relationship between cycads and Ginkgo differs in our analyses; in some analyses, cycads and Ginkgo are successive sisters to a clade of conifers and Gnetales, whereas in others, Ginkgo and cycads form a clade that is sister to the other gymnosperms.
Most of the conflict observed in our own analyses, as well as in previous analyses, involves the difference between first and second vs. third codon positions of cpDNA genes. When ML analyses of cpDNA genes are conducted, or third postions are excluded in MP analyses, a topology highly similar to those obtained with mtDNA genes emerges: gymnosperms are strongly supported as monophyletic and are sister to the angiosperms. Within the gymnosperms, cycads and Ginkgo are successive sisters to the remainder of the clade; Gnetales are closely related to conifers. However, the precise relationship of Gnetales to conifers is unclear. Some analyses of both plastid and mitochondrial genes support a placement of Gnetales within conifers as sister to Pinaceae (the "gne-pine" hypothesis; Bowe, Coat, and dePamphilis, 2000
; Chaw et al., 2000
). However, with 18S rDNA sequences (Chaw et al., 1997
; Soltis et al., 1999) and with multiple genes that include a broad representation of conifers (Rydin, Källersjö, and Friis, 2002
), Gnetales are sister to all conifers. Hence, the gne-pine hypothesis may be an artifact of inadequate taxon sampling in some analyses (Rydin, Källersjö, and Friis, 2002
).
Hypothesis tests using the LR and SH tests and the cpDNA data provide conflicting results. The LR test based on first and second positions indicates that there is a significant difference in the MP tree length between the two topologies examined, Gnetales as sister to all conifers (conifers monophyletic) and a topology in which Gnetales are sister to Pseudotsuga (conifers non-monophyletic). The preferred topology based on the LR test is that Gnetales are sister to Pseudotsuga. The SH test based on first and second codon positions and all codon positons indicated that there is no significant difference in –ln score between the best topology (Gnetales sister to Pseudotsuga) and Gnetales sister to a conifer clade. However, there is a significant difference in –ln score between the best topology and the topology in which Gnetales are sister to the rest of the seed plants and the topology in which Gnetales are sister to the angiosperms. Thus, we found that the alternative placement of Gnetales as sister to conifers (rather than as sister to Pinaceae) could not be rejected consistently, in agreement with Rydin, Källersjö, and Friis (2002)
, who, in some analyses, recovered Gnetales as sister to a monophyletic conifer clade. Lastly, a structural mutation in the chloroplast genome (Raubeson and Jansen, 1992
) also suggests that a Gnetales-conifer sister-group relationship may be a more parsimonious explanation of the data.
Future considerations
With 15 772 bp of sequence data representing all three plant genomes, relationships among the five lineages of extant seed plants are not yet fully resolved. The value of additional molecular studies of a restricted number of taxa therefore needs to be carefully considered. The precise relationship of cycads and Ginkgo—are they sister taxa or subsequent branches to the remaining gymnosperms?—has yet to be answered. Furthermore, although most molecular data suggest a close relationship between conifers and Gnetales, the exact nature of this relationship is unclear: Are Gnetales derived from within conifers or are they the sister group to conifers? To determine the placement of Gnetales relative to conifers we suggest that additional conifers be added to existing gene sequence data sets, rather than simply adding sequences of more genes.
Although molecular data can provide some additional insights into the relationships among extant seed plants, efforts at achieving a better understanding of relationships among extant seed plants would be better placed by integrating fossil taxa into comprehensive analyses of seed plant phylogeny. For example, DNA data strongly support the monophyly of extant gymnosperms, but is this a valid conclusion when fossil taxa are included? If no single extant lineage of gymnosperms is the sister group of angiosperms, can a fossil lineage be identified that fills this role? More attention must be paid to fossils and morphological features, with future efforts directed at the integration of fossils into phylogenetic analyses based on both morphological and molecular data (see also Doyle, 1998
; Donoghue and Doyle, 2000
). Such analyses will require a careful reconsideration of the morphological characters included in previous studies of seed plant phylogeny. Clearly, the homology of those characters that linked angiosperms and Gnetales must be reevaluated. Analyses that integrate fossils and morphology with molecular characters may be more fruitful than further analyses of molecular data alone.
