© 2006 Society of Systematic Biologists
Reconciling Extreme Branch Length Differences: Decoupling Time and Rate through the Evolutionary History of Filmy Ferns
Edited by Thomas Buckley: Associate Editor
Department of Biology, Duke University Durham, North Carolina, 27708, USA E-mail: ejs7{at}duke.edu; pryer{at}duke.edu
 |
Abstract |
|---|
 Top Abstract MethodsResults and DiscussionConclusions and ProspectsReferences |
|---|
The rate of molecular evolution is not constant across the Tree of Life. Characterizing rate discrepancies and evaluating the relative roles of time and rate along branches through the past are both critical to a full understanding of evolutionary history. In this study, we explore the interactions of time and rate in filmy ferns (Hymenophyllaceae), a lineage with extreme branch length differences between the two major clades. We test for the presence of significant rate discrepancies within and between these clades, and we separate time and rate across the filmy fern phylogeny to simultaneously yield an evolutionary time scale of filmy fern diversification and reconstructions of ancestral rates of molecular evolution. Our results indicate that the branch length disparity observed between the major lineages of filmy ferns is indeed due to a significant difference in molecular evolutionary rate. The estimation of divergence times reveals that the timing of crown group diversification was not concurrent for the two lineages, and the reconstruction of ancestral rates of molecular evolution points to a substantial rate deceleration in one of the clades. Further analysis suggests that this may be due to a genome-wide deceleration in the rate of nucleotide substitution.
Keywords: Bayesian analysis; divergence time estimates; molecular clock; molecular rate heterogeneity; monilophyte phylogeny; penalized likelihood; rbcL
Received August 3, 2005; Revised November 15, 2005; Accepted March 13, 2006
Phylogenetic branch length, as estimated from DNA sequence data, is a function of both the rate of nucleotide substitution and time. Length differences among individual branches within a phylogeny can therefore result from discrepancies in substitution rate, time, or a combination of these factors. All of life's extant diversity, however, ultimately shares a common ancestor, and thus a common age. Any significant differences among cumulative evolutionary path lengths, from the root of the Tree of Life to the many extant species, must purely be the result of net rate discrepancies. This holds for any phylogenetic tree, as the sampled extant taxa will always share a common root.
Phylogenetic trees with a combination of long and short cumulative evolutionary paths are common, and the phenomenon of unequal net rates of nucleotide substitution among lineages is widespread and well recognized (Britten, 1986; Bromham and Penny, 2003; Gaut et al., 1993; Langley and Fitch, 1974; Wolfe et al., 1987). Significant deviations from the constant rate of a molecular clock have been identified across the Tree of Life (vertebrates: Adachi et al., 1993; Bleiweiss, 1998; Bromham, 2002; Bulmer et al., 1991; Cantatore et al., 1994; Hoegg et al., 2004; Krieger and Fuerst, 2002; Li et al., 1987, 1990; Martin and Palumbi, 1993; Martin et al., 1992; Mooers and Harvey, 1994; Springer and Kirsch, 1989; Wu and Li, 1985; invertebrates: Castro et al., 2002; Hebert et al., 2002; Moran et al., 1995; Schön et al., 2003; seed plants: Bousquet et al., 1992, Gaut et al., 1992; Nickrent and Starr, 1994; ferns: Des Marais et al., 2003; Schneider et al., 2004b; liverworts: Lewis et al., 1997; fungi: Lutzoni and Pagel, 1997; Zoller and Lutzoni, 2003; algae: Zoller and Lutzoni, 2003; bacteria: Moran et al., 1995).
An analysis of net rate differences, although capable of providing an informative summary, cannot possibly reveal all the intricacies involved in the evolution of the rate of evolution. Any given phylogenetic path from root to tip comprises many individual branch segments, with rate and time together determining the length of each. Time intervals between successive divergences can vary along a path, as can rates of molecular evolution, and early changes in rate in one direction (either an acceleration or a deceleration) can be masked by subsequent changes in the other direction. Characterizing the net amount and direction of change in the rate of molecular evolution is important in understanding the evolutionary history of a lineage, but equally important is an evaluation of the interaction of rate and time on individual branches through the past.
Within filmy ferns (Hymenophyllaceae), phylogenetic analyses of plastid rbcL sequences have revealed considerable cumulative path length differences (Pryer et al., 2001b), but the significance of these differences has not yet been addressed nor the factors responsible identified. Filmy ferns compose one of the earliest diverging families of leptosporangiate ferns (Hasebe et al., 1995; Pryer et al., 2001a, 2004; Schneider et al., 2004a). The more than 600 described species all have extremely thin leaves, with blades generally only a single cell thick, that bear unique marginal sori—reproductive structures consisting of a short to elongate sporangia-bearing receptacle subtended by a protective indusium (Iwatsuki, 1990). The species otherwise exhibit considerable levels of both morphological and ecological diversity (Dubuisson, 1996, 2003b). Recent molecular phylogenetic studies have resolved two major lineages of filmy ferns, largely corresponding to the two traditionally recognized genera—Trichomanes and Hymenophyllum (Pryer et al., 2001b). The primary and most consistent morphological differences between these groups are related to the morphology of the sori, but other less generalized differences also exist. The Trichomanes clade is generally characterized by having sori with campanulate (bell-shaped) indusia and exserted receptacles, tends to inhabit lower latitudes and lower altitudes, and comprises terrestrial, climbing, and epiphytic species. The Hymenophyllum clade usually has bivalved indusia and included receptacles, tends to be more successful at higher latitudes and altitudes, and is composed mostly of epiphytic species, many of which are present high in the forest canopy. The aforementioned path length discrepancies are also manifested between the two major filmy fern clades, with the Trichomanes clade comprising species with relatively long path lengths and the Hymenophyllum clade comprising species with relatively short path lengths (Pryer et al., 2001b).
The major lineages of filmy ferns, as sister taxa, share a common age. Therefore, the path length differences observed in analyses of rbcL data, if significant, must ultimately be the result of net rate change at this locus in one or both of the filmy fern groups. Relative to the ancestral rate of molecular evolution there was either (1) a net acceleration in the Trichomanes lineage, (2) a net deceleration in the Hymenophyllum lineage, or (3) both a net acceleration in Trichomanes and a net deceleration in Hymenophyllum. However, a consideration of both time and rate through the evolutionary history of filmy ferns—rather than simply net changes in rate—yields countless plausible evolutionary scenarios, each of which has unique implications.
