On the Phylogenetic Position of Cystodium: It's Not a Tree Fern – It's a Polypod!

2006·5.4 MB·English

by  Petra Korall

American Fern
Journal 96[2]:45-53 (2006)
On
the Phylogenetic Position of Cystodium:
Not a Tree Fern - a Polypod!
It's It's
Petra
Korall"^
NC
Department Duko Durham, USA,
of Biology, University, 27708,
Museum
and Swedish Sweden
of Natural History, Stockholm,
Conant
David
S.
VT USA
Lyndon
Natural Science Department, State College, Lyndonville, 05851,
Harald
Schneider
Albrecht-von-Haller Institute of Plant Sciences,
Germany
Georg-August University, Gottingen,
UeDA
KUNIHTKC
Division of Life Scient^e, Graduate School of Natural Science and Technology,
Kanazawa Kanazawa,
University, Japan
Harufumi
Nishida
Chuo
Faculty of Science and Engineering, University, 1-13-27 Kasuga,
Bunkyo, Tokyo
112-8551, Japan
Kathleen M. Pryer
NC USA
Department Duke Durham,
of Biology, University, 27708,
—
Abstract. The phylogenetic position of Cystodium Sm. studied here the time using
for
is first
J.
DNA
sequence data. Based on a broad sampling of leptosporangiate ferns and two plastid genes
and we show Cystodium
(rbcL atpB], that does not belong to the tree fern family Dicksoniaceae, as
previously thought. Our results strongly support including Cystodium within the large polypod
clade, and suggest close relationship to the species-poor grade taxa the base of the polypod
its at
topology [Sphenomeris and Lonchitis, or Saccoloma in this study). Further studies, with an
expanded taxon sampling within polypods, are needed to fully understand the more precise
phylogenetic relationships of Cystodium.
C
Cystodium Sm., with single species sorbifolium (Sm.) Sm., has
its
J. J.
traditionally been included in Dicksoniaceae (Christensen, 1938; Pichi
Sermolli, 1977; Tryon and Tryon, 1982; Kramer, 1990; Stevenson and Loconte,
more
1996] and, specifically, has been considered to be closely related to
Dicksonia L'Her. (Hooker, 1844; Copeland, 1947; Holttum and Sen,
1961).
New
Cystodium grows lowland from Borneo Guinea and
in rainforests
to
adjacent islands such as the Bismarck and Louisiade Archipelagos, and the
Solomon
Islands (Holttum, 1963; Croft, 1986; Kramer, 1990). has a creeping,
It
hairy, dictyostelic rhizome with large, bipinnate leaves that are tvjo meters or
more in length (Croft, 1986). The terminal sori are covered by a true indusium
^ Corresponding author. Current address: Department of Phanerogamic Botany, Swedish
Museum SWEDEN.
Box
of Natural fiistory, 50007, SE-104 05 Stockholm, E-mail: petra. korall®
nrm.se.
^
4
m
AMERICAN FERN VOLUME NUMBER
JOURNAL:
46
2 (20(36)
and an indusium, modified This
adaxial, falso a part of the leaf (Crofl, 1986}.
indusiiim type has been highlighted as an important featnre that nnites
Cahchkwna
Cystodium with Dicksonia, Thyrsopteris Knnze, Culcita C. Presl,
(Maxon) Turner & White, and Cibotiuni Kanlf. witliin Dicksoniaceae
(Kramer,
1990).
