1582

American Journal of Botany 91(10): 15821598. 2004.

PHYLOGENY AND EVOLUTION OF FERNS(MONILOPHYTES) WITH A FOCUS ON THE EARLY

LEPTOSPORANGIATE DIVERGENCES1

KATHLEEN M. PRYER,2,6 ERIC SCHUETTPELZ,2 PAUL G. WOLF,3HARALD SCHNEIDER,2,4 ALAN R. SMITH,5 AND RAYMOND CRANFILL5

2Department of Biology, Duke University, Durham, North Carolina 27708 USA; 3Department of Biology, Utah State University,Logan, Utah 84322 USA; 4Albrecht-von-Haller-Institut fur Pflanzenwissenschaften, Abteilung Systematische Botanik,

Georg-August-Universitat, Untere Karspule 2, 37073 Gottingen, Germany; and 5University Herbarium, University of California,Berkeley, California 94720 USA

The phylogenetic structure of ferns (5 monilophytes) is explored here, with a special focus on the early divergences amongleptosporangiate lineages. Despite considerable progress in our understanding of fern relationships, a rigorous and comprehensiveanalysis of the early leptosporangiate divergences was lacking. Therefore, a data set was designed here to include critical taxa thatwere not included in earlier studies. More than 5000 bp from the plastid (rbcL, atpB, rps4) and the nuclear (18S rDNA) genomeswere sequenced for 62 taxa. Phylogenetic analyses of these data (1) confirm that Osmundaceae are sister to the rest of the leptospor-angiates, (2) resolve a diverse set of ferns formerly thought to be a subsequent grade as possibly monophyletic (((Dipteridaceae,Matoniaceae), Gleicheniaceae), Hymenophyllaceae), and (3) place schizaeoid ferns as sister to a large clade of core leptosporangiatesthat includes heterosporous ferns, tree ferns, and polypods. Divergence time estimates for ferns are reported from penalized likelihoodanalyses of our molecular data, with constraints from a reassessment of the fossil record.

Key words: Bayesian inference; divergence time estimates; ferns; fossil record; molecular systematics; monilophytes; penalizedlikelihood; vascular plant evolution.

Among green plant lineages, none are as species-rich asvascular plants (tracheophytes). The various members of thislineage, such as clubmosses, ferns, horsetails, gymnosperms,and angiosperms, are easy to recognize as monophyletic be-cause of their shared derived features. Vascular plants are char-acterized by the presence of tracheids and sieve elements (vas-cular tissue) for structural support and long-distance move-ment of water and nutrients throughout the plant body, a high-ly structured and dominant (or co-dominant) sporophyte phase,and branched (polysporangiate) sporophytes. This combinationof features has allowed vascular plants to become the mostconspicuous organisms on the planet and the dominant pri-mary producers in terrestrial ecosystems.

Traditionally, the vascular plant tree of life has been viewedas consisting of several grades of taxa. More explicitly, vas-cular plant evolution was seen as a successive series of incre-mental increases in complexity, from simple bryophytic an-

1 Manuscript received 29 February 2004; revision accepted 22 June 2004.The authors are especially grateful to Mark Chase, Jeffrey Palmer, and

Douglas Soltis for their invitation to provide a contribution on ferns to thisspecial issue and for their patience and encouragement throughout manuscriptpreparation. We thank Frank Axelrod, Julie Barcelona, David Barrington, Da-vid Conant, Jean-Yves Dubuisson, John Game, Jeffrey Hill, Barbara J. Hosh-izaki, Masahiro Kato, David Lorence, Tarek Milleron, Robbin Moran, TomRanker, Patricia Sanchez-Baracaldo, Laurens H. Smith Jr., Hanna Tuomisto,Dennis Wall, and Henk van der Werff for sending us plant material; JohannesVogel for sharing a previously unpublished rps4 sequence for Asplenium scol-opendrium; Jeffrey Hunt and Sedonia Sipes for assistance with DNA sequenc-ing; Mark Chase, Jeff Palmer, and two anonymous reviewers for useful com-ments on the manuscript; and curators and staff, especially Holly Forbes, atthe University of California Botanical Garden, and the University of GottingenBotanical Garden. This work was supported in part by NSF grants DEB-9615533 and DEB-0347840 to K. M. P., DEB-0408077 to K. M. P. and E. S.,DEB-0089909 to K. M. P. and H. S., DEB-9616260 to A. R. S., DEB-9707087to P. G. W., and DEB-0073036 to R. C.

6 E-mail: pryer@duke.edu.

cestors through vascularized spore producers, more complexseed plants, and ultimately to angiosperms. In the past 20years, however, we have witnessed unprecedented interest andinsight into the phylogenetic relationships among major groupsof living (and extinct) vascular plants, and the former pictureof vascular plant evolution, one of predominant paraphyly, ischanging as the tracheophyte tree of life begins to come intofocus (Crane, 1985a, b; Mishler and Churchill, 1985; Doyleand Donoghue, 1986a, b, 1992; Loconte and Stevenson, 1990,1991; Raubeson and Jansen, 1992; Chase et al., 1993; Garbaryet al., 1993; Doyle et al., 1994; Manhart, 1994; Mishler et al.,1994; Nixon et al., 1994; Rothwell and Serbet, 1994; Hasebeet al., 1995; Pryer et al., 1995, 2001a, 2004; Kranz and Huss,1996; Kenrick and Crane, 1997; Doyle, 1998; Wolf et al.,1998; Mathews and Donoghue, 1999, 2000; Qiu et al., 1999,2000; P. S. Soltis et al., 1999, 2004; Barkman et al., 2000;Chaw et al., 2000; Kenrick, 2000; Nickrent et al., 2000; Ren-zaglia et al., 2000; D. E. Soltis et al., 2000, 2002; Gensel andBerry, 2001; Pigg, 2001; Rydin et al., 2002; Burleigh andMathews, 2004; Soltis and Soltis, 2004). These advances to-ward a more complete understanding of vascular plant rela-tionships have not been tied to the use of any one type ofevidence, but instead have attempted to make sense of allthe available data, molecular and morphological (includingboth fossil and extant organisms).

It is now commonly agreed that a deep phylogenetic di-chotomy occurred in the early-mid Devonian (ca. 400 millionyears ago [mya]), separating a group that includes the modernlycophytes (less than 1% of extant vascular plants) from agroup that contains all other living vascular plant lineages, theeuphyllophytes (Fig. 1; Raubeson and Jansen, 1992; Kenrickand Crane, 1997; Doyle, 1998; Nickrent et al., 2000; Pryer etal., 2001a, 2004). The extant lycophytes all possess lycophylls(leaves with an intercalary meristem) and comprise three main

October 2004] 1583PRYER ET AL.PHYLOGENY AND EVOLUTION OF FERNS

Fig. 1. Consensus tree showing relationships among the major lineages of vascular plants (based on Pryer et al., 2001a). Key clades are indicated on thetree: T 5 tracheophytes, L 5 lycophytes, E 5 euphyllophytes, S 5 spermatophytes, M 5 monilophytes. A black band following a lineage name indicatesmembership of that lineage within a particular taxon listed across the top of the figure (taxa that are not monophyletic are in quotes).

clades: homosporous Lycopodiales (clubmosses), and hetero-sporous Isoetales (quillworts) and Selaginellales (spikemoss-es). Although living lycophytes are almost all relatively di-minutive plants, many fossil members of this lineage, includ-ing such well-known examples as Lepidodendron, were largearborescent forms that dominated the Carboniferous landscapeand that today are the major component of coal deposits(Stewart and Rothwell, 1993).

Euphyllophytesthe sister group to lycophytesare char-acterized by euphylls (leaves with marginal or apical meri-stems and an associated leaf gap in the vascular stele), lateralbranches that terminate in sporangia, and a distinctively lobedprimary xylem strand (Stein, 1993; Kenrick and Crane, 1997).Extant members also possess a 30-kilobase inversion in thelarge single-copy region of the plastid genome (Raubeson andJansen, 1992). Living euphyllophytes belong to two majorclades (Fig. 1): seed plants (spermatophytes) and monilophytes(Kenrick and Crane, 1997; Nickrent et al., 2000; Pryer et al.,2001). Spermatophytes are united by the presence of seeds(megasporangia surrounded by integument tissue), wood pro-duced through the activity of a secondary meristem (cambi-um), and axillary branching. Extant seed plants likely numberbetween 250 000 and 300 000 species (Thorne, 2002; but seeScotland and Wortley, 2003) distributed unequally among fivemajor clades: cycads, Ginkgo, conifers, gnetophytes, and an-giosperms. Several extinct fossil lineages also belong to thisgroup, including pteridosperms, Bennettitales, and glossopter-ids. Recent phylogenetic analyses, including those based on acombination of genes from all three genomes, have revealedconflicting signals regarding the relationships among extantspermatophyte lineages (Bowe et al., 2000; Chaw et al., 2000;Magallon and Sanderson, 2002; Rydin and Kallersjo, 2002;Rydin et al., 2002; D. E. Soltis et al., 2002). Other papers in

this special issue deal specifically with seed plant relationships(Burleigh and Mathews, 2004; Crane et al., 2004; Soltis andSoltis, 2004); the remainder of this paper will focus on thesister group to the spermatophytes, the monilophytes (Fig. 1).

The monilophytes (5 Infradivision Moniliformopses, sensuKenrick and Crane, 1997) share a distinctive vasculature, hav-ing protoxylem confined to lobes of the xylem strand (Stein,1993), therefore the Latin moniliformis appellation for neck-lace-like. The monophyly of this clade has been inferredfrom cladistic analyses of morphology including fossil taxa(Kenrick and Crane, 1997), studies of sperm ultrastructure(Renzaglia et al., 2000, 2001, 2002), and analyses of DNAsequence data (Nickrent et al., 2000; Renzaglia et al., 2000;Pryer et al., 2001a). Extant members, which number more than11 500 species, have a three-residue (nine-nucleotide) insertionin the plastid rps4 gene (Pryer et al., 2001a) and belong tofive major lineages: whisk ferns (Psilotales), ophioglossoidferns (Ophioglossales), horsetails (Equisetopsida), marattioidferns (Marattiales), and leptosporangiate ferns (Polypodiales)(Fig. 1). The Late Devonian to Early Carboniferous cladoxy-lopsids (e.g., Iridopteridales and Pseudosporochnales), whichhave characteristics of both ferns and horsetails, are almostcertainly among the stem groups (Fig. 2) of the monilophytes(Skog and Banks, 1973; Stein et al., 1984; Berry and Stein,2000; Hilton et al., 2003).

