Sei sulla pagina 1di 16

American Journal of Botany 95(5): 626641. 2008.

PHYLOGENY OF MARATTIOID FERNS (MARATTIACEAE):


INFERRING A ROOT IN THE ABSENCE OF A CLOSELY
RELATED OUTGROUP1

Andrew G. Murdock2
Department of Integrative Biology, University of California, Berkeley, 1001 Valley Life Sciences Bldg.,
California 94720-2465 USA

Closely related outgroups are optimal for rooting phylogenetic trees; however, such ideal outgroups are not always available. A
phylogeny of the marattioid ferns (Marattiaceae), an ancient lineage with no close relatives, was reconstructed using nucleotide
sequences of multiple chloroplast regions (rps4 + rps4trnS spacer, trnStrnG spacer + trnG intron, rbcL, atpB), from 88 collec-
tions, selected to cover the broadest possible range of morphologies and geographic distributions within the extant taxa. Because
marattioid ferns are phylogenetically isolated from other lineages, and internal branches are relatively short, rooting was problem-
atic. Root placement was strongly affected by long-branch attraction under maximum parsimony and by model choice under maxi-
mum likelihood. A multifaceted approach to rooting was employed to isolate the sources of bias and produce a consensus root
position. In a statistical comparison of all possible root positions with three different outgroups, most root positions were not
significantly less optimal than the maximum likelihood root position, including the consensus root position. This phylogeny has
several important taxonomic implications for marattioid ferns: Marattia in the broad sense is paraphyletic; the Hawaiian endemic
Marattia douglasii is most closely related to tropical American taxa; and Angiopteris is monophyletic only if Archangiopteris and
Macroglossum are included.

Key words: Angiopteris; Christensenia; Danaea; long-branch attraction; Marattia; Marattiaceae; rooting.

Recent phylogenetic studies have greatly improved our for the early branching events in the Euphyllophytes (Nickrent
understanding of the evolutionary relationships of ferns. In et al., 2000; Pryer et al., 2001, 2004; Wikstrm and Pryer, 2005;
particular, relationships of lineages within the diverse leptospo- Qiu et al., 2006) and genomic structural characters have also
rangiate ferns have been significantly clarified (e.g., Pryer et al., proven inconclusive (Kelch et al., 2004; Wolf et al., 2005). Pre-
2001, 2004; Schuettpelz et al., 2006; Schuettpelz and Pryer, vious studies that have attempted to integrate fossils with mo-
2007). The leptosporangiate ferns, Marattiales (marattioids), lecular data in phylogenetic analyses have found that inclusion
Ophioglossales (ophioglossoids), Psilotales (whisk ferns), of fossils can alter tree topology, resolution, and support de-
Equisetales (horsetails), and Spermatophytes (seed plants) pending on the approach, in some cases causing basal relation-
(classification following Smith et al., 2006) are all well-sup- ships in Euphyllophytes to be more poorly or differently
ported clades in phylogenetic studies; however the interrela- resolved than in analyses including only extant taxa (Rothwell
tionships among these lineages, that together constitute the and Nixon, 2006; Schneider, 2007).
Euphyllophytes, remain unresolved and controversial (Roth- The lack of a clear phylogenetic consensus at the base of the
well and Nixon, 2006; Schuettpelz et al., 2006). DNA sequence Euphyllophyte tree can be attributed to the loss of historical
data from all compartments of the genome have generated an signal through time and through extinction-limited sampling,
array of conflicting, but generally poorly supported topologies methodological difficulties arising from long branches separat-
ing the extant crown groups, and relatively high levels of evo-
1 Manuscript received 1 January 2007; revision accepted 14 February 2008. lutionary rate heterogeneity (Soltis et al. 2002), problems
The author thanks B. Mishler, A. Smith, K. Will, T. Carlson, and others common to most deep, unresolved nodes of the tree of life
for manuscript comments and plant material: M. Frantz, J. Strother, S. Lin, (Mishler, 2000; Giribet, 2002; Burleigh and Mathews, 2004;
M. Lehnert, J. Game, the Mishler and Baldwin laboratories, D. Kelch, K. Embley and Martin, 2006). All the extant members of the early-
Pryer, E. Schuettpelz, H. Schneider, J. Metzgar, M. Windham, N. diverging Euphyllophyte lineages have been sampled suffi-
Nagalingum, P. Korall, A. Grusz, G. Theseira, Forest Research Institute
Malaysia, D. Palmer, P. Bily and the Nature Conservancy (Maui), M.
ciently in previous studies to be certain that the long-branch
Christenhusz, D. Lorence and N.T.B.G., S. Stroud and the Ascension Island problem cannot be mitigated by further sampling, with one pos-
Conservation Centre, M. A. H. Mohamed and University of Malaya, R.B.G. sible exception: Christensenia, a rare and morphologically
Kew, R.B.G. Edinburgh, D. Walker, B. Weigle, R. Whitehead, K. Roux and unique marattioid fern genus that was resolved as sister to all
S.A.N.B.I., Xishuangbana Botanic Garden, University of California other marattioid ferns in a morphological cladistic analysis (Hill
Botanical Garden, T. Ranker, S. Graham, H. Rai, T. Motley, and D. and Camus, 1986), has not been included in previous molecular
Barrington. This research was a portion of the authors doctoral dissertation analyses.
research, which was supported by NSF Doctoral Dissertation Improvement Of the extant early-branching lineages of the Euphyllophytes,
Grant (DEB-0608497), NSF ATOL Grant (DEB-0228729), University of the marattioid ferns (also marattialean or marattiaceous in
California Pacific Rim Foundation, Polynesia Education and Research
Laboratories Research Fellowship, and the University of California,
some earlier literature) are perhaps the least studied. Marattioid
Berkeley Department of Integrative Biology Summer Research and Hansen ferns are an ancient lineage of eusporangiate ferns with no close
Travel Grants. relatives whose extant taxa are essentially restricted to the trop-
2 E-mail: murdock@berkeley.edu ical regions of the world. Historically, marattioids were thought
to be closely related to ophioglossoids based on a small number
doi: 10.3732/ajb.2007308 of comparable morphological characters (Campbell, 1911), but
626
May 2008] MurdockPhylogeny of Marattiaceae 627

recent results have found marattioids to be phylogenetically ing sessile synangia, most neotropical species having sessile or
isolated from other fern lineages (Smith, 1995), with most slightly raised synangia with lower pinnae more divided, and
molecular analyses resolving Ophioglossales as sister to Psi- the neotropical M. laevis Sm. complex having distinctive
lotales and several suggesting a possible affinity between Mar- stalked synangia. Angiopteris is distinguished by having spo-
attiales and Equisetales (Pryer et al., 2001, 2004; Wikstrm rangia that are fused only at the base and by having monolete
and Pryer, 2005; Qiu et al., 2006). Despite the uncertainty spores; all other genera have trilete spores.
concerning the interrelationships of early-branching fern lin- While there are some taxonomic difficulties with nearly all
eages, the hypothesis that marattioid and leptosporangiate lineages, marattioids present an especially challenging combi-
ferns are closely related lineages (if not sister lineages) has not nation of problems for a taxonomist, analogous in many ways
been questioned. to tree ferns, and these problems have contributed to a chroni-
Marattioids have one of the most extensive fossil records cally unsettled species-level taxonomy. Many marattioids grow
of any modern fern lineage, with numerous fossils of extinct to be extremely large with some species having fronds over 6 m
lineages dating from the Carboniferous, scattered Mesozoic long. The large size of marattioids presents problems for tax-
fossils that more closely resemble modern genera, but are es- onomists due to the necessarily fragmentary nature of most col-
sentially lacking from the Tertiary fossil record (Millay, 1979; lections, some of which, including type specimens, consist of
Hill and Camus, 1986; Taylor and Taylor, 1993; Collinson, only a few small pinnules. Apart from size, separating plasticity
2001). Extant marattioids can be diagnosed morphologically on from heritable differences is perhaps the greatest challenge for
the basis of multiple characters, some of which are apparent in systematists working on marattioid ferns. Within populations
putative fossil relatives: large, thick-walled sporangia that are and even within individuals one can find striking differences in
fully or partially fused into synangia, with each sporangium morphology. Light, temperature, and frond age have a strong
producing large quantities of small spores; complex polycyclic effect on all aspects of frond morphology, and irregular frond
stem vasculature; leaves basally flanked by paired, starchy, vas- division is common in the family (Backer 1929; Asama, 1960;
cularized auricles (often called stipules); leaves possessing Holttum, 1978; A. Murdock, personal observation); herbarium
prominent fleshy swellings (pulvini) at nodes and/or bases of specimens commonly present only a small picture of the overall
segments; extensive mucilage canal systems throughout the tis- variability of any given plant. Some investigators have misin-
sues of the plant; multicellular root hairs; cyclocytic multi- terpreted hypervariability as taxonomically informative and
cyclic stomata (Campbell, 1911; Hill and Camus, 1986; Rolleri have named numerous poorly differentiated species, making
et al., 2003). this group nomenclaturally burdensome. Additionally, the size,
slow growth rate, and the difficulty of obtaining and propagat-
Taxonomy Modern marattioids have most commonly been ing live plants make common garden experiments and even
treated in a single family Marattiaceae Kaulf. (e.g., Rolleri moderately large living collections impractical.
et al., 2003; Smith et al., 2006), although some authors (e.g., Because of the taxonomic difficulties in marattioid ferns,
Pichi Sermolli, 1977) have further dissected this into smaller fami- species diversity estimates have varied widely among investi-
lies, Danaeaceae C. Agardh, Angiopteridaceae Fe ex J. Bommer, gators. In Angiopteris, Ching (1959) recognized 62 species and
and Christenseniaceae Ching (= Kaulfussiaceae Campb.) with an additional nine species of Archangiopteris just for China,
varying circumscriptions (Murdock et al., 2006b). Marattiaceae while Rolleri (2003) recognized only 10 species worldwide;
s.l. is typically treated as having four genera: Danaea Sm., others have only recognized one polymorphic species in An-
restricted to the neotropics; Christensenia Maxon, restricted to giopteris (e.g., Hooker and Baker, 1868; Backer, 1929). Nearly
Southeast Asia; Marattia Sw., with a pantropical distribution; all species of Angiopteris have at one time been synonymized
and Angiopteris Hoffm., native to Madagascar, Asia, and Oceania with A. evecta (type from Tahiti), and nearly all Old World
(introduced in Hawaii and the American tropics). Some authors Marattia species have at one time been synonymized with M.
have recognized up to four poorly defined genera segregated fraxinea (type from Runion). One result of this taxonomic un-
from Angiopteris s.l.: Archangiopteris Christ and Giesenh., certainty is that assessment of conservation status of marattioid
Macroglossum Copel., Protangiopteris Hayata, and Protomar- taxa is problematic. Many species have been construed as
attia Hayata, all with highly restricted ranges in Asia. Presl highly restricted endemics (Fu and Jin, 1992; Hsu et al., 2000;
(1845) subdivided Marattia into five genera (Marattia, Dis- Kingston et al., 2004), and some are threatened by deforesta-
costegia C. Presl, Gymnotheca C. Presl, Stibasia C. Presl, and tion, plant invasion, and predation by feral pigs (Warshauer,
Eupodium J. Sm.), a taxonomy also followed by De Vriese and 1979; Pickard, 1983; Brownsey and Smith-Dodsworth, 2000).
Harting (1853) in their monograph of the family; however, These factors make a strong argument in favor of using molecu-
more recent treatments (e.g., Copeland, 1947; Rolleri et al., lar data and the resulting phylogenetic inferences to make in-
2003) have treated these genera within a more broadly circum- formed taxonomic and conservation decisions for marattioid
scribed Marattia (referred to in this paper as Marattia s.l.). ferns.
While marattioids are united as a group by multiple distinc-
tive characters, they are at the same time morphologically di- Previous work A morphological analysis by Hill and
verse. Danaea is the only reproductively dimorphic genus in Camus (1986) was the first study to examine the evolutionary
the family and is also distinguished from other genera by long relationships of marattioid ferns in a cladistic framework and
synangia that are sunken into a thickened lamina of the fertile includes one of the most detailed discussions of morphological
fronds, sporangia that dehisce via apical pores, and large peltate characters in the group to date. In addition to extant taxa, Hill
scales. Christensenia is unique among marattioids in having and Camus (1986) included several well-characterized fossil
palmately lobed to pedately compound blades, reticulate vena- taxa; Liu et al. (2000) used a similar approach, but they were more
tion, circular synangia that appear more or less scattered on the concerned with interrelationships of fossil taxa. Hill and Camus
abaxial side of the frond. Marattia is the most widespread and (1986), using Ophioglossales as an outgroup, resolved Chris-
morphologically diverse genus, the paleotropical species hav- tensenia as sister to a clade containing all other marattioids and
628 American Journal of Botany [Vol. 95