|
|
|
FOOTNOTES |
|---|
5 Author for reprint requests (dsoltis{at}botany.ufl.edu
)
|
LITERATURE CITED |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Bowe L. M. G. Coat C. W. dePamphilis 2000 Phylogeny of seed plants based on all three genomic compartments: extant gymnosperms are monophyletic and Gnetales' closest relatives are conifers. Proceedings of the National Academy of Sciences, USA 97: 4092-4097
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]
Chaw S.-M. C. L. Parkinson Y. Cheng T. M. Vincent J. D. Palmer 2000 Seed plant phylogeny inferred from all three plant genomes: monophyly of extant gymnosperms and origin of Gnetales from conifers. Proceedings of the National Academy of Sciences, USA 97: 4086-4091
Chaw S.-M. A. Zharkikh H.-M. Sung T.-C. Lau W.-H. Li 1997 Molecular phylogeny of extant gymnosperms and seed plant evolution: analysis of nuclear 18S rRNA sequences. Molecular Biology and Evolution 14: 56-58[Abstract]
Crane P. R. 1985 Phylogenetic analysis of seed plants and the origin of angiosperms. Annals of the Missouri Botanical Garden 72: 716-793[CrossRef][ISI]
Deno H. K. Shinozaki M. Sugiura 1983 Nucleotide sequence of tobacco chloroplast gene for the alpha subunit of proton-translocating ATPase. Nucleic Acids Research 11: 2185-2191
Donoghue M. J. J. A. Doyle 2000 Seed plant phylogeny: demise of the anthophyte hypothesis?. Current Biology 10: R106-R109
Doyle J. A. 1996 Seed plant phylogeny and the relationships of Gnetales. International Journal of Plant Sciences 157: S3-S39
Doyle J. A. 1998 Molecules, morphology, fossils, and the relationships of angiosperms and Gnetales. Molecular Phylogenetics and Evolution 9: 448-462
Doyle J. A. M. J. Donoghue 1986 Seed plant phylogeny and the origin of the angiosperms: an experimental cladistic approach. Botanical Review 52: 321-431
Doyle J. A. M. J. Donoghue 1992 Fossils and seed plant phlyogeny reanalyzed. Brittonia 44: 89-104[CrossRef][ISI]
Frohlich M. W. D. S. Parker 2000 The mostly male theory of flower evolutionary origins: from genes to fossils. Systematic Botany 25: 155-170
Goldman N. J. P. Anderson A. G. Rodrigo 2000 Likelihood-based tests of topologies in phylogenetics. Systematic Biology 49: 652-670
Goremykin V. V. Bobrova J. Pahnke A. Troitsky A. Antonov W. Martin 1996 Noncoding sequences from the slowly evolving chloroplast inverted repeat in addition to rbcL data do not support Gnetalean affinities of angiosperms. Molecular Biology and Evolution 13: 383-396[Abstract]
Graham S. W. R. G. Olmstead 2000 Utility of 17 chloroplast genes for inferring the phylogeny of the basal angiosperms. American Journal of Botany 87: 1712-1730
Hamby R. K. E. A. Zimmer 1992 Ribosomal RNA as a phylogenetic tool in plant systematics. In P. S. Soltis, D. E. Soltis, and J. J. Doyle [eds.], Molecular systematics of plants, 50–91. Chapman and Hall, New York, New York, USA
Hansen A. S. Hansmann T. Samigullin A. Antonov W. Martin 1999 Gnetum and the angiosperms: molecular evidence that their shared morphological characters are convergent, rather than homologous. Molecular Biology and Evolution 16: 1006-1009[ISI]
Hasebe M. et al 1995 Fern phylogeny based on rbcL nucleotide sequences. American Fern Journal 85: 134-181[CrossRef]
Hasebe M. R. Kofuji M. Ito M. Kato K. Iwatsuki K. Ueda 1992 Phylogeny of gymnosperms inferred from rbcL gene sequences. Botanical Magazine of Tokyo 105: 673-679
Hasegawa M. H. Kishino T. Yano 1985 Dating of the human-ape splitting by a molecular clock of mitochondrial DNA. Journal of Molecular Evolution 22: 32-38