Monilophytes, like lycophytes, are all spore bearing andseed-free. Because of this, members of these two lineageswere traditionally lumped under various terms, such as pte-ridophytes or ferns and fern allies (Fig. 1). These termsserved the botanical community well while there was littleresolution or understanding of the relationships among thesetaxa. Now, however, we have considerable confidence in thebroad-scale phylogenetic relationships of vascular plants; we

1584 [Vol. 91AMERICAN JOURNAL OF BOTANY

Fig. 2. Crown group and total group age definitions, as utilized in thisstudy. Crown group age refers to the time of the deepest divergence amongthe extant taxa in the total group of interest. Total group age refers to thetime of divergence of the total group (crown group 1 stem group fossils)from its sister group.

therefore prefer to use the terms monilophytes (or quite sim-ply, a more inclusive, ferns) and lycophytes, which specifyclade membership, to the terms pteridophytes and fernsand fern allies that unite paraphyletic assemblages of plants(Fig. 1). Likewise, the term eusporangiate ferns circum-scribes a nonmonophyletic unit and should be avoided (al-though using the word eusporangiate in a descriptive sense,such as a eusporangiate lineage, may still be useful in de-scribing sporangial morphology and development). On the oth-er hand, leptosporangiate ferns is a long-standing and fa-miliar term that does define a monophyletic group and contin-ues to be informative.

Within ferns (monilophytes), the first dichotomy separatesa clade consisting of whisk ferns and ophioglossoid ferns froma clade comprising horsetails, marattioid ferns, and leptospor-angiate ferns (Fig. 1). Whisk ferns and ophioglossoids are bothrelatively small lineages (just over 100 species total in twofamilies and about four generamore if one recognizes thesegregates of Botrychium at generic rank; see Hauk et al.,2003) and both have a poor fossil record. The sister relation-ship between these two lineages has only been identified re-cently (Manhart, 1995; Pahnke et al., 1996; Wolf, 1997; Nick-rent et al., 2000; Pryer et al., 2001a; but see Rothwell, 1999),and unique synapomorphies are cryptic because of the extentof morphological simplification present in both families. Somewhisk ferns, in fact, have such strikingly simplified body plans(pseudodichotomously branched sporophytes, no roots, andhighly reduced leaves) that they were long thought to be re-lated to some of the earliest fossil lineages of vascular plants(Parenti, 1980; Bremer, 1985), such as Rhynia. Reduction ofthe root system, however, does appear to be a shared trait.Ophioglossoids have simple, unbranched roots that lack roothairs, and whisk ferns lack roots altogether (Schneider et al.,2002). In addition, both groups have axial (and subterranean)gametophytes, sporangia that are adaxially attached, and a eu-sporangiate pattern of sporangial development, although theseare not unique synapomorphies.

Relationships among the horsetail, marattioid, and leptospo-

rangiate fern lineages remain elusive (Pryer et al., 2001a) andare shown in Fig. 1 as a polytomy. Horsetails are an ancientgroup of plants with fossil relatives dating back to the LateDevonian. Members of this clade have whorled appendages, acharacteristic stele, highly reduced leaves, and sporangia borneon sporangiophores that are clustered into strobili in extantmembers (Kenrick and Crane, 1997). Fossil horsetails are di-verse and include some arborescent representatives, such asCalamites, with secondary xylem. These larger forms becameextinct in the Permian, but some herbaceous representativessurvived, including the Mesozoic Equisetites. There are 15species of living horsetails (Equisetum); all are relatively smallin stature compared to their woody ancestors and have aworldwide distribution, mostly in temperate regions. A recentmolecular phylogenetic analysis of horsetails (Des Marais etal., 2003) estimated that the Equisetum crown group (Fig. 2)diversified in the early Cenozoic, which is in agreement withdates estimated from fossils (Stewart and Rothwell, 1993). Thelong branch leading to the crown group (with no other livingtaxa to sample) is a complicating factor in determining theexact relationship of Equisetum to other monilophyte lineages(Wikstrom and Pryer, unpublished manuscript).

The marattioid ferns (Fig. 1) first appeared in the middleCarboniferous. In the Late Carboniferous and Permian severallarge marattioid representatives had evolved, including Psa-ronius, which reached heights of about 8 m. Many extantmembers are also treelike, but they do not possess secondarymeristematic tissues. They have distinctive polycyclic dictyo-steles and sporangia borne abaxially on the blades. This cladeis represented by more than 200 species in about four genera(including Marattia, Danaea, and Angiopteris) and is almostexclusively restricted to tropical regions (Hill and Camus,1986).

The most familiar of the monilophytes are the leptospor-angiate ferns (Fig. 1), a monophyletic group of more than11 000 species. These ferns are characterized by sporangia thatdevelop from a single cell and have mature sporangial wallsonly one cell thick; most possess a distinctive annulus thatserves to eject the spores (usually 64). Features of the sporan-giaincluding the shape and position of the annulus, thestructure and shape of sporangial groups (sori), and whetheror not a flap of tissue (indusium) protects the sorihave fig-ured prominently in the taxonomy and classification of theseferns. The earliest-known occurrence of leptosporangiate fernsis in the Early Carboniferous (Galtier and Scott, 1985; Galtierand Phillips, 1996); by the end of the Carboniferous six fam-ilies were present. In subsequent major filicalean radiations inthe Permian, Triassic, and Jurassic, several families (e.g., Os-mundaceae, Schizaeaceae, Matoniaceae, and Dipteridaceae)with extant representatives replaced these Carboniferous fam-ilies (Rothwell, 1987). The more-derived polypod ferns, whichcomprise more than 80% of living fern species, recently wereshown to have diversified in the Cretaceous (on the basis ofmolecular age estimates), suggesting an ecological opportu-nistic response to the diversification of angiosperms (Schnei-der et al., 2004b).

Increasingly robust phylogenetic hypotheses, broadly inclu-sive of ferns and utilizing data from single or multiple sources(e.g., morphology; plastid, nuclear, and/or mitochondrialgenes), have improved confidence in the composition of andthe relationships among many taxa historically treated at fa-milial and ordinal ranks (Hasebe et al., 1994, 1995; Manhart,1995; Pryer et al., 1995, 2001a; Kranz and Huss, 1996; Pahnke

October 2004] 1585PRYER ET AL.PHYLOGENY AND EVOLUTION OF FERNS

et al., 1996; Rothwell, 1996, 1999; Schneider, 1996; Stevensonand Loconte, 1996; Wolf, 1997; Wolf et al., 1998; Vangerowet al., 1999; Schneider et al., 2004b). In this paper, we buildon our current, best estimate of monilophyte relationships(Renzaglia et al., 2000; Pryer et al., 2001a; Schneider et al.,2004b), with our primary aim to focus attention on the basalnodes of the leptosporangiate fern tree, for which we providenew DNA data and phylogenetic analyses. It should be notedthat for many of the more derived leptosporangiates (poly-pods), the overall phylogenetic picture is still equivocal (Wolfet al., 1994; Wolf, 1995; Murakami et al., 1999; Cranfill, 2001;Smith and Cranfill, 2002; Ranker et al., 2004; Schneider etal., 2004a, c) and in need of further sampling and study. Nev-ertheless, providing a robust overall framework for ferns, es-pecially at the base of the tree, will ultimately enable us toanswer some long-standing systematic questions and work to-ward understanding the patterns of character evolution thatgave rise to the Cretaceous radiation and diversification ofpolypod ferns (Schneider et al., 2004b).

MATERIALS AND METHODSTaxonomic samplingSixty-two taxa were selected to represent all major

vascular plant lineages (see Appendix 1 in Supplemental Data accompanyingthe online version of this article). Our sampling included three lycophytes(outgroup), each representing a different lineage, and six seed plants, includ-ing at least one representative from each of the five major lineages. Withinmonilophytes, the primary focus of this study, sampling was more extensiveand included at least two taxa from each of the eusporangiate lineages. Withinleptosporangiate ferns, 44 taxa were chosen to represent the major lineages,and we focused our sampling toward the basal nodes, with only a few ex-emplars from the hyperdiverse polypods.

DNA isolation, amplification, and sequencingGenomic DNA was ex-tracted from fresh, silica-dried, or herbarium leaf material using a modifiedDoyle and Doyle (1987) CTAB (cetyltrimethylammonium bromide) procedure(Dubuisson, 1997; Pryer et al., 2001b) or a DNeasy kit (Qiagen, Valencia,California, USA). For each taxon, four genes (plastid rbcL, atpB, rps4; nuclearsmall-subunit ribosomal DNA [18S]) were amplified separately using thepolymerase chain reaction (PCR), following established protocols (see Pryeret al., 2001b). PCR products were cleaned using QIAquick columns (Qiagen)according to the manufacturers protocol. Sequencing reactions were carriedout for both strands of the purified PCR products using Dye Terminator CycleSequencing or Big Dye Terminator Cycle Sequencing reagents (Applied Bio-systems, Foster City, California, USA). For information on amplification andsequencing primers see Appendix 2 in Supplemental Data accompanying theonline version of this article. All sequencing reactions were processed usingeither ABI 377 or ABI 3700 automated sequencers (Applied Biosystems), andeach sequencing read was evaluated for possible contamination using theNCBI nucleotide-nucleotide BLAST (blastn) tool (Altschul et al., 1997). Morethan one-third of the sequence data (97 new sequences) used in the analysesdescribed in this article were generated specifically for this study; all otherdata were obtained from GenBank (see Appendix 1 in Supplemental Dataaccompanying the online version of this article).