Cystodium was not included in any of the large-scale molecular phylogenetic
studies in the last decade that focused on the relationships of monilophytes and
Icptosporangiate ferns (Hasebe et 1994, 1995; Prycr et qL, 1995; Rothwoll,
cd.,
1999; Pryer et 2001, 2004; Schneider et 2004; Wikstrom and Pryer,
cd., cd.,
2005]. In several of these studies, Dicksoniaceae, represented by one or more of
was
genera (never Cystodium], consistently found to be closely related to
its
Cyatheaceae in a cladc termed the "tree ferns", where most of the taxa are
When
more was
arboresc:ent. than one representative of Dicksoniaceae in-
DNA
monophyly was
cluded in these studies, the of the family often called into
Wolf
question (Hasebe et 1994, 1995; et al, 1999; Pryer et 2004). in a
cd,, al.,
morphological by Stevenson and Loconte
strictly cladistic analysis (1996),
where Cystodium was included along with other dicksoniaceous
ferns,
Dicksoniaceae was recovered monophylctic. However, including Cystodium
as
Dicksoniaceae, has been challenged based on anatomy
in differences in stipe
its
and morphology
(Nishida, 1984; Croft, 1986) spore (Gastony, 1981). Croft (1986)
even considered Cystodium be so distinctive from other dicksonioid genera
to
DNA
own we
that merited family, Cystodiaceae. Here use sequence data
it its
from two plastid genes [rbcL and atpB] investigate the phj^logenetic
to
Cystodium and
relationship of to other leptosporangiate ferns, specifically to
whether
address belongs within
tree ferns.
it
and Methods
Matef^iaus
DNA
and
Tcixon sampling, isolation, amplification, sequencing, sequence
—
DNA
Cystodium was
alignment. Total of sorbifolium extracted using the
DNeasy plant mini kit from Qiagen (Valencia, California, USA). The voucher
DNA
specimen Christensen 1529 (record no. 2498 in Fern database
is (S) at
http://www. biology. duke. edu/pr3^erlab/ferndb). The rbcL Rnd atpB gcAics from
genome
the plastid were amplified using the polymerase chain reaction (PCR),
PCR
following standard protocols. products were cleaned using the Montage
PCR
USA)
cleanup Massachusetts, according
kit (Millipore, Billerica, to the
manufacturer's protocol. Sequencing reactions were carried out both
for
PCR
Dye
strands of the purified products using Big Terminator Cycle
Sequencing reagents (Applied Biosystcms, Foster City, California, USA). For
information on amplification and sequencing primers, see Table All
1
.
sequencing reactions were processed using either ABl 3700 or ABI 3730XL
automated sequencers (Applied Biosystems). Sequence fragments were
Ann
assembled and edited using Sequencher version (Cene Codes,
4.2.2
Arbor, Michigan, USA). The corrected consensus sequences were aligned
manually using MacClade version 4.07bl3 (Maddison and Maddison,
2005),
AMI
The d:)cL and citpB C]enbank nmnbers for C. sodiijolium are 84111 and
r-p\
KORALL
ET PHYLOGENETIC POSITION OE CYSTODIUM
AL.: 47
Table Primers used for amplifying and sequencing rbcL and atpB from Cystodium sorbifoHum.
1.
Primer Sequonco
(5' to 3'] References
rbcL
ATGTCACCACAAACGGAGACTAAAGC
ESRBCLIF
Koraii et aL, 2006
AGAYCGTTTCYTATTYGTAGGAGAAGC
ES645F
Korall aL,
et
2()()fi
TAGRAATARGAAAGGRTCTCTCCAAGG
ES663R
2006
Korall et al,
TGAGGAGTCCACTTAGTAGGTTGAGG
ESRBGL1361R
Korall ei al, 2006
atpB
TTGATAGGGGAGGYGGTGTWAGTGT
ATPB672F
Wolf, 1997
ATGGCAGAATRITTGGGAGATRTYA
ATPB1163F
Wolf, 1997
GRACATTTGGACATYTRGATGGTAG
ATPB1419F
Wolf, 1997
TGTAAGGYTGYAAAGTTTGGTTAA
ATPB1592R
Wolf, 1997
ATPB609R TGRTTDGGTTGRGGTGTAGGTTG
Pryor 2004
et
a/.,
GAATTGGAAAGTATTGGATTAGG
ATPE384R
Pryer 2004
et aL.