Hoot S. B. J. W. Kadereit F. R. Blattner K. B. Jork A. E. Schwarzbach P. R. Crane 1997 Data congruence and phylogeny of the Papaveraceae s. l. based on four data sets: atpB and rbcL sequences, trnK restriction sites, and morphological characters. Systematic Botany 22: 575-590
Huelsenbeck J. P. 2000 MrBayes. Distributed by the author. Department of Biology, University of Rochester, Rochester, New York, USA
Huelsenbeck J. P. D. M. Hillis 1996 Parametric bootstrapping in molecular phylogenetics: applications and performance. In J. D. Ferraries and S. R. Palumbi [eds.], Molecular zoology: advances, strategies, and protocols, 19–45. Wiley-Liss, New York, New York, USA
Källersjö M. V. Albert J. S. Farris 1999 Homoplasy increases phylogenetc structure. Cladistics 15: 91-93[CrossRef][ISI]
Källersjö M. J. S. Farris M. W. Chase B. Bremer M. F. Fay C. J. Humphries G. Petersen O. Seberg K. Bremer 1998 Simultaneous parsimony jackknife analysis of 2538 rbcL DNA sequences reveals support for major clades of green plants, land plants, seed plants, and flowering plants. Plant Systematics and Evolution 213: 259-287
Lewis L. A. B. D. Mishler R. Vilgalys 1997 Phylogenetic relationships of the liverworts (Hepaticae), a basal Embryophyte lineage, inferred from nucleotide sequence data of the chloroplast gene rbcL. Molecular Phylogenetics and Evolution 7: 377-393
Loconte H. D. W. Stevenson 1990 Cladistics of the Spermophyta. Brittonia 42: 197-211[CrossRef][ISI]
Malek O. K. Lattig R. Hiesel A. Brennicke V. Knoop 1996 RNA editing in bryophytes and a molecular phylogeny of land plants. EMBO 14: 1403-1411
Mort M. E. P. S. Soltis D. E. Soltis M. L. Mabry 2000 A comparison of three methods for estimating internal support on phylogenetic trees. Systematic Biology 49: 102-113
Olmstead R. G. P. A. Reeves A. C. Yen 1998 Patterns of sequence evolution and implications for parsimony analysis of chloroplast DNA. In D. E. Soltis, P. S. Soltis, and J. J. Doyle [eds.], Molecular systematics of plants II, 164–187. Kluwer, New York, New York, USA
Parkinson C. L. K. L. Adams J. D. Palmer 1999 Multigene analyses identify the three earliest lineages of extant flowering plants. Current Biology 9: 1485-1488
Pryer K. M. H. Schneider A. R. Smith R. Cranfill P. Wolf J. S. Hunt S. D. Sipes 2001 Horsetails and ferns are a monophyletic group and the closest living relatives to seed plants. Nature 409: 618-622[CrossRef][Medline]
Qiu Y.-L. M. W. Chase D. H. Les C. R. Parks 1993 Moleuclar phylogenetics of the Magnoliidae: cladistic analyses of nucleotide sequences of the plastid gene rbcL. Annals of the Missouri Botanical Garden 80: 587-606[CrossRef][ISI]
Qiu Y.-L. J. Lee F. Bernasconi-Quadroni D. E. Soltis P. S. Soltis M. Zanis E. A. Zimmer Z. Chen V. Savolainen M. W. Chase 1999 The earliest angiosperms: evidence from mitochondrial, plastid and nuclear genomes. Nature 402: 404-407
Qiu Y.-L. J. Lee F. Bernasconi-Quadroni D. E. Soltis P. S. Soltis M. Zanis E. A. Zimmer Z. Chen V. Savolainen M. W. Chase 2000 Phylogenetic analyses of basal angosperms based on five genes from all three genomes. International Journal of Plant Sciences 161: S3-S27
Rambaut A. N. C. Grassly 1997 Seq-Gen: an application for the Monte Carlo simulation of DNA sequence evolution along phylogenetic trees. Computer Applications in the Biosciences 13: 235-238
Raubeson L. A. R. K. Jansen 1992 A rare chloroplast-DNA structural mutation is shared by all conifers. Biochemical Systematics and Ecology 20: 17-24