Sequence alignmentSequence fragments obtained as chromatogramswere edited and assembled into contiguous alignments using Sequencher(Gene Codes, Ann Arbor, Michigan, USA). The resulting consensus sequencesfor each gene were aligned manually using MacClade version 4.05 (Maddisonand Maddison, 2000). The alignments for rbcL and atpB were straightforwardbecause no insertions or deletions (indels) were present. Indels, however, ex-isted in both the rps4 and 18S nrDNA alignments; translated amino acidsequences (rps4) and rRNA secondary structure (18S) were used as alignmentguides. Regions of ambiguous alignment (within rps4 and 18S nrDNA) wereexcluded from the subsequent analyses, as were portions of alignments at the59 and 39 ends that contained copious amounts of missing data.

Phylogenetic analysesTo assess the combinability of the four single-genedata sets, a procedure was invoked in which topological conflict among treesresulting from analyses of the individual data sets was examined. Each single-gene data set was analyzed independently with PAUP* version 4.0b10 (Swof-ford, 2002) using an equally weighted maximum parsimony bootstrap ap-proach to assess clade support (Felsenstein, 1985). For the plastid genes, thebootstrap analysis consisted of 1000 replicates, each with 10 random-addition-sequence replicates and tree bisection and reconnection (TBR) branch swap-ping. For 18S nrDNA, the bootstrap analysis consisted of 1000 replicates,each with one random-addition-sequence and TBR branch swapping, savinga maximum of 1000 trees per replicate (these modifications were necessaryhere to limit search time). The bootstrap consensus trees resulting from eachof the four analyses were compared visually for conflict (high support forincongruent relationships; see Appendix 3 in Supplemental Data accompa-nying the online version of this article). Using a significance threshold of70%, some topological conflict was detected with regard to the lycophyte andseed plant lineages. However, no significant conflict among the four geneswas detected within the study group (monilophytes), and for this reason, thesingle-gene molecular data sets were combined and analyzed in unison.

The combined data set was analyzed using a Bayesian Markov chain MonteCarlo (B/MCMC) approach, as implemented in MrBayes version 3.0b (Huel-senbeck and Ronquist, 2001). Each gene was assigned its own model of se-quence evolution (GTR 1 I 1 G for each), as determined using a hierarchicallikelihood ratio test in Modeltest (Posada and Crandall, 1998). Two indepen-dent B/MCMC analyses were conducted using these four models, flat priors,and four chains. Chains were run for 10 million generations, and trees weresampled every 1000 generations. Following completion, the sampled treesfrom each analysis were plotted against their likelihood to recognize the pointwhere the likelihoods converged on a maximum value. All trees prior to thisconvergence (500 trees; 500 000 generations, for each of the two analyses)were discarded as the burn-in phase. Because both analyses converged onthe same maximum, the post burn-in trees from each (19 000 total trees) werepooled, and a majority-rule consensus was calculated to obtain a topologywith average branch lengths, as well as posterior probabilities for all resolvednodes.

Phylogenetic analyses of the combined data set were also conducted usingmaximum likelihood and equally weighted maximum parsimony approaches,in PAUP* (Swofford, 2002). A heuristic search for the most likely tree wasconducted using a single model of sequence evolution (GTR 1 I 1 G; asidentified in Modeltest, using parameters as estimated in the program), 100random-addition-sequence replicates, and TBR branch swapping. A heuristicsearch for the most parsimonious tree was conducted using 1000 random-addition-sequence replicates and TBR branch swapping. Maximum likelihoodand maximum parsimony bootstrap analyses also were conducted. The max-imum parsimony bootstrap analysis consisted of 1000 replicates, each with10 random-addition-sequence replicates and TBR branch swapping. The max-imum likelihood bootstrap analysis utilized the same model of evolution se-lected for the original maximum likelihood search and consisted of 100 rep-licates, each with one random-addition-sequence and nearest neighbor inter-change (NNI) branch swapping (to limit the search time).

Divergence time estimationDivergence times were estimated using a pe-nalized likelihood approach (Sanderson, 2002) that does not require an as-sumption of rate constancy (i.e., a molecular clock). Instead, this methodcombines a parameter-rich model, allowing for a different rate of substitutionon every branch, with a roughness penalty that constrains rate fluctuationsfrom branch to branch. The relative contributions of these two componentsare controlled by a smoothing parameter that can be objectively selected usinga cross-validation procedure (Sanderson, 2002). We obtained divergence timeestimates for ferns through a penalized likelihood analysis (using the computerprogram r8s, version 1.60; Sanderson, 2003) of our Bayesian consensus tree,incorporating 21 fossil constraints from a reassessment of the fern fossil re-cord (see Table 1). In our analysis, the three lycophyte taxa were pruned fromthe tree, and the root of the resulting tree (i.e., the divergence of monilophytesfrom spermatophytes) was used as a calibration point based on the concurrentappearance of fossils belonging to each of these lineages in the Middle De-

1586 [Vol. 91AMERICAN JOURNAL OF BOTANY

vonian (see Table 1). The appropriate smoothing value was determined usingcross validation; we considered values from 0.1 to 10 000, and the value 10received the best (i.e., lowest) cross validation score. Using this smoothingvalue, the Bayesian consensus tree, and the 21 fossil constraints, we searchedfor the solution that optimized the penalized likelihood function (10 randomstarts, each with 10 random perturbations; truncated Newton algorithm).

To evaluate the effects of phylogenetic uncertainty (Pagel and Lutzoni,2002), due to both topological and branch length estimation error, divergencetimes were also estimated for each of 100 Bayesian trees randomly sampledfrom among the 16 952 trees that contained all nodes with significant support(posterior probability $95%). For each tree, we identified the appropriatesmoothing value through cross validation and used penalized likelihood toestimate divergence times (as before). For these analyses, the fossil constraintswere applied only to well-supported nodes (posterior probability $95%); thethree fossil constraints that were applied to poorly supported nodes (posteriorprobability ,95%) in the consensus analysis were, for the 100 replicate anal-yses, applied to the next deeper well-supported node. The 100 molecular ageestimates for each well-supported node were averaged and their standard de-viation calculated.

RESULTS

AlignmentsThe mean sequence length, alignment length,number of characters included after pruning regions of ambig-uous alignment, numbers of variable and parsimony informa-tive characters, and percentage missing data are given forrbcL, atpB, rps4, and 18S nrDNA in Table 2. For some taxa,rps4 and/or atpB had premature stop codons within the gene,which are likely to be corrected with RNA editing (Wolf etal., in press).

Phylogenetic analysesThe Bayesian analysis of the four-gene combined data set yielded a well-resolved and well-sup-ported topology (Fig. 3). The maximum likelihood (ML) anal-ysis of the combined data set resulted in a single most-likelytree (2lnL 5 58 123.47526; tree not shown); and likewise, themaximum parsimony (MP) analysis of the combined data setresulted in a single most parsimonious tree (11 975 steps; CI5 0.273; RI 5 0.561; tree not shown). The ML tree is iden-tical in topology to the Bayesian tree (shown in Fig. 3), withone exception within polypod ferns. The MP tree is similar tothese two trees, but with some minor topological differenceswithin the polypod and tree fern clades, a conflicting place-ment of Gnetum, and a different relationship among the ly-cophyte outgroup taxa. In the tree resulting from the Bayesiananalysis, 43 of the 52 nodes in the monilophyte study group(Fig. 3) receive significant support from all three measuresBayesian posterior probabilities (PP $ 95), and maximumlikelihood and maximum parsimony bootstrap percentages(BPML $ 70 and BPMP $ 70, respectively).

Seed plants and monilophytes are strongly supported sistergroups (PP 5 100; BPML 5 100; BPMP 5 100). Within mon-ilophytes, each of the eusporangiate lineages (whisk ferns,ophioglossoid ferns, horsetails, and marattioid ferns), as wellas the leptosporangiate ferns, are strongly supported (PP 5100; BPML 5 100; BPMP 5 100). Likewise, whisk ferns to-gether with ophioglossoid ferns are well supported (PP 5 100;BPML 5 99; BPMP 5 100) and sister to the remaining moni-lophytes. Horsetails are resolved as sister to the marattioidferns regardless of the optimization criterion used, but thisrelationship is weakly supported by all three measures (PP 582; BPML , 50; BPMP 5 76). These two eusporangiate lineagesare always together with leptosporangiate ferns in a well-sup-ported clade (PP 5 100; BPML 5 88; BPMP 5 87).

All analyses provide exceptionally robust support (PP 5100; BPML 5 100; BPMP 5 100) for the following majorgroups of leptosporangiate ferns: osmundaceous ferns, filmyferns, schizaeoid ferns, core leptosporangiates, heterosporousferns, and polypod ferns (Fig. 3). Two remaining major groupsof leptosporangiate ferns were consistently resolved as mono-phyletic, but with somewhat reduced overall support: tree ferns(PP 5 100; BPML 5 84; BPMP 5 85) and gleichenioid ferns(sensu Jarrett, 1980) (PP 5 88; BPML 5 86; BPMP 5 88).Within leptosporangiates, the osmundaceous ferns are sister tothe rest (PP 5 100; BPML 5 100; BPMP 5 100).

As sister to all other leptosporangiates (minus Osmunda-ceae), our analyses identified a clade consisting of filmy fernstogether with gleichenioid ferns, and the Bayesian analysisprovided strong support (PP 5 96) for this relationship. Al-though this clade also was resolved consistently with maxi-mum parsimony and maximum likelihood, these two methodsprovided only weak support (BPML 5 55; BPMP 5 57). Withinthe gleichenioid ferns, there is robust support (PP 5 100; BPML5 100; BPMP 5 100) for the family Gleicheniaceae (Dicran-opteris, Gleichenella, Diplopterygium, Gleichenia, and Stich-erus) to include the often-separated genus Stromatopteris(Stromatopteridaceae of Bierhorst, 1977; Wagner, 1977). Thereis also strong support for a sister relationship between Chei-ropleuria 1 Dipteris and Matonia 1 Phanerosorus (PP 5100; BPML 5 97; BPMP 5 95), with that clade in turn sister toGleicheniaceae (PP 5 88; BPML 5 86; BPMP 5 88).

