Content uploaded by Stefan Wanke
Author content
All content in this area was uploaded by Stefan Wanke
Content may be subject to copyright.
1
Evolution of the genus Aristolochia
- Systematics, Molecular Evolution and Ecology -
Evolution der Gattung Aristolochia
- Systematik, Molekulare Evolution und Ökologie -
Dissertation
zur
Erlangung des Doktorgrades (Dr. rer. nat.)
der
Mathematisch-Naturwissenschaftlichen Fakultät
der
Technischen Universität Dresden
vorgelegt von
Stefan J. U. Wanke
aus
Düren
Dresden 2006
2
Angefertigt mit Genehmigung der Mathematisch-Naturwissenschaftlichen Fakultät der
Technischen Universität Dresden, in englischer Sprache.
1. Gutachter: Prof. Dr. Christoph Neinhuis, Dresden
2. Gutachter: PD Dr. Thomas Borsch, Bonn
3. Gutachter: Prof. Dr. Thomas Speck, Freiburg
3
to my parents
to my wife
[...] the processes involved in ‚descent with modification’, to use
Darwin’s classic phrase, can be shown clearly to apply to
differentiation within species, as well as to the further
divergence of species [...] once they have become separated
from each other.
G.L. Stebbins (1950)
Variation and evolution in plants. New York: Columbia University Press
4
Table of Contents
Acknowledgements 7
Chapter 1 Introduction 9
Chapter 2 Evolution of Piperales – matK gene and trnK intron
sequence data reveal lineage specific resolution contrast. 11
Abstract 12
Introduction 13
Material & Methods 19
Results 32
Discussion 41
Acknowledgments 50
Chapter 3 Systematics of the genus Aristolochia (Aristolochiaceae)
3.1 Systematics of pipevines – Combining morphological and
fast-evolving molecular characters to investigate the
relationships within subfamily Aristolochioideae
(Aristolochiaceae). 51
Abstract 52
Introduction 53
Material & Methods 56
Results 63
Discussion 69
Acknowledgments 72
Appendix 73
5
3.2 Colonisation, phylogeography and evolution of endemism in
Mediterranean Aristolochia (Aristolochiaceae). 76
Abstract 77
Introduction 78
Material & Methods 82
Results 87
Discussion 94
Acknowledgments 106
Appendix 107
Chapter 4 Molecular Evolution
4.1 Universal primers for a large cryptically simple cpDNA
microsatellite region in Aristolochia. 108
Abstract 109
Acknowledgments 112
Chapter 5 Pollination biology
5.1 Trapped! Pollination of Aristolochia pallida Willd. in the
Mediterranean 114
Abstract 115
Introduction 116
Material & Methods 119
Results 120
Discussion 125
Acknowledgments 130
6
References 131
Curiculum Vitae 153
Publications
Peer-reviewed articles; printed, in press or submitted 155
Abstracts 158
Erklärung 159
7
Acknowledgements
This PhD thesis wouldn’t have been possible without the support, fruitful discussions
and help of numerous colleagues, cooperation partners (especially co-authors of the
publications) and research group members. Above all I would like to express my
gratitude to Prof. Dr. Ch. Neinhuis for financial support of: I) the lab work and II) many
trips to foreign countries during the last 3 years. He was always available for
discussions and raised new ideas, besides he greatly helped to develop my selfhood
and my work.
I’m indebted to Dr. Dietmar Quandt, for his encouragement and support during my
time in the working group. In addition I would like to thank Prof. Dr. T. Speck and PD
Dr. T. Borsch for taking over the co-referees of my thesis.
In detail, I would like to thank my co-authors of the present thesis in chronological
order of the chapters: Alejandra Jaramillo (University of Missouri-Columbia), Thomas
Borsch (University of Bonn), Marie-Stephanie Samain (University of Gent), Dietmar
Quandt (University of Dresden), Christoph Neinhuis (University of Dresden), Favio
Gonzalez (University of Columbia), Hafez Mahfoud (University of Dresden), Steffi
Prieskorn (University of Dresden), Kristin Petzold (University of Dresden), Björn Rulik
(Tierkundemuseum Dresden), and Matthias Nuss (Tierkundemuseum Dresden).
I’m also grateful to the following colleagues not listed above, for past, present, and
ongoing joint projects, which have resulted in several publications and will result in
several others (in alphabetical order): Mario Blanco, Andrea Costa, Holger deGroot,
Paul Goetghebeur, Stanislav Gorb, Harry Horner, Guido Mathieu, Hiroko Murata, Jin
Murata, Kai Müller, Tetsuo Ohi-Toma, Birgit Oelschlägel, Thomas Speck, Takashi
Sugawara, Liesbeth Vanderschaeve, Pablo Vargas, Susann Wicke and the research
group Plant Phylogenetics and Phylogenomics at the Institute of Botany, TU Dresden.
8
In earlier times Prof. Dr. Miguel Vences showed me unintended how big science is
working, which is gratefully acknowledged.
Most of the projects listed here (past and ongoing) were not possible, without financial
support from Erlus AG via Christoph Neinhuis, the European Community Program
SYNTHESYS, Deutscher Akademischer Austauschdienst (DAAD), and Syrian
Ministry of Academy and Agriculture. Beside this the coauthors of joint projects
obtained financial support from various other funding organisations.
Making available and maintaining the living plant collection in the Botanical Garden
Dresden, I would like to thank the staff of the garden and in particular the custodian
Dr. Barbara Ditsch.
Supply of plant material from various colleagues and the Botanical Gardens Dresden,
Berlin and Bonn is gratefully acknowledged.
Last but not least, I’m indebted to my parents, for all kind of support during the last 30
years, to my wife who accepted being away most of the last 3 years and who always
encouraged me to go ahead with research, as well as to my grandparents and my
parents-in-law.
9
Chapter 1 Introduction
Systematics and evolution are linked to each other and it is nearly impossible to
investigate one of these topics without looking at the other. In addition the study of
those two fields is not possible without understanding and studying related research
fields, such as e.g. ecology. This is the reason, why the present thesis is not only
consisting of one research topic, but also includes related fields like pollination
biology or molecular evolution.
The family Aristolochiaceae, as currently circumscribed, consists of four genera,
belonging to two subfamilies (Aristolochioideae and Asaroideae) each with two
genera (Neinhuis et al. 2005, Wanke et al. 2006b, 2006c). Saruma (a monotypic
genus) and the small genus Asarum, ~85 spp. are merged in Asaroideae. The
systematics and phylogeny of Asarum has been comprehensively studied by (Kelly
1998). Systematic and evolutionary problems have been found only on the population
and species level of closely related species within one section (Yamijdi et al., in
press). Basically, this is due to hybridisation of this populations and species. The
other subfamily Aristolochioideae consists of the genus Thottea (~30 spp., from
southeast Asia), which is only poorly studied and the species rich genus Aristolochia
(~500 spp., see Chapter 3). Besides the evolution of the genus Thottea more
questions still need to be resolved. These problems concern the monophyly of
Aristolochiaceae, since Hydnoraceae, a parasitic family, and Lactoridaceae (a
monotypic family from the Juan Fenadez Islands) cause the paraphyly of
Aristolochiaceae (Nickrent et al. 2002, Wanke et al. 2006c). These questions are
beyond the scope of the present thesis, but are currently under investigation or will be
in due course.