In this study, we explore the roles of time and rate in the evolutionary history of filmy ferns. We evaluate the significance and nature of rate differences at the rbcL locus and we separate time and rate across the filmy fern phylogeny to yield an evolutionary time scale of filmy fern diversification, as well as reconstructions of ancestral rates of molecular evolution.
|
Methods |
|---|
|
TopAbstractMethodsResults and DiscussionConclusions and ProspectsReferences |
|---|
Taxonomic Sampling and Sequence Alignment
Fifty species were selected from the Hymenophyllaceae—25 species each from Trichomanes and Hymenophyllum—representing all of the major filmy fern lineages (Dubuisson et al., 2003a; Hennequin et al., 2003). To place this family within a broader context, 60 other vascular plant species were selected: 42 additional ferns from across the leptosporangiate phylogeny, nine species representing the four major eusporangiate fern lineages (Pryer et al., 2004), six seed plants, and three lycophytes (outgroup). DNA sequences of the plastid rbcL gene were obtained for each included species from GenBank (for voucher information and GenBank accession numbers, see Table 1) and aligned manually using MacClade 4.06 (Maddison and Maddison, 2000). The 5' and 3' ends of the resulting alignment that contained copious amounts of missing data were cropped, yielding a data matrix of 1206 base pairs (402 codons) for 110 species with no missing data (only 13 ambiguities were present within the matrix). The resulting alignment is available in TreeBASE (http://www.treebase.org; study accession number S1449).
|
Phylogenetic Analyses
The rbcL data were analyzed using a Bayesian Markov chain Monte Carlo (B/MCMC) approach, as implemented in MrBayes 3.0b4 (Ronquist and Huelsenbeck, 2003), to simultaneously yield a distribution of trees, a consensus phylogenetic hypothesis, and support values for resolved nodes. Four independent B/MCMC analyses were conducted using the model of sequence evolution most applicable to the data (GTR+
+I, as selected using ModelTest 3.06; Posada and Crandall, 1998), flat priors, and four chains. Chains were run for 10 million generations, and trees were sampled every 1000 generations. Following completion of each analysis, we plotted the output parameter estimates through time using Tracer 1.3 (Rambaut and Drummond, 2005) in order to recognize the point of convergence to the stationary distribution. All generations prior to this convergence (1,000,000 generations, 1000 trees, for each of the four analyses) were discarded, conservatively, as the "burn-in" phase. Through the superimposition of the parameter plots from the four analyses and a comparison of the trees resulting from these analyses, we confirmed that all four independent runs had converged to the same stationary distribution. Therefore, we pooled the post burn-in trees from each analysis (36,000 total trees), and computed a plurality consensus (using the command: sumt contype = allcompat) to obtain a fully resolved topology with average branch lengths, as well as posterior probability estimates for all nodes.
Significance Tests for the Presence of Rate Differences
To determine whether observed branch length differences were the result of a significant departure from rate constancy (i.e., a molecular clock), two models were compared. In the simpler (null) model, a molecular clock was applied such that the rates of molecular evolution for each of the branches were constrained to be equal. In the more complex (alternative) model, each branch was allowed its own unique rate of molecular evolution. These two models were contrasted across the entire Bayesian consensus tree, as well as across several partitions pruned from this tree: filmy ferns plus their resolved sister group, filmy ferns, Hymenophyllum, and Trichomanes. For each comparison, likelihoods were calculated using the program Baseml (part of the PAML 3.14b package; Yang, 1997) with the appropriate models of sequence evolution (GTR+; either with or without a molecular clock constraint). The resulting likelihoods were compared using the likelihood ratio test statistic (Felsenstein, 1981).
In addition to these tests for the presence of molecular clocks, several tests were performed to determine whether significant differences in rate were present between (as opposed to within) partitions. For these analyses, the pruned tree comprising filmy ferns plus their resolved sister group was utilized and comparisons were again made between simple (null) and more complex (alternative) models. Three comparisons were made: (1) a two-rate model in which Hymenophyllum and Trichomanes had the same rate but the resolved filmy fern sister group had a different rate versus a three-rate model in which all three included partitions (Hymenophyllum, Trichomanes, and the resolved filmy fern sister group) had different rates; (2) a two-rate model in which the filmy fern sister group and Hymenophyllum had the same rate but Trichomanes had a different rate versus the three-rate model described above; and (3) a two-rate model in which the filmy fern sister group and Trichomanes had the same rate but Hymenophyllum had a different rate versus the three-rate model described above. Again, likelihoods were determined using the program Baseml (part of the PAML 3.14b package; Yang 1997) with the appropriate models (GTR+; branch rates as described above), and compared using the likelihood-ratio test statistic (Felsenstein, 1981).
To further characterize rate differences, pairwise relative rate comparisons were conducted. All included filmy fern species, as well as all species in the resolved sister lineage to filmy ferns, were evaluated relative to one another. For each pairwise comparison (3916 total), a three-taxon tree was constructed (outgroup = Osmunda cinnamomea) and two models were compared—one with (null) and one without (alternative) the constraint of equal rates between the two ingroup species. The likelihoods corresponding to each of these models were compared using the likelihood ratio test statistic (Felsenstein, 1981). All 3916 pairwise comparisons were made in an automated fashion using the program HyPhy 0.99 beta (Kosakovsky Pond et al., 2005), with the GTR+ +I model of sequence evolution and globally estimated parameters; no correction for multiple comparisons was incorporated.
Estimation of Divergence Times and Ancestral Rates
Ancestral rates of molecular evolution and divergence times were estimated using penalized likelihood (Sanderson, 2002). This semiparametric method separates rate and time from branch length by combining a likelihood model in which each branch has its own rate parameter with a roughness penalty that penalizes excessive rate change from branch to branch. The interplay between the likelihood model and the roughness penalty is controlled by a smoothing parameter that is objectively identified from the data using a cross-validation procedure (Sanderson, 2002). Penalized likelihood analyses of the Bayesian consensus tree as well as of the 100 randomly sampled trees from the Bayesian posterior (to evaluate the effects of phylogenetic uncertainty due to both topological and branch length estimation error) were conducted using the program r8s version 1.60 (Sanderson, 2003). In all analyses, the three lycophyte outgroup taxa were pruned, and the resulting root—the divergence of monilophytes (ferns) from spermatophytes (seed plants)—was used as the fixed calibration point (380 Ma, node 001, Table 1) based on the earliest appearance of fossils belonging to each of these lineages in the Middle Devonian. In addition to this fossil calibration point, 16 minimum fossil age constraints from a previous reassessment of the fern fossil record (Pryer et al., 2004; Schneider et al., 2004a) were incorporated (Table 1; fossil constraints were applied only to nodes receiving high posterior probability support, 
0.99). The appropriate smoothing value was independently identified for each of the 101 trees (100 randomly sampled trees plus the consensus tree) using cross validation (smoothing values from 1 to 10,000 were considered; for most trees, a value of 100 was found to be the most appropriate). Searches for solutions that optimized the penalized likelihood function were conducted using the truncated Newton algorithm with 10 random starts, each with 10 random perturbations.
Examining the Influence of Selection
To identify whether significant selectional differences exist between the Hymenophyllum and Trichomanes clades, or between either of these lineages and the filmy fern sister group, a series of tests was conducted using a subtree pruned from the Bayesian plurality consensus (comprising filmy ferns plus their resolved sister group). A total of three comparisons were made, analogous to those used to test for rate differences between partitions: (1) a two-ratio model in which Hymenophyllum and Trichomanes had the same nonsynonymous/synonymous substitution (dn/ds) ratio but the filmy fern sister group had a different ratio versus a three-ratio model in which all three included partitions (Hymenophyllum, Trichomanes, and the resolved filmy fern sister group) had different dn/ds ratios; (2) a two-ratio model in which the filmy fern sister group and Hymenophyllum had the same dn/ds ratio but Trichomanes had a different ratio versus the three-ratio model described above; and (3) a two-ratio model in which the filmy fern sister group and Trichomanes had the same dn/ds ratio but Hymenophyllum had a different ratio versus the three-ratio model described above. Likelihoods were calculated for each of these models using the program Codeml (part of the PAML 3.14b package; Yang, 1997). Codon frequencies were estimated from the average nucleotide frequencies at the three codon positions and equal dn/ds ratios and rates were assumed among sites. Resulting likelihoods were compared using the likelihood ratio test statistic (Felsenstein, 1981).