The schizaeoid ferns are sister to a robustly supported cladethat we here refer to as the core leptosporangiates (PP 5100; BPML 5 100; BPMP 5 100), which includes the hetero-sporous ferns, tree ferns, and polypods. Each of these lineagesis clearly monophyletic, but the relationships among them areequivocal. Within heterosporous ferns, two clades are wellsupported (PP 5 100; BPML 5 100; BPMP 5 100); within treeferns, some clades are also well supported, but the relation-ships among these remain ambiguous. Within polypods is agrade of enigmatic and species-poor genera (Saccoloma, Lon-chitis, Sphenomeris) leading to a well-supported (PP 5 100;BPML 5 100; BPMP 5 97), hyperdiverse clade that containsover 80% of all living fern species.

Divergence time estimatesThe results of our penalizedlikelihood analysis of the Bayesian consensus tree are pre-sented as a chronogram plotted against the geologic time scalein Fig. 4. Age estimates for all nodes, as well as mean agesand standard deviations (resulting from the 100 replicate anal-yses) for all well-supported nodes (nodes with black symbolsin Fig. 4), are presented in Table 1. Our divergence time es-timates are generally older than those implied by the fossilrecord, but they show relatively small standard deviations (Ta-ble 1).

According to our analyses, the initial divergence amongmonilophyte lineages (node 07, Fig. 4) occurred in the LateDevonian (;364 mya). All four eusporangiate lineages, aswell as the leptosporangiate fern lineage, were present by theend of the Carboniferous. We estimate that the whisk andophioglossoid fern lineages diverged from one another in theLate Carboniferous (;306 mya, node 08), with their crowngroup divergences in the Late Cretaceous (;88 mya, node 09)and Middle Jurassic (;162 mya, node 10), respectively (Fig.4; Table 1). As indicated by the fossil record, horsetails andmarattioid ferns had diverged from one another by the end ofthe Devonian (;354 mya, node 12, Fig. 4); however, the

October 2004] 1587PRYER ET AL.PHYLOGENY AND EVOLUTION OF FERNS

crown group divergences within these groups appear to bemore recent phenomena. Extant horsetails are estimated tohave diversified in the Tertiary (;38 mya, node 13); extantmembers of the marattioid ferns (node 14) began to diversifyin the Middle Triassic (;237 mya, Fig. 4; Table 1).

Within leptosporangiate ferns, we estimate the earliest di-vergences to have occurred in the Carboniferous and Permian.These divergences gave rise to the osmundaceous, filmy, glei-chenioid, and schizaeoid ferns, as well as to the core leptospo-rangiates (Fig. 4). The initial divergence within the osmun-daceous ferns is estimated to have occurred by the end of theTriassic, and our analyses support an origin of the two majorfilmy fern lineages in the Jurassic (;163 mya, node 21, Fig.4). The earliest divergences within the gleichenioid ferns(nodes 22 and 23, Fig. 4), giving rise to the three extant glei-chenioid families (Gleicheniaceae, Dipteridaceae, and Maton-iaceae), occurred in the Permian (;263 mya) and Triassic (227mya), but we estimate divergences within each of these fam-ilies to be more recent (Cretaceous). The initial divergencewithin schizaeoid ferns (node 32, Fig. 4) is estimated to haveoccurred in the Triassic (;212 mya).

A Late Triassic diversification gave rise to the three majorlineages of core leptosporangiates (Fig. 4)heterosporousferns, tree ferns, and polypod ferns. The earliest divergenceswithin each of these lineages occurred in the Jurassic. Themost species-rich groups of polypod ferns, namely the eupo-lypods and pteridoids (nodes 55 and 57, respectively, Fig. 4),comprise more than 80% of extant fern species and are esti-mated to have diversified in the Cretaceous (consistent withthe findings of Schneider et al., 2004b).

DISCUSSION

Phylogenetic relationshipsThere has been a steady trendin recent years of increasing resolution at the base of the fernphylogeny. The results of our analyses of a four-gene (rbcL,atpB, rps4, 18S) combined data set provide the highest levelsof resolution and support to date across the backbone of theleptosporangiate tree (Fig. 3). This, combined with our exten-sive sampling within the early leptosporangiate divergences(including representatives from the majority of currently rec-ognized genera), allows us to draw several important conclu-sions.

MonilophytesOur sampling of the four major eusporan-giate lineages of monilophytes in this study was identical tothat of Pryer et al. (2001a). However, increased taxonomicsampling within the early leptosporangiate divergences, andthe addition of Bayesian analytical tools to the maximum par-simony and maximum likelihood methods used in Pryer et al.(2001a), allow for improved understanding of the relationshipsamong the eusporangiate lineages. In agreement with that ear-lier study, all measures used here provided the highest degreeof support for the monophyly of monilophytes, for each of thefour lineages of eusporangiate monilophytes (whisk ferns,ophioglossoid ferns, horsetails, and marattioid ferns), and forleptosporangiate ferns (Fig. 3). As was demonstrated previ-ously (Nickrent et al., 2000; Pryer et al., 2001a), whisk fernsare most closely related to ophioglossoids, which contradictspreviously accepted relationships (Wagner, 1977; Rothwell,1999), and together they form the sister group to the rest ofmonilophytes with robust support. Horsetails are resolved assister to marattioid ferns in both the current study and Pryer

et al. (2001a), as well as in Wikstrom and Pryer (unpublishedmanuscript), but usually with weak support. The precise re-lationship of horsetails does remain subject to further study.

Leptosporangiate fernsEstablishing a robust phylogenetichypothesis of monilophyte relationships, especially among thebasal leptosporangiate nodes, is critical to understanding char-acter evolution and early diversification in ferns. Recent stud-ies (Hasebe et al., 1995; Pryer et al., 1995; Schneider, 1996;Stevenson and Loconte, 1996) focused on leptosporangiateferns have provided a framework of higher-level relationships,replacing former intuitive estimates founded largely on con-cepts of overall similarity (see Smith, 1995 for a review). Inthose studies, however, only a few internal nodes received ro-bust support, and most nodes at the base of the fern topology(especially those along the backbone) were weakly supported,making it impossible to say with certainty how any of thesegroups are related to one another. Our current study improvesgreatly on this situation.

Osmundaceous fernsThese ferns were considered by near-ly all earlier workers to be an isolated and basal group, oreven a transitional form, between the eusporangiate andleptosporangiate condition because they possess a mixture ofsporangial features from both types (Smith, 1995). In all anal-yses that have included a reasonable sampling of leptospor-angiate ferns, the osmundaceous ferns have been shown to besister to the rest of the leptosporangiates (Fig. 3; Hasebe et al.,1994, 1995; Pryer et al., 1995; Wolf et al., 1998), althoughnot always with strong support. This phylogenetic position isconsistent with the fossil record for Osmundaceae (Fig. 4; Ta-ble 1) and is resolved here with clear support (PP 5 100; BPML5 100; BPMP 5 100).

Gleichenioid fernsThe phylogenetic affinities of the pe-culiar New Caledonian genus Stromatopteris have been hotlycontested (Bierhorst, 1977; Wagner, 1977). Most recently, thegenus was recognized as a member of Gleicheniaceae, but ina separate subfamily (Stromatopteridoideae) from the othergenera (subfamily Gleichenioideae: Dicranopteris, Gleiche-nella, Diplopterygium, Gleichenia, Sticherus) (Kramer andGreen in Kubitzki, 1990). Our sampling includes a represen-tative from each of these six genera, and our analyses clearlyestablish that Stromatopteris nests well within Gleicheniaceae,sister to Gleichenia. Several earlier studies, most importantlyJarrett (1980), suggested that Cheiropleuriaceae, Dipterida-ceae, and Matoniaceae might be allied to one another and toGleicheniaceae. Nonetheless, in previous single-gene studies(Hasebe et al., 1995; Pryer et al., 1995), representative taxafrom these groups were usually variously displayed as a gradeat the base of the leptosporangiate ferns. We show here (Fig.3) that Cheiropleuria 1 Dipteris and Matonia 1 Phaneroso-rus are strongly supported as sister taxa (PP 5 100; BPML 597; BPMP 5 95) and that there is substantial support for a sisterrelationship between this group and Gleicheniaceae (PP 5 88;BPML 5 86; BPMP 5 88). Thus, we see considerable merit inreferring to this entire clade as the gleichenioid ferns. Thisresolves the long-running controversy surrounding the rela-tionships of Matoniaceae, a family once thought to be taxo-nomically isolated from all other fern groups (Klavins et al.,2004), and Cheiropleuria and Dipterisonce incorrectly as-sumed to be more closely related to more-derived members ofPolypodiaceae (see Smith, 1995 for a review). Gleichenioid

1588 [Vol. 91AMERICAN JOURNAL OF BOTANYT

AB

LE1.

Mol

ecul

arage

est

imat

esan

dfo

ssil

age

co

nst

rain

tsfo

rv

asc

ula

rpl

ant

no

des

show

nin

Fig.