Although the issue of big genera is a hot topic in systematics, attempts to study these
groups comprehensively are rare, mainly because a detailed knowledge on the
specific group is an essential precondition which is nearly impossible to achieve by
only one scientist. It is generally recognized (e.g. Frodin 2004) that botanists should
10
not consider these big genera as “black holes” but should focus on their resolution.
Large genera pose high taxonomic challenges as well as unparalleled opportunities to
study phenomena such as character evolution, changes in evolutionary diversification
rates, adaptive radiations, rapid speciation, key innovations and chromosomal
rearrangement (Berry et al. 2005).
The present study deals with the species rich genus Aristolochia and tries to resolve
the relationships within the genus (Chapter 3). Beside this, a kind of case study on
molecular evolution, ecology, and biogeography has been performed on some
specific topics raised during the investigation of the main clades (Chapter 3, 4, 5).
These case studies will lead to further and more detailed investigations, and will be
applicable to similar problems in other clades or geographical areas, as well as open
the possibility to look at a specific topic from a different viewpoint.
The presented thesis starts with a detailed circumscription of Aristolochiaceae and its
relatives of the order Piperales (Chapter 2) followed by a more intensive investigation
of the subfamily Aristolochioideae (Chapter 3). Finally, a detailed study of the
subgenus Aristolochia focusing especially on the Old World representatives is
presented. Chapters 3 & 4 report on findings, which raised several new questions and
ideas for further studies. Each subchapter has its own introduction and abstract
resulting in a short generall introduction here, to avoid too much redundant
information.
11
Chapter 2 Evolution of Piperales – matK gene
and trnK intron sequence data reveal lineage
specific resolution contrast
This study has been published as:
Wanke, S., Jaramillo, MA., Borsch, T., Samain, MS., Quandt, D.
Neinhuis, C., 2006. Molecular Phylogenetics and Evolution
doi.org/10.1016/j.ympev.2006.07.007.
12
Abstract
Piperales are one of the largest basal angiosperm orders with a nearly worldwide
distribution. The order includes three species rich genera, Piper (ca. 1,000 species),
Peperomia (ca. 1,500-1,700 species), and Aristolochia s. l. (ca. 500 species).
Sequences of the matK gene and the non-coding trnK group II intron are analysed for
a dense set of 105 taxa representing all families (except Hydnoraceae) and all
generic segregates (except Euglypha within Aristolochiaceae) of Piperales. A large
number of highly informative indels are found in the Piperales trnK/matK dataset.
Within a narrow region approximately 500 nt downstream in the matK coding region
(CDS), a length variable simple sequence repeat (SSR) expansion segment occurs,
in which insertions and deletions have led to short frame-shifts. These are corrected
shortly afterwards, resulting in a maximum of 6 amino acids being affected.
Furthermore, additional non-functional matK copies were found in Zippelia
begoniifolia, which can easily be discriminated from the functional open reading frame
(ORF). The trnK/matK sequence data fully resolve relationships within Peperomia,
whereas they are not effective within Piper. The resolution contrast is correlated with
the rate heterogenity between those lineages. Parsimony, Bayesian and likelihood
analyses result in virtually the same topology, and converge on the monophyly of
Piperaceae and Saururaceae. Lactoris gains high support as sister to
Aristolochiaceae subf. Aristolochioideae, but the different tree inference methods
yield conflicting results with respect to the relationships of subfam. Asaroideae. In
Piperaceae, a clade formed by the monotypic genus Zippelia and the small genus
Manekia (=Sarcorhachis) is sister to the two large genera Piper and Peperomia.
13
Introduction
The order Piperales is one of the most species rich clades among basal angiosperms,
comprising about 3,300 species, with three genera that include more than 500
species so-called “big genera” (Frodin, 2004,). Nearly all types of growth and life
forms are represented, such as geophytes, herbs, succulents, lianas, shrubs, trees,
parasites and epiphytes. Members of Piperales exhibit a diverse spectrum of
specializations in floral morphology and pollination. On the one hand,
Aristolochiaceae attract insects with their highly specialized flowers, on the other
hand Piperaceae and Saururaceae possess perianthless reduced flowers that are
pollinated by flies and bees (Semple, 1974; de Figuereido and Sazima, 2000;
Marquis, 1988; Bornstein, 1989). Therefore, Piperales are an important lineage for
understanding early angiosperm diversification. Piperales also comprise some
economically important plants like Piper nigrum (black pepper) used as a spice and
several Peperomia species that are widely used as ornamental plants. In addition,
secondary metabolites of Aristolochiaceae such as aristolochic acids are important
compounds in pharmacology (e.g. Nortier et al., 2000).
Molecular data provide evidence for a sister relationship between Piperales and
Canellales (e.g. Qiu et al., 1999; Savolainen et al., 2000; Qiu et al., 2005; Zanis et al.,
2002; Hilu et al., 2003; Borsch et al., 2003; Löhne and Borsch, 2005; Kim et al., 2004;
Borsch et al., 2005). The close relationship of all four orders, Piperales, Canellales,
Magnoliales and Laurales to form the magnoliid clade is meanwhile well supported by
substitution based tree inferences (Borsch et al., 2005; Borsch et al., 2003; Qiu et al.,
2005; Zanis et al., 2002) and chloroplast genome microstructural changes (Löhne and
Borsch, 2005). Phylogenetic analyses with a dense taxon sampling in Magnoliales
(Sauquet et al., 2003), Laurales (Renner, 2005; Renner and Chanderbali, 2000;
Renner, 1999) and Canellales (Karol et al., 2000; Suh et al., 1993) have been
published, whereas Piperales lack a thorough molecular analysis.
14
Characterisation and utility of the trnK intron including the matK gene
The matK coding region has a length of 1500-1600 bp in most angiosperms, and is
located within domain V of the trnK UUU group II intron (Neuhaus and Link, 1987; Ems
et al., 1995; Fig. 1). Based on structural similarities, this ORF has been suggested to
code for a maturase (Neuhaus and Link, 1987; Mohr et al., 1993), and in fact matK is
the only maturase of higher plant plastids (Vogel et al., 1997). Transcription
experiments have shown that trnK including matK are co-transcribed (Chieba et al.,
1996) and there is accumulating evidence for the expression of the gene (Du Jardin
et al., 1994; Barthet and Hilu, pers. comm.). Sequence variation is considerable,
however, the reading frame of matK has been found intact even in extremely fast
evolving lineages, such as Lentibulariaceae (Müller and Borsch, 2005a). Frameshift
mutations in matK have been reported from Poaceae (Hilu and Alice, 1999) and
Lentibulariaceae (Müller and Borsch, 2005a), although only near the 3’ end of the
CDS where they apparently have minimal impact on function. Moreover, small
inversions (2-4 codons) have been encountered in Amaranthaceae (Müller and
Borsch, 2005b), again with minimal impact on the amino acid composition of the
gene.