resolved Angiopteris as monophyletic only if defined in a broad MorphologyThe previously published morphological analysis by Hill and
sense to include Archangiopteris, Macroglossum, and Proto- Camus (1986) provided the starting point for building a new morphological
matrix for marattioid ferns. Many characters used by Hill and Camus were at
marattia. Both Danaea and Marattia were resolved as mono- least partially nonindependent, and character states were questionably discrete
phyletic; however, relationships among Danaea, Marattia, and in some cases, so the data set was critically reviewed and extensively recoded.
Angiopteris s.l. remained unresolved. However, Hill and Many characters useful for taxonomic purposes were excluded due to a lack of
Camus (1986, p. 276) stated that Marattia, Macroglossum, An- independence or a nondiscrete nature. While there are several cladistic analyses
giopteris, and Archangiopteris were all found to be paraphyletic of fossils that included some marattioid ferns (e.g., Rothwell, 1999; Rothwell
in their analysis, despite the fact that their figures show only and Nixon, 2006), these contained few additional characters that were useful
for resolution within marattioid ferns. Additional characters have been added
Angiopteris to be paraphyletic, and then only if narrowly cir- from a variety of literature sources and from observation of specimens and liv-
cumscribed. Additionally, they implied that Marattia laevis ing material.
should be segregated as a more narrowly defined genus based
on cladistic principles and then discussed multiple alternatives Phylogenetic markersMultiple regions of the chloroplast genome, in-
to recognizing a paraphyletic Marattia; however, these conclu- cluding rps4 plus the rps4trnSGGA intergenic spacer (rps4trnS) (~900 bp),
sions and discussions are not consistent with their trees, where and the region from trnSGCU to trnGUUC (including the trnG intron) (trnSGG)
Marattia s.l. is shown as monophyletic. (~16002100 bp) were sequenced to provide resolution within the ingroup.
Apart from recent phylogenetic work on Danaea (Christen- Eighty-eight accessions were sequenced for rps4trnS and trnSGG for maximal
husz et al., 2008) and the complete plastid genome sequence of resolution within marattioid ferns. The chloroplast region rps4trnS was first
used in fern phylogenetics by Smith and Cranfill (2002) and has been widely
Angiopteris evecta (Roper et al., 2007), relatively few marat- used in other groups of green plants. Apart from Small et al. (2005), a study that
tioid ferns have been sequenced. Species that have been se- assessed the potential utility of numerous chloroplast noncoding regions for
quenced have mostly been part of large-scale phylogenetic fern phylogenetics, trnSGG has not been widely used in fern phylogenetic stud-
studies (e.g., Hasebe et al., 1993; Manhart, 1994, 1995; Pryer ies. It appears, however, to be a highly promising phylogenetic marker for
et al., 2001a, 2004; Wikstrm and Pryer, 2005; Qiu et al., 2006) lower-level phylogeny reconstruction in many groups of plants (Hamilton,
with minimal sampling of marattioids. There have also been 1999; Shaw et al., 2005, 2007; Szvnyi et al., 2006). In most nonflowering
plants, trnSGG includes multiple short coding regions (trnSGCU, psaM, ycf12,
several recent papers focused specifically on Chinese and Tai- and trnGUUC), which serve as positions for primer design and assist with se-
wanese species of Angiopteris or Archangiopteris (Hsu et al., quence alignment. Three other chloroplast regions with relatively slow se-
2000; Chiang et al., 2002; Li and Lu, 2007; S. Lin, Shenzhen quence evolution, atpB (~1270 bp), rbcL (~1370 bp), and rps4 (~600 bp) were
Fairy Lake Botanical Gardens, unpublished manuscript), which sequenced to provide resolution at deeper levels and assist with outgroup root-
have uncovered very little genetic differentiation between the ing. Because rbcL, atpB, and rps4 had limited variability in the ingroup, a
many named species in the region. No molecular analysis has smaller data set of 36 taxa (with 22 ingroup taxa selected to represent all major
clades, and 14 outgroup taxa) was assembled for outgroup rooting.
included the genus Christensenia, apart from Christenhusz
et al. (2008), where it was used as an outgroup to Danaea. In-
PrimersPrimers used in this study are shown in Table 1, with correspond-
clusion of Christensenia in nucleotide sequence data sets could ing primer locations for trnSGG shown in Fig. 1. Novel primers were designed
potentially improve the resolution of the placement of marat- when necessary after initial sequencing efforts using published primer se-
tioid ferns in relation to other early-diverging fern lineages. quences and comparison with sequences in GenBank. The forward PCR primer
For this study, a phylogeny of marattioid ferns was inferred for trnSGG from Shaw et al. (2005) is not optimal for pteridophytes because of
using nucleotide sequence data and morphology. This phylog- base mismatches, but the reverse primer is generally applicable. Nor are the
eny helps to clarify the poorly understood taxonomy of marat- updated primers in Shaw et al. (2007), designed for angiosperms, not for pteri-
dophytes because of base mismatches in both the forward and reverse primers
tioids and to interpret the morphological evolution and and the location of the forward primer over an insertion found in the trnS gene
biogeographic history of the group. The understanding of polar- in Equisetum. For trnSGG, forward and reverse PCR primers trnS-GCU-F1 and
ity of morphological evolution in marattioid ferns is crucial for 3-trnG-UUC, and internal sequencing primers 5-trnG2G and 5-trnG2S can be
understanding relationships of extant taxa to putative fossil used for most nonflowering plants; additional internal primers were designed
relatives, dating divergence times, and assessing morphological specifically for use in Marattiaceae and may work in related fern families; with
homologies with related lineages. Because confidence in root- minor modifications, these could be used in more distantly related ferns and
other nonflowering plant lineages. Locations of primer sites in trnSGG are
ing is essential for determining polarity and rooting can be dif- shown in Fig. 1. Internal primers for rbcL were designed specifically for this
ficult in groups that are phylogenetically isolated from other study and will likely have limited applicability in other lineages; other primers
lineages, rooting hypotheses and outgroup relationships were used for rbcL and atpB are variations on those used in other fern studies (e.g.,
explored using a multifaceted approach designed to isolate po- Hasebe et al., 1993; Wolf, 1997; Pryer et al., 2001a, b) but have been rede-
tential sources of bias and produce a consensus root position. signed for use in Marattiaceae and other early-diverging fern lineages.

SequencingExtraction of genomic DNA was performed using the Qiagen


MATERIALS AND METHODS DNeasy Plant Mini Kit (Valencia, California, USA) from fresh, silica-dried, or
recent herbarium specimen leaf material. In general, fresh or silica-dried mate-
SamplingSampling for this study represented the broad range of mor- rial was needed; amplification of PCR products from herbarium specimens was
phology and geographic distribution in marattioids, including multiple indi- often problematic. PCR was performed on 1 : 1 to 1 : 50 dilutions of total ge-
viduals from all currently recognized genera (Christensenia, Danaea, nomic extraction, using Bioneer PCR Premix Tubes (Bioneer, Alameda, Cali-
Angiopteris, Archangiopteris, Macroglossum, and Marattia) as well as indi- fornia, USA), Amersham PCR Beads, or AmpliTaq Gold (Applied Biosystems,
viduals from all previously recognized genera and subgenera within Marattia Foster City, California, USA) (specific reaction mixes available on request
s.l. (i.e., Discostegia C. Presl, Eupodium J. Sm., Gymnotheca C. Presl, Stibasia from the author). PCR was performed on an MJ Research (Waltham, Massa-
C. Presl, and subgenus Mesosorus Rosenst.). Generic names are used in the chusetts, USA) PTC-200 thermal cycler with protocols following standard PCR
sense of Copeland (1947). Species names used are based on regional floristic protocols for atpB and rbcL, the protocol of Smith and Cranfill (2002) for
treatments, taxonomic revisions, and the taxonomic opinions of the author rps4trnS, and the protocol of Shaw et al. (2005) for trnSGG. Touchdown PCR
based on examination of types and numerous other collections. Names shown (Don et al., 1991) was used to amplify several samples that were problematic
in quotes are those that are used in regional floras but are likely to change in under standard protocols. PCR product was visualized and quantified on aga-
upcoming revisions (e.g., M. Christenhusz, University of Turku, unpublished rose gels stained with ethidium bromide and subsequently purified using USB
manuscript, for Danaea). ExoSAP-IT or USB PCR Pre-Sequencing Kit (USB Corp., Cleveland, Ohio, USA)
May 2008] MurdockPhylogeny of Marattiaceae 629

Table 1. Primers used in this study. PCR amplification primers are in boldface; direction indicated by F (forward) or R (reverse). Main primers used for
sequencing trnSGG indicated by an asterisk (*). Primer location for trnSGG, as shown in Fig. 1, is in boldface followed in brackets by the coordinate
within trnSGG of the 5 primer end using the Angiopteris cp genome (Roper et al., 2007) as the standard.
Primer (Source) a Direction Sequence Region amplified [5 coordinate]

trnS-GCU-F1* () F GAG AGA GGG ATT CGA ACC CTC GG trnSGG 1 [5]
trnS-GCU-F2 () F GAG GGA TTC GAA CCC TCG GTA C trnSGG 1 [9]
psaM-F1 () F GCA CTT ATA AGG GCT AAG AT trnSGG 3 [399]
psaM-R1 () R ATC TTA GCC CTT ATA AGT GC trnSGG 2 [399]
psaM-F2 () F GAC TGT CTG AAA TTG ATG TC trnSGG 3 [424]
psaM-R2 () R GAC ATC AAT TTC AGA CAG TC trnSGG 2 [424]
ycf12-F1 () F CGT ATC AGG ACC ATT AGT AAT TGC trnSGG 5 [791]
ycf12-R1 () R GCA ATT ACT AAT GGT CCT GAT ACG trnSGG 4 [791]
5-trnG2G* (1) F GCG GGT ATA GTT TAG TGG TAA AA trnSGG 7 [1001]
5-trnG2S* (1) R TTT TAC CAC TAA ACT ATA CCC GC trnSGG 6 [1001]
3-trnG-R1 () R ACC CGC ATC ACT AGC TTG GAA GGC trnSGG 8 [1737]
3-trnG-UUC* (1) R GTA GCG GGA ATC GAA CCC GCA TC trnSGG 8 [1752]
rps5 (2) F ATG TCC CGT TAT CGA GGA CCT rps4trnS
trnS R (3) R GAT TCG AAC CCT CGG TA rps4trnS
rbcL-F1 () F ATG TCA CCA CAA ACG GAG ACB AAA RC rbcL
rbcL-F2-Int () F AAG CTG AGA CAG GCG AAG TAA AAG G rbcL
rbcL-R3-Int () R ATT TGC AGT AAA TCC KCC TGT CAG A rbcL
rbcL-R4 () R TCA CAA GCA GCA GCY ART TCA GGA CTC rbcL
atpB-F1 () F TTG ATA CGG GAG CTC CTC TWA GTG T atpB
atpB-F2 () F ATG GCT GAA TAT TTY MGA GAT GTT A atpB
atpB-R3 () R TTC CTG TAT RGA TCC CAT TTC TGT atpB
atpE 384R (4) R GAA TTC CAA ACT ATT CGA TTA GG atpB
aSources: , Novel primer designed for this study; 1, Shaw et al., 2005; 2, Nadot et al., 1994; 3, Smith and Cranfill, 2002; 4, Wolf, 1997.