AM184112,
and
respectively. All rbcL atpB sequences for the other 62 taxa
No
included here were taken from Pryer aL
et (2004). insertions or deletions
were
(indels) required align rbcL or atpB.
to
—
Phylogenetic analyses. The rbcL and atpB were analysed with
data
sets
Bayesian Markov Chain Monte approach (B/MCMC)
a Carlo using the parallel
version of MrBayes 3.0B4 (Ronquist and Huclscnbeck, 2003], and equally
maximum
weighted parsimony (MP) with PAUP""
version 4.0 beta 10
CSEM/OIT
(Swofford, 2002). All analyses were performed on the high-
performance, shared computing cluster Duke University (Durham, North
at
USA]. were
Carolina, Trees rooted with three lycopods: Huperzia Bernh.,
and
Selaginella Pal. Beauv., Isoetes L. (Pryer et 2004).
al.,
Nucleotide models Bayesian were chosen
substitution the analyses using
for
MrModeltest
modified version Modeltest Posada and
2.2 of Crandall,
(a 3.6;
Each was
gene
1998). treated as a single partition (following Pryer et aL, 2004),
and for both rbcL and atpB the general time reversible model (GTR) with
gamma
and
a distribution of substitution rates (F) a proportion of invariant sites
was
suggested based on the Akaike information criterion (AlC). Separate
(1]
analyses of the two datasets were run for 3 million generations, on six parallel
chains, with the temperature parameter heating the chains) The
(for set to 0.1.
sampled values for different parameters were examined using MrBayes and
Drummond,
Tracer v. 1.2.1 (Rambaut and 2005] to see if the parameters had
We
converged. also checked for a proper mixing of the chains. For each analysis
every 1000* was sampled and, analysing parameter 300
tree after values, trees
were Each was
discarded as ''burn-in". analysis repeated four times ascertain
to
was The
that apparent stationarity reached. majority rule consensus trees from
the replicates were compared and no differences were found. Trees from each
analysis were then pooled (a total of 10800 trees, excluding those discarded as
was
burn-in] before majority consensus our Bayesian
a rule tree calculated. In
we
analyses, consider branches with a posterior probability (PP) of 1.00 as well
supported, PP between 0.95-0.99 moderately supported, and PP <0.95
a as a of
low
support.
as
VOLUME NUMBER
AMERICAN FERN JOURNAL:
48 96 2 (2nOB]
Adiantum
0.67/69
+/+
Ceratopteris pteridoid ferns
Coniogramme
56/51
Blechnum
0.67/
Thelypteris eupolypod
ferns
+/96 m
Asplenia
Dennstaedtia
Microlepia
0.87/-
Pteridium
Monachosorum
0,86/-
polypod
"basal ferns"
polypods
+/+ Lonchitis
0.98/82 —
1/821
99/- Sphenomeris
'
I Cystodium
'
0.83/-
Saccoloma
A
z a
oil
Salvinia
heterosporous
ferns
Marsilea
+/+
core Pilularia
+/+
Dicksonia
—
leptosporangiate 0.93/-I
§3l Lophosoria
I
ferns
Cafochlaena
'
+/-
Metaxya
Cyathea
tree ferns
+/93
0.99/95
HymenophyHopsis
Loxoma
Loxsomopsis
Plagiogyha
Anemia
+/99
Schizaea schizaeoid ferns
Lygodium
+/97
Dicranopteris
GleicheneHa
Diplopterygium
Gieichenia
Stromatopteris
gleichenioid ferns
0.83/73
Sticherus
Cheiropleuria
+/+
leptosporangiate
Dipteris
90/-
ferns
Matonia
Phanerosorus
HymenophyHum
filmy ferns
Trichomanes
0.93/52
Lep
top
ten's
osmundaceous
Todea ferns
Osmund
a
Angiopteris
monilophytes Ma
rataa marattioid ferns
+/92
096/60 Danaea
0,54/-
Equisolum
horsetails
HIppochaete
Botrychium
ophioglossoid ferns
Ophioglossum
Psilotum
whisk
ferns
Tmesipteris
Cycas
Ginkgo
Gnetum
+/91
seed
plants
Pinus
Austrobaileya
Chloranthus
Isoetes
+/99
Huperzia lycophytes
0.85/76 Selaginelfa
50% (B/MCMC)
Fig. 1. majnrity-nile consensus treo resulting from Bayesian analyses of the
combined (rbcL and atpB) data Numbers on branches denote support values from Bayesian
set.
