American Journal of Botany 93(2): 217–225. 2006.

PHYLOGENY AND BIOGEOGRAPHY OF THE STAGHORN FERN

1
GENUS PLATYCERIUM (POLYPODIACEAE, POLYPODIIDAE)

HANS-PETER KREIER AND HARALD SCHNEIDER2

Albrecht-von-Haller Institute of Plant Sciences, Georg-August Universty Göttingen, Untere Karspüle 2,
37073 Göttingen, Germany

The genus Platycerium is one of the few pantropical epiphytic fern genera with six species in Afro-Madagascar, 8–11
Australasian species, and a single species in tropical South America. Nucleotide sequences of four chloroplast DNA markers are
employed to reconstruct the phylogeny of these ferns and to explore their historical biogeography. The data set was designed to
resolve conflicting hypotheses on the relationships within the genus that were based on previous phylogenetic studies exploring
morphological evidence. Our results suggest a basal split of Platycerium into two well-supported clades. One clade comprises
species occurring in Africa, Madagascar, and South America, whereas the second clade contains exclusively Australasian species.
The latter clade is further divided into a clade corresponding to P. bifurcatum and its putative segregates and a clade of seven
species occurring from Indochina throughout the Malesian region to New Guinea and Australia. The Afro-Madagascan clade
includes a clade of two species found in tropical Africa and a clade of four species that includes three species endemic to
Madagascar. The single neotropical species of this genus, P. andinum, is nested within the Afro-Madagascan clade but is not
closely related to any extant species.

Key words: Africa; historical biogeography; Madagascar; Malesia; phylogeny; polygrammoid ferns; Polypodiaceae; South
America.

In several recent studies, the evolution of epiphytic ferns Moran, 2001). The genus consists of 15 to 18 species growing
was explored with phylogenetic approaches using nucleotide predominantly as epiphytes or sometimes on rocks in
sequence data (e.g., Schneider et al., 2002, 2004a, c; Haufler et subtropical to tropical lowland forests. They are among the
al., 2003; Janssen and Schneider, 2005). This approach most frequent vascular plant epiphytes in those forests and
provided new insights in the relationships of these ferns, predominantly grow in relatively open conditions (Boyer,
biogeography, and morphological evolution. In this context, 1964; Hoshizaki, 1972; Kornas, 1979; Hennipman and Roos,
a similar study on the staghorn fern genus, Platycerium Desv., 1982; Benzing, 1990; Gardette, 1996; Leon and Young, 1996;
is of particular interest because we may solve a longstanding Hoshizaki and Moran, 2001; Zotz and Hietz, 2001).
controversy about the relationships within this genus and its Occurrence of crassulacean-acid-metabolism (CAM) and
biogeographic history (Hoshizaki, 1972; Hennipman and formation of water-storage tissue are further conspicuous
Roos, 1982: Hoshizaki and Price, 1990; Hoshizaki and Moran, characters suggesting adaptation to open conditions in dry,
2001). semideciduous forests (e.g., Hennipman and Roos, 1982;
The staghorn fern genus, Platycerium Desv., stands out Benzing, 1990; Burrows, 1990; Ong et al., 1995; Zotz and
among ferns as one of the most commonly grown ornamental Hietz, 2001).
ferns (Hoshizaki and Moran, 2001; Poremski and Biedinger, Platycerium is one of the few genera of polygrammoid ferns
2001; Darnaedi and Praptosuwiryo, 2003). Both the economic with a pantropical distribution although only a single species,
value and the accessibility of living plants is reflected in the P. andinum Baker, is found in tropical South America
frequent use of these ferns in studies on cytological, (Hoshizaki, 1972; Hennipman and Roos, 1982: Hoshizaki
developmental, and physiological features of vascular plants and Price, 1990; Hoshizaki and Moran, 2001). Roughly one
(e.g., Camloh and Gogala, 1992; Kwa et al., 1995; Ong et al., half (six species) of all species occurs in Africa and
1995; Brumme et al., 1998; Camloh et al., 1999; Holtum and Madagascar, while 8–11 species are found in subtropical to
Winter, 1999; Abrozoic-Dolinsek et al., 2002). These attractive tropical Asia, Malesia, and Australia. None of these species
plants are distinct from other ferns by the presence of stellate occurs in all three regions distinguished here, and the species
hairs (shared with Pyrrosia Mirbel) on the leaf lamina, the diversity in the Afro-Madagascan region is notable because all
differentiation of the leaves into litter collectors—often called other genera of polygrammoid ferns show a much lower rate of
mantle leaves or base fronds—and dichotomously forked species diversity in Africa when compared to either the
trophosporophylls with coenosoroid to acrosoroid patches of neotropics or the Australasian region.
sporangia (Hoshizaki, 1970, 1972; Hennipman and Roos, Two studies explored the relationships among the species of
1982; Lee, 1989; Hoshizaki and Price, 1990; Hoshizaki and Platycerium using morphological evidence and a cladistic
1
approach to analyze these data (Hoshizaki, 1972; Hennipman
Manuscript received 9 August 2005; revision accepted 3 November 2005. and Roos, 1982), but both analyses generated strongly
We thank C. Alford who kindly shared several less frequently grown
divergent hypotheses. They disagree in the proposed relation-
species and the curator/gardeners of the Old Botanical Garden Göttingen
for their continuing efforts to grow these ferns. The project was supported ships of several species, although their data sets are based on
financially by a German Science Foundation grant to H.S. (SCHN 785/2- more or less the same morphological characters. They differ
1) under the Schwerpunkt Programm SPP 1127 ‘‘Radiations—Origin of mainly in the homology assessment of some characters and by
Biological Diversity.’’ differently weighing the information of other characters
2
Author for correspondence (e-mail: [email protected]) (Hennipman and Roos, 1982; Hoshizaki and Price, 1990). In
217
218 A MERICAN J OURNAL OF B OTANY [Vol. 93