Figure 1. The trnK/matK region. Coding regions are represented by enlarged black boxes, highly
length-variable regions by small black boxes (H1-H7). Location of the trnK intron domains (DI – DVI) as
well as domain X (DX) in matK are indicated. For further reference of the internal primers see Table 2.
Size and position of length variable regions change with study group. Length of the region is presented
proportional based on the situation found in Aristolochia reticulata.
A peculiarity of matK are substitution rates in first and second codon positions
approaching those in the third (Hilu and Liang, 1997), which contribute to the high
overall evolutionary rate of matK in contrast to other chloroplast genes. MatK seems
15
to be single copy in the vast majority of plants, although some additional copies of
pseudogenic nature have been discovered in Valerianaceae (Hidalgo et al., 2004;
Bell et al., 2001), Nepenthaceae (Meimberg et al., 2006), as well as in some
bryophyte lineages (Jankowiak et al., 2004). In all these cases, pseudogenic copies
were easily identified and grouped together in phylogenetic reconstructions either
based on distances or characters.
The matK gene has become one of the most frequently used chloroplast gene
markers in angiosperm phylogenetic studies. Since matK can easily be co-amplified
with the flanking non-coding intron parts, the complete trnK intron is increasingly
used, expanding the dataset to 2400-2700 bp. As a consequence, the utility of this
region could be extended to the inter- and intra-species level (e.g., Müller and Borsch,
2005a; Wanke et al., 2006b, Wanke et.al., 2006b).
Circumscription of Piperales
Piperales as considered here include the families Piperaceae, Aristolochiaceae,
Saururaceae and Lactoridaceae as well as parasitic Hydnoraceae (Nickrent et al.,
2002; APG, 2003). However, Aristolochiaceae and Lactoridaceae were sometimes
placed in their respective own orders Aristolochiales and Lactoridales (Takhtajan,
1992, 1997). Hydnoraceae had been placed with other parasitic plants in Rafflesiales
(Cronquist, 1988). Monophyly of Piperales is strongly supported by sequence (e.g.
Chase et al., 1993; Graham and Olmstead, 2000; Mathews and Donoghue, 2000;
Soltis et al., 2000; Borsch et al 2003; Hilu et al 2003; Jaramillo et al., 2004; Löhne
and Borsch, 2005; Neinhuis et al., 2005; Borsch et al., 2005). A set of phenotypic
characters has been suggested as synapomorphic for Piperales such as two-ranked
leaves, sheathing leaf base, nuclear endosperm, a single adaxial prophyll, swollen
nodes, distinct vascular bundles, wood with broad rays, vessel elements with simple
perforation and secondary metabolism products like alkaloids from benzyl-isoquinolin
and aporphine type (Doyle and Endress, 2000). All Piperales lineages show strong
trends towards reduction and fusion of flower organs, with the genus Peperomia
being considered to have the most reduced flowers in Piperales (Jaramillo et al.,
2004).
16
Major lineages within Piperales
According to Jaramillo et al. (2004), Piperaceae comprise four genera, the large
genera Peperomia Ruiz and Pavon (about 1700 species; Wanke et al., 2006d) and
Piper L. (more than 1000 species; Jaramillo and Manos 2001) constituting the core of
Piperaceae, and the small genera Manekia Trel. (considered as an earlier name of
Sarcorhachis, Arias et al., in press), and Zippelia Blume. The distinctiveness of
Peperomia has long been recognized, either as subfamily Peperomioideae (Thorne
1992), or, alternatively, as a separate family Peperomiaceae (e.g. Burger, 1977). The
most detailed study on Peperomia was primarily based on fruit morphology by
Dahlstedt (1900), dividing Peperomia into nine subgenera and seven sections.
However trnK/matK data only support the monophyly: of subgenera Micropiper,
Sphaerocarpidium (Wanke et al., 2006d). Monophyly of the genus has been
substantiated by molecular approaches (Neinhuis et al., 2005, Jaramillo et al., 2004),
flower morphology (Jaramillo et al., 2004) and pollen ultrastructure (Mathew and
Mathew, 2001). Piper species are mostly shrubs, trees or lianas. Several generic
names are included as synonyms within Piper, e.g. Arctottonia Trel., Macropiper Miq.,
Pothomorphe Miq., Ottonia Spreng. (Jaramillo and Manos 2001), and Trianaeopiper
Trel. (Jaramillo and Callejas 2004). Analysis of ITS sequence data provides support
for three major clades within Piper, which have diversified in Asia, the South Pacific
and the Neotropics, respectively. Zippelia has been included in Piper or has been
regarded as independent genus within Piperaceae or even Saururaceae. Recently,
Zippelia together with Manekia has been proposed to be distantly related to Piper
(Jaramillo and Manos, 2001), or to form a clade which is sister group to Piper and
Peperomia (Jaramillo et al., 2004).
Saururaceae contain six species in four genera with an East Asian – North American
disjunction: Saururus, Gymnotheca, Anemopsis and Houttuynia (Liang, 1995).
Several studies have been conducted to reveal phylogenetic relationships in this
small family. Most of these studies dealt with ontogenetic/morphological data as
characters (e g. Tucker et al., 1993; Liang and Tucker, 1995; Liang and Tucker, 1990;
17
Tucker, 1975; Tucker, 1981; Meng et al., 2003). Two hypotheses on relationships
were formulated: 1. Anemopsis + Houttuynia and Saururus + Gymnotheca; 2.
Saururus branching first, followed by Gymnotheca and Anemopsis + Houttuynia. This
is also substantiated by chloroplast sequence data (Meng et al., 2003; Jaramillo et al.,
2004; Neinhuis et al., 2005), but contrasting with the signal obtained form the nuclear
genome, which placed Anemopsis as sister to all other Saururaceae (Meng et al.,
2003).
Aristolochiaceae are distributed worldwide. and are generally divided into two
subfamilies, Asaroideae and Aristolochioideae (Huber, 1993). Asaroideae comprise
about 85 species and occur mainly in northern temperate regions with a centre of
diversity in Asia (Kelly, 1998), whereas Aristolochioideae have ca. 500 species and
occur predominantly in tropical-subtropical regions (Ma, 1990). The most recent
classification for Asaroideae is based on Kelly (1997, 1998) who recognises two
genera (Asarum and Saruma) with several segregates at the subgenus or section
level in Asarum. González and Stevenson (2002) provide a detailed discussion of the
various systematic treatments of Aristolochiaceae. They recognize five genera within
Aristolochioideae: Thottea, Isotrema, Endodeca, Pararistolochia and Aristolochia.
Holostylis (= Aristolochia holostylis (Duchartre) F.Gonzalez) and Euglypha have been
included in a broadly circumscribed genus Aristolochia. Similarly, the most
comprehensive analysis of morphological characters of Aristolochiaceae by Kelly and
González (2003) recognizes seven genera: Asarum and Saruma in Asaroideae, in
addition to the five in Aristolochioideae. Studies on Aristolochiaceae based on partial
matK sequences (Murata et al., 2001) examined only four recognized genera. Other
molecular inference (trnL-F) (Neinhuis et al. 2005) is largely congruent with the
results of González and Stevenson (2002) and Kelly and González (2003). However,
discrepancy exists among relationships between segregates within Aristolochia.