4.L

inea

gen

am

es,

both

cro

wn

grou

p(C

G)and/

orto

tal

grou

p(T

G),a

repr

ovid

edw

here

app

licab

le(se

eFi

g.2

for

term

inol

ogy)

.Am

ole

cula

rage

est

imat

e,de

rived

from

the

pena

lized

likel

ihoo

dan

aly

sis

of

the

conse

nsu

str

ee,

as

well

as

am

ean

mole

cula

rage

6st

anda

rdde

viat

ion

(SD)

,der

ived

from

the

pena

lized

likel

ihoo

dan

aly

sis

of

100

repl

icat

eB

ayes

ian

tree

s,are

als

opr

ovid

edfo

reach

node

,as

app

licab

le.F

ossi

lage

const

rain

tsw

ere

app

lied

tonode

su

sing

an

apo

mor

phy-

base

dapp

roac

h.A

foss

ilco

nst

rain

twas

ass

igne

dto

an

ode

on

lyif

the

app

licab

lefo

ssil

shar

eda

deriv

edfe

atur

ew

itha

linea

gede

scen

ded

from

that

node

(crow

ngr

oup

age

);if

we

were

un

abl

eto

rejec

tthe

poss

ibili

tyth

ata

repo

rted

cha

ract

erst

ate

was

asy

napo

mor

phy

for

the

tota

lgr

oup

(crow

ngr

oup

1st

emgr

oup)

,the

foss

ilage

const

rain

twas

ass

igne

dto

the

nex

tde

eper

no

de(to

talgr

oup

age

).m

ya5

mill

ion

year

sago

,NA

5fo

ssil

co

nst

rain

tnot

app

lied.

Nod

eLi

neag

enam

e(s)

Mol

ecul

arage

est

imat

e(m

ya)

Mea

nm

ole

cula

rage

6SD

(mya

)Fo

ssil

age

const

rain

tC

omm

ents

01E

uphy

lloph

ytes

(CG)

Sper

mato

phyt

es(T

G)M

onilo

phyt

es(T

G)

380.

0038

0.00

60.

0038

0.00

The

app

eara

nce

of

Ibyk

aan

dC

ross

iam

ark

sth

edi

verg

ence

of

the

two

exta

ntlin

eage

sof

eu

phyl

loph

ytes

inth

eM

iddl

eD

evon

ian

(Eife

lian)

.We

accept

Ibyk

ato

beth

eold

estm

on-

iloph

yte

foss

il(b

utn

ot

ass

igna

ble

toth

eho

rset

aill

inea

ge)a

nd

Cro

ssia

tobe

the

old

est

sper

mato

phyt

efo

ssil,

base

do

nth

epr

otox

ylem

posi

tion

inth

em

atu

rest

ele

(mes

arch

inm

on

iloph

ytes

,en

darc

hin

sper

mato

phyt

es(K

enric

kand

Cra

ne,1

997)

.02

Sper

mato

phyt

es(C

G)32

1.66

321.

566

2.92

NA

03A

ngio

sper

ms

(CG)

121.

0012

1.20

62.

0412

1.00

The

old

esta

ngi

ospe

rmcro

wn

grou

pfo

ssils

date

toth

eE

arly

Cre

tace

ous

(Vala

ngin

ian;

Bre

nner

,19

96;M

agal

lon

an

dSa

nder

son,

2001

),bu

tthe

seare

not

ass

igna

ble

toeith

erlin

-eage

sam

pled

here

.We

ther

efor

euse

dth

eold

estf

ossi

lsass

igna

ble

toC

hlor

anth

acea

e(E

arly

Cre

tace

ous:

Bar

rem

ian;

Friis

et

al.,

1994

,199

9)to

const

rain

this

node

.04

Gym

nosp

erm

s(C

G)C

onife

rs(T

G)31

0.00

310.

006

0.00

310.

00C

onife

rco

nes

an

dtw

igs

from

the

Car

boni

fero

us(P

enns

ylvan

ian;M

iller

,19

99)a

reth

eold

-est

un

equ

ivoc

alre

main

so

fan

exta

ntsp

erm

ato

phyt

elin

eage

and

indi

cate

the

time

of

di-

verg

ence

of

co

nife

rsfr

omo

ther

seed

plan

tlin

eage

s.05

27

1.98

272.

166

6.85

NA

06

292.

4428

4.95

617

.90

NA

07M

onilo

phyt

es(C

G)36

4.43

359.

976

1.82

NA

08W

hisk

fern

s(T

G)O

phio

glos

soid

fern

s(T

G)

305.

5729

2.62

613

.15

NA

Afo

ssil

reco

rdfo

rth

ew

hisk

fern

sha

snot

yet

been

docu

men

ted

(Ken

rick

and

Cra

ne,

1997

).T

heo

ldes

tfos

sila

ssig

nabl

eto

the

oph

iogl

osso

idfe

rns

isB

otry

chiu

mw

ight

onii

(Late

Pale

ocen

e:T

hane

tian;

Rot

hwel

land

Stoc

key,

1989

).09

Whi

skfe

rns

(CG)

88.4

781

.66

69.

84N

A

10O

phio

glos

soid

fern

s(C

G)16

1.96

153.

456

11.9

8N

A

11L

epto

spor

angi

ate

fern

s(T

G)35

9.56

354.

006

0.04

NA

We

reco

gniz

em

em

bers

of

Tede

leac

eae

and

Bot

ryop

terid

acea

e,da

ting

back

toth

eE

arly

Car

boni

fero

us(G

altier

an

dPh

illip

s,19

96),

as

am

ong

the

old

estl

epto

spor

angi

ate

fern

s.T

heir

rela

tions

hips

toan

yex

tant

linea

geare

uncle

ar.

The

foss

ilconst

rain

tfro

mnode

12(3

54m

ya)w

as

app

lied

ton

ode

11fo

rth

e10

0re

plic

ate

analy

ses

use

dto

calc

ulat

eth

em

ole

cula

rage

mean

an

dst

anda

rdde

viat

ion.

12H

orse

tails

(TG)

Mar

attio

idfe

rns

(TG)

354.

00

354.

00Ar

chae

ocal

amite

san

dre

lativ

esfr

omth

eL

ate

Dev

onia

n(F

amen

nian;

Stei

net

al.,

1984

;B

atem

an,1

991;

Ken

rick

an

dC

rane

,199

7)are

accept

edhe

reto

beth

eold

estu

nequ

ivoc

alm

em

bers

of

the

hors

etai

llin

eage

,bas

edon

stem

anato

mic

alcha

ract

ers.

Mar

attio

idfe

rns,

such

as

Psa

roni

usan

dSc

olec

opte

ris,

were

abu

ndan

tin

the

Lat

eC

arbo

nife

rous

(Pen

nsyl-

van

ian)

an

dPe

rmia

n(H

illan

dC

amus

,198

6;L

iuet

al.,

2000

);th

eref

ore,

this

linea

gew

as

likel

ypr

esen

tin

the

Ear

lyC

arbo

nife

rous

(Miss

issipp

ian).

13H

orse

tails

(CG)

37.5

137

.65

63.

87N

A

14M

arat

tioid

fern

s(C

G)23

6.61

230.

526

5.45

NA

15

206.

0020

6.00

60.

0020

6.00

Foss

ilsass

igna

ble

toM

arat

tia(i.

e.,be

arin

gsy

nang

iasi

mila

rto

thos

eof

exta

ntM

arat

tia)

date

back

toth

eL

ate

Tria

ssic

(Hill

and

Cam

us,1

986;

Liu

et

al.,

2000

).16

Lep

tosp

oran

giat

efe

rns

(CG)

Osm

unda

ceou

sfe

rns

(TG)

323.

1031

9.60

65.

0828

2.00

The

old

esto

smu

nda

ceou

sfe

rnfo

ssil,

ass

igna

ble

base

don

stel

eorg

aniz

atio

n,is

Gra

mm

atop

-te

ris

from

the

Ear

lyPe

rmia

n(A

sselia

n;Sk

og,2

001;

Roe

ssle

rand

Gal

tier,

2002

).

17O

smun

dace

ous

fern

s(C

G)20

6.00

206.

006

0.00

206.

00B

ased

on

stem

an

ato

mic

alcha

ract

ers,

Mill

er(1

971)

consi

dere

dO

smun

daca

ulis

from

the

Lat

eTr

iass

icto

bea

mem

ber

of

the

osm

unda

ceou

scro

wn

grou

p(T

idw

ella

nd

Ash

,199

4;C

ollin

son,

1996

).18

98

.57

81.2

56

22.9

7N

A

October 2004] 1589PRYER ET AL.PHYLOGENY AND EVOLUTION OF FERNST

AB

LE1.

Con

tinue

d.

Nod

eLi

neag

enam

e(s)

Mol

ecul

arage

est

imat

e(m

ya)

Mea

nm

ole

cula

rage

6SD

(mya

)Fo

ssil

age

const

rain

tC

omm

ents

19

286.

2428

4.54

66.

5826

9.00

The

obl

ique

an

nu

lus

an

dsp

ore

wall

ultr

astru

ctur

eof

Olig

ocar

pia

and

Szea

(Perm

ian:

Sak-

maria

n;W

an

get

al.,

1999

;Yao

and

Tayl

or,

1988

)ind

icat

eth

atth

ese

foss

ilsare

mem

bers

of

the

linea

geth

atin

clud

esfil

my

and

glei

chen

ioid

fern

s.T

hele

afm

orp

holo

gyof

thes

efo

ssils

does

no

tin

dica

tea

clo

sere

latio

nshi

pto

any

exta

ntm

em

ber

of

this

linea

ge;t

here

-fo

re,w

eco

nsi

der

them

tobe

stem

grou

pm

em

bers

.20

Gle

iche

nioi

dfe

rns

(TG)

Film

yfe

rns

(TG)

272.

8927

1.16

66.

06N

AT

heo

ldes

tgle

iche

nioi

dfe

rnfo

ssils

are

Mid

dle

Tria

ssic

inage

(Coll

inson

,199

6;Sk

og,

2001

)an

dbe

long

toth

edi

pter

idoi

d/m

aton

ioid

linea

ge(se

enode

23).

The

old

estfi

lmy

fern

foss

il,H

opet

edia

,is

from

the

Lat

eTr

iass

ic(C

arnia

n;A

xsm

ithet

al.,

2001

).21

Film

yfe

rns

(CG)

163.

2315

7.07

611

.26

NA

22G

leic

heni

oid

fern

s(C

G)G

leic

heni

acea

e(T

G)

263.