prior to sequencing. Nucleotide sequencing was done on Applied Biosystems PAUP*. BI analyses were continued until apparent stationarity, diagnosed us-
96 capillary 3730xl DNA Analyzers or an Applied Biosystems 377 sequencer. ing the program Tracer version 1.3 (Rambaut and Drummond, 2003). The BI
analysis of all data sets was based on 3 000 000 generations, four parallel chains,
Sequence lignment and phylogenetic analysisContigs of sequences were sampled every 1000 generations, with a burn-in of 500 000 generations ex-
constructed and proofread using the program Sequencher (Gene Codes, Ann cluded from the final analysis. Tests for the presence of a molecular clock in the
Arbor, Michigan, USA). The rooting data sets were supplemented with se- ingroup were performed using a likelihood ratio test (Felsenstein, 1981) for all
quences from GenBank (see Appendix 1); outgroup taxa were mostly single genes separately and combined in PAUP*.
species, but in some cases were a composite of sequences from two related spe-
cies when necessary due to availability of sequences and plant material. Coding RootingOne hundred random outgroup sequences used in rooting trials
regions were confidently alignable by eye. Where indels occurred in coding were generated in MacClade using the average base composition of the ingroup
regions, genes were translated into amino acids using the program MacClade (which did not significantly differ from the outgroups). Each of the 100 random
version 4.08 (Maddison and Maddison, 2005) to accurately identify the triplets sequences was used individually as an outgroup to the rooting data set {3} and
involved in the indels. Noncoding regions were aligned first by the program analyzed under MP, and the first 20 sequences were analyzed under ML, to as-
ClustalW (http://www.ebi.ac.uk/clustalw/) and then fixed by eye where obvi- sess tendency of random outgroups to attach to particular ingroup branches. In
ous alignment errors occurred. Areas of uncertain alignment and complex both MP and ML random rooting trials, ingroup topology was constrained to
nested indels were excluded in final analyses. Microstructural characters were match that of Fig. 2 using a backbone constraint tree, and random sequence
coded as present/absent where discrete, following Graham et al. (2000). Identi- addition was used. ShimodairaHasegawa tests (Shimodaira and Hasegawa,
cal taxa (i.e., with a pairwise difference of 0 in informative characters), of 1999; Goldman et al., 2000) comparing the likelihood scores, and Wilcoxon
which there were several, particularly in Angiopteris, were merged prior to signed-rank tests (Templeton, 1983) comparing the parsimony tree length
analysis when necessary for longer analyses. scores of 41 constraint trees corresponding to all possible root attachment
Data from each gene region were analyzed separately and combined into points within the atpB+rbcL+rps4 data set were performed with three possible
three data sets: {1} rps4trnS+trnSGG; {2} rps4trnS+trnSGG with morphol- outgroups (Equisetaceae, Ophioglossaceae, Osmundaceae) following the gen-
ogy and coded indels; and {3} atpB+rbcL+rps4 (for rooting analysis)these eral procedures of Graham et al. (2002) and Buckley et al. (2001) in PAUP*
data sets will be referred to later by their respective numbers inside the braces. using RELL estimation. Constraint trees for all root positions analyzed were
Maximum parsimony (MP), maximum likelihood (ML), and Bayesian in- generated using MacClade. Lundberg root vectors were created by reconstruct-
ference using a Markov chain Monte Carlo approach (BI) were used to analyze ing the nodal states for a polytomy containing all outgroup taxa in MacClade
phylogenetic data in this study using PAUP* version 4.0b10 (Swofford, 2002) with equivocal reconstructions coded as unknown. Midpoint and Lundberg
and MrBayes version 3.1 (Huelsenbeck and Ronquist, 2001; Ronquist and rootings were calculated using PAUP*.
Huelsenbeck, 2003). The programs ModelTest version 3.7 (Posada and Crandall,
1998) and MrModelTest version 2.2 (Nylander, 2004) were used to select
models for ML and BI analyses, respectively, with model selection based on
the Akaike information criterion (AIC). For MP and ML analyses, all data sets
were analyzed using a heuristic search with 10 random sequence addition rep-
licates and tree-bisection-reconnection (TBR) branch swapping. Starting trees
for full ingroup MP analyses were obtained using 100 000 random addition se-
quence replicates with no swapping and MulTrees unselected; the best trees
found were then used as starting trees for complete TBR branch swapping with Fig. 1. Map of the chloroplast trnSGG region and location of primers
MulTrees selected. MP bootstrap analyses were run using heuristic searches used in this study. Primer locations and direction correspond with Table 1;
with 10 random addition sequence replicates or were limited to 10 million rear- black circles indicate PCR primers; white circles indicate internal sequenc-
rangements per replicate when necessary to complete an analysis. Decay analy- ing primers. Location of a gene above or below the line indicates relative
ses used commands generated in MacClade, which were then executed in direction of transcription.
630 American Journal of Botany [Vol. 95

Fig. 2. Phylogeny of Marattiaceae (ingroup only) based on maxiumum likelihood (ML) analysis of combined trnSGG and rps4trnS {1}. Support
values: Bayesian posterior probabilities / maxiumum parsimony (MP) bootstrap / decay index. + = 100; - = <50 (0 for decay index values). The branch
indicated by *, supported only by MP analysis, collapses in the ML tree but is preserved here for explanatory purposes. Genera recognized in a taxonomic
revision by Murdock (2008) indicated by branches with support values in ovals. Hypothesized root placement indicated by a dashed line (see Discussion,
Rooting).
May 2008] MurdockPhylogeny of Marattiaceae 631

RESULTS all regions used in this study proved to be much less informa-
tive than would be expected given the variation found in nu-
Phylogenetic relationships Sequences of trnSGG proved merous studies of much younger lineages using the same
to be the most informative within the 88 taxon ingroup data set, markers. Tests for the presence of a molecular clock in the in-
with 483 variable characters (367 parsimony informative) of an group were rejected by a likelihood ratio test for all genes sepa-
aligned length of 1992 characters; rps4trnS had 246 variable rately and combined. The relatively short length of branches, in
characters (213 parsimony informative) of an aligned length of combination with the distance from the nearest outgroups,
992 characters. Both regions contained microstructural charac- likely contribute to the difficulty in rooting the phylogeny (see
ters (indels and repeat motifs): 56 in trnSGG and 14 in rsp4 section Rooting).
trnS (with only one of these in the coding region of rps4). In the
36-taxon rooting data set {3} (22 ingroup taxa and 14 outgroup Morphology The MP analysis of the morphological data
taxa), rbcL had 566 variable characters (432 parsimony infor- set produced a topology that is incongruent with molecular
mative) of an aligned length of 1371 characters; atpB had 585 trees. While many elements agree with the molecular results
variable characters (426 parsimony informative) of an aligned (e.g., Marattia paraphyletic; Danaea, Christensenia, Eupo-
length of 1269 characters; rps4 had 331 variable characters dium, Ptisana, and Angiopteris monophyletic), the interrela-
(266 parsimony informative) of an aligned length of 523 tionships of these groups differed, albeit with very few
characters. characters supporting any branch. Character descriptions, mor-
Support for the monophyly of Marattiaceae was consistently phological matrix, and cladogram are included in Appendix S1
high in all analyses that included outgroups (100% Bayesian (see Supplemental Data with online version of this article).
posterior probability and MP bootstrap), regardless of root
placement within the family (described later). Support for inter- Rooting Rooting of the phylogeny was inconsistent be-
nal branches was variable: in general, clades of closely related tween data sets, model choice, and optimality criteria. Addi-
(and morphologically similar taxa) were highly supported on tionally, both rooting and ingroup topology were affected by
long branches with 100% Bayesian posterior probability, 100% outgroup choice and analytical method. Figure 4 summarizes
MP bootstrap, and very high decay index values 43; however, the rooting positions resolved by various analyses on the com-
support for relationships between and within these clades was bined atpB+rbcL+rps4 data set. As a convention, root posi-
often much lower (Fig. 2). Major clades that received high sup- tions in this paper are referred to by the clade subtended by the
port included Danaea, Christensenia, and Angiopteris s.l. (i.e., branch to which the root attaches, e.g., the branch subtending
including Archangiopteris and Macroglossum). Marattia s.l. Danaea and will be followed by the root position number in
was resolved as paraphyletic in all analyses, regardless of opti- Fig. 5 in brackets, e.g., [1].
mality criterion or inclusion of morphology. Marattia s.s., i.e.,
the clade that includes the type Marattia alata Sw., was re- Parsimony rootingWith the exception of atpB analyzed
solved as sister to Angiopteris s.l. (91% Bayesian posterior separately, MP consistently resolved the root on the branch
probability, 81% MP bootstrap, decay index = 6). Two groups subtending Danaea [1], the longest internal branch, in analyses
of taxa generally treated in Marattia s.l., the neotropical Marat- of data set {1}, data set {3}, rbcl alone, and rps4 alone. To test
tia laevis complex, which corresponds to Eupodium J. Sm., and for the possibility of long-branch attraction (LBA), the ingroup
a clade of the paleotropical taxa, were both strongly supported portion of data set {3} was rooted with 100 random outgroups
as monophyletic; these two clades (referred to as Marattia A with base composition equal to the average composition of the
and Marattia B, respectively) were resolved as sisters in all ingroup taxa. In all 100 trials, MP attached the root to the branch
molecular analyses; however, support for this relationship was subtending Danaea [1]. In 76% of the random outgroup rooting
lower (97% Bayesian posterior probability, 68% MP bootstrap, trials, Christensenia was resolved on the branch between Dan-
decay index = 6). aea and the rest of the taxa (i.e., sister to all other taxa), but in
While MP analyses of data set {1} (with or without inclusion 24% of the trials Christensenia was separated from Danaea and
of outgroup taxa) and trnSGG alone were congruent with the placed in the same position recovered by ingroup analyses (i.e.,
topology produced by all ML and BI analyses, other MP analy- in a clade with Marattia s.s. and Angiopteris). When Danaea
ses produced an alternate topology. In MP analyses of data set was excluded from the analysis, MP resolved the root on the
{2}, and rps4trnS alone, Christensenia was resolved as sister branch subtending Christensenia [24] in 97% of the trials. Lun-
to a clade of Marattia s.l. + Angiopteris s.l. Relaxation of parsi- dberg rooting (Lundberg, 1972) agrees with all other MP root-
mony by only a single step is all that is required to produce the ing results, attaching the root to the branch subtending Danaea
ML topology, indicating the low level of support for either [1], as do results of analyses in which third position nucleotides
placement of Christensenia under MP. Apart from the change were downweighted or excluded.
in position of Christensenia, all other relationships were con- Wilcoxon signed-rank tests (Templeton, 1983), comparing
gruent with the topology shown in Fig. 2, with bootstrap sup- the parsimony tree length scores of 41 constraint trees corre-
port values varying slightly but without showing any trend for sponding to all possible root attachment points within data set
increased or decreased support. {3}, were performed with three possible outgroups (Equiseta-
Relative to related lineages, even the longest branches within ceae, Ophioglossaceae, Osmundaceae) (Figs. 5, 6). In the three
Marattiaceae are short (Fig. 3), suggesting either recent radia- analyses, only two branches were found to be not significantly
tion or reduced rates of molecular evolution in the family rela- less parsimonious (p 0.05) than the MP root placement (the
tive to other lineages, as proposed by Soltis et al. (2002). While branch subtending Christensenia [24] and the branch subtend-
trnSGG proved to be the most variable region used in this study, ing Marattia A [7]).
this region is far more conserved in Marattiaceae than in other
groups, including the filmy ferns (Hymenophyllaceae) (J. Nitta, Likelihood and Bayesian rootingThe unconstrained ML
University of California, Berkeley, unpublished manuscript); analysis of data set {3} under a GTR+G+I model of evolution
632 American Journal of Botany [Vol. 95