Lactoris fernandeziana, the only species in Lactoridaceae, is confined to the Juan
Fernandez Islands, Chile. Its systematic position has been controversial among
members of basal angiosperms (for a review see Stuessy et al., 1998; González and
Rudall, 2001) partly caused by a difficult interpretation of convergent morphological
18
characters e.g. pollen with a saccus (apart from Lactoris only known from
gymnosperms; Carlquist, 1964; Zavada and Taylor, 1986). Molecular data also
support the position within Piperales (Qiu et al., 2005; Hilu et al., 2003; Borsch et al.,
2003; Borsch et al., 2005), or even within Aristolochiaceae (Borsch et al., 2005;
Neinhuis et al., 2005) making the family paraphyletic, or close to Aristolochiaceae and
Hydnoraceae (Nickrent et al. 2002).
This study aims to: A) characterize patterns of molecular evolution for the trnK/matK
region in Piperales, using a dense taxon sampling, B) investigate patterns of
microstructural changes within the trnK intron and the matK gene, C) resolve major
clades within Piperales, investigate the relationships between small and large genera
(Asarum/Saruma versus Aristolochioideae, Manekia/Zippelia versus
Piper/Peperomia).
19
Material and methods
Sampling strategy
A total of 105 accessions of Piperales are sampled, including a representative
number of species/taxonomic groups for each genus in the order. Hydnoraceae
(Hydnora and Prosopanche) have not been sampled as they lack the matK region in
the cp genome (Nickrent, pers. comm.). In Piperaceae, 30 taxa are selected from the
genus Piper (following the sampling of Jaramillo and Manos 2001)), 27 taxa of
Peperomia (following the sampling of Wanke et al. (2006d), representing many
subgenera recognised by Dahlstedt (1900)), as well as the monotypic genus Zippelia
and two species of the genus Manekia. All four genera with in total five species of
Saururaceae are sampled. Aristolochiaceae sampling includes 36 taxa from both
subfamilies, representing all genera of Aristolochioideae accepted by González and
Stevenson (2002) and Neinhuis et al. (2005) and a selection of Asaroideae.
Lactoridaceae is represented by its only species Lactoris fernandeziana. Three
genera of Canellales are used as outgroup.
Plant material has either been collected in the field or has derived from plants
cultivated in botanical gardens (collections of the Botanical Gardens Bonn and
Dresden, Germany) as well as herbarium specimens. A list of the sampled species,
along with collection localities, vouchers and GenBank accession numbers is
provided in Table 1. For generic segregates, of which the taxonomy is until know
unclear, the most common name is used, e.g. Macropiper within Piper (Smith, 1975).
20
Table 1. Taxa used in the present study, including information about the correspondence to clades (for outgroup, family affiliation is given; for
Aristolochia, Piper, Peperomia affiliation correspond to different taxonomic hierarchical ranks, depending on author), the origin of the studied
material (field or collection), voucher information and the herbarium where the voucher is deposited, as well as GenBank accession numbers
(incl. source), are given. GenBank accessions indicated with * are updated with the trnK intron from previously published data, based on matK
gene only (*1 Hilu et al. 2003, *2 Müller et al. 2006).
Taxon Corresponding
to clade
Origin /
Garden accession
Voucher (Herbarium) GenBank accession
(incl. source)
Aristolochiaceae Adans.
Aristolochia L.
A. acuminata Lamk. Aristolochia s.str. BG Bonn, 17417 Wanke & Neinhuis 146 (DR) DQ 532063
A. albida DUCH. Aristolochia s.str. BG Bonn, 17419 Neinhuis 92 (DR) DQ 532064
A. arborea LINDEN Isotrema BG Bonn, 02560 Neinhuis 93(DR) DQ 532044
A. baetica L. Aristolochia s.str. BG Bonn, 14517 Neinhuis 95 (DR) DQ 882189
A. bracteolata LAM. Aristolochia s.str. BG Bonn, 16714 Neinhuis 94 (DR) DQ 532059
A. californica TORR. Isotrema BG Dresden, s.n. Wanke & Neinhuis 143 (DR) DQ 532039
A. chilensis BRIDGES “Howardia” BG Bonn, 12878 Neinhuis 97 (DR) DQ 882185
A. cf. cordiflora MUTIS EX KUNTH “Howardia” BG Lankester s.n. Holst 8602 DQ 532056
A. cruenta BARRINGER “Howardia” Costa Rica: La Selva Blanco 767 (USJ) DQ 882186
A. erecta L. Einomeia Travis, Texas, USA privat coll. Westlund DQ 882188
A. eriantha MART & ZUCC. “Howardia” BG Bonn, 12952 Neinhuis 99 (DR) DQ 532054
A. gigantea MART. & ZUCC. “Howardia” BG Bonn, 02099 Neinhuis 101 (DR) DQ 882187
A. gorgona M.A.BLANCO “Howardia” Heredia: Puerto Viejo
de Sarapiquí, Costa
Rica
Blanco 1752 (USJ) DQ 532051
A. holostylis (DUCHARTRE) F.
GONZALEZ
Holostylis BG Bonn, 02193 Neinhuis 116 (DR) DQ 532057
A. labiata WILD. “Howardia” BG Bonn, 09867 Neinhuis 96 (DR) DQ 532055
A. macrophylla LAM. Isotrema BG Dresden, s.n. Neinhuis s.n. (DR) DQ 882193
A. manshuriensis KOMAROV Isotrema BG Bonn 13085 Neinhuis 104 (DR) DQ 532040
A. micranthaDUCH. Einomeia priv. coll. B.Westlund
Texas, USA
Neinhuis 103 (DR) DQ 532046
A. parvifolia SIBTH. & SM. Aristolochia s.str. BG Bonn, 15274 Neinhuis 105 (DR) DQ 882190
A. pentandra JACQ. Einomeia Cola de Caballo,
Mexico
privat coll. B.Westlund DQ 532045
A. pichinchensis PFEIFER “Howardia” Rio Paleque, Prov.
Los Rios, Ecuador
Moran 6928 (QCA, NY) DQ 532050
21
A. pistolochia L. Aristolochia s.str. France, Cassis,
Calenque d'En Veau
Wanke 037 (DR 025372) DQ 296652
A. promissa MAST. Pararistolochia BG Bonn, 13014 Neinhuis 118 (DR) DQ 532065
A. reticulata NUTT. Endodeca priv. coll. B.Westlund
Texas, USA
Neinhuis 108 (DR) DQ 532037
A. salvadorensis LINDEN Isotrema BG Bonn, 10720 Neinhuis 109 (DR) DQ 882191
A. serpentaria L. Endodeca priv. coll. B.Westlund
Texas, USA
Neinhuis 112 (DR) DQ 532038
A. tomentosa SIMS. Isotrema BG Bonn, 02682 Neinhuis 113 (DR) DQ 532041
A. westlandii HEMSL. Isotrema BG Bonn, 14211 Neinhuis 115 (DR) DQ 882192
A. triactina HOOK. f. Pararistolochia BG Bonn, 12767 Neinhuis 119 (DR) DQ 532066
Thottea ROTTB.