contrast to Hennipman and Roos (1982), the phylogenetic Uffelen, 1991; Collinson, 2001) nor divergence time estimates
hypothesis of Hoshizaki (Hoshizaki, 1972; Hoshizaki and (Schneider et al., 2004b) support an origin of these ferns before
Price, 1990) suggested a strong correlation between clade the breakup of Gondwana, which is required in the vicariance
differentiation and geographical distribution. The Afro- hypothesis. In general, historical biogeography of these ferns
Madagascan species together with the neotropical taxa form may provide unique insights into the evolution of epiphytic
one clade, whereas the Australasian species form its sister vegetation in extant subtropical and tropical forests.
clade. This second clade was divided into two putative natural
groups in which one corresponds to the P. bifurcatum (Cav.) C.
Chr. complex consisting of either a single species with four MATERIALS AND METHODS
varieties or of four morphological quite similar species with
partly overlapping distributions (Hoshizaki, 1972; Hennipman Samples of leaf material were collected of all currently accepted species of
and Roos, 1982; Bostock and Spokes, 1998; Hoshizaki and Platycerium from the Old Botanical Garden of the University of Göttingen and
Moran, 2001). The phylogenetic hypothesis of Hoshizaki from the garden of Charles Alford (Florida). At least a single representative was
(1972) suggests an early split of Platycerium into an African collected of each species including representatives of all four putative
clade and an Asiatic clade, whereas Hennipman and Roos segregates of the P. bifurcatum complex, both variations of P. stemaria
(1982) hypothesize a more complex pattern created by repeated (Beauv.) Desv., and three collections of P. alcicorne Desv. Sequences of other
polygrammoid ferns were predominantly used in previous studies (Schneider et
long-distance dispersal and/or extinction events. So far, the al., 2002, 2004a, c; Janssen and Schneider 2005; Schneider et al., in press), but
phylogeny of staghorn ferns has not been explored using DNA some additional sequences, especially for Pyrrosia, were added in this study.
sequence data, which may provide new evidence in the Voucher specimens were deposited at the herbarium of the University
ongoing controversy. Göttingen (GOET), whereas all new sequence data were submitted to GenBank
The introduction of phylogenetic studies based on cpDNA (see Appendix for voucher information and GenBank accession numbers).
sequence data has strongly improved our understanding of the Genomic DNA was extracted using Invisorb Plant Mini kit (Invitek, Berlin,
evolution of polygrammoid ferns (Schneider et al., 2002, Germany). Sequences of four chloroplast regions, rbcL, rps4, rps4-trnS
intergenic spacer (IGS), and trnL-F IGS, were generated using PCR primers
2004a, c; Ranker et al., 2004). These studies have confirmed and protocols identical to those in previous studies (Haufler et al., 2003;
some of the proposed relationships but challenged other Schneider et al., 2004a, c; Janssen and Schneider, 2005). PCR products were
generally accepted or controversial hypotheses of natural prepared for sequencing with a BigDye Terminator Cylce Sequencing kit,
relationships. Whereas most of these studies focused on global version 3.1 (ABI Prism, Applied Biosystems, Foster Coty, California, USA)
relationships, some data sets were designed to explore the and sequenced using an ABI capillary sequencer 3100 (Applied Biosystems).
relationships in particular lineages/genera of polygrammoid All sequences were assembled and manually aligned using BioEdit (http://
ferns, such as the grammitid fern Adenophorus (Ranker et al., www.mbio.ncsu.edu/BioEdit/bioedit.html), and the final alignment was
adjusted manually in MacClade 4.0 (Maddison and Maddison, 2000).
2003), the ant-fern Lecanopteris (Haufler et al., 2003), the Ambiguously aligned regions were excluded from all analyses. Taxonomically
non-natural genus Polypodium and its relatives (Haufler and informative indels were identified visually.
Ranker, 1995; Haufler et al., 2000; Schneider et al., in press), Maximum parsimony (MP) and maximum likelihood (ML) analyses were
and the drynarioid ferns (Janssen and Schneider, 2005). The performed with PAUP* version 4.0b10 (Swofford, 2000). Model and
study on drynarioids is of particular interest for the study of parameters were selected using the hierarchical likelihood ratio test and the
Platycerium because both lineages were studied with a cladistic Akaike information criterion as implemented in Modeltest (Posada and
approach using morphological evidence and share the Crandall, 1998). MP and ML analyses were calculated using the heuristic
mode with 1000 respectively 100 random-addition-sequence replicates, tree-
occurrence of a single, well-studied sister lineage. The cpDNA bisection-reconnection (TBR) branch swapping to completion, and MULPARS
data supported most of the relationships suggested by on. Bootstrap support values were estimated by calculating 100 or 1000
morphological evidence (Roos, 1985), improved our un- respectively 100 bootstrap replicates, each with 10 random-addition-sequence
derstanding of their evolution based on the clearly defined replicates, TBR branch swapping, and MULPARS on. NeighborNet (Bryant
sister clade, and increased information about the phylogenetic and Moultan, 2004) and split decomposition analyses (Huson, 1998), based on
distance among species (Janssen and Schneider, 2005; H. a distance matrix calculated using the general-time-reversible (GTR) model,
Schneider and T, Janssen, in preparation). The sister lineage of were applied to estimate alternative relationships within Platycerium, indicated
by subset of characters.
Platycerium, the genus Pyrrosia, was studied in a comprehen- The distribution ranges of ancestral nodes were inferred using the software
sive monographic study using morphological evidence DIVA (Ronquist, 1997), which allows us to estimate ancestral distributions by
(Hovenkamp, 1986). The relationships of both genera were taking dispersal and vicariance events into account. In addition, a maximum
assumed based on the shared character of stellate hairs on the parsimony approach was applied to reconstruct ancestral distributions with both
lamina, which was subsequently supported in global phylo- ACCTRAN and DELTRAN optimization of character state changes as
genetic studies of polygrammoids (Schneider et al., 2002, implemented in MacClade 4.0. Likelihood estimates of ancestral area estimates
2004c). were calculated using the one-parameter Markov K-state model (Lewis, 2001)
as implemented in the ancestral state reconstruction packages for Mesquite 9.0
In this study, we explore the phylogeny of Platycerium using (Maddison and Maddison, 2003; website, http://mesquiteproject.org). This
sequences of four cpDNA regions for all currently accepted approach maximizes the probability of arriving at the distribution of terminal
species. Using this evidence, we want to infer the hypotheses taxa, given a stochastic model of evolution (Schluter et al., 1997; Pagel, 1999).
of Hoshizaki (1972) and Hennipman and Roos (1982) as well Information about species distributions was obtained from herbarium speci-
as reconstruct the biogeographic history of this fern lineage. It mens and literature (Hennipman and Roos, 1982; Johns, 1991; Bostock, 1998;
has been proposed that the neotropical species P. andinum is Roux, 2001; Fernandez, 2003). The influence of the hypothetical basal lineage
closely related to the Madagascan P. quadridichotomum of African Pyrrosia, as suggested in a phylogenetic analysis of this genus
(Bonap.) Tardieu (Hoshizaki, 1972; Hennipman and Roos, (Hovenkamp, 1986), was inferred by generating a hypothetical phylogeny in
which such a lineage is placed as the sister to all other lineages within Pyrrosia.
1982; Hoshizaki and Price, 1990; Moran and Smith, 2001), and Patterns of character evolution were reconstructed using MacClade 4.0 by
this putative species pair was included in a list of possible plotting the character matrices of previous studies (Hoshizaki, 1972; Hennip-
vicariance events between South America and Africa (Moran man and Roos, 1982) onto the phylogeny recovered using the combined
and Smith, 2001). However, neither the fossil record (van cpDNA data set.
February 2006] K REIER AND S CHNEIDER —P HYLOGENY OF P LATYCERIUM 219