29

NA

Foss

ilsass

igna

ble

toG

leic

heni

acea

e(e.

g.,G

leic

heni

tes,

Gle

iche

noid

es,G

leic

heno

psis

;C

ol-

linso

n,19

96;S

kog,

2001

)are

know

nfr

omth

eJu

rass

icand

Cre

tace

ous,

butt

hese

cannot

beass

igne

du

nequ

ivoc

ally

toany

exta

ntlin

eage

and

are

likel

yst

emgr

oup

mem

bers

.Pe

rmia

nto

Tria

ssic

foss

ils(O

ligoc

arpi

a,Sz

ea)t

hata

reofte

ndi

scus

sed

as

Gle

iche

niac

eae

(Yao

an

dTa

ylor

,19

88;T

idw

ella

nd

Ash

,199

4;C

ollin

son,

1996

;Wang

et

al.,

1999

)may

infa

ctbe

stem

grou

pm

em

bers

of

Gle

iche

niac

eae

butc

ould

als

obe

stem

grou

pm

em

bers

of

Mat

onia

ceae

,Dip

terid

acea

e,or

film

yfe

rns

(seen

ode

19).

23M

aton

iace

ae(T

G)D

ipte

ridac

eae

(TG)

227.

0022

7.00

60.

0022

7.00

Dip

terid

oids

an

dm

ato

nioi

dsare

abu

ndan

tin

the

foss

ilre

cord

from

the

Lat

eTr

iass

icto

Ear

-

lyC

reta

ceou

s(S

kog,

2001

).T

heold

estf

ossi

ls,i

nclu

ding

Phl

ebop

teri

sand

Tom

anio

pter

is,

date

back

toth

eM

iddl

eTr

iass

icand

shar

ele

afand

sora

lcha

ract

ers

with

exta

ntge

nera

such

as

Dip

teri

san

dM

aton

ia(T

idw

ella

nd

Ash

,199

4;C

ollin

son,

1996

;Kla

vins

et

al.,

2004

).24

Mat

onia

ceae

(CG)

114.

7711

3.75

68.

92N

A

25D

ipte

ridac

eae

(CG)

70.8

968

.67

66.

62N

A

26G

leic

heni

acea

e(C

G)12

4.22

120.

616

6.15

NA

27

103.

5310

1.98

64.

60N

A

28

89.0

089

.04

60.

4589

.00

Gan

dolfo

et

al.

(199

7)de

mon

stra

ted

asi

ster

grou

pre

latio

nshi

pbe

twee

nth

efo

ssil

Boo

dlep

-te

ris

(Late

Cre

tace

ous:

Turo

nia

n)and

the

exta

ntge

nus

Stro

mat

opte

ris;

ther

efor

e,a

min

i-m

um

age

of

89m

yaw

as

ass

igne

dto

the

dive

rgen

ceof

Stro

mat

opte

ris

from

oth

erexta

ntta

xa.

29

99.6

395

.42

69.

07N

A

30

36.5

834

.11

65.

22N

A

31Sc

hiza

eoid

fern

s(T

G)26

6.25

265.

306

7.89

NA

The

old

ests

chi

zaeo

idfe

rnfo

ssils

are

from

the

Mid

dle

Jura

ssic

(seen

ode

32).

32Sc

hiza

eoid

fern

s(C

G)21

1.61

212.

506

12.6

016

9.00

Wik

stro

met

al.

(200

2)de

mon

stra

ted

asi

ster

grou

pre

latio

nshi

pbe

twee

nth

efo

ssil

Stac

hyp-

teri

s(M

iddle

Jura

ssic

:Bajo

cian)

and

the

exta

ntge

nus

Lygo

dium

;th

eref

ore,

am

inim

umage

of

169

mya

was

app

lied

toth

edi

verg

ence

of

Lygo

dium

from

oth

erexta

ntsc

hiza

eoid

gene

ra.T

here

latio

nshi

pso

fo

ther

Jura

ssic

schi

zaeo

idfe

rns,

such

as

Klu

kia

and

Klu

kiop

-si

s,are

un

know

n.33

13

5.29

138.

186

11.8

612

1.00

The

Ear

lyC

reta

ceou

s(N

eoco

mian

)gen

usR

uffor

diais

gene

rally

accept

edto

bea

rela

tive

of

the

ex

tant

genu

sAn

emia

(Wik

stro

met

al.,

2002

);th

eref

ore,

am

inim

umage

of

121

mya

was

ass

igne

dto

the

dive

rgen

ceof

Anem

iafr

omSc

hiza

ea.

34H

eter

ospo

rous

fern

s(T

G)22

0.02

219.

236

6.52

NA

The

old

estf

ossi

lass

igna

ble

toth

ehe

tero

spor

ous

fern

linea

geis

Cry

belo

spor

ites

berb

erio

i-de

sfr

omth

eL

ate

Jura

ssic

(Lup

iaet

al.,

2000

).35

Het

eros

poro

usfe

rns

(CG)

Mar

sile

acea

e(T

G)Sa

lvin

iace

ae(T

G)

173.

3217

1.27

67.

7613

7.00

Bas

edo

nle

afan

dst

emm

orp

holo

gy,

Yam

ada

and

Kat

o(2

002)

dem

onst

rate

dth

atth

efo

ssil

Reg

nelli

tes

na

gash

imae

(Earl

yC

reta

ceou

s:B

erria

sian

)is

ass

igna

ble

toM

arsi

leac

eae;

ther

efor

e,w

eaccept

this

foss

ilas

am

inim

umconst

rain

tfor

the

time

of

dive

rgen

cebe

-tw

een

Mar

sile

acea

ean

dSa

lvin

iace

ae.

36M

arsi

leac

eae

(CG)

93.0

289

.64

67.

41N

A

37Sa

lvin

iace

ae(C

G)89

.00

89.1

76

0.69

89.0

0M

egas

pore

so

fAz

olla

date

back

toth

eL

ate

Cre

tace

ous:

Turo

nia

n(C

ollins

on,1

991)

;we

use

thes

eto

mark

the

time

of

dive

rgen

cebe

twee

nAz

olla

and

Salv

inia

.

1590 [Vol. 91AMERICAN JOURNAL OF BOTANYT

AB

LE1.

Con

tinue

d.

Nod

eLi

neag

enam

e(s)

Mol

ecul

arage

est

imat

e(m

ya)

Mea

nm

ole

cula

rage

6SD

(mya

)Fo

ssil

age

const

rain

tC

omm

ents

38Tr

ee

fern

s(T

G)Po

lypo

dfe

rns

(TG)

210.

79

NA

We

accept

Cya

thoc

aulis

an

dre

late

dfo

ssils

(Midd

leJu

rass

ic;L

antz

et

al.,

1999

)as

the

old

-est

un

equ

ivoc

altr

eefe

rns

(seen

ode

42).

Ass

ignm

ents

of

Tria

ssic

and

Ear

lyJu

rass

icfe

rnfo

ssils

totr

eefe

rnge

nera

such

as

Dic

kson

iaare

am

bigu

ous

and

cannot

beaccept

edw

ith-

ou

tfu

rther

crit

ical

ev

alu

atio

n.T

heold

estu

nequ

ivoc

alpo

lypo

dfe

rnfo

ssils

are

from

the

Low

erC

reta

ceou

s(N

eoco

mian

;Che

net

al.,

1997

;Den

g,20

02)o

fnort

hern

Chi

na(se

en

ode

47).

39Tr

ee

fern

s(C

G)18

2.86

181.

786

6.58

NA

40L

oxom

atac

eae

(TG)

154.

8215

0.52

611

.03

112.

00B

ased

on

rhiz

ome

an

ato

my

an

dha

irst

ruct

ure,

Loxs

omop

teri

s(E

arly

Cre

tace

ous:

Apt

ian;

Skog

,200

1)is

co

nsi

dere

dto

bea

mem

ber

of

Lox

omat

acea

eand

isuse

dto

const

rain

itstim

eo

fdi

verg

ence

from

Pla

giog

yria

.

41L

oxom

atac

eae

(CG)

34.6

132

.52

65.

96N

A

42

159.

0015

9.89

63.

0715

9.00

We

accept

Cya

thoc

aulis

an

dre

late

dfo

ssils

(Midd

leJu

rass

ic;L

antz

et

al.,

1999

;Sko

g,20

01)a

sbe

long

ing

toth

eC

yath

ea1

Hym

enop

hyllo

psis

linea

ge.W

euse

thes

efo

ssils

toco

nst

rain

the

time

of

dive

rgen

ceof

Cya

thea

1H

ymen

ophy

llops

isfr

omoth

ertr

eefe

rnge

nera

.43

50

.23

41.8

06

11.7

9N

A

44

152.

73

NA

45

133.

5012

9.47

615

.54

NA

46

123.

85

NA

47Po

lypo

ds(C

G)15

9.54

162.

686

8.96

121.

00V

ario

uscro

wn

grou

ppo

lypo

dfo

ssils

(with

well-

pres

erved

poly

pod

spor

angi

a)are

know

nfr

omth

eL

ower

Cre

tace

ous

(Neo

comi

an;C

hen

et

al.,

1997

;Den

g,20

02)o

fN

orth

ern

Chi

naan

dare

use

dhe

reto

co

nst

rain

the

age

of

this

node

.48

15

3.48

N

A

49L

inds

aeoi

dfe

rns

(TG)

133.

1713

6.75

68.

7899

.00

Bas

edo

ncha

ract

eris

ticro

ot

an

ato

my,

afo

ssil

from

the

Ear

lyC

reta

ceou

s(A

lbian

;Sch

neid

eran

dK

enric

k,20

01)p

rovi

des

unequ

ivoc

alevid

ence

for

the

pres

ence

of

linds

aeoi

dfe

rns

(Sph

enom

eris

inFi

g.4)

at

this

time.

50D

enns

taed

tiace

ae(T

G)12

4.92

128.

796

9.85

NA

The

foss

ilco

nst

rain

tfro

mn

ode

54(9

3.5m

ya)w

as

app

lied

toth

isnode

for

the

100

repl

i-cate

an

aly

ses

use

dto

calc

ulat

eth

em

ole

cula

rage

mean

and

stan

dard

devi

atio

n.51

Den

nsta

edtia

ceae

(CG)

102.