|
Results and Discussion |
|---|
|
TopAbstractMethodsResults and DiscussionConclusions and ProspectsReferences |
|---|
Phylogeny
Phylogenetic analysis of the plastid rbcL data yielded a reasonably well-supported hypothesis of relationships (Fig. 1); 60 of 107 nodes received high posterior probability (PP) support (
0.99). In the Bayesian consensus tree, monilophytes (MON) are strongly supported as monophyletic (PP = 1.00), consistent with earlier analyses of morphological and multigene DNA sequence data sets (Kenrick and Crane, 1997; Renzaglia et al., 2000; Nickrent et al., 2000; Pryer et al., 2001a, 2004; Wikström and Pryer, 2005). Within monilophytes, whisk ferns (WHI) and ophioglossoid ferns (OPH) form a well-supported clade (PP = 1.00), sister to the remaining monilophyte lineages. Subsequently, marattioid ferns (MAR) are resolved as sister to a clade consisting of horsetail ferns (HOR) and leptosporangiate ferns (LEP). These basal monilophyte relationships based on analyses of rbcL alone are not in complete agreement with those identified in other studies (Pryer et al., 2001a, 2004), but are also not supported by high posterior probability values (Fig. 1). This region of the monilophyte phylogeny has proven difficult in the past; even multigene analyses have so far failed to fully clarify these relationships (Pryer et al., 2001a, 2004; Wikström and Pryer, 2005).
|
Within leptosporangiate ferns (LEP), osmundaceous ferns (OSM) are sister to a large, well-supported clade (PP = 1.00) containing all other leptosporangiate lineages: gleichenioid (GLE), schizaeoid (SCH), tree (TRE), heterosporous (HET), polypod (POL), and filmy (FIL) ferns (Fig. 1). Each of these lineages, with the exception of gleichenioid ferns, is strongly supported as monophyletic by our analyses (PP = 1.00). Tree, heterosporous, and polypod ferns form a clade of core leptosporangiates (COR; PP = 1.00) sister to schizaeoid ferns (PP = 1.00); and gleichenioid ferns are paraphyletic to this core leptosporangiate + schizaeoid fern clade. Filmy ferns are resolved as sister to the assemblage of gleichenioid, schizaeoid, tree, heterosporous, and polypod ferns (Fig. 1). These results are mostly consistent with earlier analyses of leptosporangiate fern relationships (Hasebe et al., 1995; Pryer et al., 2001a, 2004), with the areas of uncertainty in this study also equivocal in the earlier studies.
Two major filmy fern lineages are resolved and well supported (PP = 1.00), corresponding to the two traditionally defined filmy fern genera: Hymenophyllum (HYM) and Trichomanes (TRI). The composition of these two clades is identical to that found in previous studies (Dubuisson et al., 2003a; Ebihara et al., 2002; Hennequin et al., 2003; Pryer et al., 2001b), with the monotypic segregate genera (Cardiomanes, Hymenoglossum, Rosenstockia, and Serpyllopsis) and species of Microtrichomanes, all nested within the Hymenophyllum clade. High posterior probability support is present for 9 of 23 nodes resolved within the Hymenophyllum clade and 12 of 23 nodes within the Trichomanes clade. The relationships of species within each of these two major clades are essentially in agreement with previous studies (Dubuisson et al., 2003a; Ebihara et al., 2002; Hennequin et al., 2003; Pryer et al., 2001b).
Significant Rate Differences
Considerable branch length differences were evident between the two major filmy fern lineages, as well as within each of these lineages and across the phylogeny as a whole (Fig. 1). Likelihood ratio test comparisons of null models of rate constancy versus alternative models with unique rates of molecular evolution for each branch revealed statistically significant departures from rate constancy across the entire tree and within all partitions examined (comparisons 1 to 5, Table 2). Even within the Hymenophyllum clade, where branch lengths appear to reflect clock-like evolution (Fig. 1), a molecular clock could be rejected (P < 0.001). Nevertheless, although these tests do indicate that the branch length differences observed are ultimately the result of significantly different rates of molecular evolution, they do not identify particular branches with aberrant rates. Within filmy ferns specifically, these tests alone do not reveal where an acceleration or deceleration in rate occurred, nor do they even distinguish between inter- and intrageneric differences.
|
Likelihood ratio tests to identify significant differences in rate among, as opposed to within, partitions yielded more meaningful results (Table 2). A significant rate difference between the Trichomanes and Hymenophyllum clades (P < 0.001; comparison 6, Table 2) was uncovered when we compared a null two-rate model in which filmy ferns were assigned a single rate of evolution (and their sister lineage a second rate) versus an alternative three-rate model in which the two filmy fern genera were each allowed to have a unique rate (and their sister lineage a third rate). Comparisons of two additional two-rate models (one in which Hymenophyllum was assigned the same rate as the filmy fern sister lineage and Trichomanes a unique rate; the other in which Trichomanes was assigned the same rate as the filmy fern sister lineage and Hymenophyllum a unique rate) versus the three-rate model described above, further clarified the nature of the rate differences. These tests revealed that the rate within Hymenophyllum is significantly different from that of the filmy fern sister lineage (P = 0.001; comparison 7, Table 2), whereas the rate within Trichomanes is not (P = 0.830; comparison 8, Table 2).
Pairwise relative rate comparisons among all included filmy fern species and all species in the lineage resolved as sister to filmy ferns were consistent with the above results. Of the 3916 total comparisons, 1875 were significant (P < 0.05; colored boxes in Fig. 2), many at higher significance thresholds (P < 0.01 or P < 0.001; darker colors in Fig. 2). Clearly, rates vary across the phylogeny. Significant differences were observed within the Trichomanes and Hymenophyllum clades, as well as among species in the sister group to filmy ferns (Fig. 2). The most striking differences, however, were found in comparisons between the Hymenophyllum and Trichomanes clades (537 of 625 comparisons were significant, P < 0.05). These differences were consistently unidirectional; when a significant difference existed, the Hymenophyllum lineage evaluated was always slower than the Trichomanes lineage. This result indicates that considerable intergeneric rate differences exist. It does not, however, inherently reveal the phylogenetic extent of the rate differences (i.e., whether a rate discrepancy was maintained throughout a filmy fern clade or only present along the branch leading to it), nor does it reveal which of the two lineages contains the aberrant rate.
|
In theory, relative rate comparisons between species in the sister group to filmy ferns and species in the Trichomanes and Hymenophyllum clades could be used to identify the aberrant filmy fern lineage. This assumes that one filmy fern lineage would show a large number of significant differences in just one direction relative to the filmy fern sister group, but the other filmy fern lineage few. Our pairwise comparisons between the filmy fern and sister lineage taxa did result in a relatively large proportion of significant outcomes—876 of 1950 comparisons were significant. These differences, however, were not entirely unidirectional or restricted to a single filmy fern lineage. Of the 975 comparisons between species of Hymenophyllum and species from the sister group to filmy ferns, 537 were significant; all but one of the significant comparisons indicated that the net rate along the branch leading to the species of Hymenophyllum was slower. Of the 975 comparisons between species of Trichomanes and species from the sister group to filmy ferns, 339 were significant; most of these significant comparisons (261) indicated that the net rate along the branch leading to the Trichomanes species examined was faster, but others (78) indicated that this rate was slower. Although there are more significant comparisons relative to the Hymenophyllum clade, and the vast majority of these are in one direction, the lack of a uniform background rate (as evidenced by the results relative to filmy ferns and comparisons among the sister group taxa) makes it difficult to definitively identify the aberrant filmy fern group using pairwise comparisons alone.