KORALL
ET PHYLOGENETIC POSITION OF CYSTODIUM
AL.: 49
MP
analyses for each data set included a heuristic search for the most
TBR
parsimonious with 1000 random-sequence-addition and
trees replicates,
branch swapping. Support nodes was by
for calculated a bootstrap analysis,
with 3000 each with 10 random-sequence-addition and
replicates,
replicates,
MP
TBR
we
branch swapping. In our analyses, considered branches with
>90%
70-89%
a bootstrap percentage (BP) of as well-supported, as moderately
<70%
supported, and weakly
supported.
as
—
and
Comhinahility rbcL atpB data To evaluate combinability
of sets. of
data the resultant topologies from each of the single-gene analyses were
sets,
compared and examined Comparisons were made
for potential conflicts. be-
B/MCMC
tween topologies produced by the same analytical method, was
i.e.,
MP MP
B/MCMC,
compared with and with MP. Incongruence supported by
a
bootstrap percentage of 70 or higher, or a Bayesian posterior probability of 0.99
was considered Because no were
or higher, a conflict. conflicts observed, the
two were combined
data sets into a single data
set.
—
Analyses combined The combined was
the data data analysed
of set. set
B/MCMC
with two
settings as for the separate data In the analyses, the
sets.
genes were each treated as one partition a total of two partitions) and each
(i.e.,
GTR
was + F + model
assigned the of nucleotide substitution in the
(as
I
separate data set analyses].
Results
—
Tree All three data atpB, and combined] included 63
statistics. sets [rhcL,
taxa. The rhcL data included 1320 characters of which 559 were parsimony
set
informative, the atpB data set 1150 characters (562 parsimony informative],
and combined 2470 parsimony
the data set characters (1121 informative). In
MP,
the rhcL analysis yielded two most parsimonious with length
trees, a tree
of 4788 steps, found in one island; the atpB analysis yielded 47 trees of 4088
steps in one island; the combined analysis resulted in eight trees of 8895 steps
found in two islands not shown).
(trees
—
Conflicting topologies. Conflicts in leptosporangiate fern relationships
(i.e.,
incongruence supported by BP > 70% PP > were found when
not
a or 0.99]
MP
topologies from the separate analyses of atpB and rhcL were compared
to
B/MCMC
when
each other, or analyses were compared to each other. The only
when
found was comparing between methods
conflict analytical within the
combined Metaxya
analysis: in the tree fern clade, sister to the other tree
is
MP
=
ferns in (BP 89%), but nested witliin the clade with strong support in
is
B/MCMC =
(PP
1.00].
A
(posterior probaljility; PP) and maximujn parsimony [bootstrap pcnoontages; BP) analyses. plus
= ^ A
represents a PP 1.00, or BP 100%. dash represents a bootsti^ap percentage <50%.
[+)
)
[
=
Thickened branches are well supported (PP 1.00, BP > 90%). Names of clades follow Pryer et
al.
and Schneider al
(2004) et (2004).
VOLUME NUMBER
AMERICAN WAIN JOURNAL:
50 2 (2000)
9(i
—
Because
Phylogenetic relationships of leptosporangiate ferns. the resultant
topologies leptosporangiate relationships from each of the separate gene
for
analyses were not in conflict with one another, the phylogenetic relationships
presented here are based on analyses of the combined data set (Fig.
1].
polypod
Cystodium, the focus of our study, resolved as part of the fern
is
- =
BP 100%
clade with high support (PP Within polypods, there
1.00, Fig.
1].
;
= <
a weakly supported basal split (PP 0.87, BP 50%) with Saccoloma Kaulf.
is
The where Cystodium
sister to the rest. subsequent divergence yields a clade
^ <
+ Spbenomeris Maxon BP and
Lonchitis (PP 0.99, 50%],
sister to L.
is
together these taxa are sister to a large clade that includes the well-supported
moi
=
BP
g6%].