T. corymbosa (GRIFF.) HOU Thottea Malaysia Weber & Anthonysamy
870519-1/1 (WU)
DQ 532036
T. siliquosa (LAM.) HOU Thottea BG Bonn,
09037;India, Kerala
(Bogner 86-3421)
Neinhuis 121 (DR) DQ 532035
T. dependens (PLANCH)
KLOTZSCH
Thottea field origin A. Weber (WU) DQ 882194
Asarum L.
A. caudatum LINDL. Asarum Oregon, Mt Hood Neinhuis 88 (DR) DQ 532034
A. chingchengense CHENG
&YANG
Asarum BG Bonn, 02680 Neinhuis 90 (DR) DQ 882196
A. yakusimense MASAM. Asarum BG Bonn, 14276 Neinhuis 91 (DR) DQ 882197 *1
Saruma OLIVER
S. henryi OLIV. Saruma BG Bonn, 02618 Borsch 3456 (BONN) DQ 532033
Lactoridaceae ENGLER
L. fernandeziana PHIL. Lactoris Juan Fernandez
Island, Chile
Crawford & Stuessy 11950 DQ 882195
Saururaceae RICH.
A. californica (NUTT.) HOOK &
ARN.
Anemopsis BG Bonn, 06422 Wanke 002 (DR) DQ 882198
22
G. chinensis DECNE. Gymnotheca BG Bonn, 17072 Wanke 004 (DR) DQ 882199
H. cordata THUNB. Houttuynia BG Bonn, 08120 Borsch 3481 (BONN) DQ 212712, Wanke et al.
(2006d)
S. cernuus L. Saururus Florida, USA Borsch & Wilde 3108 (VPI,
FR)
DQ 882200 *2
S. chinensis (LOUR.) BAILL. Saururus BG Bonn, 00312 Wanke 001 (DR) DQ 212713, Wanke et al.
(2006d)
Piperaceae C. AGARDH
Peperomia RUIZ & PAVON
P. argyreia (MIQ.) MORR. Tildenia BG Berlin, 062-56-74-
83
GH 13462 (B) DQ 212734, Wanke et al.
(2006d)
P. bicolor SODIRO Sphaerocarpidium BG Berlin, 107-84-74-
83
GH 13436 (B) DQ 212761, Wanke et al.
(2006d)
P. blanda (JACQ.) KUNTH Sphaerocarpidium BG Berlin, 054-24-74-
73
GH 13453 (B) DQ 212763, Wanke et al.
(2006d)
P. clusiifolia (Jacq.) Hook. Rhynchophorum BG Berlin, 062-58-74-
83
Schwerdtfeger GH13433 (B) DQ 212753, Wanke et al.
(2006d)
P. cuspidilimba C.DC. Micropiper BG Berlin, 054-48-74-
83
GH 23129 (B) DQ 212733, Wanke et al.
(2006d)
P. fagerlindii YUNCK. Rhynchophorum BG Berlin, 107-73-74-
83
GH 13431 (B) DQ 212742, Wanke et al.
(2006d)
P. fraseri C.DC. Panicularia BG Berlin, 285-64-89-
80
GH 27028 (B) DQ 212719, Wanke et al.
(2006d)
P. glabella (SW.) A. DIETR. Sphaerocarpidium BG Bonn, 18749 Wanke 061 (DR) DQ 212757, Wanke et al.
(2006d)
P. gracillima S. WATSON Geophila BG Bonn, 06005 Wanke 060 (DR) DQ 212716, Wanke et al.
(2006d)
P. graveolens RAUH &
BARTHLOTT
Panicularia El Oro, Ecuador Rauh & Barthlott 35122
(HEID)
DQ 212722, Wanke et al.
(2006d)
P. hylophila C.DC. Sphaerocarpidium BG Berlin, 173-23-95-
33
Horich (B) DQ 212758, Wanke et al.
(2006d)
P. inaequalifolia RUIZ & PAV. Sphaerocarpidium BG Berlin, 213-35-00-
80
GH 39585 (B) DQ 212749, Wanke et al.
(2006d)
P. macrostachya (VAHL) A.
DIETR.
Rhynchophorum BG Berlin, 039-82-89-
23
Leuenberger GH 26073 (B) DQ 212744, Wanke et al.
(2006d)
P. magnoliifolia (JACQ.) A. Rhynchophorum BG Berlin, 039-34-89- Leuenberger & Hagemann GH DQ 212752, Wanke et al.
23
DIETR. 20 26231 (B) (2006d)
P. marmorata HOOK. f. Tildenia BG Bonn, 17527 Wanke 064 (DR) DQ 212725, Wanke et al.
(2006d)
P. maypurensis KUNTH Tildenia BG Bonn, 11132 Wanke 006 (DR) DQ 212735, Wanke et al.
(2006d)
P. metallica LINDEN & RODIGAS. Tildenia BG Bonn, 16189 Wanke 066 (DR) DQ 212740, Wanke et al.
(2006d)
P. pereskiifolia (JACQ.) KUNTH Micropiper BG Berlin, 140-28-74-
83
Schwertfeger GH 13455 (B) DQ 212726, Wanke et al.
(2006d)
P. pernambucensis MIQ. Rhynchophorum BG Berlin, 062-62-74-
83
Schwertfeger GH 13432 (B) DQ 212751, Wanke et al.
(2006d)
P. pitcairnensis C.DC. Sphaerocarpidium BG Bonn, 17744 Wanke 007 (DR) DQ 212762, Wanke et al.
(2006d)
P. ppucuppuccu TREL. Micropiper BG Berlin Wanke 043 (DR) DQ 212728, Wanke et al.
(2006d)
P. rhombea RUIZ & PAV. Micropiper BG Berlin, 224-08-95-
80
GH 39582 (B) DQ 212731, Wanke et al.
(2006d)
P. rotundifolia (L.) KUNTH Sphaerocarpidium BG Berlin, 166-08-83-
20
Leuenberger GH 23064 (B) DQ 212754, Wanke et al.
(2006d)
P. spec. Sphaerocarpidium Zaire, Irangi Fischer s.n. DQ 212760, Wanke et al.
(2006d)
P. trifolia (L.) DIETR. Micropiper BG Berlin, 078-06-97-
80
GH 37007 (B) DQ 212727, Wanke et al.
(2006d)
P. tuisana C.DC. Sphaerocarpidium BG Berlin, 173-24-95-
33
Horich GH 35526 (B) DQ 212756, Wanke et al.
(2006d)
P. vinasiana C.DC. Rhynchophorum BG Berlin, 173-25-95-
33
Horich & San Jose GH 35591
(B)
DQ 212743, Wanke et al.
(2006d)
Piper s.l. L.