Divergence Time Estimates and Reconstructions of Ancestral Rates of Evolution
In identifying the aberrant filmy fern lineage, ancestral rate reconstructions can be considerably more informative than the previous tests. These reconstructions more realistically assign a unique rate of molecular evolution to each branch in the phylogeny and incorporate time. Therefore, they can provide an absolute rate estimate for each internal branch, including the filmy fern stem branch (i.e., the branch immediately subtending the initial divergence between the Hymenophyllum and Trichomanes clades), that is indicative of the ancestral filmy fern rate. Such reconstructions can thus also supply a distribution of the rates outside of filmy ferns, as well as a rate distribution for each of the two filmy fern lineages for comparison. The simultaneous estimation of divergence times provides additional insight. However, it should be noted that, as with reconstructing phylogenetic relationships, reconstructing ancestral rates and divergence times are dependent on models, and the results are contingent on the use of an appropriate model with valid assumptions. Because filmy ferns themselves lack a solid fossil record to constrain our analyses, the reconstructions of rates and dates within this group are especially dependent on the penalized likelihood model of rate change. Although this model is biologically plausible, and without a doubt more realistic than simply assuming a molecular clock, the results below should be interpreted with caution.
The chronological results of our penalized likelihood analysis of the Bayesian consensus tree are presented as a chronogram (i.e., a tree in which internode lengths are proportional to time) in Figure 3. For all nodes resolved in the Bayesian consensus tree, age estimates from the analysis of the consensus phylogeny, as well as mean ages and standard deviations resulting from the 100 replicate analyses, are presented in Table 1. Because not all nodes resolved in the consensus phylogeny were present in each of the 100 randomly sampled trees, some mean ages and standard deviations are based on fewer than 100 samples. According to our analyses, the initial divergence among monilophyte lineages (node 007, Fig. 3) occurred in the Late Devonian (ca. 360 Ma). The whisk fern (WHI), ophioglossoid fern (OPH), marattioid fern (MAR), horsetail fern (HOR), and leptosporangiate fern (LEP) lineages were all present by the end of the Carboniferous (290 Ma). Within leptosporangiate ferns, we estimate the earliest divergences to have occurred in the Carboniferous and Permian. These divergences gave rise to the osmundaceous (OSM), filmy (FIL), gleichenioid (GLE), and schizaeoid (SCH) ferns, as well as to the core leptosporangiate lineage (COR, node 033, Fig. 3). A Late Triassic diversification gave rise to the three major lineages of core leptosporangiates—heterosporous ferns (HET), tree ferns (TRE), and polypod ferns (POL) (Fig. 3). Our estimates for the temporal origins of the major fern lineages are largely in accord with previous ideas (Collinson, 1996; Pryer et al., 2004; Rothwell, 1987, 1996; Skog, 2001; Soltis et al., 2002; Tidwell and Ash, 1994). The major diversification of polypod ferns is estimated to have occurred in the Cretaceous, also consistent with recent analyses (Schneider et al., 2004a).
|
Within filmy ferns, the initial divergence (node 058, Fig. 3), yielding the Hymenophyllum and Trichomanes clades, is estimated to have occurred near the Triassic-Jurassic boundary (206 Ma). The oldest of the scarce filmy fern fossils, which are not definitively assignable to one of the two extant clades, are from the epoch immediately preceding this boundary (Late Triassic) (Axsmith et al., 2001). Based on our analyses, the major diversification within the two extant filmy fern clades was not concurrent. The initial divergence within the Trichomanes clade (node 083) is estimated to have occurred in the Middle Jurassic, with major diversification in the Late Jurassic and throughout the Cretaceous. Diversification within the Hymenophyllum clade, however, appears to be a much more recent phenomenon. The initial divergence within this clade (node 059) is estimated to have occurred in the Early Cretaceous, with subsequent major divergences in the Late Cretaceous and Tertiary.
Reconstructions of ancestral rates of molecular evolution for the rbcL locus are presented graphically as a ratogram (i.e., a tree in which internode lengths are proportional to rate) in Figure 4. The rate estimates for all internodes, including means and standard deviations resulting from the 100 replicate analyses are presented in Table 1. As was previously suggested by the significance tests for the presence of rate differences and the pairwise relative rate comparisons, considerable differences in absolute rate are present across the resolved phylogeny, ranging from about 0.00005 to 0.00150 substitutions/site/Ma (Table 1). Some groups are consistent in their relatively slow rate of molecular evolution, including osmundaceous ferns (OSM) and tree ferns (TRE). Others are consistent in their relatively fast rate, including schizaeoid ferns (SCH) and polypod ferns (POL). Within filmy ferns (FIL), a clear difference in rate can be observed between the Trichomanes (TRI) and Hymenophyllum (HYM) clades, with the rate within Trichomanes estimated to be about twice that of Hymenophyllum (Fig. 4). Rate heterogeneity is of course present within each of these lineages, but is much less pronounced than the resolved intergeneric differences.
|
When rates within the Hymenophyllum and Trichomanes crowns (i.e., for each clade, the set of branches subsequent to the initial divergence) and along the Hymenophyllum and Trichomanes stems (i.e., for each clade, the branch immediately subtending the initial divergence) are compared to the rates outside of filmy ferns, and to the rate of the filmy fern stem (i.e., the branch immediately subtending the initial divergence between Hymenophyllum and Trichomanes), it appears as though the rates within Hymenophyllum are aberrant (see ratogram, right side of Fig. 4). The rates within Trichomanes, on the other hand, appear to be rather consistent with the rates outside of filmy ferns and those along the filmy fern stem. The similarities and differences among the various partitions become even more apparent when the pools of estimates corresponding to each of the partitions (resulting from analyses of the 100 sampled trees) are graphed (left side of Fig. 4). Rates outside of filmy ferns, while varied from slow to fast, form a distribution centered at about 0.00053 substitutions/site/Ma, with 50% of the estimates falling between about 0.00036 and 0.00075 substitutions/site/Ma (Fig. 4). The distributions of rates along the filmy fern stem, the Trichomanes stem, and within the Trichomanes crown, do not show a substantial deviation from the centrality displayed in the distribution of rates outside of filmy ferns (medians of 0.00057, 0.00055, and 0.00055 substitutions/site/Ma, respectively). The rates along the Hymenophyllum stem and within the Hymenophyllum crown, however, are considerably slower (medians of 0.00035 and 0.00023 substitutions/site/Ma, respectively). The Hymenophyllum clade (including its stem), with a substantially different rate of molecular evolution, clearly represents the aberrant filmy fern lineage, as suggested by the tests for rate differences and the pairwise relative rate comparisons.
Based on the results of this study, the overall path length differences observed between the two major filmy fern lineages are inferred to ultimately have resulted from a rate deceleration in the Hymenophyllum lineage. However, to fully appreciate individual branch length similarities and differences, a consideration of both time and rate is necessary. Following the initial divergence within filmy ferns and the origin of the two extant lineages, there was maintenance of the ancestral rate of evolution along the Trichomanes stem, but a deceleration in rate along the Hymenophyllum stem (Fig. 4). Yet, because a greater amount of time elapsed between the origin of the Hymenophyllum lineage (node 058) and the first divergence among extant Hymenophyllum species (node 059) than between the origin of the Trichomanes lineage (node 058) and the first divergence among extant Trichomanes species (node 083, Fig. 3), there is only a minor length difference between these two branches (Fig. 1). The discrepancy between the initial divergence times in Hymenophyllum and Trichomanes had the opposite effect on branch lengths within the two crowns. Less time for the accumulation of substitutions in the Hymenophyllum crown (relative to the Trichomanes crown, Fig. 3), combined with the slower rate of molecular evolution (Fig. 4), resulted in strikingly shorter branches in the crown of Hymenophyllum (Fig. 1).