Fig. 3. Maxiumum likelihood (ML) phylogram of Marattiaceae and related lineages based on combined analysis of atpB+rbcL+rps4 {3}, showing
short branch lengths within Marattiaceae. Root position shown is recovered when Equisetum is included in the outgroup and when the GTR+I+G model is
implemented in ML; root position within Marattiaceae varies with outgroup and model choice.
May 2008] MurdockPhylogeny of Marattiaceae 633

Fig. 4. Unrooted network, with topology and branch lengths from maxiumum likelihood (ML) analysis of ingroup taxa. Root positions reconstructed
by various analyses of atpB+rbcL+rps4 {3} with three possible outgroups (Equisetaceae [Equ], Ophioglossaceae [Oph], or Osmundaceae [Osm]) are in-
dicated with arrows. Numbers correspond to root positions in Fig. 5. Rooting with Equisetaceae or any larger group containing Equisetum produces the
same rooting results.

attached the root to the branch subtending Christensenia [24] three possible outgroups (Equisetaceae, Ophioglossaceae, Os-
when Equisetum is included in the outgroup. However, when mundaceae); the results are shown in Figs. 5 and 6. In the SH
Equisetum is excluded from the analysis, rooting changes. In tests the majority of root positions were not significantly less
global analyses, ML and BI analyses of data set {3} resolved optimal than the ML root position. Analyses rooted with the
Marattiaceae and Equistaceae as a clade (99% posterior proba- nearest outgroup, Osmundaceae, could exclude 14 of 41 possi-
bility) sister to leptosporangiate ferns (98% posterior probabil- ble root positions (p 0.05); Ophioglossaceae and Equiseta-
ity), in agreement with other previous studies (Pryer et al., ceae excluded six and four root positions, respectively, all of
2001, 2004; Wikstrm and Pryer, 2005). Despite the fact that which were redundant with those excluded by rooting with Os-
Equisetaceae and Marattiaceae share a most recent common an- mundaceae. In all tests, the root position resolved by MP was
cestor in these analyses, the Equisetum crown group is on such not significantly less optimal than the ML root position.
a long branch that Osmundaceae, Ophioglossaceae, and Psilot-
aceae are closer to Marattiaceae in terms of branch length.
When either Osmundaceae or Ophioglossaceae are used as out- DISCUSSION
groups, unconstrained ML analyses (under a GTR+G+I model
of evolution) attach the root to the branch subtending Marattia Phylogenetic relationships While it was hoped that se-
s.s. [26]. Under simpler models of evolution, however, rooting quence data from Christensenia might break up the long branch
became consistent, with the root attached to the branch below leading to marattioid ferns and help resolve relationships among
Danaea (in agreement with MP) in all outgroup combinations the early-diverging fern lineages, this was not the case. Chris-
with both gene data sets. Midpoint rooting and enforcement of tensenia could be sister to all other marattioid ferns, depending
a molecular clock assumption on both gene data sets also re- on root placement (see Discussion, section Rooting) however,
solved the root in agreement with MP. The BI analyses of both nucleotide sequences from Christensenia are highly similar to
gene data sets resolved a basal polytomy (not shown). those from other marattioids, and Christensenia is in no way
As a comparison to the MP random rooting analysis, the in- a genetic missing link between marattioids and other early-
group portion of data set {3} was rooted with 20 random out- diverging fern lineages.
groups with base composition equal to the average composition of Long-branch attraction, but not long-branch repulsion (LBR),
the ingroup taxa and analyzed under ML. In all trials, multiple appears to be affecting the analyses: the ML ingroup topology
equally likely trees were recovered, resulting in a total of 313 trees was consistent with or without outgroups, and the MP topology
for the 20 trial runs. Each of the 41 possible root positions (Fig. 5) was either congruent with the ML topology or produced an al-
was resolved at least once; frequency of root attachment resolu- ternate (but equally poorly supported) topology when outgroups
tion was nonrandom and scaled approximately to branch length. were included. If LBR were affecting the ML analysis, MP
ShimodairaHasegawa (SH) tests, comparing the likelihood analysis would not be expected to produce the same topology
scores of 41 constraint trees corresponding to all possible root because parsimony has the opposite bias. Additionally, remov-
attachment points within data set {3}, were performed with ing Danaea and/or Christensenia from the analysis had no
634 American Journal of Botany [Vol. 95

waii), Marattia A, which corresponds to Eupodium (the neo-


tropical Marattia laevis complex with stalked synangia), and
Marattia B, a clade comprising the paleotropical taxa. A tax-
onomic revision of Marattiaceae based on the phylogenetic re-
sults presented here, including new combinations and a more
detailed discussion of nomenclature can be found in a forth-
coming paper (Murdock, 2008).
The resolution of Danaea as sister to the other marattioid
ferns (i.e., the consensus root position shown in Fig. 2) agrees
with all previous molecular phylogenies (Pryer et al., 2001,
2004; Des Marais et al., 2003; Murdock, 2005; Wikstrm and
Pryer, 2005; Rothwell and Nixon, 2006; Qiu et al., 2006;
Schuettpelz et al., 2006), but disagrees with the morphological
analysis of Hill and Camus (1986). The sampling in this study
is insufficient to make any clear conclusions about relationships
of species groups within Danaea; Christenhusz et al. (2008),
contains more comprehensive sampling within Danaea than is
presented here.
Christensenia has been previously described as the most
primitive of the marattioid ferns (e.g., Campbell, 1911; Hill and
Camus, 1986); the results of this study do not support this
hypothesis. Christensenia is certainly distinctive, and the unique
morphology of this genus is mirrored in the nucleotide sequences:
while similar to Angiopteris and Marattia s.l., the sequences of
Christensenia contain multiple autapomorphies, mostly in the
form of unique indels that provide no grouping information
within the family. The origin of Christensenia from within a
clade of pinnately compound, free-veined taxa may seem sur-
prising, but is the inevitable conclusion unless Christensenia was
Fig. 5. The 41 possible root attachment points compared in Fig. 6. resolved as sister to all other marattioid taxa. Asama (1960) and
Root position 1, shown in a black circle, is the consensus root position. Hill and Camus (1986) hypothesized that the leaf morphology
and venation of Christensenia resulted from the compression of
a rachis of a complex pinnate frond; the position in the marattioid
phylogeny and the disposition of sori in irregular rows on both
effect on the ML topology. Under MP, real outgroups per- sides of main vein branches support this hypothesis. Christense-
formed no differently than random outgroups. While this result nia has likely undergone significant morphological change as
does not indicate that the root chosen by MP is incorrect, it does adaptation to extreme low-light environments: low, creeping
provide evidence that LBA may be occluding the true signal. In habit; large, raised stomata; and high levels of chlorophyll con-
contrast, when random outgroups were used to root the ingroup centrated in cells on the adaxial leaf surface (with abaxial surface
under ML, each of the 41 possible root positions was selected at cells possibly providing internal light reflection) (Nasrulhaq-
least once, with a weak tendency to attach the root to longer Boyce and Mohamed, 1987; Rolleri, 1993).
internal branches. While LBA may be biasing MP analyses, the Because of the more or less free sporangia and trilete
root position selected by MP corresponds to that chosen by spores, it might be presumed that Angiopteris represents a
midpoint rooting, enforcement of molecular clock assumptions plesiomorphic form among extant marattioids. Bierhorst
under ML, MP Lundberg rooting, and morphology (Rothwell (1971, p. 368) hypothesized the opposite case, arguing that
and Nixon, 2006; Murdock et al., 2006a); it is this consensus the semifree state seen in Angiopteris sporangia is second-
rooting that is shown in Fig. 2. arily derived from a fully synangiate state: The sporangia of
The phylogeny of marattioid ferns presented here has several Angiopteris, although discrete at maturity, have no morpho-
important impacts on the understanding of the taxonomy of logical integrity when initiated. In the Marattiales, the dif-
marattioids. Marattiaceae s.l. is resolved as monophyletic with ference between free sporangia and synangia is directly
strong support, in agreement with earlier phylogenies (e.g., attributable to differential growth of the intersporangial ar-
Pryer et al., 2001; Des Marais et al., 2003; Pryer et al., 2004; eas. The phylogeny presented here supports Bierhorsts hy-
Wikstrm and Pryer, 2005). Marattia s.l. is paraphyletic based pothesis: partially free sporangia and trilete spores are a
on the evidence presented in this study, and is resolved in all derived state in extant marattioids. Fossil marattioids had
analyses (excluding morphology alone) as three well-supported both monolete and trilete spores (some fossils showing a mix-
clades: Marattia s.s. (neotropical taxa and M. douglasii of Ha- ture of the two in the same sporangium), and fused to free