P. aduncum L. Radula Valle, Kolumbien MAJ076 (DUKE) DQ 882201
P. aduncum L. Radula Samar, Philippinen MAJ200 (DUKE) DQ 882202
P. arieianum C.DC. Schilleria /
Steffensia
Choco, Kolumbien MAJ069 (DUKE) DQ 882204
P. augustum RUDGE Schilleria /
Steffensia
Choco, Kolumbien MAJ122 (DUKE) DQ 882203
P. auritum KUNTH Steffensia Choco, Kolumbien MAJ063 (DUKE) DQ 882205
P. bavinum C.DC. Piper s.str. Prov. Ha Tinh,
Vietnam
MAJ392 (DUKE) DQ 882210
24
P. caninum BLUME Piper s.str. Prov. Surigao,
Philippinen
MAJ218 (DUKE) DQ 882213
P. cf. longum L. Macrostachys BG Köln, s.n. Neinhuis s.n. (DR) DQ 882218
P. cf. magnificum Hort. ex TREL.Schilleria /
Steffensia
BG Bonn, 05020 Wanke 069 (DR) DQ 882209
P. cinereum C.DC. Steffensia Choco, Kolumbien MAJ066 (DUKE) DQ 882216
P. confertinodum (TREL. &
YUNCK.) M.A. JARAM. &
CALLEJAS
Confertinodum
group
Choco, Kolumbien MAJ054 (DUKE) DQ 882227
P. crocatum RUIZ & PAV. Piper s.str. BG Bonn, 18143 Wanke 070 (DR) DQ 212714, Wanke et al.
(2006d)
P. decumanum L. Piper s.str. Prov. Leyte,
Philippinen
MAJ210 (DUKE) DQ 882212
P. flagellicuspe TREL. & YUCK. Radula Choco, Kolumbien MAJ065 (DUKE) DQ 882206
P. hispidum SWARTZ Radula Choco, Kolumbien MAJ053 (DUKE) DQ 882219
P. michelianum C.DC. Arctottonia Edo. Jalisco, Mexico MAJ537 (DUKE) DQ 882217
P. munchanum C.DC. Schilleria /
Steffensia
Choco, Kolumbien MAJ120 (DUKE) DQ 882207
P. nigrum L. Piper s.str. Prov. Lagunas,
Philippinen
MAJ181 (DUKE) DQ 882215
P. ornatum N.E.BR. Piper s.str. BG Bonn, 18144 Wanke 005 (DR) DQ 882211
P. parvulum M. A. Jaram. &
Callejas
Trianae group Choco, Kolumbien MAJ055 (DUKE) DQ 882226
P. peltatum L. Pothomorphe Choco, Kolumbien MAJ045 (DUKE) DQ 882208
P. penninerve C.DC. Piper s.str. Prov. Surigao del
Norte, Philippiene
MAJ213 (DUKE) DQ 882214
P. pulchrum C.DC. Macrostachys Antioquia, Kolumbien MAJ100 (DUKE) DQ 882222
P. reticulatum L. Enckea Choco, Kolumbien MAJ128 (DUKE) DQ 882220
P. reticulatum L. Enckea Antioquia, Kolumbien MAJ062 (DUKE) DQ 882221
P. spec BG Bonn, 00854 Borsch 3475 (BONN) DQ 882225
P. spoliatum TREL. & YUNCK. Macrostachys Choco, Kolumbien MAJ060 (DUKE) DQ 882223
P. subpedale TREL. & YUNCK. Schilleria /
Steffensia
Choco, Kolumbien MAJ057 (DUKE) DQ 882224
Macropiper MIQ.
M. excelsum (FORST. f.) MIQ. Macropiper BG Bonn, 17450 Wanke 071 (DR) DQ 882229
M. hooglandii HUTTON & GREEN Macropiper Cult. Auckland
Museum, Neuseeland
ROG 8496 (AK) DQ 882228
25
Manekia TREL.
M. naranjoana (C.DC.)
CALLEJAS
Manekia Costa Rica O. Vargas s.n. (DUKE) DQ 882239
M sydowii (TREL.) ARIAS,
CALLEJAS & BORNSTEIN
Manekia Antioquia, Kolumbien MAJ038 (DUKE) DQ 882238
Zippelia BLUME
Z. begoniifolia BLUME Zippelia BG Kunming, s.n. Wanke & Neinhuis s.n. (DR) DQ 882230
Z. begoniifolia D12 clone 2 DQ 882231
Z. begoniifolia E12 clone 3 DQ 882232
Z. begoniifolia F12 clone 4 DQ 882233
Z. begoniifolia C12 clone 1 DQ 882234
Z. begoniifolia F11 clone 5 DQ 882235
Z. begoniifolia G12 clone 5 DQ 882236
Z. begoniifolia D11 clone 4 DQ 882237
Outgroup
Canella winterana Gaertn. Canellaceae BG Bonn, 15293 Borsch 3466 (BONN) DQ 882240 *2
Drimys lanceolata (POIR.)
BAILL.
Winteraceae BG Bonn, 00769 Borsch 3484 (BONN) DQ 882241 *2
Pseudowintera colorata
(RAOUL) DANDY
Winteraceae BG Bonn, 00770 Borsch 3490 (BONN) DQ 882242 *1
26
Methods
DNA-isolation, amplification and sequencing
Total genomic DNA was isolated from fresh material, silica gel dried leaves or
herbarium specimens. A modified CTAB procedure (with a triple-extraction was
conducted as described in Borsch et al. (2003).
The trnK/matK region was generally amplified in two parts with an overlap of 250 to
400 bp, using the primers listed in Table 2. In some species, the trnK/matK region
was amplified in three parts due to long insertions of AT rich microsatellites.
Amplification profiles differed only with respect to annealing temperatures for the
specific primer combination used, and were otherwise: 3 min at 96°C, 3 min at 50°C
(48°C), 3 min at 72°C, 34 cycles (39 cycles) of 1 min at 94°C, 1.50 min at 48/50/52°C,
3 min at 72°C and a final extension 20 min at 72°C. Reactions of 25 µl containing 15
µl DNA template (2 ng/µl), 3.3 µl dNTP mix (1.25 mM each), 0.5 µl of each primer (20
pmol/µl) and 1 U Taq Polymerase (Promega) were conducted. After gel
electrophoresis the PCR products were purified using a QiaQuick gel extraction kit
(QIAGEN). Direct sequencing used the ABI PrismTM BigDye Terminator Cycle
Sequencing Ready Reaction Kit (Perkin Elmer) with subsequent electrophoresis on
ABI 310 or 377 automated sequencers, or the CEQ DTCS Quick Start Kit (Beckman
Coulter) with the CEQ 8000 sequencer, following standard protocols for each kit. In
some cases, the PCR products were cloned using the pGEMT-easy vector kit
(Promega) and sequenced with the amplification primers after plasmid isolation and
purification through GFX microplasmid kit (AmerSham). Cloning followed standard
procedures with 1 µl vector, 1 µl ligase, 5 µl Buffer (all provided with each kit) and 3 µl
PCR product.
27
Table 2. Amplification and sequencing primers used in the present study.