1.00,
—
The
polypods are included in the core leptosporangiate ferns (PP 1.00,
= = =
BP 100%) together with the tree ferns (PP 0.99, BP 95%) and the
= =
BP The
heterosporous ferns (PP 1.00, 99%). core leptosporangiate ferns
=
BP^lOO^^)
are part of the well-supported leptosporangiate ferns (PP 1.00, as
— =
BP
sister to the strongly supported schizaeoid ferns (PP 1.00, lOO^/o). Filmy
=
BP
ferns and gleichenioid ferns group together with low support (PP 0.90,
50%) and usmundaceous supported
the ferns are strongly as sister to the rest
= =
BP
97%;
of leptosporangiate ferns (PP 1.00, Fig.
1).
Discussion
—
combined
Conflicting topologies. In the analysis, the conflicting positions
B/MCMC)
(MP
Metaxya between methods
of within the tree ferns analytical vs.
was also recovered in a study on tree ferns (Korall el ciL. 2006) and is disciLSsed
there as Ukoly being due to a long loranch attraction artefact [Felsenstein, 1978)
MP
in the analysis.
Phylogenetic relationships leplospomngiale Cystodium
of ferns, in
—
shown
particular. Tlie overall leptosporangiate fern relationships in Fig. 1
Our
are not in conflict witli the results of Pryer et ah (2004). results clearly
—
demonstrate Cystodium not polypod
that is a tree fern it is a (l"ig. 1).
Cystodiunj likely closely related to Sphenomeris and Lonchitis (or perhaps
is
and and
Saccoloma], not included in the well-supported species-rich clade
is
that includes the pteridoid, eupolypod, and a more exclusive group of basal
polypod names sensu Schneider
ferns (clade et 2004).
al..
Several morphological and anatomical studies have revealed distinctive
differences between Cystodium and other dicksonioid genera (Sen and Mittra,
19G6; Gastony, 1981; Nishida, 1984; Croft, 1986; Tryon and Lugardon, 1991),
some between Cystodium and
loading authors to suggest a close affinity
Saccoloma and/or Dennstaedtia Bernh., example.
for
Sen and Cystodium from
Mittra (1966) listed several features to separate
and
Dicksonia (including receptacle shape, cortex structure, orientation of
subsidiary and guard-mother but they nevertheless considered be
cells), to
it
part of the dicksonioid-cyathaeoid group.
KORALL ET PHYLOGENETIC POSITION OF CYSTODfUM
AL.:
51
Their conclusion was by and who
contradicted Nisliida (1984) Croft (1986]
thoroughly studied the stipe vasculature of Cystodium and showed major
differences with other dicksonioid genera. In cross section the stipe
Cystodium
vasculature of consists of several strands that unite form
to
first
two branched and two
bilateral arcs, then, closer to the lamina, dorsiventral,
branched arcs. In Dicksonia several vascular bundles form one abaxial and two
which
lateral arcs, unite into a single incurved (U-shaped) strand. Other
dicksonioid genera have vascular strands similar to Dicksonia. This led Croft
Cystodium monotypic
(1986] to transfer to a separate, family, Cystodiaceae,
and Cystodium compared
suggest that be groups with 2-lipped
to to ''other
and
indusia hairy rhizomes, such as the Dennstaedtiaceae" (in the broad sense,
following Kramer,
1990].
e.g.,
noted annulus Cystodium
Croft (1986] also that the of sporangia not fully
is
and
oblique continuous, as typical of other dicksonioid genera. should be
is
It
how
noted, however, that there no general agreement on to interpret the tree
is
fern annulus; one morphological cladistic study considers the annulus to be
oblique (Stevenson and Loconte, whereas another
1996], treats as vertical to
it
slightly oblique as in the polypods; Pryer et 1995).
(i.e., al.,
Further supporting a relationship with the Dennstaedtiaceae the
sA.
is
Cystodium
striated perine of the spores. shares spore ornamentation with
this
Saccoloma and some species of Dennstaedtia (Gastony, 1981; Tryon and
Lugardon,
1991].