RESULTS two studied chloroplast spacer regions. None of the clades

showed more than one indel. These signatures allow us to
Phylogeny of Platycerium—Maximum parsimony and recognize the major phylogenetic groups using these two IGS
maximum likelihood analyses of the combined data sets found regions.
the same general topology (Figs. 1, 2). The maximum parsimony
analyses resulted in 36 most parsimonious trees of a length 3729 Phylogenetic biogeography of Platycerium—The phylog-
steps (consistency index [CI] ¼ 04.421, homoplasy index eny suggests an early split into an Afro-Madagascan lineage
[HI] ¼ 0.579, retention index [RI] ¼ 0.727, rescaled consistency and an Australasian lineage (Fig. 3). The first one splits into
index [RC) ¼ 0.364), whereas the maximum likelihood found a tropical African lineage with two species and an East
a single tree with a
ln ¼ 24 041.06 using the GTR model with African-Madagascan lineage with four species of which three
the parameters based on a base-pair distribution of A ¼ 0.3035, are endemic to Madagascar. Platycerium alcicorne is sister to
C ¼ 0.1863, G ¼ 0.2087, T ¼ 0.3017, rate of invariable sites an exclusively Madagascan clade and its occurrence in both
I ¼ 0.3067, and a gamma distribution of C ¼ 1.200. Pyrrosia East Africa and Madagascar may indicate a dispersal from East
was found to be sister to a monophyletic genus Platycerium, Africa to Madagascar. However, likelihood reconstruction
which consists of three lineages with 100% bootstrap support in allows us to suggest an origin of this species in Madagascar
maximum parsimony and maximum likelihood analyses. These with a subsequent expansion of its distribution to East Africa,
clades correspond to the Afro-American group, Javan-Australian the Comoros, and Mascarenes. The ancestor of the neotropical
group, and Malayan-Asian group proposed by Hoshizaki (1972). P. andinum likely occurred in the Afro-Madagascan region, but
All analyses recovered with considerable support the Afro- the exact pattern is ambiguous. The Australasian sister lineage
Madagascan-neotropical clade as sister to the other two clades. of the Afro-Madagascan clade may have originated either in
Analyses of the single-gene data sets did not give any evidence Australia or SE Asia as indicated by likelihood and maximum
for conflicting information among the four cpDNA data sets. parsimony reconstructions of ancestral distributions. The
The Afro-American clade showed an internal split into Javan-Australian lineage may have originated in Australia
a tropical African clade comprising P. elephantotis Schweinf. and expanded its distribution towards Java. The alternative
and P. stemaria (Beauv.) Desv. and a Madagascan clade concept of an origin in Java and migration to Australia cannot
including all species occurring in Madagascar. Sister to three be excluded based on current data. Its sister lineage, the
exclusively Madagascan species (P. ellisii Baker, P. mada- Australasian clade, may have its ancestral distribution either in
gascariense Baker, P. quadridichotomum (Bonap.) Tardieu) is Australia or Indochina þ western Malesia, but likelihood
P. alcicorne Desv., which is the only species occurring in estimates gave slightly higher probabilities for the second
Madagascar and along the coast of East Africa. Samples of P. hypothesis. The ancestors of the clade containing P. suberbum
alcicorne from East Africa and Madagascar possessed 100% and P. wandae were growing in central or western Malesia,
identical trnL-F IGS sequences. The sole New World species, whereas the alternative hypothesis of origin in eastern Malesia
P. andinum Baker, comes out as sister to the Madagascan to Australia is less likely, according to the phylogenetic
subclade but without bootstrap support. Neither the single reconstructions. The clade comprising P. coronarium and P.
cpDNA marker data sets nor the combined data set found ridleyi is more or less restricted to western Malesia þ
evidence for the hypothesis of P. andinum being sister species Philippines. Likewise, P. wallichii occurs in Thailand, Burma,
to P. quadridichotomum. and northern India. The other two putative species pairs may be
The Javan-Australian clade is synonymous to the P. the result of disruptions of their ancestral distribution.
bifurcatum complex, which occurs throughout the Sunda Platycerium grande from the Philippines is sister to P.
Islands and eastern Australia. The four cpDNA markers were holttumii occurring in Indochina, whereas P. superbum from
more or less identical in the four samples that correspond to eastern Australia is sister to P. wandae occurring in New
one representative of each putative species [P. bifurcatum Guinea. In general, the ancestral distributions for many basal
(Cav.) C. Chr., P. hillii T. Moore, P. veitchii (Underw.) C. nodes are poorly resolved.
Chr., P. willinckii T. Moore].
Among the seven species of the Malayan-Asian clade, P. DISCUSSION
coronarium (J. König ex O.F. Müll.) Desv. and P. ridleyi forms
a strongly supported subclade. These two species occur more Hypotheses for Platycerium phylogeny—The Malayan-
or less sympatric in western Malesia (Malay Peninsula, Asian clade of Platycerium is easily recognized by the soral
Sumatra, Java, Borneo), the Philippines, and the southern part patches, of which there are only one or two per frond. In the
of Indochina (only P. coronarium). The remaining five species sister species P. coronarium and P. ridleyi, these patches are
are also grouped in a strongly supported clade, but their localized at specialized lobes of the lamina, whereas in the
internal relationships were poorly resolved. Two species pairs other five species of this lineage these patches are on
were found without bootstrap support, P. grande (Fée) Kunze conspicuously broadened sinuses (Hoshizaki, 1972; Hoshizaki
þ P. holttumii de Jonch. & Hennipman, P. superbum de Jonch. and Price, 1990). Hennipman and Roos (1982) did not accept
& Hennipman þ P. wandae Racib., leaving P. wallichii Hook. the homology of the soral patches and found those taxa nested
as an isolated taxon. in different clades. Their cpDNA data support the homology
The neighborNet and split-decomposition analyses found no assessment of Hoshizaki (1972) by providing evidence for
alternatives to the strongly supported clades but indicate that a putative transformation of the sorus location at broadened
the current data set is insufficient to distinguish between sinuses to specialized lobes of the lamina. The two species with
alternative relationships among the P. wallichii group and the the specialized lobes, P. coronarium and P. ridleyi, were found
clade comprising three species endemic to Madagascar. to be sister species, and their position as sister to the other
Several clades of Platycerium possess clade-specific deletion species with soral patches leaves it unclear if this condition
or insertion with a length of more than two base pairs in the evolved from those patches as found in the P. wallichii clade.
220 A MERICAN J OURNAL OF B OTANY [Vol. 93
February 2006] K REIER AND S CHNEIDER —P HYLOGENY OF P LATYCERIUM 221