Rothwell G. W. R. Serbet 1994 Lignophyte phylogeny and the evolution of spermatophytes: a numerical cladistic analysis. Systematic Botany 19: 443-482
Rydin C. M. Källersjö E. M. Friis 2002 Seed plant relationships and the systematic position of Gnetales based on nuclear and chloroplast DNA: conflicting data, rooting problems, and the monophyly of conifers. International Journal of Plant Sciences 163: 197-214
Sanderson M. J. M. F. Wojciechowski J.-M. Hu T. S. Khan S. G. Brady 2000 Error, bias, and long-branch attraction in data for two chloroplast photosystem genes in seed plants. Molecular Biology and Evolution 17: 782-797
Savolainen V. M. W. Chase C. M. Morton S. B. Hoot D. E. Soltis C. Bayer M. F. Fay A. De Bruijn S. Sullivan Y.-L. Qiu 2000 Phylogenetics of flowering plants based upon a combined analysis of plastid atpB and rbcL gene sequences. Systematic Biology 49: 306-362
Shimodaira H. M. Hasegawa 1999 Multiple comparisons of log-likelihoods with applications to phylogenetic inference. Molecular Biology and Evolution 16: 1114-1116[ISI]
Soltis D. E. et al 1997 Angiosperm phylogeny inferred from 18S ribosomal DNA sequences. Annals of the Missouri Botanical Garden 84: 1-49
Soltis D. E. et al 2000 Angiosperm phylogeny inferred from a combined data set of 18S rDNA, rbcL and atpB sequences. Botanical Journal of the Linnaean Society 133: 381-461
Soltis P. S. D. E. Soltis M. W. Chase 1999 Angiosperm phylogeny inferred from multiple genes as a tool for comparative biology. Nature 402: 402-404
Soltis P. S. D. E. Soltis V. Savolainen P. R. Crane T. G. Barraclough 2002 Integration of molecular and fossil data and evidence for molecular living fossils. Proceedings of the National Academy of Sciences, USA 99: 4430-4435
Soltis P. S. D. E. Soltis P. G. Wolf D. L. Nickrent S.-M. Chaw R. L. Chapman 1999 The phylogeny of land plants inferred from 18S rDNA sequences: pushing the limits of rDNA signal?. Molecular Biology and Evolution 16: 1774-1784[Abstract]
Swofford D. L. 1999 PAUP*: phylogenetic analysis using parsimony (*and other methods). Version 4.0b4a. Sinauer, Sunderland, Massachusetts, USA
Swofford D. L. G. J. Olsen P. Waddell D. M. Hillis 1996 Phylogenetic inference. In D. M. Hillis, C. Moritz, and B. K. Mable [eds.], Molecular systematics, 407–425. Sinauer, Sunderland, Massachusetts, USA
Winter K.-U. A. Becker T. Munster J. T. Kim H. Saedler G. Thiessen 1999 The MADS-box genes reveal that gnetophytes are more closely related to conifers than to flowering plants. Proceedings of the National Academy of Sciences, USA 96: 7342-7347
Wolf P. G. 1997 Evaluation of atpB nucleotide sequences for phylogenetic studies of ferns and other pteridophytes. American Journal of Botany 84: 1429-1440[Abstract]
Wolf P. G. K. M. Pryer A. R. Smith M. Hasebe 1998 Phylogenetic studies of extant pteridophytes. In D. E. Soltis, P. S. Soltis, and J. J. Doyle [eds.], Molecular systematics of plants II, 541–556. Kluwer, Boston, Massachusetts, USA
Yang Z. 1994 Phylogenetic analysis using parsimony and likelihood methods. Journal of Molecular Evolution 39: 294-307
Yoshinaga K. Y. Kubota T. Ishii K. Wada 1992 Nucleotide sequence of atpB, rbcL, trnR, dedB and psaI chloroplast genes from a fern Angiopteris lygodiifolia: a possible emergence of Spermatophyta lineage before the separation of Bryophyta and Pteridophyta. Plant Molecular Biology 18: 79-82[CrossRef][ISI][Medline]
This article has been cited by other articles:
|
|
 |