In summary, several morphological and anatomical studies have suggested
between Cystodium and
significant differences other genera traditionally
some
assigned to Dicksoniaceae. In addition, these studies propose similiar-
ities to Dennstaedtiaceae In order to gain a better understanding of the
s.I,
precise phylogenetic position of Cystodium, the taxon sampling of early
diverging lineages of polypods will need to bo greatly expanded. Until then,
we
advocate the tentative inclusion of Cystodium in Lindsaeaceae, following
new
a classification for extant ferns (Smith et 2006], as certainly not
al., it is
member
a of Dicksoniaceae.
Acknowledgments
was
This by
stTidy financially supportoci a postdoctoral fellowship to Petra Korall from the
NSF CAREER
Swedish Research Council [2003-2724] and grant Kathleen Pryer (DEB-0347840].
a to
Herbarium curators at S are thanked for their help.
Literature Cited
Manual
Christenskn, C. 1938. Filiranae. In F. Verdoorn, ed., of Pteridology, 522-550. Martinus
The
Hague.
Nijhnff,
CoPELANu, E. B. 1947. Genera Filicum. Chronica Botanica Company, WaUham, MA, USA.
The
Croft, R. 198fi. stipe and rachis vasculature of the dii:ksoniaid fern, Cystodium sorbifolium
J.
Kew
41:789-803.
(Cj'stodiaceae). Bull,
1978. Cases in which parsimony or compatibility methods will be positively mis-
Fr:i.sr;\S]'L;iN,
J.
SysE 27:401-410.
leading. Zool.
VOLUME NUMBER
AMERICAN FERN JOURNAL:
5Z 9() 2 [2006]
C
Ga.s'iony, inSL Spore niorhology in Lho Dicksoniat:oae. L Tho genera Cystodium, Thyrsoptcn's,
J.
and Amer. 68:808-819.
CulcitQ. Bot.
J.
Hasebe, M., T. Omoki, M. Nakazavva, T. Sano and M. Kato. 1994. rbcL gene stujnences provide
USA
NatL Acad.
evidence for evolutionary lineages of leptosporangiate ferns. Proc. Sci.
tht^
91:57:U)-5734.
Haskbe, M., p. G. Wolf, K. M. Pryer, K. Ueda, M. Lro, R. Sano, G. Gastony, Yc^koyama, R.
J. J. J.
Maniiart, N. Murakami, E. H. Crane, C. IL Hauiler and VV. U. Hauk. 1995. Fern pliylogeny
based on rbcL nucleulide sequences. Amer. Fern 85:134-181.
J,
HoLTTUM, R. E. and Sen. 1961. Morpliology and classification the tree ferns. Phytomt)r]}hology
LJ. (jf
11:400-420.
HuLiruM, R. E. 19G3. Cyalheaceae. In C. G. G. Van Steeiiis and R. E. Holttum, eds., Flora
}.
The
Molesiana. Marlinus NijhofL Dr. VV. Junk Publishers, Hague, Boston, Lt)ndon.
W. William Pamplin, London.
Hcx^ER, 1844. Species Filicum. Vol. IL
J.
KoRALL, P., K. M. Pryer, S. Mi-t/car, H. SciiNi:ii)i-:R and D. S. Gonant. 2006. Tree ferns:
}.
monophyletic groups and thcnr relationships as revealed l)y foiu^ protein-coding plastid loci.
Molec. Phylogenet. Evol. 39:830-845
Kramer and
Kkamer, K. U. 1990. Dicksoniaceae. In K. Kubilzki, ed., [Vol. eds. K. U. P. S. Green].
The families and genera of vascular plants. Vol. Ptcridophyies and Gymnospenns.
1
Germany.
Springer- Verlag, Berlin.
W. MacClade
Mauimson, D, R. and P. Mauuison. 2005. version 4.07. Sinauer Associates,
USA.