Fig. 2. Phylogram of the maximum likelihood (ML) analyses of the combined data set. Only platyceriods are shown. Length of branches corresponds
to the like number of substitutions. Thick branches indicate support of 95% in ML bootstrap analyses. Sketches of fertile leaves illustrate the typical
position of the coenosori (gray areas) in the major lineages of Platycerium. The sketches are based on Hoshizaki (1972). Numbers above or below branches
correspond to putative synapomorphic character states. Unambiguous reconstructions are given above branches, whereas those characters with alternative
optimization of character reconstructions under ACCTRAN and DELTRAN are given below branches. 1 ¼ presence of stellate hairs on the lamina (a); 2 ¼
differentiation of mantle leaves (b, trophosporophyll on the left, mantle leave on the right); 3 ¼ one to two soral patches that are localized relatively basally
on the fertile frond þ dark sclerenchymatous cells in the cortex of the stipe (illustrated c); 4 ¼ root buds present or absent (present in all species of
Platycerium except the members of the clade including P. ridleyi to P. grande); 5 ¼ scales pseudopeltate to peltate; 6 ¼ sporangia located on a stalked lobe
of the lamina þ only eight spores per sporangium þ sporangium reduced þ large scales with flabelloid margin; 7 ¼ large scales without flabelloid margin.

The sister pair P. coronarium and P. ridleyi share several because the character states are too variable in this fern genus
unique morphological characters, such as reduced sporangia (Hoshizaki and Price, 1990). The structure of the stipe cortex
containing only eight spores instead of sixty-four, the shape of and the soral patches indicates the alternative relationships with
the paraphyses, and the ornamentation of the spores members of the Malayan-Asian clade that are strongly
(Hoshizaki, 1972; Hennipman and Roos, 1982). The relation- supported by cpDNA evidence. In contrast to conflicting
ships among the remaining taxa of the Malyan-Asian clade are hypotheses concerning P. wallichii, close relationships among
still not fully discovered. Platycerium wallichii is isolated the remaining four species of this lineage were undisputed,
morphologically by several autapomorphic character states based on the presence of more or less sinuate base frond
such as the lack of marginal scale indument. Reduction of one margin with a conspicuous fringe and the very dense set of
main branch in the foliage frond, horizontally exposed soral hairs on the rhizome scales (Hoshizaki, 1972; Hennipman and
patches, and the long paraphyses rays have been interpreted as Roos, 1982; Hoshizaki and Price, 1990).
character states that suggest relationships to the African P. The Javan-Australian clade is easily recognized by the
elephantotis. These homology assessments were criticized occurrence of subpeltate to peltate rhizome scales. The

‹
Fig. 1. Bootstrap consensus tree obtained in a maximum parsimony analyses of the combined data set. Numbers above branches correspond to
bootstrap values. The three major clades, Afro-American (AA), Javan-Australian (JA), and Malayan-Asian (MA), are indicated with parentheses. The other
abbreviations indicate taxa occurring at least predominantly in: AU ¼ Australia, IN ¼ Indochina, MD ¼Madagascar; ME ¼ Madagascar and east Africa; TA
¼ tropical Africa, NG ¼ New Guinea; NT ¼ neotropics; PH ¼ Philippines; and WM ¼ western Malesia. Genera are abbreviated as follows: Ade. ¼
Adenophorus, Agl. ¼ Aglomorpha, Art. ¼ Arthromeris, Cam. ¼ Campyloneurum, Dic. ¼ Dictymia, Dry. ¼ Drynaria, Gon. ¼ Goniophlebium, Lec. ¼
Lecanopteris, Lem. ¼ Lemmaphyllum, Lpt. ¼ Leptochilus, Lox. ¼ Loxogramme, Mic. ¼ Microsorum, Mgr. ¼ Microgramma, P. ¼ Platycerium, Ple. ¼
Pleopetis, Pol. ¼ Polypodium, Pyr. ¼ Pyrrosia, Sel. ¼ Selliguea, Ter. ¼ Terpsichore; Thy. ¼ Thylacopteris. Taxa marked with a star are composed of
sequences from different species of the same genus. The Latin numbers below the branches indicate the presence of clade-specific insertions-deletions of
more than two base pairs. I ¼ insertion in the trnL-F IGS of TTTATTTAACCCCA; II ¼ insertion in the trnL-F IGS of TTCTCA; III ¼ deletion in the rps4-
trnS IGS of ATTTATGC; IV ¼ deletion in the rps4-trnS IGS of AAAAGACTTAG; V ¼ deletion in the rps4-trnS IGS of CTTGTTCATT; VI ¼ insertion in
the rps4-trnS IGS of CAAATT(GA)A; VII ¼ deletion in the rps4-trnS IGS.
222 A MERICAN J OURNAL OF B OTANY [Vol. 93

Fig. 3. Distribution and phylogeny of the genus Platycerium including reconstructed distribution of each ancestor. Gray areas indicate current
distribution of the genus. Reconstructed ancestral areas are given to each node: AA ¼ Australasian, AM ¼ Afro-Madagascan region, AU ¼ Australia, EA ¼
east Africa, IN ¼ Indochina, IM ¼ Indochina þ Western and Central Malesia; JV ¼ Java, MD ¼ Madagascar, ME ¼ Madagascar þ East Africa, NG ¼ New
Guinea, PH ¼ Philippines, SA ¼ South America, NA ¼ New Guinea þ Australia, NT ¼ neotropics, TA ¼ tropical Africa, WA ¼ western Malesia.
Probability of ancestral areas was calculated using likelihood approaches for nodes 1 to 14, but pie charts are only given for 1, 2, 5, 6, 7, 8, 10, and 12. Two
versions are given for node 7 marks as 7a and 7b. Pie chart node 1, 2, 7a: gray ¼ Afro-Madagascan, black ¼ Australasian. Pie chart node 5: gray ¼ tropical
Africa or neotropics, black ¼ Madagascan. Pie chart node 7b, 8, 11, 12: gray ¼ Indochina, western and central Malesia, Black ¼ Australia and eastern
Malesia.