Fig. 6. The 41 possible root positions for the ingroup topology based on atpB+rbcL+rps4 {3} tested with three possible outgroups (AC). Values on the
x-axis correspond to the 41 root positions in Fig. 5 ranked by likelihood. Root positions that were significantly less optimal than the ML root position using
a ShimodairaHasegawa test are marked with an asterisk (*). Root positions not significantly less parsimonious than the most parsimonious position using a
Wilcoxon signed-rank test are marked with a plus-sign (+). ML and MP indicate the maximum likelihood and most parsimonious root positions.
May 2008] MurdockPhylogeny of Marattiaceae 635
636 American Journal of Botany [Vol. 95

sporangia. Because monolete, ellipsoid spores are generally netically distinct from M. fraxinea as well as nearby Austral-
held to be a derived state in fern lineages, with the tetrahedral asian species.
spores of the lycophyte and bryophyte lineages (and the tri-
lete marks found in the majority of these lineages) assumed to Angiopteris evecta This species, the type of Angiopteris, is
be the plesiomorphic state, the trilete, tetrahedral spores in extremely widespread and encompasses a wide range of mor-
Angiopteris may represent a reversal to the ancestral state phological variation. The morphological plasticity of this genus
(Tryon and Lugardon, 1991; Kenrick and Crane, 1997). How- is such that nearly all geographic and taxonomic distinctions
ever, given that basal fern lineages have both monolete and break down upon close scrutiny. While there are indications of
trilete spores, and fossil marattioids also produce both types phylogenetic and biogeographic structure within the A. evecta
of spores, the polarity of the character state change in spore complex, the markers used for this study are insufficiently
morphology is difficult to interpret. The finding that Angiopteris informative to clearly elucidate relationships in this clade.
is monophyletic only if it is circumscribed to include Macro- Angiopteris longifolia, treated in most floras as a synonym of
glossum and Archangiopteris (which also encompasses the A. evecta, is distinguished mainly by having sori extending to
rarely recognized Protomarattia and Protangiopteris) agrees the apex of the pinnule (as opposed to stopping well below the
with the results of Hill and Camus (1986), and the morpho- apex in A. evecta). In the Society Islands, these two forms can
logical arguments presented in Camus (1989) and Chang be seen in mixed populations with no other apparent differences
(1975). (A. Murdock, personal observation). The form introduced to
the Hawaiian Islands, purportedly from Samoa, is comparable
Marattia douglasii Hooker and Arnott (1832) originally to A. longifolia. While there is little basis for maintaining these
identified Marattia douglasii as M. alata, known only from the as separate species, particularly given the range of morphology
American tropics. Later authors either considered this species that A. evecta encompasses (as currently understood), it is inter-
to represent an entirely new genus, Stibasia (e.g., Presl, 1845; esting to note that MP analysis groups A. longifolia from
de Vriese and Harting, 1853), or to be more closely related to Tahaa (Society Islands) with the three accessions from the Ha-
the M. melanesica complex in Southeast Asia (Kuhn, 1889; waiian islands, albeit with poor support (Fig. 2).
Copeland, 1947). Nucleotide sequence data support Hooker
and Arnotts original interpretation that M. douglasii is related Morphology Morphology suffers from the same general
to M. alata. The most likely biogeographic scenario that led to problem as the molecular data: the major clades are highly sup-
this distribution pattern is that the progenitor of M. douglasii ported by multiple characters; however, there are very few
dispersed to Hawaii from the American tropics, possibly car- characters that can inform interrelationships of these clades,
ried by a tropical storm track (Ramp Neale et al., 2007). How- and there is often homoplasy in these few characters. While
ever, given the current sampling and topology, it remains a inclusion of morphology in phylogenetic analyses has increased
remote possibility that Marattia s.s. originated in Hawaii and phylogenetic resolution in other plant groups (e.g., Doyle, et al.,
dispersed to the neotropics. Further sampling within Marattia 1994; Mishler et al., 1994), in Marattiaceae the clades that re-
s.s. and clarification of fossil affinities may help answer this ceived strong support from the morphology and microstructural
question. data are clades that were already strongly supported by se-
quence data.
Marattia purpurascens One of the rarest species in the
family, M. purpurascens, is endemic to Ascension Island and is Rooting Rooting can be problematic in cases where the
known from only a few remaining populations (Cronk, 1980; clade in question is distantly related to the closest available out-
Gray et al., 2005). While morphologically similar to the Afri- group and when internal branches of the clade are relatively
can M. fraxinea, the sequences from this species are clearly dis- short (Graham et al., 2002; Schuettpelz and Hoot, 2006); Mar-
tinct from the South African M. fraxinea sequenced in this attiaceae is an excellent example of just such a clade. Because
study. It is likely, given the wide range of morphological varia- accurate root placement is essential for making conclusions
tion in M. fraxinea, that several species are subsumed under this about character evolution, estimating divergence dates, and for
name; while this has been previously proposed, species delimi- testing most other tree-based hypotheses, it is prudent to check
tation has proven difficult due to intermediate and overlapping for possible rooting biases even when relatively close outgroups
morphologies (Pichi Sermolli, 1969). Because Ascension Is- are available.
land is quite young, ~1 million years old, and the habitat for M. Multiple methods for dealing with problematic rooting and
purpurascens on Ascension Island is probably much younger long-branch attraction (LBA) have been proposed by previous
than that, given the relatively recent volcanic activity (Nielson authors (see e.g., Wheeler, 1990; Nixon and Carpenter, 1993,
and Sibbett, 1996; Ashmole and Ashmole, 2000), it would be 1996; Mishler, 1994; Sanderson et al., 2000; Sanderson and
remarkable (albeit not impossible) for the observed sequence Shaffer, 2002; Graham et al., 2002; Huelsenbeck et al., 2002).
divergence to have arisen in that time. Additional sampling Lundberg (1972) suggested a method in which the ingroup is
from within the M. fraxinea complex, particularly from West rooted using a vector statement representing a hypothetical an-
African populations, is needed to clarify the origin of M. pur- cestor. Lundberg rooting is of particular utility when outgroups
purascens and other lineages in the M. fraxinea complex. are readily identifiable but are so distant that they are affecting
the topology of the ingroup taxa; however, this is not really the
Marattia howeana Marattia howeana, another rare, en- case in the current study, as the ingroup topology is consistent
dangered species, is endemic to Lord Howe Island. Threatened with different outgroups, only the root position changes. More-
due to habitat destruction and grazing by introduced pigs, the over, the composition of the Lundberg vector has a strong effect
few remaining known populations of M. howeana are being on rooting, and the best method of determining this vector is not
protected and monitored. While treated in the past as a variety always clear when there are multiple outgroups to choose from
of M. fraxinea, M. howeana is both morphologically and ge- and their character composition is highly heterogeneous.
May 2008] MurdockPhylogeny of Marattiaceae 637

If a data set is clock-like, midpoint rooting or enforcing a mo- parameter-rich models used in ML select alternative root posi-
lecular clock in a ML analysis are robust methods for selecting tions, these positions are never significantly more optimal than
a root position. Huelsenbeck et al. (2002) found that even if data root position [1] on the branch subtending Danaea.
violates clock assumptions, enforcing a molecular clock may Other possible data sources for determining the root of any
still identify the correct root. Midpoint rooting may choose the clade include fossil morphology and external evidence such as
correct root under even wider ranges of rate heterogeneity stratigraphy and historical biogeography. While not always
(Swofford et al., 1996). However, with both midpoint rooting available, these are valuable sources of information on rooting
and molecular clock enforcement, if clock assumptions are vio- and character polarity that should not be overlooked. In the case
lated by the data, confidence in these rootings is impossible to of Marattiaceae, fossil morphology, as currently understood,
assess, and alternative criteria are always desirable. While non- does not provide clear rooting information because of large
reversible character change models and asymmetric step-matri- amounts of missing data, seemingly random distribution of
ces have been used for rooting trees, these methods do not appear character states in fossil taxa, and the absence of Tertiary fossils
to provide consistently accurate results and often require unjus- that could be more directly compared with extant taxa. If the
tifiable assumptions (Yang, 1994; Huelsenbeck et al., 2002). relationships of fossil marattioids and polarity of character
Neither MP nor ML is immune from the effects of LBA; MP change were more clearly understood, it is likely that more in-
tends to be more strongly affected, although ML can suffer formation on the rooting of the extant marattioid ferns could be
similarly if models are poorly suited to the data due to violated obtained. Additionally, genomic structural data or gene dupli-
assumptions or mismatch in the selection process (Sullivan and cations may provide clearer rooting information for the family
Swofford, 2001; Swofford et al., 2001; Steel, 2005). Addition- in the future.
ally, ML might in some cases produce positively misleading
results in the form of long-branch repulsion (LBR) (Siddall,
LITERATURE CITED
1998; Siddall and Whiting, 1999). While simulation studies
have disagreed on the question of whether MP or ML are better Asama, K. 1960. Evolution of leaf forms through the ages explained by
suited to dealing with the challenges of LBA, the general con- the successive retardation and neoteny. Science Reports, Tohoku
sensus is that MP will outperform ML when long-branch taxa University, Sendai, Japan, 2nd series (Geology), special vol. 4:
are sister in the true phylogeny (Kolaczkowski and Thornton, 252280.
2004; Bergsten, 2005); as Pol and Siddall (2001) stated, it is not Ashmole, P., and M. Ashmole. 2000. St. Helena and Ascension Island:
clear whether parsimony outperforms likelihood for reasons A natural history. Oswestry, Shropshire, UK.
other than succumbing to its own vices at the right time. The Backer, C. A. 1929. The problem of Krakatao as seen by a botanist.
Visser and Co., Weltevreden, Java; Martinus Nijhoff, The Hague,
most effective measures for preventing LBA and LBR are to (1) Netherlands.
use a model-based method with a good-fitting model, (2) in- Bergsten, J. 2005. A review of long-branch attraction. Cladistics 21:
crease taxon sampling as much as possible to break up the long 163193.
branches, and (3) choose appropriately varying character sets in Bierhorst, D. W. 1971. Morphology of vascular plants. Macmillan, New
which the branches are not particularly long (Mishler, 2005). York, New York, USA.
While model-based methods (e.g., ML and BI) are readily Brownsey, P. J., and J. C. Smith-Dodsworth. 2000. New Zealand
available and widely applied, we currently have very few tools Ferns and Allied Plants, 2nd ed. David Bateman, Auckland, New
to accurately assess a priori the fit of models to data and to de- Zealand.
termine how model mismatch might affect the results (but see Buckley, T. R., C. Simon, H. Shimodaira, and G. K. Chambers.
2001. Evaluating hypotheses on the origin and evolution of the New
Goremykin and Hellwig, 2006). Additionally, increasingly pa- Zealand alpine cicadas (Maoricicada) using multiple-comparison tests
rameter-rich and better-fitting models are not necessarily better of tree topology. Molecular Biology and Evolution 18: 223234.
for phylogenetic inference (Tarrio et al, 2000; Steel, 2005; Burleigh, J. G., and S. Mathews. 2004. Phylogenetic signal in nucle-
Kelchner and Thomas, 2006). Increased sampling will be maxi- otide data from seed plants: Implications for resolving the seed plant
mally effective for rooting if additional taxa fall along long tree of life. American Journal of Botany 91: 15991613.
branches; additional sampling within crown groups will likely Campbell, D. H. 1911. The Eusporangiatae: The comparative morphol-
have little effect. Unfortunately, in many cases, including Mar- ogy of the Ophioglossaceae and Marattiaceae. Carnegie Institution of
attiaceae, extinction has removed the possibility of sampling Washington, Publication 140, New York, New York, USA.
below the crown group for molecular analyses. Camus, J. M. 1989. The limits and affinities of Marattialean fern gen-
The rooting problem encountered in Marattiaceae is similar to era in China and the west Pacific. In K. H. Shing and K. U. Kramer
[eds.], Proceedings of the International Symposium on Systematic
that found in several previous studies (e.g., Tarrio et al., 2000; Pteridology, Beijing, China, 1988, 3137. China Science and
Wiens and Hollingsworth, 2000; Graham et al., 2002; Schuettpelz Technology Press, Beijing, China.
and Hoot, 2006); rooting techniques used in this paper may Chang, C.-Y. 1975. Morphology of Archangiopteris Christ. & Gies. and
prove useful in similar situations regardless of the organisms in its relationship with Angiopteris Hoffm. Acta Botanica Sinica 15:
question. In Marattiaceae, the rate heterogeneity and the absence 235247 [English translation of Acta Botanica Sinica 15: 261276.
of any near outgroup are the most severe issues; nucleotide com- 1973].
positional differences between the ingroup and the outgroup do Chiang, T. Y., Y. C. Chiang, C. H. Chou, Y. P. Cheng, and W. L.
not appear to be a factor as they were in Drosophila (Tarrio Chiou. 2002. Phylogeography and conservation of Archangiopteris
et al., 2000). Because model selection and outgroup composition somai Hayata and A. itoi Shieh based on nucleotide variation of the
atpB-rbcL intergenic spacer of chloroplast DNA. Fern Gazette 16:
have a strong effect on root placement, methods such as mid- 335 (Abstract).
point rooting, enforcement of a molecular clock, or Lundberg Ching, R. C. 1959. Pteridophyta: OphioglossaceaeOleandraceae. In S.
rooting may be the best options for rooting the family. All three Chien and W. Chun [eds.], Flora Reipublicae Popularis Sinicae, vol.
of these methods resolve the same root position, the branch 2. Academia Sinica, Peking [Beijing], China.
subtending Danaea [1], which is the same position found by Christenhusz, M. J. M., H. Tuomisto, J. S. Metzgar, and K. M.
MP and ML under simpler models. In addition, when the more Pryer. 2008. Evolutionary relationships within the neotropical,
638 American Journal of Botany [Vol. 95