Factors Influencing Evolutionary Rate
The slowed rate of rbcL sequence evolution in the Hymenophyllum clade may have resulted from intensified purifying selection, relaxed positive selection, or simply a genome-wide deceleration in the rate of nucleotide substitution (at either the plastid or organismal level). The rbcL gene is protein coding, and selection would act on nonsynonymous amino-acid replacement substitutions but not on synonymous silent substitutions. Therefore, if either intensified purifying selection or relaxed positive selection were responsible for the rate deceleration in the Hymenophyllum clade, we would expect to find a lower frequency of nonsynonymous substitutions in Hymenophyllum relative to Trichomanes and the filmy fern sister group, but no difference among these clades in the frequency of synonymous substitutions. This bias would result in a significantly smaller nonsynonymous to synonymous (dn/ds) ratio in Hymenophyllum, relative to Trichomanes and the filmy fern sister group.
To determine whether a significant difference in selection was present between the two filmy fern lineages, or between either of these lineages and the filmy fern sister group, we conducted a series of three tests. Specifically, we compared three two-ratio models in which two of the clades were constrained to have the same dn/ds ratio to a three-ratio model in which each clade was allowed a unique dn/ds ratio. The results of these tests revealed that dn/ds ratios were significantly different among the three clades (P < 0.001; comparisons 9 to 11, Table 2). However, the Hymenophyllum clade had the highest dn/ds ratio, contrary to what would be expected if selection were responsible for the deceleration in rate. Within Hymenophyllum, the deceleration in synonymous substitution rate appears to have been more substantial than the deceleration in nonsynonymous substitution rate. For this reason, a genome-wide deceleration of nucleotide substitution rate (at either the plastid or organismal level) seems more probable.
Several factors have been generally proposed to possibly influence the genome-wide rate of DNA substitution, including: (1) generation time or replication rate—more frequent replication will yield an increased number of mutations (Brunsfeld et al., 1994; Conti et al., 1993; Gaut et al., 1992, 1996, 1997; Kohne, 1970; Laird et al., 1969; Laroche and Bousquet, 1999; Laroche et al., 1997; Li et al., 1987; Wu and Li, 1985; but see Whittle and Johnston, 2003); (2) replication/repair efficiency—a less efficient system will allow more mutations to occur (Britten, 1986; Friedberg et al., 1995; Schön et al., 1998; Wu and Li, 1985); (3) exposure to mutagens—increased exposure to DNA-damaging mutagens will result in a higher mutation rate (Friedberg et al., 1995; Hebert et al., 2002; Lutzoni and Pagel, 1997; Martin and Palumbi, 1993); (4) population size—lineages with smaller effective population sizes will experience faster rates of evolution due to the amplified effects of genetic drift (Moran, 1996; Ohta, 1972, 1992); and (5) speciation rate—increased cladogenesis will result in more genetic change due to rapid evolution associated with speciation events (Barraclough and Savolainen, 2001; Bousquet et al., 1992; Mayr, 1954; Mindell et al., 1989; Webster et al., 2003). Although some of these factors involve the supply of DNA mutations, others center on the fixation of introduced mutations; many of the factors are intrinsic to evolutionary lineages, but a few are extrinsic. Overall, the hypothesized mechanisms are confounded and their contributions poorly understood.
Almost any of the proposed factors could ultimately provide an explanation for the evolutionary rate deceleration detected in Hymenophyllum. A slower replication rate, a more efficient replication/repair system, decreased exposure to mutagens, or larger population sizes in Hymenophyllum, relative to Trichomanes, are all plausible. However, with our current knowledge of filmy fern life history, it is impossible to even begin to discriminate among the possibilities. Filmy ferns are primarily tropical in distribution and have not been the focus of intensive ecological studies more common for plants of temperate regions. It is known that members of both the Trichomanes and Hymenophyllum clades possess long-lived gametophytes capable of vegetative reproduction (rare in ferns; usually only the sporophyte is long-lived and capable of vegetative reproduction); some populations in North America and Europe exist exclusively as gametophytes (Farrar, 1967; Rumsey et al., 1998). However, the relative contributions of the gametophyte and sporophyte stages in filmy fern life cycles have not been explored. Life history differences related to these relative contributions may certainly exist between the clades, but none have yet been described. With regard to potential mutagens, ultraviolet (UV) radiation, a leading candidate for mutagenesis (Lutzoni and Pagel 1997), cannot be invoked as being responsible for the rate discrepancy in filmy ferns. Because Hymenophyllum comprises species that generally inhabit more exposed niches, one would expect it to be more susceptible to UV mutagensis; yet, as we have shown here, it displays a deceleration in evolutionary rate. A speciation rate effect is also difficult to justify, as the two major extant filmy fern clades are approximately equally diverse (Pryer et al., 2001b). Virtually nothing is known about the replication/repair systems or even population size in filmy ferns.
|
Conclusions and Prospects |
|---|
|
TopAbstractMethodsResults and DiscussionConclusions and ProspectsReferences |
|---|
The cumulative rbcL branch length differences observed between the major lineages of filmy ferns are the result of significant differences in molecular evolutionary rate. Significance tests for the presence of rate differences and pairwise relative rate comparisons both revealed substantial disparity between the Trichomanes and Hymenophyllum clades, and suggested that the rate of evolution within Hymenophyllum is aberrant. The estimation of divergence times indicated that the Hymenophyllum clade diversified much more recently than did the Trichomanes clade, and the simultaneous reconstruction of ancestral rates of molecular evolution supported the notion that the rate of rbcL evolution slowed in the evolutionary history of Hymenophyllum. Thus, the extremely short branches in the Hymenophyllum crown are the result of a short duration combined with a slow evolutionary rate. Further analyses indicated that selection was not responsible for the decreased rate of rbcL evolution in Hymenophyllum, and instead implied that this observed slow-down might be due to a genome-wide deceleration in the rate of nucleotide substitution. If this deceleration is truly genome-wide, additional molecular markers, both in the chloroplast and in the other genomic compartments, should yield similar results to those of rbcL in this study. Although data are not currently available to properly address the universality of this hypothesis, additional studies to examine molecular evolution of the plastid, mitochondrial, and nuclear genomes in filmy ferns are underway. These, combined with detailed ecological, morphological, and ecophysiological studies to identify rate-influencing mechanisms, will provide even greater insight into the evolutionary history of this group of plants.
|
|
Acknowledgement |
|---|
We thank Thomas Buckley, David Hearn, Sabine Hennequin, Petra Korall, François Lutzoni, Jordan Metzgar, Spencer Muse, Nathalie Nagalingum, Roderic Page, Harald Schneider, Jeff Thorne, Marcy Uyenoyama, Stefan Zoller, and two anonymous reviewers for helpful comments, criticism, and advice at various stages in the completion of this project. Support for this study was provided in part by the U.S. National Science Foundation (grants DEB-0408077 to E.S. and DEB-0347840 to K.M.P.), the Society of Systematic Biologists (Graduate Student Research Award to E.S.), and the A. W. Mellon Foundation Fund to Duke University for Plant Systematics.
|
References |
|---|
|
TopAbstractMethodsResults and DiscussionConclusions and ProspectsReferences |
|---|
-
Adachi J., Cao Y., Hasegawa M. Tempo and mode of mitochondrial DNA evolution in vertebrates at the amino acid sequence level: Rapid evolution in warm-blooded vertebrates. J. Mol. Evol. (1993) 36:270–281.[CrossRef][Web of Science][Medline]
Axsmith B. J., Krings M., Taylor T. N. A filmy fern from the Upper Triassic of North Carolina (USA). Am. J. Bot. (2001) 88:1558–1567.