presented data does not provide further insights into the soral patches are either homoplastic or plesiomorphic. The two
differentiation of this clade because sampling is insufficient tropical African species, P. elephantotis and P. stemaria, are
(only four specimens), and the markers are not variable enough characterized by the occurrence of mixed sporangia (intermin-
to resolve the internal phylogeny of this clade. gling of sporangia of different maturity in the same sporangia
The Afro-American clade is characterized by the presence of patch) and distantly inserted receptacles (0.6–0.9 mm apart), but
well-developed central meristeles in the stipes and by the ability these features occur also in some Asiatic taxa (Hoshisaki, 1972).
to proliferate through root buds, although root buds are also Inference of Platycerium phylogeny using morphological
found in P. bifurcatum s.s. Hoshizaki (1972) regards P. characters is difficult because of the variability of some
andinum as a lineage apart within the Afro-American clade features and the absence of many critical characters in the sister
sharing only plesiomorphic traits with the others, especially P. clade Pyrrosia (Hovenkamp, 1986). Denser sampling is not
quadridichotomum. Our results strongly support the lack of any likely to smooth out this gap because none of the extant species
morphological synapomorphic character state distinguishing P. of Pyrrosia resembles closely the morphology of Platycerium.
andinum and P. quadridichotomum. Characters such as brown, Hovenkamp (1986) stressed out the occurrence of asymmetric
mature base fronds and the position of the fertile fronds and leaves in Pyrrosia africana (Kunze) F. Ballard as a single
February 2006] K REIER AND S CHNEIDER —P HYLOGENY OF P LATYCERIUM 223

character suggesting close relationships between this species conclusions seem to be possible. This region evolved as a result
and the Platycerium lineage, but we did not have access to of the collision between Southeast Asia and the Australian
DNA from this species. Some putatively interesting traits, such croton in the Miocene (Hall, 2002). In a recent study on
as the evolution of CAM, are impossible to investigate based drynarioid ferns, evidence was found for a correlation of the
on the paucity of observations (Ong et al., 1995; Holtum and distribution pattern and species diversity of these ferns with the
Winter, 1999; Zotz and Hietz, 2001). formation of the Malesian region in the Miocene (H. Schneider
A further issue is the limitation of available DNA sequence and T. Janssen, unpublished manuscript). Whereas drynarioid
data to the chloroplast genome, which is likely inherited ferns often grow in wet rainforests dominating the core of
maternally in these ferns (Vogel et al., 1998). These data make Malesia, Platycerium occurs predominantly in seasonally dry
it impossible to detect misleading information that is caused by forests found in Indochina, the Sunda Islands, and eastern
introgression and/or hybrid speciation (Linder and Rieseberg, Australia (Hoshizaki, 1972; Hennipman and Roos, 1982). The
2004). However, the few hybrids reported for these ferns are only phylogenetic patterns may, therefore, reflect the unfolding of
known to arise only in cultivation (Hoshizaki, 1975; Hoshizaki these habitats since the drift of Australia from the southern
and Price, 1990; Hoshizaki and Moran, 2001), and no hybrids Pacific towards SE Asia. Long distance dispersal events may
were ever reported to occur in the wild. This is similar to other have been an important factor, but we also need to consider the
groups of polygrammoid ferns in which nearly all known natural transformation of the vegetation in Malesia during and after the
hybrids are members of the temperate Polypodium s.s. lineage. Pleistocene glaciations. Currently available information is
Based on a study of the epiphytic polygrammoid fern genus insufficient to fully explore these hypotheses.
Pleopeltis in Mexico, Haufler et al. (2000) suggested different In summary, the current distribution of Platycerium appears
speciation mechanisms in Pleopeltis vs. temperate Polypodium. to be created by the combination of long-distance dispersals
Similar results were found for the epiphytic asplenioid fern, and subsequent speciation(s) in Madagascar and Australasia.
Asplenium nidus, in Australasia (Yatabe et al., 2001). This
hypothesis implies that hybridization and introgression are rare LITERATURE CITED
or at least less common in tropical epiphytic ferns.
ABROZOIC-DOLINSEK, J., H. CAMLOH, B. BOHANEC, AND J. ZEL. 2002.
Scenarios of historical biogeography—The reconstruction Apospory in leaf culture of staghorn fern (Platycerium bifurcatum).
of the history of Platycerium is hindered by the absence of Plant Cell Reports 20: 791–796.
fossils allowing us to estimate the distribution of this genus in BENZING, D. H. 1990. Vascular epiphytes: general biology and related
space and time. Inference of divergence times is problematic biota. Cambridge University Press, Cambridge, UK.
for the whole lineage of polygrammoid ferns due to their poor BOSTOCK, P. D., AND T. M. SPOLES. 1998. Polypodiaceae. Flora of
fossil record and equivocal assignment of these few fossils to Australia 48: 468–495.
BOYER, Y. 1964. Contribution á l’étude de l’écophyosiologe de deix fougères
extant lineages (van Uffelen, 1991; Collinson, 2001; Kvacek et épiphytes: Platycerium stemaria (Beauv.) Desv. et P. angolense Welw.
al., 2004). However, the current fossil record of derived ferns Annales des Sciences Naturelles: Botanique 12: 87–228.
and existing divergence time estimates indicate that this genus BRUMME, S., V. KRUFT, U. K. SCHMITZ, AND H. P. BRAUN. 1998. New
did not originate before the Eocene (Schneider et al., 2004b). insights into the co-evolution of cytochrome c reductase and the
The observed split into an Afro-American and Australasian mitochondrial processing peptidase. Journal of Biological Chemistry
lineage is unlikely to be the result of the breakup of Gondwana. 273: 13143–13149.
Similar, the proposed vicariance between P. andinum and other BRYANT, D., AND V. MOULTON. 2004. Neighbor-Net: an agglomerative
members of the Afro-American clade does not fit in with the method for the construction of phylogenetic networks. Molecular
time of the Gondwana break-up. Long-distance dispersal with Biology and Evolution 21: 255–265.
subsequent speciation events is therefore the preferred BURROWS, J. E. 1990. Southern African ferns and fern allies. Frandsen,
explanation for the observed origin of the Southern American Sandton, South Africa.
species from Afro-Madagascan ancestors. A similar problem CAMLOH, M., AND N. GOGALA. 1992. In vitro culture of Platycerium
bifurcatum gametophytes. Scientia Horticulturae (Amsterdam) 51:
arises with the separation of the Afro-Madagascan and
343–346.
Australasian clades of Platycerium. The sister clade, Pyrrosia, CAMLOH, M., B. VILHAR, J. ZEL, AND M. RAVNIKAR. 1999. Jasmonic acid
also occurs in both regions but has a considerably higher stimulates development of rhizoids and shoots in fern leaf culture.
species diversity in Australasia than in the Afro-Madagascan Journal of Plant Physiology 155: 798–801.
region. Unfortunately, current sampling of Pyrrosia is in- COLLINSON, M. 2001. Cainozoic ferns and their distribution. Brittonia 53:
sufficient to test the hypothesis of an African clade as sister to 173–235.
the remaining clades of Pyrrosia, as suggested by Hovenkamp DARNAEDI, D., AND T. N. PRAPTOSUWIRYO. 2003. Platycerium bifurcatum. In
(1986). To test the putative influence of this putative sister W. P. de Winter and V. B. Amoroso [eds.], Plant resources of South-
clade to the remaining species of Pyrossia missing in this East Asia, vol. 15(2), Cryptogams: ferns and fern allies, 157–159.
study, we reconstructed ancestral areas using DIVA, MP, and Backhuys Publishers, Leiden, Netherlands.
ML analyses with and without inclusion of this putative clade. FERNANDEZ, B. 2003. New records for Platycerium andinum Baker in Peru.
However, both analyses failed to provide unambiguous support American Fern Journal 93: 160–163.
for either an Afro-Madagascan or Australasian origin. The GARDETTE, E. 1996. Microhabitats of epiphytic fern communities in large
lowland rain forest plots in Sumatra. In J. M. Camus, M. Gibby, and
pattern recovered within the Afro-Madagascan clade fits best R. J. Johns [eds.], Pteridology in perspective, 655–658. Royal Botanic
with the hypothesis of a colonization of Madagascar by species Gardens, Kew, UK.
occurring in Africa followed by a subsequent speciation HALL, R. 2002. Cenozoic geological and plant tectonic evolution of SE
event(s). The three species endemic to Madagascar occur Asia and the SW Pacific: computer-based reconstructions, models and
mainly in the northern habitats that may have evolved since the animations. Journal of Asian Earth Science 20: 353–431.
Eocene/Oigocene (Wells, 2003). The diversification of these HAUFLER, C. H., W. A. GRAMMER, E. HENNIPMAN, T. A. RANKER, A. R.
ferns in the Australasian region is mysterious, but a few SMITH, AND H. SCHNEIDER. 2003. Systematics of the ant-fern genus
224 A MERICAN J OURNAL OF B OTANY [Vol. 93