Primer name Direction Sequence (5’-3’) Taxonomic Group used for Design
MG1 rev. AAC TAG TCG GAT GGA GTA GAT Piperales1 (excl. Zippelia +
Sarcorhachis) Liang & Hilu (1996)
MG15 for. ATC TGG GTT GCT AAC TCA ATG Piperales1 Liang & Hilu (1996)
NYmatK 480F for. CAT CTG GAA ATC TTG STT C outgroup + Asaroideae +
Aristolochia
Borsch (2000)
trnK 3914Fdi for. GGG GTT GCT AAC TCA ACG G Piperales1 Johnson & Soltis (1995)
psbA-R for. CGC GTC TCT CTA AAA TTG CAG TCA T Piperales1 Steele (1994)
AR-matK-660R rev. A(CT)G GAT TCG CAT TCA TA Aristolochiaceae + Saururaceae this study
Ca-matK-1690R rev. AGA GGA TTG TTT ATG GAG outgroup this study
Pi-matK-470F for. TTC AAA CCC TTC G(CT)T (AG)CT GG Piper + Saururaceae Wanke et al. (2006a)
Pi-matK-730R rev. ATA GAA ATG GA(CT) TCG TTC AAG Saururaceae + Peperomia Wanke et al. (2006a)
Pi-matK- 500F for. TTT GCA TTT ATT GCG AAT C Piperaceae + Saururaceae this study
AS-matK-670R rev. GA(AG) AGG ATT GTT TAC G(AG)A G
Thottea + Asaroideae +
Saururaceae + outgroup
Wanke et al. (2006b)
Pi-matK-700R rev. AT(AG) AGA AGA TTG TTT ACG G Piper + Zippelia + Sarcorhachis this study
AS-matK-460F for. TAC TTC CCT TTT T(ACT)G AGG
Asaroideae + Thottea +
Saururaceae
Wanke et al. (2006a)
Pi-matK-560F for. TGG ATA CAA GAT GTT CCC Peperomia + Saururaceae this study
AR-matK-080R rev. ACT CCT GAA A(AG)A GAA GTG G Aristolochiaceae + Saururaceae Wanke et al. (2006b)
Pi-matK-1060F for. ACT T(AG)T GGT CTC AAC (CT)G Piperales Wanke et al. (2006a)
Pi-matK-370R rev. TTT (CT)CC TAT AAT TGG AGC Piperaceae Wanke et al. (2006a)
Pi-matK-1480F for. TCG TAA ACA (CT)AA AAG TAC
Piperaceae + Saururaceae +
Asaroideae + Aristolochia
Wanke et al. (2006a)
TH-matK-420F for. AAC TGA ATA AAT GGA TAG AGC Thottea Wanke et al. (2006b)
AR-matK-420F for. AAG TGA ATA AAT GGA TAG AGC Aristolochia Wanke et al. (2006b)
AR-matK-1850R rev. CCA GGC AAG ATA CTA AT Aristolochia this study
ZiSa-matK-480F for. AGT TCA AAA CAT TCG CTA CTG G Zippelia + Sarcorhachis this study
Pi-matK-1820R rev. ACA CTA ATT GGA AGG AGA ATG G Piper + Peperomia this study
ZiSa-trnK-F rev. AAC CGT GCT TGC ATT TTT CAT TG Zippelia + Sarcorhachis Wanke et al. (2006a)
Pi-matK-950R rev. CCT ATC GCT CTT TTG ATT TTG GAA Piper + Peperomia Wanke et al. (2006a)
AR-matK-1200F for. TTC CAA AGT CAA AAG AGC G
Lactoris + Aristolochiaceae +
outgroup
this study
Pi-matK-2800F for. AAT CTT TCT CAT TAT TAC AGT GG
Piper + Saururaceae +
Aristolochioideae
this study
Pi-matK-2030F for. CCT CTT TGC ATT TAT TGC G Piperaceae + Saururaceae + Wanke et al. (2006a)
28
Asaroideae
AR-matK-1510R rev. TAG ACT CCT GAA A(AG)A GAA GTG G Aristolochia this study
AR-matK-960R rev. AAC CTT TTC CCG CAT CAG G Aristolochia this study
TH-matK-960R rev. AAC CTT TTC CCG CAT TAG A Thottea Wanke et al. (2006b)
TH-matK-930F for. TAA TGC GGG AAA AGG TTC Thottea Wanke et al. (2006b)
AR-matK-930F for. TAT TAG TAC CTG ATG CGG G Aristolochia this study
ZiSa-trnK-R for. AAT CCG TAT TCC TTT TTC TCC G Zippelia + Sarcorhachis this study
AR-matK-780R rev. GGT CTT CTG AAA ATG ATT AC Aristolochia this study
AR-matK-680R rev. CCG AGA AAA ACG AAT ATG GAT T Aristolochia this study
AR-matK-1400F for. CTC TTT CAG GAG TCT ATC TAT G Aristolochia this study
AR-matK-1450R rev. CGT TAG AGT TGC ACG TTA Aristolochia this study
AR-matK-1510R rev. TAG ACT CCT GAA ARA GAA GTG G Aristolochia this study
AR-matK-2100R rev. TGA AAA TGA TTA CAA AGC ACT AC Aristolochia this study
AR-matK-2400R rev. ATT TTC TAG CAT TTG ACT CC Aristolochia this study
AR-matK-2510R rev. AAA AAT CTC AAT AAA TGY AA Aristolochia this study
AR-matK-3500R rev. ATC CAA ATA CCA AAT ASA TTC C Aristolochia this study
AR-trnK-1320R rev. ATC GCT CTT TTG ACT TTG G Aristolochiaceae + Lactoris Wanke et al. (2006b)
Pe-matK-2000F for. TTC CTT ACG AAT CCA TAG A Piper + Peperomia Wanke et al. (2006a)
Pe-matK-2500R rev. TTC GCA ATA AAT GCA AAG AGG Piperaceae + Saururaceae Wanke et al. (2006a)
Pe-matK-2700F for. AAA CAA TCT TTT CAT TTA CG Peperomia Wanke et al. (2006a)
Pi-matK-1820R rev. ACA CTA ATT GGA AGG AGA ATG G Piper + Peperomia Wanke et al. (2006a)
trnK-med-150F for. AGA GAA TAC TTC CAT CCT TAC CG Aristolochia this study
trnK-med-440R rev. ATT CGT CTT TAC TCA CTC CGT A Aristolochia this study
1 Also used for outgroup amplification and sequencing
29
Sequence alignment and treatment of microstructural changes
Sequences were manually aligned using PhyDE® (Müller et al., 2005) following
alignment rules proposed by Borsch et al. (2003) and Löhne and Borsch (2005) and
guided by secondary structures of DNA especially for length mutations and
inversions. Secondary structures and the resulting free energy (∆G) of hairpins was
calculated using RNAstructure 4.2 (Mathews et al., 2005). The aforementioned
alignment rules have been compiled to account for mutational events other than
nucleotide substitutions. The observed motifs are largely the result of simple
sequence repeats (one or several copies), deletions, and inversions (see also Benson
et al., 1997; Graham et al., 2000; Kelchner, 2000; Löhne and Borsch, 2005; Müller
and Borsch, 2005a). Seven mutational hotspots were excluded from the final matrix
(Table 3), especially microsatellites, because of uncertain primary homology. An indel
matrix was calculated using the “simple indel coding” approach (SIC, Simmons and
Ochoterena, 2000) as implemented in SeqState (Müller 2005a). The alignment and
the indel matrix are available from TreeBASE (www.treebase.org).