eusporangiate fern genus Danaea (Marattiaceae). Molecular Phylo- Marattiaceae, and Ophioglossaceae (chiefly derived from the Kew
genetics and Evolution 46: 3448. Herbarium). Robert Hardwicke, London, UK.
Collinson, M. E. 2001. Cainozoic ferns and their distribution. Brittonia Hsu, T. W., S. J. Moore, and T. Y. Chiang. 2000. Low RAPD poly-
53: 173235. morphism in Archangiopteris itoi, a rare and endemic fern in Taiwan.
Copeland, E. B. 1947. Genera filicum: The genera of ferns. [Annales Botanical Bulletin of Academia Sinica 41: 1518.
Cryptogamici et Phytopathologici 5]. Chronica Botanica, Waltham, Huelsenbeck, J. P., and F. Ronquist. 2001. MRBAYES, Bayesian in-
Massachusetts, USA. ference of phylogenetic trees. Bioinformatics (Oxford, England) 17:
Cronk, Q. C. B. 1980. Extinction and survival in the endemic vascular 754755.
flora of Ascension Island. Biological Conservation 17: 207219. Huelsenbeck, J. P., J. P. Bollback, and A. M. Levine. 2002. Inferring
De Vriese, W. H., and P. Harting. 1853. Monographie des Marattiaces. the root of a phylogenetic tree. Systematic Biology 51: 3243.
Noothoven van Goor, Arnz Co., Leiden, Netherlands. Kelch, D. G., A. Driskell, and B. D. Mishler. 2004. Inferring phy-
Des Marais, D. L., A. R. Smith, D. M. Britton, and K. M. Pryer. logeny using genomic characters: A case study using land plant plas-
2003. Phylogenetic relationships of extant horsetails, Equisetum, tomes. In B. Goffinet, V. Hollowell, and R. Magill [eds.], Molecular
based on chloroplast DNA sequence data (rbcL and trnL-F). systematics of bryophytes. Monographs in systematic botany from the
International Journal of Plant Sciences 164: 737751. Missouri Botanical Garden, vol. 98, 311. Missouri Botanical Garden
Don, R. H., P. T. Cox, B. J. Wainwright, K. Baker, and J. S. Mattick. Press, St. Louis, Missouri, USA.
1991. Touchdown PCR to circumvent spurious priming during Kelchner, S. A., and M. A. Thomas. 2006. Model use in phylogenetics:
gene amplification. Nucleic Acids Research 19: 4008. Nine key questions. Trends in Ecology & Evolution 22: 8794.
Doyle, J. A., M. J. Donoghue, and E. A. Zimmer. 1994. Integration of Kenrick, P., and P. R. Crane. 1997. The origin and early diversification
morphological and ribosomal RNA data on the origin of angiosperms. of land plants: A cladistic study. Smithsonian Institution, Washington,
Annals of the Missouri Botanical Garden 81: 419450. D.C., USA.
Embley, T. M., and W. Martin. 2006. Eukaryotic evolution, changes Kingston, N., S. Waldren, and N. Smyth. 2004. Conservation genet-
and challenges. Nature 440: 623630. ics and ecology of Angiopteris chauliodonta Copel. (Marattiaceae), a
Felsenstein, J. 1981. Evolutionary trees from DNA sequences: A critically endangered fern from Pitcairn Island, South Central Pacific
maximum likelihood approach. Journal of Molecular Evolution 17: Ocean. Biological Conservation 117: 309319.
368376. Kolaczkowski, B., and J. W. Thornton. 2004. Performance of maxi-
Fu, L. K., and J. M. Jin [eds.]. 1992. China plant red data book: Rare and mum parsimony and likelihood phylogenetics when evolution is het-
endangered plants, vol. 1. Science Press, Beijing, China. erogeneous. Nature 431: 980984.
Giribet, G. 2002. Current advances in the phylogenetic reconstruction Kuhn, M. 1889. Farne. In H. G. A. Engler [ed.], Die Forschungsreise
of metazoan evolution. A new paradigm for the Cambrian explosion? S.M.S. Gazelle in den Jahren 1874 bis 1876 unter Kommando des
Molecular Phylogenetics and Evolution 24: 345357. Kapitn zur See Freiherrn von Schleinitz herausgegeben von dem
Goldman, N., J. P. Anderson, and A. G. Rodrigo. 2000. Likelihood- Hydrographischen Amt des Reichs-Marine-Amts. IV. Theil. Botanik.
based tests of topologies in phylogenetics. Systematic Biology 49: Ernst Siegfried Mittler und Sohn, Berlin, Germany.
652670. Li, C.-X., and S.-G. Lu. 2007. Phylogeny and divergence of Chinese
Goremykin, V. V., and F. H. Hellwig. 2006. A new test of phyloge- Angiopteridaceae based on chloroplast DNA sequence data (rbcL and
netic model fitness addresses the issue of the basal angiosperm phy- trnL-F). Chinese Science Bulletin 52: 9197.
logeny. Gene 381: 8191. Liu, Z.-H., J. Hilton, and C.-S. Li. 2000. Review on the origin, evo-
Graham, S. W., R. G. Olmstead, and S. C. H. Barrett. 2002. Rooting lution and phylogeny of Marattiales. Chinese Bulletin of Botany 17:
phylogenetic trees with distant outgroups: a case study from the 3952.
commelinoid monocots. Molecular Biology and Evolution 19: Lundberg, J. G. 1972. Wagner networks and ancestors. Systematic
17691781. Zoology 21: 398413.
Graham, S. W., P. A. Reeves, A. C. E. Burns, and R. G. Olmstead. Maddison, D. R., and W. P. Maddison. 2005. MacClade, version 4.08.
2000. Microstructural changes in noncoding chloroplast DNA: Sinauer, Sunderland, Massachusetts, USA.
Interpretation, evolution, and utility of indels and inversions in basal Manhart, J. R. 1994. Phylogenetic analysis of green plant rbcL se-
angiosperm phylogenetic inference. International Journal of Plant quences. Molecular Phylogenetics and Evolution 3: 114127.
Sciences 161 (supplement): S83S96. Manhart, J. R. 1995. Chloroplast 16S rDNA sequences and phyloge-
Gray, A., T. Pelembe, and S. Stroud. 2005. The conservation of the netic relationships of fern allies and ferns. American Fern Journal
endemic vascular flora of Ascension Island and threats from alien spe- 85: 182192.
cies. Oryx 39: 15. Millay, M. A. 1979. Studies of Paleozoic Marattialeans: A monograph of
Hamilton, M. B. 1999. Tropical tree gene flow and seed dispersal: the American species of Scolecopteris. Palaeontographica. Abteilung
Deforestation affects the genetic structure of the surviving forest frag- B 169: 169.
ments. Nature 401: 129130. Mishler, B. D. 1994. Cladistic analysis of molecular and morphological
Hasebe, M., M. Ito, R. Kofuji, K. Ueda, and K. Iwatsuki. data. American Journal of Physical Anthropology 94: 143156.
1993. Phylogenetic relationships of ferns deduced from rbcL gene Mishler, B. D. 2000. Deep phylogenetic relationships among plants
sequence. Journal of Molecular Evolution 37: 476482. and their implications for classification. Taxon 49: 661683.
Hill, C. R., and J. M. Camus. 1986. Evolutionary cladistics of marat- Mishler, B. D. 2005. The logic of the data matrix in phylogenetic analy-
tialean ferns. Bulletin of the British Museum (Natural History). sis. In V. A. Albert [ed.], Parsimony, phylogeny, and genomics, 5770.
Botany. 14: 219300. Oxford University Press, Oxford, UK.
Holttum, R. E. 1978. The morphology and taxonomy of Angiopteris Mishler, B. D., L. A. Lewis, M. A. Buchheim, K. S. Renzaglia, D.
(Marattiaceae) with description of a new species. Kew Bulletin 32: J. Garbary, C. F. Delwiche, F. W. Zechman, T. S. Kantz, and
587594. R. L. Chapman. 1994. Phylogenetics of the green algae and
Hooker, W. J., and G. A. W. Arnott. 1832. The botany of Captain bryophytes..Annals of the Missouri Botanical Garden 81: 451483.
Beecheys voyage; comprising an account of the plants collected by Murdock, A. G. 2005. Molecular evolution and phylogeny of marattioid
Messrs. Lay and Collie, and other officers of the expedition, during the ferns, an ancient lineage of land plants. Botany 2005: Annual Meeting
voyage to the Pacific and Behrings Strait, performed in His Majestys of the Botanical Society of America, Austin, Texas, USA [online
ship Blossom, under the command of Captain F. W. Beechey in the abstract]. Website http://www.2005.botanyconference.org/engine/
years 1825, 26, 27, and 28, part 3. H. G. Bohn, London, UK. search/index.php?func=detail&aid=41.
Hooker, W. J., and J. G. Baker. 1868. Synopsis filicum; or a synop- Murdock, A. G. In press. A taxonomic revision of the eusporangiate fern
sis of all known ferns, including the Osmundaceae, Schizaeaceae, family Marattiaceae, with description of a new genus Ptisana. Taxon.
May 2008] MurdockPhylogeny of Marattiaceae 639