Lecanopteris: testing phylogenetic hypotheses with DNA sequences. POREMSKI, S., AND N. BIEDINGER. 2001. Epiphytic ferns for sale: influence
Systematic Botany 28: 217–227. of commercial plant collection on the frequency of Platycerium
HAUFLER, C. H., E. A. HOOPER, AND J. P. THERRIEN. 2000. Modes and stemaria (Polypodiaceae) in coconut plantations on the southeastern
mechanisms of speciation in pteridophytes: implications of contrast- Ivory Coast. Plant Biology (Stuttgart) 3: 72–76.
ing patterns in ferns representing temperate and tropical habitats. POSADA, D, AND K. A. CRANDALL. 1998. Modeltest: testing the model of
Plant Species Biology 15: 223–236. DNA substitution. Bioinformatics 14: 817–818.
HAUFLER, C. H., AND T. A. RANKER. 1995. rbcL sequences provide RANKER, T. A., J. M. O. GEIGER, S. C. KENNEDY, A. R. SMITH, C. H.
phylogenetic insights among sister species of the fern genus HAUFLER, AND B. S. PARRIS. 2003. Molecular phylogenetics and
Polypodium. American Fern Journal 85: 361–374. evolution of the endemic Hawaiian genus Adenophorus (Grammiti-
HENNIPMAN, E., AND M. C. ROOS. 1982. A monograph of the fern genus daceae). Molecular Phylogenetics and Evolution 26: 337–347.
Platycerium (Polypodiaceae). Verhandelingen der Koninklijke Neder- RANKER, T. A., A. R. SMITH, B. S. PARRIS, J. M. O. GEIGER, C. H. HAUFLER,
lands Akademie van Wetenschappen, Afdeeling Natuurkunde, Tweede S. C. K. STRAUB, AND H. SCHNEIDER. 2004. Phylogeny and evolution of
Reeks 80: 1–126. grammitid ferns (Grammitidaceae): a case of rampant morphological
HOLTUM, J. A. M., AND K. WINTER. 1999. Degrees of crassulacean acid homoplasy. Taxon 53: 415–428.
metabolism in tropical epiphytic and lithophytic ferns. Australian RONQUIST, F. 1997. Dispersal-vicariance analysis: a new approach to the
Journal of Plant Physiology 26: 749–757. quantification of historical biogeography. Systematic Biology 46:
HOSHIZAKI, B. J. 1970. Rhizome scales of Platycerium. American Fern 195–203.
Journal 60: 144–160. ROOS, M. C. 1985. Phylogenetic systematics of the Dynarioideae
HOSHIZAKI, B. J. 1972. Morphology phylogeny of Platycerium species. (Polypodiaceae). Verhandelingen der Koninklijke Nederlands Akade-
Biotropica 4: 93–117. mie van Wetenschappen, Afdeling Natuurkunde, Tweede Reeks 85: 1–
HOSHIZAKI, B. J. 1975. A staghorn fern (Platycerium) hybrid. American 138.
Fern Journal 65: 99–101. ROUX, J. P. 2001. Conspectus of southern African pteridophyta. Southern
HOSHIZAKI, B. J., AND R. MORAN. 2001. Fern grower’s manual, 2nd ed. African Botanical Diversity Network Report 13: 1–223.
Timber Press, Portland, Oregon, USA. SCHNEIDER, H., T. JANSSEN, P. HOVENKAMP, A. R. SMITH, R. CRANFILL, C. H.
HOSHIZAKI, B. J., AND M. C. PRICE. 1990. Platycerium update. American HAUFLER, AND T. A. RANKER. 2004a. Phylogenetic relationships of the
Fern Journal 80: 53–69. enigmatic Malesian fern Thylacopteris (Polypodiaceae, Polypodii-
HOVENKAMP, P. H. 1986. A monograph of the fern genus Pyrrosia. Leiden dae). International Journal of Plant Sciences 165: 1077–1087.
Botanical Series 9: 1–310. SCHNEIDER, H., H.-P. KREIER, R. WILSON, AND A. R. SMITH. In press. The
HUSON, D. 1998. SplitsTree: a program for analyzing and visualizing Synammia enigma: evidence for a temperate lineage of polygrammoid
evolutionary data. Bioinformatics 14: 68–73. ferns (Polypodiaceae, Polypodiidae) in southern South America.
JANSSEN, T., AND H. SCHNEIDER. 2005. Exploring the evolution of humus Systematic Botany.
collecting leaves in drynarioid ferns (Polypodiaceae, Polypodiidae) SCHNEIDER, H., E. SCHUETTPELZ, K. M. PRYER, R. CRANFILL, S. MAGALLÓN,
based on phylogenetic evidence. Plant Systematics and Evolution AND R. LUPIA. 2004b. Ferns diversified in the shadow of angiosperms.
252: 175–197. Nature 428: 553–557.
JOHNS, R. J. 1991. Pteridophytes of tropical east Africa, a preliminary SCHNEIDER, H., A. R. SMITH, R. CRANFILL, C. H. HAUFLER, T. A. RANKER,
check-list of the species. Royal Botanical Gardens, Kew, UK. AND T. HILDEBRAND. 2002. Gymnogrammitis dareiformis is a poly-
KORNAS, J. 1979. Distribution and ecology of the pteridophytes in Zambia. grammoid fern (Polypodiaceae)—resolving an apparent conflict
Polska Akademia Nauk Wydzial II Nauk Biologiczynch, Krakow, between morphological and molecular data. Plant Systematics and
Poland. Evolution 234: 121–136.
KWACEK, Y., J. DASKOWAM, AND R. ZETTER. 2004. A re-examination of SCHNEIDER, H., A. R. SMITH, R. CRANFILL, T. J. HILDEBRAND, C. H. HAUFLER,
Cenozoic Polypodium in North America. Review of Palaeobotany AND T. A. RANKER. 2004c. Unravelling the phylogeny of poly-
and Palynology 128: 219–227. grammoid ferns (Polypodiaceae and Grammitidaceae): exploring
KWA, S. H., Y. C. WEE, P. P. KUMAR. 1995. Role of ethylene in the aspects of the diversification of epiphytic plants. Molecular
production of sporophytes from Platycerium coronarium (Koenig) Phylogenetics and Evolution 31: 1041–1063.
Desv. Frond and rhizome pieces cultured in vitro. Journal of Plant SCHLUTER, D., T. PRICE, A. O. MOOERS, AND D. LUDWIG. 1997. Likelihood
Growth Regulation 14: 183–139. of ancestor states in adaptive radiations. Evolution 51: 1699–1711.
LEE, Y. H. 1989. Development of mantle leaves in Platycerium bifurcatum SWOFFORD, D. L. 2000. PAUP*: phylogenetic analysis using parsimony
(Polypodiaceae). Plant Systematics and Evolution 165: 199–209. (*and other methods), version 4. Sinauer, Sunderland, Massachusetts,
LEON, B., AND K. R. YOUNG. 1996. Distribution of pteridophyte diversity and USA.
endemism in Peru. In J. M. Camus, M. Gibby, and R. J. Johns [eds.], VAN UFFELEN, G. 1991. Fossil Polypodiaceae and their spores. Blumea 36:
Pteridology in perspective, 77–91. Royal Botanic Gardens, Kew, UK. 253–272.
LEWIS, P. O. 2001. A likelihood approach to estimating phylogeny from VOGEL, J. C., S. J. RUSSELL, F. J. RUMSEY, J. A. BARRETT, AND M. GIBBY. 1998.
discrete morphological character data. Systematic Biology 50: 913– Evidence for maternal transmission of chloroplast DNA in the genus
925. Asplenium (Aspleniaceae, Pteridophyta). Botanica Acta 111: 247–249.
LINDER, C. R., AND L. H. RIESEBERG. 2004. Reconstructing patterns of reticulate WELLS, N. A. 2003. Some hypotheses on the Mesozoic and Cenozoic
evolution in plants. American Journal of Botany 91: 1700–1708. paleoenvironmental history of Madagascar. In S. Goodman and J. P.
MADDISON, D. R., AND W. P. MADDISON. 2000. MacClade 4.05. Sinauer, Renstead [eds.], Natural history of Madagascar, 16–34. University of
Sunderland, Massachusetts, USA. Chicago Press, Chicago, Illinois, USA.
MORAN, R., AND A. R. SMITH. 2001. Phytogeographic relationships YATABE, Y., S. MASUYAMA, D. DARNAEDI, AND N. MURAKAMI. 2001.
between neotropical and Africa-Madagascan pteridophytes. Brittonia Molecular systematics of the Asplenium nidus complex from Mt.
53: 304–351. Halimum National Park, Indonesia: evidence for reproductive
ONG, B. L., S. L. NG, AND Y. C. WEE. 1995. Photosynthesis in Platycerium isolation among three sympatric rbcL sequence types. American
coronacrium (Koenig ex Mueller) Desv. Photosynthetica 31: 59–69. Journal of Botany 88: 1517–1522.
PAGEL, Y. H. 1989. The maximum likelihood approach to reconstructing ZOTZ, G., AND P. HIETZ. 2001. The physiological ecology of vascular
ancestral character states of discrete characters on phylogenies. epiphytes: current knowledge, open questions. Journal of Experi-
Systematic Biology 48: 612–622. mental Botany 52: 2967–2978.
February 2006] K REIER AND S CHNEIDER —P HYLOGENY OF P LATYCERIUM 225