Barraclough T. G., Savolainen V. Evolutionary rates and species diversity in flowering plants. Evolution (2001) 55:677–683.[CrossRef][Web of Science][Medline]
Bleiweiss R. Slow rate of molecular evolution in high-elevation hummingbirds. Proc. Natl. Acad. Sci. USA (1998) 95:612–616.
Bousquet J., Strauss S. H., Doerksen A. H., Price R. A. Extensive variation in evolutionary rate of rbcL gene sequences among seed plants. Proc. Natl. Acad. Sci. USA (1992) 89:7844–7848.
Britten R. J. Rates of DNA sequence evolution differ between taxonomic groups. Science (1986) 231:1393–1398.
Bromham L. Molecular clocks in reptiles: Life history influences rate of molecular evolution. Mol. Biol. Evol. (2002) 19:302–309.
Bromham L., Penny D. The modern molecular clock. Nat. Rev. Genet. (2003) 4:216–224.[CrossRef][Web of Science][Medline]
Brunsfeld S. J., Soltis P. S., Soltis D. E., Gadek P. A., Quinn C. J., Strenge D. D., Ranker T. A. Phylogenetic relationships among the genera of Taxodiaceae and Cupressaceae: evidence from rbcL sequences. Syst. Bot. (1994) 19:253–262.[CrossRef]
Bulmer M., Wolfe K. H., Sharp P. M. Synonymous nucleotide substitution rates in mammalian genes: implications for the molecular clock and the relationship of mammalian orders. Proc. Natl. Acad. Sci. USA (1991) 88:5974–5978.
Cantatore P., Roberti M., Pesole G., Ludovico A., Milella F., Gadaleta M. N., Saccone C. Evolutionary analysis of cytochrome b sequences in some perciformes: Evidence for a slower rate of evolution than in mammals. J. Mol. Evol. (1994) 39:589–597.[CrossRef][Web of Science][Medline]
Castro L. R., Austin A. D., Dowton M. Contrasting rates of mitochondrial molecular evolution in parasitic Diptera and Hymenoptera. Mol. Biol. Evol. (2002) 19:1100–1113.
Chase M. W., Soltis D. E., Olmstead R. G., Morgan D., Les D. H., Mishler B. D., Duvall M. R., Price R. A., Hills H. G., Qiu Y.-L., Kron K. A., Rettig J. H., Conti E., Palmer J. D., Manhart J. R., Sytsma K. J., Michaels H. J., Kress W. J., Karol K. G., Clark W. D., Hedren M., Gaut B. S., Jansen R. K., Kim K.-J., Wimpee C. F., Smith J. F., Furnier G. R., Strauss S. H., Xiang Q.-Y., Plunkett G. M., Soltis P. S., Swensen S. M., Williams S. E., Gadek P. A., Quinn C. J., Eguiarte L. E., Golenberg E., Learn G. H. Jr., Graham S. W. S., Barrett C. H., Dayanandan S., Albert V. A. Phylogenetics of seed plants: An analysis of nucleotide sequences from the plastid gene rbcL. Ann. Mo. Bot. Gard. (1993) 80:528–580.[CrossRef]
Collinson M. E. "What use are fossil ferns?"—20 years on: With a review of the fossil history of extant pteridophyte families and genera. In: Pteridology in Perspective—Camus J. M., Gibby M., Johns R. J., eds. (1996) Kew, UK: Royal Botanic Gardens. 349–394.
Conti E., Fischbach A., Sytsma K. J. Tribal relationships in Onagraceae: Implications from rbcL sequence data. Ann. Mo. Bot. Gard. (1993) 80:672–685.[CrossRef]
Des Marais D. L., Smith A. R., Britton D. M., Pryer K. M. Phylogenetic relationships and evolution of extant horsetails, Equisetum, based on chloroplast DNA sequence data (rbcL and trnL-F). Int. J. Plant Sci. (2003) 164:737–751.[CrossRef][Web of Science]
Dubuisson J.-Y. Systématique du genre Trichomanes (Filicopsida, Hymenophyllaceae): Confrontation et combinaison des données anatomo-morphologiques, cytologiques et moléculaires (1996) France: Université Montpellier 2. Ph.D. dissertation.
Dubuisson J.-Y. rbcL sequences: A promising tool for the molecular systematics of the fern genus Trichomanes (Hymenophyllaceae)? Mol. Phylogenet. Evol. (1997) 8:128–138.[CrossRef][Web of Science][Medline]
Dubuisson J.-Y., Hennequin S., Douzery J. P. E., Cranfill R. B., Smith A. R., Pryer K. M. rbcL phylogeny of the fern genus Trichomanes (Hymenophyllaceae), with special reference to neotropical taxa. Int. J. Plant Sci. (2003a) 164:753–761.[CrossRef][Web of Science]
Dubuisson J.-Y., Hennequin S., Rakotondrainibe F., Schneider H. Ecological diversity and adaptive tendencies in the tropical fern Trichomanes L. (Hymenophyllaceae) with special reference to climbing and epiphytic habits. Bot. J. Linn. Soc. (2003b) 142:41–63.[CrossRef][Web of Science]
Ebihara A., Iwatsuki K., Kurita S., Ito M. Systematic position of Hymenophyllum rolandi-principis Rosenst. or a monotypic genus Rosenstockia Copel. (Hymenophyllaceae) endemic to New Caledonia. Acta Phytotaxon. Geobot (2002) 53:35–49.
Farrar D. R. Gametophytes of four tropical fern genera reproducing independently of their sporophytes in the southern appalachians. Science (1967) 155:1266–1267.
Felsenstein J. Evolutionary trees from DNA sequences: A maximum likelihood approach. J. Mol. Evol. (1981) 17:368–376.[CrossRef][Web of Science][Medline]
Friedberg E. C., Walker G. C., Siede W. DNA Repair and Mutagenesis (1995) Washington, DC: American Society Microbiology.
Gaut B. S., Clark L. G., Wendel J. F., Muse S. V. Comparisons of the molecular evolutionary process at rbcL and ndhF in the grass family (Poaceae). Mol. Biol. Evol. (1997) 14:769–777.[Abstract]
Gaut B. S., Morton B. R., McCaig B. C., Clegg M. T. Substitution rate comparisons between grasses and palms: synonymous rate differences at the nuclear gene Adh parallel rate differences at the plastid gene rbcL. Proc. Natl. Acad. Sci. USA (1996) 93:10274–10279.