Murdock, A. G., A. R. Smith, B. D. Mishler, and K. S. Renzaglia. Rolleri, C. H. 2003. Caracteres diagnsticos y taxonoma en el gnero
2006a. Evolutionary origin of the leptosporangiate ferns: Sorting Angiopteris Hoffm. (Marattiaceae Bercht. & J. S. Presl) II, Sinopsis de
through conflicting evidence. Botany 2006: Annual Meeting of the las especies. Revista del Museo de La Plata. Seccin Botnica 16: 123.
Botanical Society of America, Chico, California, USA [online ab- Rolleri, C. H., M. C. Lavalle, A. Mengascini, and M. Rodrguez.
stract]. Website http://www.2006.botanyconference.org/engine/ 2003. Sistemtica de los helechos maratiaceae [sic] (Marattiales-
search/index.php?func=detail&aid=769. Marattiaceae). Revista del Museo de La Plata. Seccin Botnica 16:
Murdock, A. G., J. L. Reveal, and A. Doweld. 2006b. (1746) Proposal 121.
to conserve the name Marattiaceae against Danaeaceae (Pteridophyta). Ronquist, F., and J. P. Huelsenbeck. 2003. MrBayes 3: Bayesian phy-
Taxon 55: 10401042. logenetic inference under mixed models. Bioinformatics (Oxford,
Nadot, S., R. Bajon, and B. Lejeune. 1994. The chloroplast gene rps4 England) 19: 15721574.
as a tool for the study of Poaceae phylogeny. Plant Systematics and Roper, J. M., S. K. Hansen, P. G. Wolf, K. G. Karol, D. F. Mandoli,
Evolution 191: 2738. K. D. E. Everett, J. Kuehl, and J. L. Boore. 2007. The complete
Nasrulhaq-Boyce, A., and M. A. Mohamed. 1987. Photosynthetic plastid genome sequence of Angiopteris evecta (G. Forst.) Hoffm.
and respiratory characteristics of Malayan sun and shade ferns. New (Marattiaceae). American Fern Journal 97: 95106.
Phytologist 105: 8188. Rothwell, G. W. 1999. Fossils and ferns in the resolution of land plant
Nickrent, D. L., C. L. Parkinson, J. D. Palmer, and R. J. Duff. phylogeny. Botanical Review 65: 188218.
2000. Multigene phylogeny of land plants with special reference Rothwell, G. W., and K. C. Nixon. 2006. How does the inclusion of
to bryophytes and the earliest land plants. Molecular Biology and fossil data change our conclusions about the phylogenetic history
Evolution 17: 18851895. of euphyllophytes? International Journal of Plant Sciences 167:
Nielson, D. L., and B. S. Sibbett. 1996. Geology of Ascension Island, 737749.
South Atlantic Ocean. Geothermics 25: 427448. Sanderson, M. J., and H. B. Shaffer. 2002. Troubleshooting molecular
Nixon, K. C., and J. M. Carpenter. 1993. On outgroups. Cladistics 9: phylogenetic analyses. Annual Review of Ecology and Systematics 33:
413426. 4972.
Nixon, K. C., and J. M. Carpenter. 1996. On simultaneous analysis. Sanderson, M. J., M. F. Wojciechowski, J.-M. Hu, T. S. Khan, and S.
Cladistics 12: 221241. G. Brady. 2000. Error, bias, and long-branch attraction in data for
Nylander, J. A. A. 2004. MrModeltest v2. Program distributed by the two chloroplast photosystem genes in seed plants. Molecular Biology
author. Evolutionary Biology Centre, Uppsala University, Sweden. and Evolution 17: 782792.
Available at website www.abc.se/~nylander/. Schuettpelz, E., and S. B. Hoot. 2006. Inferring the root of Isotes:
Pichi Sermolli, R. E. G. 1969. Adumbratio Florae Aethiopicae. 16. Exploring alternatives in the absence of an acceptable outgroup.
Marattiaceae. Webbia 23: 329351. Systematic Botany 31: 258270.
Pichi Sermolli, R. E. G. 1977. Attempt at the classification of pterido- Schuettpelz, E., P. Korall, and K. M. Pryer. 2006. Plastid atpA data
phytes based on phylogeny. Webbia 31: 313512. provide improved support for deep relationships among ferns. Taxon
Pickard, J. 1983. Rare or threatened vascular plants of Lord Howe Island. 55: 897906.
Biological Conservation 27: 125139. Schuettpelz, E., and K. M. Pryer. 2007. Fern phylogeny inferred
Pol, D., and M. E. Siddall. 2001. Biases in maximum likelihood and from 400 leptosporangiate species and three plastid genes. Taxon 56:
parsimony: A simulation approach to a 10-taxon case. Cladistics 17: 10371050.
266281. Schneider, H. 2007. Plant morphology as the cornerstone to the inte-
Posada, D., and K. A. Crandall. 1998. Modeltest: Testing the model gration of fossils and extant taxa in phylogenetic analyses. Species.
of DNA substitution. Bioinformatics (Oxford, England) 14: 817818. Phylogeny and Evolution 1: 6571.
Presl, C. B. 1845. Supplementum tentaminis pteridographiae, continens Shaw, J., E. B. Lickey, J. T. Beck, S. B. Farmer, W. Liu, J. Miller,
genera et species ordinum dictorum Marattiaceae, Ophioglossaceae, K. C. Siripun, C. T. Winder, E. E. Schilling, and R. L. Small.
Osmundaceae, Schizaeaceae et Lygodiaceae. Haase, Prague, Czech 2005. The tortoise and the hare II: Relative utility of 21 noncoding
Republic. chloroplast DNA sequences for phylogenetic analysis. American
Pryer, K. M., H. Schneider, A. R. Smith, R. Cranfill, P. G. Wolf, J. Journal of Botany 92: 142166.
Hunt, and S. D. Sipes. 2001. Horsetails and ferns are a monophyl- Shaw, J., E. B. Lickey, E. E. Schilling, and R. L. Small.
etic group and the closest living relatives to seed plants. Nature 409: 2007. Comparison of whole chloroplast genome sequences to choose
618622. noncoding regions for phylogenetic studies in angiosperms: The tor-
Pryer, K. M., E. Schuettpelz, P. G. Wolf, H. Schneider, A. R. toise and the hare III. American Journal of Botany 94: 275288.
Smith, and R. Cranfill. 2004. Phylogeny and evolution of ferns Shimodaira, H., and M. Hasegawa. 1999. Multiple comparisons of log-
(monilophytes) with a focus on the early leptosporangiate divergences. likelihoods with applications to phylogenetic inference. Molecular
American Journal of Botany 91: 15821598. Biology and Evolution 16: 11141116.
Qiu, Y. L., L. Li, B. Wang, Z. Chen, V. Knoop, M. Groth-Malonek, Siddall, M. E. 1998. Success of parsimony in the four-taxon case: Long-
O. Dombrovska, J. Lee, L. Kent, J. Rest, G. F. Estabrook, T. A. branch repulsion by likelihood in the Farris Zone. Cladistics 14:
Hendry, D. W. Taylor, C. M. Testa, M. Ambros, B. Crandall- 209220.
Stotler, R. J. Duff, M. Stech, W. Frey, D. Quandt, and C. C. Siddall, M. E., and M. F. Whiting. 1999. Long-branch abstractions.
Davis. 2006. The deepest divergences in land plants inferred from Cladistics 15: 924.
phylogenomic evidence. Proceedings of the National Academy of Small, R. L., E. B. Lickey, J. Shaw, and W. J. Hauk. 2005. Amplification
Sciences, USA 103: 1551115516. of noncoding chloroplast DNA for phylogenetic studies in lycophytes
Rambaut, A., and A. Drummond. 2003. Tracer v.1.3: MCMC trace file and monilophytes with a comparative example of relative phyloge-
analyser. Oxford University, Oxford, UK. Available at website http:// netic utility from Ophioglossaceae. Molecular Phylogenetics and
tree.bio.ed.ac.uk/. Evolution 36: 509522.
Ramp Neale, J. M., R. B. Neale, and T. A. Ranker. 2007. Identifying Smith, A. R. 1995. Non-molecular phylogenetic hypotheses for ferns.
the dominant source regions for the colonization of Hawaiian fern American Fern Journal 85: 104122.
species using a trajectory and dispersal model. Botany and Plant Smith, A. R., and R. B. Cranfill. 2002. Intrafamilial relationships of
Biology 2007 Joint Congress, Chicago, Illinois, USA [online abstract]. the thelypteroid ferns (Thelypteridaceae). American Fern Journal 92:
Website http://www.2007.botanyconference.org/engine/search/index. 131149.
php?func=detail&aid=1722 Smith, A. R., K. M. Pryer, E. Schuettpelz, P. Korall, H. Schneider,
Rolleri, C. H. 1993. Revision of the genus Christensenia. American and P. G. Wolf. 2006. A classification for extant ferns. Taxon 55:
Fern Journal 83: 319. 705731.
640 American Journal of Botany [Vol. 95

Soltis, P. S., D. E. Soltis, V. Savolainen, P. R. Crane, and T. G. Templeton, A. R. 1983. Phylogenetic inferences from restriction endo-
Barraclough. 2002. Rate heterogeneity among lineages of tra- nucleases cleavage site maps with particular reference to the evolution
cheophytes: Integration of molecular and fossil data and evidence of humans and apes. Evolution 37: 221244.
for molecular living fossils. Proceedings of the National Academy of Tryon, A. F., and B. Lugardon. 1991. Spores of the Pteridophyta:
Sciences, USA 99: 44304435. Surface, wall structure, and diversity based on electron microscope
Steel, M. 2005. Should phylogenetic models be trying to fit an ele- studies. Springer-Verlag, New York, New York, USA.
phant? Trends in Genetics 21: 307309. Warshauer, F. R. 1979. An overview of the feral pig problem in the
Sullivan, J., and D. L. Swofford. 2001. Should we use model-based Hawaii Volcanoes National Park. In R. M. Linn [ed.], Proceedings
methods for phylogenetic inference when we know that assumptions of the 2nd Conference of Scientific Research in National Parks, 15.
about among-site rate variation and nucleotide substitution pattern are National Parks and Wildlife Service, San Francisco, California,
violated? Systematic Biology 50: 723729. USA.
Swofford, D. L. 2002. PAUP*: Phylogenetic analysis using parsimony Wheeler, W. C. 1990. Nucleic acid sequence phylogeny and random
(*and other methods). Sinauer, Sunderland, Massachusetts, USA. outgroups. Cladistics 6: 363367.
Swofford, D., G. Olsen, P. Waddell, and D. M. Hillis. Wiens, J. J., and B. D. Hollingsworth. 2000. War of the igua-
1996. Phylogenetic inference. In D. Hillis, C. Moritz, and B. Mable nas: Conflicting molecular and morphological phylogenies and
[eds.], Molecular systematics, 2nd ed., 407511. Sinauer, Sunderland, long-branch attraction in iguanid lizards. Systematic Biology 49:
Massachusetts, USA. 143159.
Swofford, D. L., P. J. Waddell, J. P. Huelsenbeck, P. G. Foster, P. O. Wikstrm, N., and K. Pryer. 2005. Incongruence between primary
Lewis, and J. S. Rogers. 2001. Bias in phylogenetic estimation and sequence data and the distribution of a mitochondrial atp1 group
its relevance to the choice between parsimony and likelihood meth- II intron among ferns and horsetails. Molecular Phylogenetics and
ods. Systematic Biology 50: 525539. Evolution 36: 484493.
Szvnyi, P., Z. Hock, E. Urmi, and J. J. Schneller. 2006. Contrasting Wolf, P. G. 1997. Evaluation of atpB nucleotide sequences for phyloge-
phylogeographic patterns in Sphagnum fimbriatum and Sphagnum netic studies of ferns and other pteridophytes. American Journal of
squarrosum (Bryophyta, Sphagnopsida) in Europe. New Phytologist Botany 84: 14291440.
172: 784794. Wolf, P. G., K. G. Karol, D. F. Mandoli, J. Kuehl, K.
Tarrio, R., F. Rodrguez-Trelles, and F. J. Ayala. 2000. Tree root- Arumuganathan, M. W. Ellis, B. D. Mishler, D. G. Kelch, R.
ing with outgroups when they differ in their nucleotide composition G. Olmstead, and J. L. Boore. 2005. The first complete chloroplast
from the ingroup: The Drosophila saltans and willistoni groups, a genome sequence of a lycophyte, Huperzia lucidula (Lycopodiaceae).
case study. Molecular Phylogenetics and Evolution 16: 344349. Gene 350: 117128.
Taylor, T. N., and E. L. Taylor. 1993. The biology and evolution of Yang, Z. 1994. Estimating the pattern of nucleotide substitution. Journal
fossil plants. Prentice Hall, Englewood Cliffs, New Jersey, USA. of Molecular Evolution 39: 105111.