Sunderland,
Massac:husetts,
NisuiUA, H. 1984. Anatomic:al studies of the frcjnd axis of the C^.yatheaceae, witli a r(?vision of
s./.,
permineralized frond axes from the GrtWaceous of Japan. In M. NisJiida, ed., Contributions
Botany Andes Academia Book, Tokyo.
to the in the 5-80. Scientific
/,
PiCHi Sermolli, R. G. 1977. Tentamen Pteridophytorum. Webbia 31:315—512.
v..
i^osADA, D. and K. A. Cranoall. 1998. Modeltest: testing the model of DN/\ substiluti(jn. Bio-
informatics 14:817-818,
Pryer, K. M., A. R. Smith and E. Skcx;. 1995. Phylogenetic relationships of extant ferns based on
J.
evidence morphology and rbcL sequences. Amer. Fern 85:205—282.
fj'um
J.
Hunt
l^RYi'iR, K. M.. H. SciiNi':iDER, A. R. Smii'ii, R. Graneiee, p. G. Woei', S. and S. D. Sires, 2001.
J.
HorsfJlails and ferns are a monophyletic group and ihe closest living relatives to seed plants.
Nature 409:618-022.
Pryer, K. M., E. Scuui-rTPiiEZ. P. G. Woi.f, H. Souneioi-k, A. R. Smith and R. Graneile. 2004. iMiylogeny
and evolution of ferns (monilophytes) with focus on the early leptosporangiate divtn'gences.
a
Amer, BoL 91:1582-1598.
J.
Ramraut, a. and A. Drummond. Tracer version Computer program distributed by authors.
1.2.1. the;
Deparlmeni of Zoology, University of Oxford. UK.
Runqets]', V. and P. Huels!:m3ECK. 2003. MrBayes 3: Bayesian j)h)'logenetic: infenMice under mixcul
J.
Bioinformatics 19:1572-1574.
niodt;Is.
RcrrnwELE, G. W. 1999. Fossils and ferns in die resolution of land plant phylogeny. Bot. Rev.
65:188-218.
Sc:hneider, H., E. Si::iiuetti'elz, K. M. I^ryek, R. Craniiel, S. Magaleon and R. Lupia. 2004. Ferns
shadow
Nature 428:553-557.
div(;rsiried in the of ajigiosperms.
Sen, and D. Mitika. 19G6. The anatomy of Cystodium. Amer. Fern 50:97-101,
LI.
J.
A
SMrni, A. R., K. M. Pryer, E. Sc:htiettpelz. P. Koraee, IL Schneider and P. G. Woei-. 2006.
Taxon
classification for extant ferns. 55, in press.
Stevenson, D. W. and H. L(x:onte. 1996. Ordinal and familial n^Iationshi])S of pteridophyte gentM'a.
Pp. 435—467 in M. Gamns, M. Gibby and R. Johns, eds., PteridoJogy in perspective. Royal
].
J.
Kew.
Botanic Gardens,
PAUP^
SwoEFORi;), 2002. phylogenetic cmalysis using parsimony ('^and other methods).
L.
13.
:
Sinauer Associates, Sunderland, Massachusetts, LISA.
and
Tryon, a. F. and B. Lugardon. 1991. Spores of the Pteridophyta: surface, wcdl structure,
New
diversity based on edectron microscope studies. Springer- Verlag, York.
Tryon, R. M. and A. F. Tryon. 1982. Ferns and allied plants, with special reference to tropical
New
America. Springor-Verlag, York.
KORALL ET PHYLOGENETIC POSITION OF CYSTODIUM
AL.: 53
WiK.s'['R[)M, N. and K. M. Pkyi-r. 2005. Incongruence between primary sequence data and the
among
and
distribution of a mitochondrial atpl group intron ferns horsetails. Mol.
II
Phylogenet. Evol. 36:484-493.
and
Wolf, P. G. 1997. Evaluation of otpB nucleotide sequences for phylogenetic studies of ferns
Amer. 84:1429-1440,
other pteridophytes. Bot.
J.
WoLF-\ P. G., S. D. SiPES, M. R. WiiiTt:, M. L. Martines, K. M. Pryer, A. R. Smith and K. Ueda. 1999.
Hymenophyllopsidaceae and
Phylogenetic relationships of the enigmatic fern families
Lophosoriaceae: evidence from rbcL nucleotide sequences. PL Syst. Evol. 219:263-270.