Gaut B. S., Muse S. V., Clark W. D., Clegg M. T. Relative rates of nucleotide substitution at the rbcL locus of monocotyledonous plants. J. Mol. Evol. (1992) 35:292–303.[CrossRef][Web of Science][Medline]
Gaut B. S., Muse S. V., Clegg M. T. Relative rates of nucleotide substitution in the chloroplast genome. Mol. Phylogenet. Evol. (1993) 2:89–96.[CrossRef][Medline]
Hasebe M., Ito M., Kofuji R., Ueda K., Iwatsuki K. Phylogenetic relationships of ferns deduced from rbcL gene sequence. J. Mol. Evol. (1993) 37:476–482.[CrossRef][Web of Science][Medline]
Hasebe M., Kofuji R., Ito M., Kato M., Iwatsuki K., Ueda K. Phylogeny of gymnosperms inferred from rbcL gene sequences. Bot. Mag. Tokyo (1992) 105:673–679.[CrossRef]
Hasebe M., Omori T., Nakazawa M., Sano T., Kato M., Iwatsuki K. rbcL gene sequences provide evidence for the evolutionary lineages of leptosporangiate ferns. Proc. Natl. Acad. Sci. USA (1994) 91:5730–5734.
Hasebe M., Wolf P. G., Pryer K. M., Ueda K., Ito M., Sano R., Gastony G. J., Yokoyama J., Manhart J. R., Murakami N., Crane E. H., Haufler C. H., Hauk W. D. Fern phylogeny based on rbcL nucleotide sequences. Am. Fern J. (1995) 85:134–181.[CrossRef][Web of Science]
Hebert P. D. N., Remiglio E. A., Colbourne J. K., Taylor D. J., Wilson C. C. Accelerated molecular evolution in halophilic crustaceans. Evolution (2002) 56:909–926.[CrossRef][Web of Science][Medline]
Hennequin S., Ebihara A., Ito M., Iwatsuki K., Dubuisson J.-Y. Molecular systematics of the fern genus Hymenophyllum s.l. (Hymenophyllaceae) based on chloroplastic coding and noncoding regions. Mol. Phylogenet. Evol. (2003) 27:283–301.[CrossRef][Web of Science][Medline]
Hoegg S., Vences M., Brinkmann H., Meyer A. Phylogeny and comparative substitution rates of frogs inferred from sequences of three nuclear genes. Mol. Biol. Evol. (2004) 21:1188–1200.
Iwatsuki K. Hymenophyllaceae. In: The families and genera of vascular plants, Pteridophytes and Gymnosperms—Kramer K. U., Green P. S., eds. (1990) volume I. Berlin, Germany: Springer-Verlag. 157–163.
Kenrick P., Crane P. R. The origin and early diversification of land plants: A cladistic study (1997) Washington: Smithsonian Press.
Kohne D. E. Evolution of higher organism DNA. Q. Rev. Biophys. (1970) 33:327–375.
Kosakovsky Pond S. L., W S. D., Muse S. V. HyPhy: Hypothesis testing using phylogenies. Bioinformatics (2005) 21:676–679.
Krieger J., Fuerst P. A. Evidence for a slowed rate of molecular evolution in the order Acipenseriformes. Mol. Biol. Evol. (2002) 19:891–897.
Laird C. D., McConaughy B. L., McCarthy B. J. Rate of fixation of nucleotide substitutions in evolution. Nature (1969) 224:149–154.[CrossRef][Medline]
Langley C. H., Fitch W. M. An examination of the constancy of the rate of molecular evolution. J. Mol. Evol. (1974) 3:161–177.[CrossRef][Web of Science][Medline]
Laroche J., Bousquet J. Evolution of the mitochondrial rps3 intron in perennial and annual angiosperms and homology to nad5 intron 1. Mol. Biol. Evol. (1999) 16:441–452.[Abstract]
Laroche J., Li P., Maggia L., Bousquet J. Molecular evolution of angiosperm mitochondrial introns and exons. Proc. Natl Acad. Sci. USA (1997) 94:5722–5727.
Lewis L. A., Mishler B. D., Vilgalys R. Phylogenetic relationships of the liverworts (Hepaticae), a basal embryophyte lineage, inferred from nucleotide sequence data of the chloroplast gene rbcL. Mol. Phylogenet. Evol. (1997) 7:377–393.[CrossRef][Web of Science][Medline]
Li W.-H., Gouy M., Sharp P. M., O'hUigin C., Yang Y.-W. Molecular phylogeny of Rodentia, Lagomorpha, primates, Artiodactyla, and Carnivora and molecular clocks. Proc. Natl Acad. Sci. USA (1990) 87:6703–6707.
Li W.-H., Tanimura M., Sharp P. M. An evaluation of the molecular clock hypothesis using mammalian DNA sequences. J. Mol. Evol. (1987) 25:330–342.[CrossRef][Web of Science][Medline]
Lutzoni F., Pagel M. Accelerated evolution as a consequence of transitions to mutualism. Proc. Natl. Acad. Sci. USA (1997) 94:11422–11427.
Maddison D., Maddison W. MacClade 4: Analysis of phylogeny and character evolution (2000) Sunderland, Massachusetts: Sinauer Associates.
Manhart J. R. Phylogenetic analysis of green plant rbcL sequences. Mol. Phylogenet. Evol. (1994) 3:114–127.[CrossRef][Medline]
Martin A. P., P G. J., Palumbi S. R. Rates of mitochondrial DNA evolution in sharks are slow compared with mammals. Nature (1992) 357:153–155.[CrossRef][Medline]
Martin A. P., Palumbi S. R. Body size, metabolic rate, generation time, and the molecular clock. Proc. Natl. Acad. Sci. USA (1993) 90:4087–4091.
Masuyama S., Yatabe Y., Murakami N., Watano Y. Cryptic species in the fern Ceratopteris thalictroides (L.) Brongn. (Parkeriaceae). I. Molecular analyses and crossing tests. J. Plant Res. (2002) 115:87–97.[CrossRef][Web of Science][Medline]
Mayr E. Change of genetic environment and evolution. Evolution as a Process—Huxley J., Hardy A. C., Ford E. B., eds. (1954) London: George, Allen, and Unwin. 157–180.
Mindell D. P., Sites J. W., Graur D. Speciational evolution—a phylogenetic test with allozymes in Sceloporus (Reptilia). Cladistics (1989) 5:49–61.[CrossRef][Web of Science]
Mooers A. O., Harvey P. H. Metabolic rate, generation time, and the rate of molecular evolution in birds. Mol. Phylogenet. Evol. (1994) 3:344–350.[CrossRef][Medline]
Moran N. A. Accelerated evolution and Muller's ratchet in endosymbiotic bacteria. Proc. Natl. Acad. Sci. USA (1996) 93:2873–2878.
Moran N. A., von Dohlen C. D., Baumann P. Faster evolutionary rates in endosymbiotic bacteria than in cospeciating insect hosts. J. Mol. Evol. (1995) 41:727–731.[Web of Science]
Nickrent D. L., Parkinson C. L., Palmer J. D., Duff R. J. Multigene phylogeny of land plants with special reference to bryophytes and the earliest land plants. Mol. Biol. Evol. (2000) 17:1885–1895.
Nickrent D. L., Starr E. M. High rates of nucleotide substitution in nuclear small-subunit (18S) rDNA from holoparasitic flowering plants. J. Mol. Evol. (1994) 39:62–70.[Web of Science][Medline]
Ohta T. Population size and rate of evolution. J. Mol. Evol. (1972) 1:305–314.[CrossRef][Web of Science][Medline]
Ohta T. The nearly neutral theory of molecular evolution. Annu. Rev. Ecol. Syst. (1992) 23:263–286.[CrossRef][Web of Science]
Posada D., Crandall K. A. ModelTest: Testing the model of DNA substitution. Bioinformatics (1998) 14:817–818.