2398

.59

68.

33N

A

52

92.0

3

NA

53

47.1

145

.02

66.

29N

A

54Pt

erid

acea

e(T

G)E

upol

ypod

s(T

G)11

9.85

93

.50

Lea

fsh

ape

an

dth

epr

esen

ceo

fps

eudo

indu

sia

allo

wfo

rth

eass

ignm

ento

fa

foss

ilP

teri

sfr

omth

eL

ate

Cre

tace

ous

(Cen

oman

ian;K

rass

ilov

and

Bac

chia

,200

0)to

the

pter

idoi

dfe

rns.

Bec

ause

this

foss

ilis

no

tre

adi

lyass

igna

ble

toany

cro

wn

grou

ppt

erid

oid

linea

ge,

we

use

ithe

reto

co

nst

rain

the

time

of

dive

rgen

cebe

twee

npt

erid

oid

and

eupo

lypo

dfe

rns.

55E

upol

ypod

s(C

G)77

.19

75.4

96

7.66

NA

The

foss

ilco

nst

rain

tfro

mn

ode

56(6

5m

ya)w

as

app

lied

toth

isnode

for

the

100

repl

icat

ean

aly

ses

use

dto

calc

ulat

eth

em

ole

cula

rage

mean

and

stan

dard

devi

atio

n.56

65

.00

65

.00

Lea

fim

pres

sion

ssh

owin

gke

yfe

atur

esof

both

Ono

clea

and

Woo

dwar

dia

(blec

hnoid

fern

s)are

know

nfr

omth

eL

ate

Cre

tace

ous

(Maa

strich

tian;

Rot

hwel

land

Stoc

key,

1991

;Up-

chu

rch

an

dM

ack,

1998

;Pig

gand

Rot

hwel

l,20

01)a

nd

are

use

dhe

reto

const

rain

the

time

of

dive

rgen

ceo

fbl

echn

oid

and

thel

ypte

ridoi

dfe

rns.

57Pt

erid

acea

e(C

G)76

.62

83.8

56

9.45

NA

58

69.8

877

.02

69.

1565

.00

Foss

ilsass

igna

ble

toAc

rost

ichu

m(th

esis

ter

genu

sto

Cer

atop

teri

s;Sc

hnei

der

et

al.,

2004

b)are

pres

enti

nth

eL

ate

Cre

tace

ous

(Maa

strich

tian;

Bon

deand

Kum

aran

,200

2)and

are

use

dto

co

nst

rain

the

time

of

dive

rgen

cebe

twee

nC

erat

opte

ris

and

Adia

ntum

.

October 2004] 1591PRYER ET AL.PHYLOGENY AND EVOLUTION OF FERNS

TABLE 2. Summary of DNA sequence data and alignments for each molecular region used in this study of monilophyte relationships.

Marker

Meansequencelengtha

Alignmentlength

Includedcharacters

Variablecharacters

Parsimonyinformativecharactersb

% Missingdatac

rbcLatpBrps418S

13201141

5911695

14021150

7681729

13201150

4741670

680635368371

558562312225

1.150.805.552.90

Total 4747 5049 4614 2054 1657 2.15a Completely missing sequences (one 18S; two rps4) were not included in this calculation.b Number of characters informative for equally weighted parsimony analyses.c Completely missing sequences (one 18S; two rps4) were included in this calculation.

ferns were diverse and abundant in the Mesozoic, and our newunderstanding of the relationships among extant members isimportant to consider when interpreting the fossil record(Skog, 2001; Klavins et al., 2004).

Gleichenioid/filmy fernsA sister relationship between glei-chenioid and filmy ferns was resolved here regardless of theoptimization criterion applied; however, the support for thisrelationship was strong only in the Bayesian analysis (Fig. 3;PP 5 96; BPML 5 55; BPMP 5 57). To the best of our knowl-edge, this sister relationship has never before been proposed,despite both gleichenioid and filmy ferns having sporangia thatpossess an annulus with an oblique to transverse aspect. Al-though the interpretation of the annulus aspect in Cheiropleu-ria and Dipteris (only slightly oblique) is somewhat ambigu-ous, it appears reasonable to suggest this as a morphologicalsynapomorphy uniting the gleichenioid 1 filmy fern clade. Atleast one interesting evolutionary and ecological implicationshould be considered if this sister relationship holds true.Many extant gleichenioids are scrambling or subscandent andpreferentially grow in open habitats and readily colonize sec-ondary habitats. Most filmy ferns are epiphytes, a trait thatmay have evolved in this lineage through a transitional stepinvolving scrambling or hemi-epiphytism (Dubuisson et al.,2003b). A scrambling habit as the putative plesiomorphic con-dition (as suggested by the sister group relationship of filmyand gleichenioid ferns) is consistent with this hypothesis.

Although it remains to be seen if our grouping of gleichen-ioid and filmy ferns will stand the test of time, it is worthnoting that what were once thought to be distantly related andsuccessive grades of taxa at the base of the leptosporangiateferns, may in fact be closely related members of a single cladethat diversified in the Late Paleozoic/Early Mesozoic (Fig. 4).The interpretation of the fossil record needs to be reevaluated,taking these results into account. Several Late Paleozoic/EarlyMesozoic fossils assumed to be related to Gleicheniaceae, suchas Szea and Oligocarpia, are more likely to be stem group(Fig. 2) members of the gleichenioid 1 filmy fern lineagebecause critical character states displayed by the fossils, suchas sporangial type and spore wall ultrastructure (Wang et al.,1998; Yao and Taylor, 1988), are plesiomorphic for this clade.

Schizaeoid fernsThe genera Lygodium, Schizaea, andAnemia (including Mohria) have almost always been consid-ered to represent a related group. Although this group of fernshas sometimes been considered to be among the more basalleptosporangiate nodes by some workers, Smith (1995) point-ed out that several researchers in the latter part of the lastcentury had suggested, largely on the basis of rather tenuouscharacters, that the schizaeoid ferns may have been a possible

point of origin for several higher leptosporangiate families,such as Pteridaceae, . . . and Marsileaceae. This was a ratherastute observation by these earlier workers, given that we haverelatively strong support (PP 5 100; BPML 5 83; BPMP 5 88)for stating that the schizaeoid ferns are the sister to the coreleptosporangiates (Fig. 3).

Core leptosporangiatesThis clade includes heterosporousferns, tree ferns, and polypods, each of which is clearly mono-phyletic (Fig. 3). The relationships among these three lineagesremain equivocal, and their circumscriptions, as presentedhere, differ considerably from traditional views. In the past,the two groups of heterosporous ferns (Marsilea/Pilularia andSalvinia/Azolla) were considered to be independently derived,but a sister group relationship has been confirmed here onceagain (Rothwell and Stockey, 1994; Hasebe et al., 1995; Pryeret al., 1995, 2001a).

The association of several genera, such as Plagiogyria, Hy-menophyllopsis, Loxoma, and Loxsomopsis, all lacking distincttrunks, among the tree ferns also conflicts with previously pro-posed relationships (Smith, 1995). These four genera had beenvariously aligned with families now seen to be only remotelyrelated: Hymenophyllopsis, Loxoma, and Loxsomopsis with thefilmy ferns (Hymenophyllaceae) or dennstaedtioid ferns (Den-nstaedtiaceae) and Plagiogyria with the osmundaceous ferns(Osmundaceae). Other tree fern genera often treated in mon-ogeneric families, e.g., Metaxya (Metaxyaceae) and Lophoso-ria (Lophosoriaceae), are shown here to be closely allied to,or embedded within, dicksonioid tree ferns (Fig. 3). To somefern systematists, one of the biggest surprises of all may bethe monophyly of the polypod ferns as circumscribed here(Fig. 3), which was not generally appreciated or accepted inearlier phylogenetic studies (Holttum, 1973; Mickel, 1974; Pi-chi Sermolli, 1977; see also Smith, 1995), despite the presenceof an obvious synapomorphythe vertical annulus of thesporanagium. Polypods comprise anywhere from 1530 fam-ilies (depending on the classification followed) and accountfor greater than 80% of extant species diversity.

Divergence time estimates and the fossil recordThe di-vergence time estimates resulting from our penalized likeli-hood analyses (Fig. 4; Table 1) are largely in accord withprevious ideas about the times of origin and diversification ofmajor fern clades (Rothwell, 1987; Tidwell and Ash, 1994;Collinson, 1996; Rothwell, 1996; Skog, 2001; P. S. Soltis etal., 2002). However, several clades with sparse fossil recordsare seen with this approach to have originated much earlierthan their meager fossil data would imply.

The earliest diverging of the 11 major lineages of monilo-phytes also have the oldest fossil records, with the exception

1592 [Vol. 91AMERICAN JOURNAL OF BOTANY

Fig. 3. Phylogenetic relationships at basal nodes of the monilophyte tree. The topology is the result of a Bayesian analysis of four genes (plastid rbcL,atpB, rps4, and nuclear 18S rDNA); average branch lengths are shown. Measures of support are given at the nodes: Bayesian posterior probability/maximumlikelihood bootstrap/maximum parsimony bootstrap; support values equal to 100 are abbreviated (1) and support values less than 50 are not reported (2).Moderately thickened lines indicate significant support from one or two measures (posterior probability $ 95; maximum likelihood bootstrap $ 70; or maximumparsimony bootstrap $ 70). Heavily thickened lines (most nodes) indicate significant support from all three measures.