APPENDIX. Voucher specimens and GenBank accession numbers. A dash indicates that the region was not sampled. Collected material was obtained either
from field collections or more often from cultivated material from one of the following Botanical Gardens: BGG ¼ Old Botanical Garden of the
Univeristy of Göttingen, Germany; BGH ¼ Botanical Garden of the University of Heidelberg, Germany; CAN ¼ Charles Alford Plant Nursery, Vero
Beach, Florida, USA; RBGE ¼ Royal Botanic Garden Edinburgh, UK. Country of origin is given for material from field collections. Voucher
specimens are deposited in one of the following herbaria: DSM ¼ University of Dar es Salaam, E ¼ Royal Botanic Garden Edinburgh, GOET ¼
Herbarium of the University of Göttingen, LPB ¼ Herbario Nacional de Bolivia, Bolivia, UC ¼ Univiersty herbarium, Univeristy of California.

(A) Voucher information for newly generated sequences

Taxon; GenBank accession: rbcL, rps4 þ rps4-trnS, trnL-F; Source; Voucher specimen.
Dictymia mckeei Tindale; DQ164441, DQ164462, DQ164504; Cult. DQ164458, DQ164489, DQ164522; Cult. BGG, Kreier CG0411
RBGE, Coll. Nr. 19842659 (E). (GOET). P. superbum De Jonch. & Hennipman; DQ164459,
Goniophlebium argutum (Hook.) J. Sm.; DQ16442, DQ164473, DQ164490, DQ164523; Cult. BGG, Kreier CG0412 (GOET). P.
DQ164505; Taiwan, Cranfill TW075 (UC). veitchii (Underw.) C. Chr.; DQ164490, DQ164491, DQ164524; Cult.
Loxogramme abyssinica (Baker) M. G. Price; DQ164443, DQ164474, BGG, Kreier CG0413 (GOET). P. wallichii Hooker; DQ164461,
DQ164506; Tanzania, Hemp 3638 (DSM). DQ164492, DQ164525; Cult. CAN, Alfrod s.n. (GOET). P. wandae
Microgramma percussa (Cav.) de la Sota; —, —, DQ164507; Cult. BGG, Racib.; DQ164462, DQ164493, DQ164426; Cult. BGG, Kreier
Schneider s.n. (GOET). Microsorum punctatum (L.) Copel.; CG0414 (GOET). P. willinckii (T. Moore) Domin; DQ164463,
DQ164444, DQ164474, DQ164508; cult. BGH, Schneider s.n. DQ164494, DQ164527; Cult. CAN, Alfrod s.n. (GOET).
(GOET). Pyrrosia assimilis (Baker) Ching; DQ164464, DQ164495, DQ164528;
Platycerium alcicorne Desv.; DQ164445, DQ164476, DQ164509; Cult. Cult. RBGE, Coll. Nr. 728/12, Sino-Amer. Bot. Exped. 805 (E). P.
BGG, Kreier CG0401 (GOET). P. andinum Baker; DQ164446, christii (Giesenh.) Ching; DQ164465, DQ164496, DQ164529; Cult.
DQ164477, DQ164510; Cult. BGG, Kreier CG0402 (GOET). P. BGG, Schneider s.n. (GOET). P. porosa (Persl.) Hovenkamp;
bifurcatum (Cav.) C. Chr.; DQ16447, DQ164478; DQ164511; Cult. DQ164466, DQ164497, DQ164430; Taiwan, Cranfill TW094 (UC).
BGG, Kreier CG0403 (GOET). P. coronarium (J. König ex O. F. P. lanceolata (L.) Farw.; DQ164467, DQ164498, DQ164431; Malay
Müller) Desv.; DQ164448, DQ164479, DQ164512; Cult. Bot. Gard. Penninsula, Cranfill BF10 (UC). P. linaerifolia (Hooker) Ching;
Göttingen, Kreier CG0404 (GOET). P. elephantotis Schweinf.; DQ164468; DQ164499; DQ164532; Taiwan, Cranfill TW101 (UC). P.
DQ164449, DQ164480, DQ164513; Cult. BGG, Kreier CG0405 lingua (Thunbg.) Farw.; —, DQ164500, —; Cult. RBGE, Coll. Nr.
(GOET). P. ellisii Baker; DQ164450, DQ164481, DQ164514; Cult. 19992206 (E). P. longifolia (Burm.) Morton; DQ164469, DQ164501,
BGG, Kreier CG0406 (GOET). P. grande (Fée) Kunze; DQ164451, DQ164533; Cult. BGG, Schneider s.n. (GOET). P. polydactyla
DQ164482, DQ164515; Cult. CAN, Alfrod s.n. (GOET). P. hillii T. (Hance) Ching; DQ164470, DQ164502, DQ164534; Cult. BGG,
Moore; DQ164452, DQ164483, DQ164516; Cult. BGG, Kreier Schneider s.n. (GOET). P. serpens (G. Forst) Ching; DQ164471;