Price R. A. Systematics of the Gnetales: A review of morphological and molecular evidence. Int. J. Plant Sci. (1996) 157:S40–S49.[CrossRef][Web of Science]
Pryer K. M., Schneider H., Smith A. R., Cranfill R., Wolf P. G., Hunt J. S., Sipes S. D. Horsetails and ferns are a monophyletic group and the closest living relatives to seed plants. Nature (2001a) 409:618–622.[CrossRef][Medline]
Pryer K. M., Schuettpelz E., Wolf P. G., Schneider H., Smith A. R., Cranfill R. Phylogeny and evolution of ferns (monilophytes) with a focus on the early leptosporangiate divergences. Am. J. Bot. (2004) 91:1582–1598.
Pryer K. M., Smith A. R., Hunt J. S., Dubuisson J.-Y. rbcL data reveal two monophyletic groups of filmy ferns (Filicopsida: Hymenophyllaceae). Am. J. Bot. (2001b) 88:1118–1130.
Qiu Y.-L., Chase M. W., Les D. H., Parks C. R. Molecular phylogenetics of the Magnoliidae: cladistic analyses of nucleotide sequences of the plastid gene rbcL. Ann. Mo. Bot. Gard. (1993) 80:587–606.[CrossRef]
Rambaut A., Drummond A. (2005) Tracer version 1.3. Computer program distributed by the authors. Department of Zoology, University of Oxford, UK.
Renzaglia K. S., Duff R. J., Nickrent D. L., Garbary D. J. Vegetative and reproductive innovations of early land plants: implications for a unified phylogeny. Philos. Trans. R. Soc. Lond. B Biol. Sci. (2000) 355:769–793.
Ronquist F., Huelsenbeck J. P. MrBayes 3: Bayesian phylogenetic inference under mixed models. Bioinformatics (2003) 19:1572–1574.
Rothwell G. W. Complex Paleozoic Filicales in the evolutionary radiation of ferns. Am. J. Bot. (1987) 74:458–461.[CrossRef][Web of Science]
Rothwell G. W. Pteridophytic evolution: An often underappreciated phytological success story. Rev. Palaeobot. Palynol. (1996) 90:209–222.[CrossRef][Web of Science]
Rumsey F. J., Jermy A. C., Sheffield E. The independent gametophytic stage of Trichomanes speciosum Willd. (Hymenophyllaceae), the Killarney Fern and its distribution in the British Isles. Watsonia (1998) 22:1–19.
Sanderson M. J. Estimating absolute rates of molecular evolution and divergence times: A penalized likelihood approach. Mol. Biol. Evol. (2002) 19:101–109.
Sanderson M. J. r8s: inferring absolute rates of molecular evolution and divergence times in the absence of a molecular clock. Bioinformatics (2003) 19:301–302.
Schneider H., Schuettpelz E. K., Pryer M., Cranfill R., Magallón S., Lupia R. Ferns diversified in the shadow of angiosperms. Nature (2004a) 428:553–557.[CrossRef][Medline]
Schneider H., Smith A. R., Cranfill R., Hildebrand T. J., Haufler C. H., Ranker T. A. Unraveling the phylogeny of polygrammoid ferns (Polypodiaceae and Grammitidaceae): Exploring aspects of the diversification of epiphytic plants. Mol. Phylogenet. Evol. (2004b) 31:1041–1063.[CrossRef][Web of Science][Medline]
Schön I., Butlin R. K., Griffiths H. I., Martens K. Slow molecular evolution in an ancient asexual ostracod. Proc. R. Soc. Lond. B Biol. Sci. (1998) 265:235–242.[CrossRef]
Schön I., Martens K., Van Doninck K., Butlin R. K. Evolution in the slow lane: Molecular rates of evolution in sexual and asexual ostracods (Crustacea: Ostracoda). Biol. J. Linn. Soc. (2003) 79:93–100.[CrossRef][Web of Science]
Skog J. E. Biogeography of Mesozoic leptosporangiate ferns related to extant ferns. Brittonia (2001) 53:236–269.[Web of Science]
Smith A. R., Tuomisto H., Pryer K. M., Hunt J. S., Wolf P. G. Metaxya lanosa, a second species in the genus and fern family Metaxyaceae. Syst. Bot. (2001) 26:480–486.
Soltis P. S., Soltis D. E., Savolainen V., Crane P. R., Barraclough T. G. Rate heterogeneity among lineages of tracheophytes: Integration of molecular and fossil data and evidence for molecular living fossils. Proc. Natl. Acad. Sci. USA (2002) 99:4430–4435.
Springer M. S., Kirsch J. A. W. Rates of single-copy DNA evolution in phalangeriform marsupials. Mol. Biol. Evol. (1989) 6:331–341.[Abstract]
Tidwell W. D., Ash S. R. A review of selected Triassic to Early Cretaceous ferns. J. Plant Res. (1994) 107:417–442.[CrossRef][Web of Science]
Webster A. J., Payne R. J. H., Pagel M. Molecular phylogenies link rates of evolution and speciation. Science (2003) 301:478.
Whittle C.-A., Johnston M. O. Broad-scale analysis contradicts the theory that generation time affects molecular evolutionary rates in plants. J. Mol. Evol. (2003) 56:223–233.[CrossRef][Web of Science][Medline]
Wikström N., Pryer K. M. Incongruence between primary sequence data and the distribution of a mitochondrial atpl group II intron amon, ferns and horsetails. Mol. Phylogenet. Evol. (2005) 36:484–493.[CrossRef][Web of Science][Medline]
Wolf P. G. Phylogenetic analyses of rbcL and nuclear ribosomal RNA gene sequences in Dennstaedtiaceae. Am. Fern J. (1995) 85:306–327.[CrossRef][Web of Science]
Wolf P. G., Sipes S. D., White M. R., Martines M. L., Pryer K. M., Smith A. R., Ueda K. Phylogenetic relationships of the enigmatic fern families Hymenophyllopsidaceae and Lophosoriaceae: Evidence from rbcL nucleotide sequences. Plant Syst. Evol. (1999) 219:263–270.[CrossRef]
Wolf P. G., Soltis P. S., Soltis D. E. Phylogenetic relationships of dennstaedtioid ferns: evidence from rbcL sequence variation. Mol. Phylogenet. Evol. (1994) 3:383–392.[CrossRef][Medline]
Wolfe K. H., Li W.-H., Sharp P. M. Rates of nucleotide substitution vary greatly among plant mitochondrial, chloroplast, and nuclear DNAs. Proc. Natl. Acad. Sci. USA (1987) 84:9054–9058.
Wu C.-I., Li W.-H. Evidence for higher rates of nucleotide substitution in rodents than in man. Proc. Natl. Acad. Sci. USA (1985) 82:1741–1745.
Yang Z. H. PAML: A program package for phylogenetic analysis by maximum likelihood. Comp. App. Biosci. (1997) 13:555–556.
Yoshinaga K., Kubota Y., Ishii T., Wada K. 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 Mol. Biol. (1992) 18:79–82.[CrossRef][Web of Science][Medline]
Zoller S., Lutzoni F. Slow algae, fast fungi: Exceptionally high nucleotide substitution rate differences between lichenized fungi Omphalina and their symbiotic green algae Coccomyxa. Mol. Phylogenet. Evol. (2003) 29:629–640.[CrossRef][Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||