October 2004] 1593PRYER ET AL.PHYLOGENY AND EVOLUTION OF FERNS

Fig. 4. Phylogenetic chronogram for monilophytes, plotted against the geologic time scale. The topology is the result of a Bayesian analysis of four genes(cf., Fig. 3). Molecular age estimates are indicated by node positions and were estimated by means of a penalized likelihood analysis of the Bayesian consensustree (with average branch lengths), incorporating 21 fossil constraints (see Table 1). Black symbols indicate nodes with significant support (posterior probability$ 95%), and white symbols indicate nodes without significant support; circles indicate nodes that were not constrained, whereas diamonds indicate that a fossilconstraint was applied at that node (position of the diamond corresponds to the age of the fossil constraint applied). All fossil ages were applied as minimumage constraints, except at node 01, which was fixed at 380 million years ago (mya). Node numbers correspond to those in Table 1, where lineage names,molecular age estimates, molecular age means, and standard deviations (for well-supported nodes), as well as fossil age constraints, are provided. Gray shadowingon branches highlights 11 major monilophyte lineages; the molecular age estimate for a lineage is indicated by the tapered point on the left side, whereas theoldest unambiguous fossil age for the lineage is indicated by the juncture line between dark gray and light gray (note that there is no fossil record for whiskferns, and the fossil record for horsetails coincides with the molecular age estimate). Pie charts show the relative contributions of the various lineages to totalextant monilophyte species diversity (.11 500 species); approximate species numbers (modified from Hassler and Swale, 2004) precede the lineage names.

1594 [Vol. 91AMERICAN JOURNAL OF BOTANY

of the whisk ferns and ophioglossoids (Fig. 4). A fossil recordfor whisk ferns is lacking, and what is available for ophiog-lossoids is only Cenozoic in age. We estimate that these twolineages diverged from one another in the Late Carboniferous(;306 mya, node 08, Fig. 4; Table 1), whereas the extantmembers diversified only in the mid to Late Mesozoic (;88mya and 162 mya, nodes 09 and 10, Fig. 4; Table 1). One ofour remaining challenges will be to identify the Carboniferousancestors of the whisk fern 1 ophioglossoid lineage. The re-maining eusporangiate monilophyte lineages, namely marat-tioid ferns and horsetails, have excellent fossil records extend-ing into the Carboniferous (Fig. 4; Table 1). Living horsetailsin the genus Equisetum appear to have diversified in the earlyCenozoic (;38 mya), which is in agreement with Des Maraiset al. (2003), whereas the extant lineages of the marattioidferns date back to the Triassic (;237 mya and 206 mya, nodes14 and 15, Fig. 4; Table 1).

The fossil record indicates that leptosporangiate ferns firstappeared in the earliest Carboniferous (Tournaisian) and soonafter diversified to give rise to roughly six independent line-ages, all of which have ambiguous relationships to extant lin-eages. Some of these Carboniferous leptosporangiate fernshave been put forward as putative stem groups (Fig. 2) ofextant lineages (Stewart and Rothwell, 1993; Galtier and Phil-lips, 1996; Rothwell, 1999); for example, Anachoropteridaceaefor osmundaceous ferns, Tedeleaceae for schizaeoid ferns, andSermyaceae for gleichenioid ferns (Collinson, 1996; but seeRothwell, 1999). Further studies are needed to elucidate rela-tionships among Carboniferous and extant leptosporangiateferns. Our divergence time estimates are consistent with thehypothesis of a major replacement of the Carboniferous lep-tosporangiate ferns with new (extant) lineages at the end ofthe Paleozoic (Rothwell, 1987; Stewart and Rothwell, 1993;Rothwell, 1996).

According to our estimates, the osmundaceous ferns, sistergroup to all other leptosporangiate ferns, arose in the LateCarboniferous (;323 mya, node 16, Fig. 4; Table 1). The old-est fossils unequivocally assignable to this lineage are fromthe Permian. The gleichenioid 1 filmy fern total lineage (node20) and the schizaeoid fern 1 core leptosporangiates total lin-eage (node 31) are estimated to have originated in the Permian(;273 mya and ;266 mya, respectively); the oldest fossilsassignable to extant families of these lineages are from theTriassic and Jurassic, respectively (Fig. 4, Table 1; Collinson,1996; Skog, 2001).

The major lineages within the core leptosporangiates(node 34, Fig. 4)heterosporous ferns, tree ferns, and polypodfernsare shown here to have diverged in the Triassic (;220mya and 211 mya), even though the oldest fossils unequivo-cally assignable to one of these lineages only date from theMiddle Jurassic. There are older fossils that have been as-signed to the tree fern lineage, especially Dicksoniaceae(Stewart and Rothwell, 1993; Skog, 2001), but these fossilsare more likely to be stem group (Fig. 2) members of the coreleptosporangiates. Therefore, these earlier assignments needreevaluation in the light of our improved understanding of theevolution of leptosporangiate ferns. The two living families ofheterosporous ferns are estimated here to have diverged in theMiddle Jurassic (;173 mya, node 35, Fig. 4), although theoldest fossil we can attribute to them is from the Early Cre-taceous (Table 1). The tree fern and polypod lineages (nodes39 and 47, respectively, Fig. 4, Table 1) also began to diversifyin the Jurassic (;183 mya and ;160 mya). The most species-

rich groups of living polypods, the pteridoid ferns and theeupolypods (nodes 57 and 55, respectively, Fig. 4, Table 1),are shown here to have diversified in the Late Cretaceous,which is consistent with the dates of divergence provided bySchneider et al. (2004b). This observation of a more recentfern diversification in the Late Mesozoic/Early Cenozoic, co-inciding with the radiation of angiosperms, is echoed in Glei-cheniaceae (node 26, Fig. 4, Table 1) and suggests that fernsmay have experienced successive replacement events, ratherthan an absolute decline, after the appearance of angiosperms(Schneider et al., 2004b). Further evidence is needed to ex-plore the implications of these findings, especially for thoselineages for which we have limited available phylogeneticdata.

CONCLUSIONS AND PROSPECTS

All generally recognized extant fern families and nearly allmonilophyte genera at the early-diverging nodes have nowbeen sampled in published molecular phylogenetic studies,with few exceptions. Although our study has an extensivesampling from the major basal fern nodes, the sampling is stillinadequate with respect to the diversity within these groups.Some of these lineages have received some attention in thisregard; e.g., Ophioglossaceae (Hauk et al., 2003), Equisetaceae(Des Marais et al., 2003), Osmundaceae (Yatabe et al., 1999),filmy ferns (Dubuisson, 1997; Pryer et al., 2001b; Dubuissonet al., 2003a; Hennequin et al., 2003), Cheiropleuriaceae (Katoet al., 2001), Matoniaceae (Kato and Setoguchi, 1998), andSchizaeaceae (Skog et al., 2002; Wikstrom et al., 2002). Manyother important groups, such as Gleicheniaceae and tree ferns,are critically in need of further study. The polypods, the mostdiverse fern lineage, are especially in need of additional atten-tion, although several ongoing studies promise much progressin the near future. Furthermore, better resolution of relation-ships among and within certain clades, is still possible withadditional molecular data, including data on genome structure(Pryer et al., 2002). Better estimates of divergence times willlikely come with new fossil findings, and improved interpre-tations of the fossil record will occur in light of additionalevidence from gene sequencing.

Our study is one of several that provide a foundation forfuture family and genus level taxonomic studies in ferns,which will, in turn, allow for improved biogeographical, eco-logical, and evolutionary interpretations. Reinterpretation ofmorphological data and developmental information can nowbe attempted on a sounder footing, especially as they pertainto deep branches in the vascular plant tree. Using the bestestimate of phylogeny available at the time (Pryer et al.,2001a), Schneider et al. (2002) were able to provide severalcritical insights into morphological evolution within ferns. Theincreased taxonomic sampling presented here will allow us toreconstruct the evolution of other critical characters (e.g., spo-rangium structure) and to determine their implications for thebiology and systematics of ferns. Another promising field willbe the reconstruction of the evolution of genome size in fernsand the correlation of size, if any, with the variation in chro-mosome number, or other features, exhibited by ferns. We cannow begin to ask what key events might have led to the manylarge, species-rich radiations in the long history of fern life onEarth.

October 2004] 1595PRYER ET AL.PHYLOGENY AND EVOLUTION OF FERNS

LITERATURE CITED

ALTSCHUL, S. F., T. L. MADDEN, A. A. SCHA FFER, J. ZHANG, Z. ZHANG, W.MILLER, AND D. J. LIPMAN. 1997. Gapped blast and psi-blast: a newgeneration of protein database search programs. Nucleic Acids Research25: 33893402.

AXSMITH, B. J., M. KRINGS, AND T. N. TAYLOR. 2001. A filmy fern fromthe Upper Triassic of North Carolina (USA). American Journal of Botany88: 15581567.

BARKMAN, T. J., G. CHENERY, J. R. MCNEAL, J. LYONS-WEILER, W. ELISENS,G. MOORE, A. D. WOLFE, AND C. W. DE PAMPHILIS. 2000. Independentand combined analyses of sequences from all three genomic compart-ments converge on the root of flowering plant phylogeny. Proceedingsof the National Academy of Sciences, USA 97: 1316613171.

BATEMAN, R. M. 1991. Palaeobiological and phylogenetic implications ofanatomically-preserved Archaeocalamites from the Dinantian of OxroadBay and Loch Humphrey Burn, southern Scotland. Palaeontographica B223: 159.

BERRY, C. M., AND W. E. STEIN. 2000. A new iridopteridalean from theDevonian of Venezuela. International Journal of Plant Sciences 161:807827.

BIERHORST, D. W. 1977. The systematic position of Psilotum and Tmesipteris.Brittonia 29: 313.

BONDE, S. D., AND K. P. N. KUMARAN. 2002. The oldest macrofossil recordof the mangrove fern Acrostichum L. from the Late Cretaceous DeccanIntertrappean beds of India. Cretaceous Research 23: 149152.

BOWE, L. M., G. COAT, AND C. W. DE PAMPHILIS. 2000. Phylogeny of seedplants based on all three genomic compartments: extant gymnospermsare monophyletic and Gnetales closest relatives are conifers. Proceed-ings of the National Academy of Sciences, USA 97: 40924097.

BREMER, K. 1985. Summary of green plant phylogeny and classification.Cladistics 1: 369385.