CG0407 (GOET). P. holttumii De Jonch. & Hennipman; DQ164453, DQ164503, DQ164535; Cult. RBGE, Coll. Nr. 19734551 (E).
DQ164484, DQ164517; Cult. CAN, Alford s.n. (GOET). P. Polypodium crystalloneuron Rosenst.; DQ151907, DQ151932,
madagascariense Baker; DQ164454, DQ164485, DQ164518; cult. DQ151959; Bolivia, Jimenez 1176 (LPB). P. fraxinifolium Jacq.;
Bot. Gard. Göttingen, Kreier CG0408 (GOET). P. quadridichotomum DQ151909, DQ151934, DQ151961; Bolivia, Jimenez 1035 (LPB). P.
(Bonap.) Tardieu; DQ164455, DQ164486, DQ164519; Cult. CAN, maritimum Hieron.; DQ151918, DQ151943, DQ151970; Cult. CAN,
Alfrod s.n. (GOET). Alfrod s.n. (UC). P. sessilifolium Desv.; DQ151923, DQ151948,
Platycerium ridleyi Christ; DQ164456; DQ164487; DQ164520; Cult. DQ151975; Bolivia, Jimenez 755 (LPB). P. triseriale Sw.; DQ151926,
BGG, Kreier CG0409 (GOET). DQ151953, DQ151980; Bolivia, Jimenez 1994 (LPB).
P. stemaria (Beauv.) Desv.; DQ164457, DQ164488, DQ164521; Cult. Synammia feuillei (Bertero) Copel.; DQ168808, DQ168813, DQ168819;
BGG, Kreier CG0410 (GOET). P. stemaria var. laurentii De Wild.; Chile, Juan Fernandez Islands, Danton and Perrier s.n. (UC).

(B) Genbank accession numbers of sequences used in previous studies (Haulfer and Ranker, 1995; Schneider et al., 2002, 2004a, c; Haufler et al., 2003;
Ranker et al., 2003, 2004; Janssen and Schneider, 2005)
Taxon; GenBank accession: rbcL, rps4 þ rps4-trnS, trnL-F.
Adenophorus oahuensis (Copel.) L. E. Bishop; AY057382, AY096236, Oleandra Cav. (O. cuminghii J. Sm. þ O. pistillaris (Sw.) C. Chr.);
AF469789. Aglaomorpha acuminata (Willd.) Hovenkamp; AY093597, AY096209, AY093596.
AY529147, AY529172, AY459176. A. coronans (wall. ex Mett.) Pecluma ptilodon (Kunze) M. G. Price; AY362588, AY362661,
Copel.; AF470349, AY459184, AY529463. A. meyeniana Schott; AF159193. Pleopeltis angusta Humb. & Bonpl. ex Willd.;
AY529153, AY459185, AY529470. Arthromeris lehmannii (Mett.) AY362590, AY362664, AF159199. P. polypodioides (L.) E. G.
Ching; AY096216, AY096198, AY459177. Andrews & Windham; AY362592, AY362665, AF159196.
Campyloneurum angustifolium (Sw.) Fée; AF470344, AY362645, Polypodium guttatum Maxon; AY362606, AY362678, AF159195.
AY083648. C. chlorolepis Alston; AF470345, AY362646, P. pellucidum Kaulf; U21149, AY096234, AF159190. P. rosei
AY093647. Maxon; AY362608, AY362680, AF159197. P. vulgare L. s.l. (P.
glycyrrhiza D.C. þ Eaton, P. vulgare L.); U21146, X84137,
Drynaria rigidula (Sw.) Bedd.; AY529166, AY529188, AY529481.
AF159188. Pyrrosia lingua (Thunbg.) Farw.; AF470343; —;
Lecanopteris carnosa (Reinw.) Blume; AF470322, AY362634, AY083646.
AY083625. L. sinuosa (Wall. ex Hook.) Copel.; AF470321, Selliguea enervis (Cav.) Ching; AY096200, AY096218, AY459178. S.
AY096227, AY083624. Lemmaphyllum carnosum C. Presl; feei Bory; AY459179, AY529192, AY459179.
AF470332, AY362631, AY083635. Leptochilus macrophyllus S. heterocarpa (Blume) Blume; AY362619, AY459180, AY459172. S.
(Blume) Noot.; AF470340, AY362639, AY083643. lanceolata (Mett.) E. Fourn.; AY459186, AY459173, AY459181. S.
Microgramma percussa (Cav.) de la Sota; AY362574, AY362648, —. triloba (Houtt.) M. G. Price; AY459174, AY459187, AY459182.
Microsorum lucidum (Roxb.) Copel.; AF470335, AY096230, Terpsichore eggersii (Baker ex Hook.) A. R. Sm.; AF468209, AY362694,
AY083638. M. musifolium (Copel.) Blume; AF470333, AY362636, AF469798. Thylacopteris papillosa (Blume) Kunze ex J. Sm.;
AY083636. AY459175, AY459188, AY459183.