Appendix 1. Voucher information and GenBank accession numbers for taxa used in this study. A dash () indicates the region was not sampled. Voucher specimens
are deposited in the following herbaria: UC = University of California Berkeley; TUR = University of Turku, Finland; NY = New York Botanic Garden; PTBG =
National Tropical Botanical Garden, Lauai, Kauai; K = Royal Botanic Gardens, Kew, E = Royal Botanic Gardens, Edinburgh. Place of origin indicated for all
taxa, including plants collected from cultivation for this study. Plants with AGM collection numbers were collected by the author in the field or are from the
living collections of the University of California Botanical Garden, the private collections of R. Whitehead, or Humboldt State University (HSC).

Taxon GenBank accessions: rps4trnS (*=rps4 only), trnSGG, atpB, rbcL; voucher, source, herbarium.
Angiopteris angustifolia C. Presl EU439158, EU439236, , ; AGM UC. A. fokiensis Hieron. EU439154, EU439232, , ; AGM 337,
1060, Malaysia, UC. A. angustifolia EU439159, EU439237, , ; Vietnam, UC. A. hypoleuca de Vriese EU439155, EU439233, ,
AGM 1061, Malaysia, UC. A. angustifolia EU439161, EU439239, , Chase 23373, in cult., Java (?), K. A. lygodiifolia Rosenst. EU439157,
; AGM 335, in cult., Philippines (?), UC. A. angustifolia EU439160, EU439235, , ; AGM 158, Taiwan, UC. A. lygodiifolia EU439156,
EU439238, , ; AGM 336, in cult., Philippines (?), UC. A. boninensis EU439234, , ; Cubey 170, Taiwan, E. A. lygodiifolia EU439163,
Hieron. EU439162, EU439240, , ; D. Lorence 9441, Bonin EU439241, , ; D. Lorence 9440, in cult., Taiwan (?), PTBG. A.
Islands, PTBG. A. caudatiformis Hieron. EU439168, EU439245, , rapensis E.D. Br. EU439153, , , ; Wood 9703, Rapa, NY. A.
; Cubey 168, China, E. A. caudatiformis EU439167, EU439244, sp. China A EU439165, , , ; E. S. J. Harris 777a, China, UC.
, ; E. S. J. Harris 772, China, UC. A. durvilleana de Vriese A. sp. China B EU439166, EU439243, , ; E. S. J. Harris 777b,
EU439155, EU439233, , ; D. Lorence 9439, Guam (?), PTBG. A. China, UC.
evecta (G. Forst.) Hoffm. EU439144, EU439223, , ; A. W. Larsen Archangiopteris caudata Ching EU439172, , , ; E. S. J. Harris 778,
04-112, American Samoa, UC. A. evecta EU439145, EU439224, , China, UC. A. hokouensis Ching EU439171, , , ; E. S. J. Harris
; J. Game s.n., Cook Islands (Rarotonga), UC. A. evecta EU439139, 779, China, UC. A. itoi W.C. Shieh EU439170, EU439247, EU439073,
EU439218, EU439071, EU439092; A. E. Hinkle 288, Fiji, UC. A. evecta EU439094; AGM 356, Taiwan, UC.
EU439140, EU439219, , ; A. E. Hinkle 311, Fiji, UC. A. evecta
EU439143, EU439222, , ; AGM 389, Hawaii (Maui), UC. A. evecta Christensenia aesculifolia (Blume) Maxon EU439102, EU439184,
EU439141, EU439220, , ; AGM 340, Hawaii (Hawaii), UC. A. EU439057, EU439079; AGM 354, Malaysia, UC. C. aesculifolia
evecta EU439142, EU439221, , ; AGM 387, Hawaii (Oahu). EU439103, EU439185, , ; AGM 1072, Malaysia, UC.
A. evecta EU439148, EU439227, , ; AGM 347, Malaysia, UC.
Danaea elliptica Sm. EU439096, EU439178, EU439054, ; AGM
A. evecta EU439149, EU439228, , ; AGM 349, Malaysia,
405, Puerto Rico, UC. D. leprieurii Kunze EU439097, EU439179,
UC. A. evecta EU439150, EU439229, , ; AGM 1058, Malaysia, UC.
EU439055, EU439077; M. Christenhusz 2128, Peru, TUR. D. nodosa
A. evecta EU439147, EU439226, , ; AGM 348, Borneo (Sabah),
(L.) Sm. EU439100, EU439182, EU439056, EU439078; M.
UC. A. evecta EU439146, EU439225, , ; AGM 339, New
Christenhusz 2565, Suriname, TUR. D. nodosa EU439098, EU439180,
Caledonia, UC. A. evecta EU439152, EU439231, , ; AGM 345,
, ; AGM 331, Puerto Rico, UC. D. nodosa (L.) Sm. EU439099,
Papua New Guinea, UC. A. evecta EU439138, EU439217, , AGM
EU439181, , ; AGM 404, Puerto Rico, UC. D. wendlandii Rchb.f.
156, Society Islands (Moorea), UC. A. evecta EU439136, EU439215,
EU439101, EU439183, , ; R. Moran 6326, Costa Rica, NY.
, ; AGM 153, Society Islands (Raiatea), UC. A. evecta EU439137,
EU439216, EU439070, EU439091; AGM 142, Society Islands (Tahaa), Equisetum ramosissimum Desf. subsp. debile (Roxb. ex Vaucher) Hauke
UC. A. evecta EU439151, EU439230, , ; AGM 346, Thailand, EU439173*, , EU439074, ; AGM 1068, Malaysia, UC.
May 2008] MurdockPhylogeny of Marattiaceae 641

Helminthostachys zeylanica (L.) Hook. EU439176*, , , EU439095; EU439113, EU439195, EU439063, EU439085; Chase 23771, New
AGM 360, Borneo (Sarawak), UC. Zealand, K. M. salicina EU439114, EU439196, , ; K. R. Wood
10237, Marquesas, PTBG. M. sambucina Blume EU439116, , ,
Macroglossum smithii (Racib.) Campb. EU439169, EU439246, EU439072,
; Lambert s.n., Vietnam, UC. M. smithii Mett. ex Kuhn EU439123,
EU439093; AGM 363, Malaysia, UC.
EU439204, , ; A. E. Hinkle 312, Fiji, UC. M. smithii EU439122,
Marattia alata Sw. EU439108, EU439190, EU439060, EU439082; M. EU439203, , ; A. E. Hinkle 289, Fiji, UC. M. sp. (M. salicina Sm.
Christenhusz 3272, Jamaica, TUR. M. attenuata Labill. EU439126, vel aff.) EU439124, EU439205, , ; AGM 333, New Caledonia,
EU439207, , ; R. Schmid s.n., New Caledonia, UC. M. attenuata UC. M. squamosa Christ EU439119, EU439200, , ; Cubey 169,
EU439127, EU439208, EU439065, ; AGM 358, New Caledonia, UC. M. New Guinea, E. M. sylvatica Blume EU439117, EU439198, , ; A.
attenuata EU439125, EU439206, , ; AGM 332, New Caledonia, W. Larsen 05-210, Indonesia (Sulawesi), UC. M. sylvatica EU439118,
UC. M. douglasii (C. Presl) Baker EU439109, EU439191, EU439061, EU439199, , ; A. W. Larsen 05-211, Indonesia (Sulawesi), UC.
EU439083; AGM 388, Hawaii (Maui), UC. M. douglasii EU439110,
EU439192, , ; AGM 390, Hawaii (Kauai), UC. M. fraxinea Sm. Psilotum nudum (L.) P. Beauv. EU439174*, , EU439075, ; AGM 330,
EU439131, EU439212, EU439067, EU439088; K. Roux s.n., South in cult., UC.
Africa, UC. M. howeana (W.R.B. Oliver) P.S. Green EU439128, Ophioglossum reticulatum L. EU439177*, , , ; AGM 1069,
EU439209, , ; AGM 359, Lord Howe Island, UC. M. kaulfussii J. Malaysia, UC.
Sm. EU439106, EU439188, , ; A. R. Smith 2907, Brazil, UC. M.
kaulfussii EU439105, EU439187, , ; J. M. Silva 3793, Brazil, Tmesipteris obliqua Chinnock EU439175*, , EU439076, ; AGM 355,
UC. M. laevis Sm. EU439107, EU439189, EU439059, EU439081; Australia (Victoria), UC.
AGM 405-B, Puerto Rico, UC. M. laevis EU439104, EU439186, Additional sequences obtained from GenBank for use in this study
EU439058, EU439080; A. R. Smith 727, in cult. NYBG, Costa Rica, UC.
M. laxa Kunze EU439111, EU439193, EU439062, EU439084; M. Angiopteris evecta, rps4trnS, trnSGG, atpB, rbcL: DQ821119; A.
Christenhusz 1313, Mexico, TUR. M. laxa EU439112, EU439194, lygodiifolia, atpB: X58429; rbcL: X58429; Botrychium dissectum,
, ; R. Kirkpatrick 1393, Mexico, UC. M. melanesica Kuhn rbcL: AY138401; Botrychium lunaria, atpB: U93826; rps4: AY870429;
EU439134, EU439214, EU439069, EU439090; Chase 2841, in cult. as M. Danaea elliptica, rbcL: AF313578; Equisetum arvense, atpB: U93824;
werneri, Papua New Guinea, K. M. melanesica EU439135, , , ; rbcL: L11053; rps4: AJ583677; Equisetum ramosissimum, rbcL:
Cubey 164, in cult. as M. werneri, Papua New Guinea, E. M. mertensiana AY226132; Equisetum telmateia, atpB: AF313542; rbcL: AF313580;
(C. Presl) C. Chr. EU439120, EU439201, , ; Howlie s.n., Caroline rps4: AJ583690; Helminthostachys zeylanica, atpB: DQ646095;
Islands (Kosrae), UC. M. oreades Domin EU439129, EU439210, , Huperzia lucidula, atpB, rbcL, rps4: AY660566; Isotes engelmannii,
; AGM 157, Australia (Queensland), UC. M. oreades EU439130, atpB: AF313544; rps4: AF313592; Isotes lacustris, rbcL: AJ010855;
EU439211, EU439066, EU439087; AGM 357, Australia (Queensland), Leptopteris hymenophylloides, rps4: AF313602; Leptopteris wilkesiana,
UC. M. pellucida C. Presl EU439121, EU439202, , ; AGM 1073, atpB: AF313539; rbcL: AB024949; Marattia attenuata, rbcL: AF313581;
Malaysia, UC. M. purpurascens de Vriese EU439132, EU439213, Ophioglossum reticulatum, atpB: U93825; rbcL: AF313582; Osmunda
EU439068, EU439089; Stedson Stroud s.n., Ascension Island, UC. M. banksiifolia, rbcL: AF313602; Osmunda cinnamomea, atpB: AF313539;
salicifolia Schrad. EU439133, , , ; F. Venter s.n., South Africa rps4: AF313602; Psilotum nudum, rbcL: L11059; Selaginella uncinata,
(Transvaal), UC. M. salicina Sm. EU439115, EU439197, EU439064, atpB, rbcL, rps4: AB197035; Tmesipteris oblanceolata, rbcL: U30836;
EU439086; John Game s.n., Cook Islands (Rarotonga), UC. M. salicina Todea barbara, atpB: AY612714; rbcL: AY612686; rps4: AY612676.

Potrebbero piacerti anche