Phylogenetic analyses
Phylogenetic reconstructions using heuristic searches under maximum parsimony
(MP) were performed using PAUP* 4.0b10 (Swofford 2002). The strict consensus tree
was inferred with command files for PAUP* 4.0b10generated by PRAP (Müller, 2004),
implementing the Parsimony Ratchet (Nixon, 1999). The following ratchet settings
were employed: 10 random addition cycles of 500 iterations each with a 25% of
upweighting of the characters in the iterations. In addition, indels were analyzed
employing SIC (Simmons and Ochoterena, 2000) as implemented in SeqState (Müller
2005a). SeqState generates a ready-to-use Nexus file containing the sequence
alignment with an automatically generated indel matrix appended. The evaluation of
the MP tree was performed using the Bootstrap approach (Felsenstein, 1985),
conducting 1000 replicates and random addition searches with 10 iterations per cycle.
Decay values as further measurement of support for the individual clades were
30
obtained using PRAP in combination with PAUP* and the same options in effect as in
the ratchet.
Maximum likelihood analyses were executed assuming a general time reversible
model (GTR), and a rate variation among sites following a gamma distribution (four
categories represented by mean). GTR+G+I was chosen as the model that best fits
the data by Modeltest v3.6 (Posada & Crandall 1998) employing the interface MTgui
(Nuin, 2005). The settings proposed by Modeltest v3.6 [BaseFreq=(0.3384 0.1510
0.1455), Nst=6, Rmatrix=(1.0672 1.7089 0.2936 0.6604 1.7089), Shape=1.2190,
Pinvar=0.0621] were executed in PAUP.
For Bayesian inference (BS) the program MrBayes v3.1 (Ronquist and Huelsenbeck,
2003) was used. To acknowledge possible deviating substitution models for the
coding and non-coding regions the data set was divided into two partitions. For both
partitions, the GTR model of nucleotide substitution was assigned, assuming site-
specific rate categories following a gamma distribution. Two runs (106 generations
each) with four chains each were run simultaneously, starting from random trees.
Chains were sampled every 10 generations and the respective trees were written to a
tree file. Calculation of the consensus tree and the posterior probability (PP) of clades
was done based upon the trees sampled after the chains converged (25 %). Only
PP’s of 95 and higher were considered significant (alpha = 0.05). Trees were
compiled and drawn using TreeGraph (Müller and Müller, 2004).
Relative rate test
Relative rate tests according to Sarich and Wilson (1967) were used to quantify the
degree of rate divergence between taxon sets (e.g. clades).. Relative rate differences
were calculated between the main Piperales groups (Asaroideae, Aristolochioideae
(excl. Thottea), Thottea, Saururaceae, Lactoridaceae, Manekia/Zippelia, Piper s.l.,
Peperomia). As in the phylogenetic analyses Canella, Drimys, and Pseudowintera
were chosen as reference taxa. Calculations of differences in substitutional rates
between groups were based on ML estimates of distances (GTR +G +I model).
Calculations were performed with help of GRate (Müller, 2002; see Müller et al.,
2004) that allows to compare average rates of previously defined taxon sets.
31
Table 3. Hotspots excluded, due to ambiguous homology assessments. The location of the hotspots, length and their composition is shown.
length distribution bp composition %* hotspot position in the alignment
Aristolochiaceae Lactoridaceae Piperaceae Saururaceae A T C G
H1 472-916 14-437 5 4-5 5 34.8 56.0 4.2 5.0
H2 1049-1068 8-17 9 9-11 9-10 0.9 68.8 29.7 0.6
H3 1249-1260 2-5 3 2-12 3-9 93.9 2.3 2.1 1.7
H4 1409-1426 8-12 8 6-8 7 64.8 3.2 11.4 20.6
H5 1525-1541 0-17 0 0-1 1 2.1 97.9 0 0
H6 3821-3847 0-27 5 4-9 6-8 72.0 14.6 7.0 6.4
H7 3872-3899 5-8 7 1-18 7 12.7 57.8 16.9 12.7
* calculated for each hotspot from the whole dataset
Table 4. Sequence statistics for the coding and noncoding regions calculated with SeqState (Müller 2004). % divergence = overall sequence
dissimilarity (uncorrected p-distance * 100)
region
characters (bp)1 length range (bp) 1 characters* (bp) % divergence* variable* parsimony informative*% GC
contents1
trnK intron incl. matK 4256 2413-3278 3683 13.8 43.8 % 32.1 % 33:8
trnK intron excl. 2575 886-1754 2002 14.8 33.2 % 22.1 % 33.5
trnK 5’ intron 1874 677-1538 1355 14.4 34.8 % 24.4 % 33.9
trnK 3’ intron 701 179-701 647 16.8 29.8 % 17.3 % 32.3
matK coding 1681 1509-1557 1681 13,3 56.6 % 44.0 % 33.7
1 calculated with hotspot regions included
* calculated with hotspot regions excluded
32
Results
Variability in the trnK intron and the matK gene
In Piperales, the trnK intron (including the matK ORF) is 2533 bp on average, ranging
from 2412-3258 bp. The matK gene itself has a length of 1509-1557 bp (mean 1534
bp). Frequent microstructural changes lead to a considerably longer aligned dataset
of 4256 characters, including hotspots (Fig. 1, Table 4). All hotspots are located in the
trnK intron and their position and extension are given in Table 3. Sequences in
hotspots are microsatellites consisting of stretches of poly-As or Ts except H1 that
comprises long and highly-variable AT-rich sequence parts inserted in
Aristolochioideae (excl. Thottea) sequences (cryptic simple microsatellite, Wanke et
al, 2006a). Table 5 summarises lineage specific characteristics for partitioned
datasets (Piper, Peperomia, Aristolochia s.l.). It is clearly shown that the number of
parsimony informative characters, compared to outgroup, is similar between the three
large genera. But parsimony informative characters within the clades are considerably
different. Piper displays only ~1/3 to ~1/2 the amount of informative sites compared to
Aristolochia or Peperomia, respectively. Generally, the number of variable characters
is twice as high within Peperomia and Aristolochia compared to Piper.
Table 5. Characterisation of the maximum parsimony trees obtained for the complete alignment
(hotspots excluded) and partitioned sets representing the “giant genera”, to evaluate the lineage
specific resolution contrast based on alignment characteristics.
without
indels*
with indels* Aristolochia Peperomia Piper
trees found# 440 3434 1 6 609
steps# 4303 5065 1371 1522 1029
CI# 0.559 0.555 0.761 0.787 0.835
RI# 0.913 0.906 0.842 0.810 0.863
RC# 0.511 0.503 0.641 0.637 0.720
HI# 0.441 0.445 0.239 0.213 0.165
total char. # 3684 4089 3684 3684 3684
constant char. # 2066 2066 2822 2695 2925
uninformative char. # 432 605 313 370 255
pars. informative char. # 1186 1418 549 619 504
pars. informative char. %$ - - 10.365 6.743 3.740
variable char. %$ - - 18.227 18.051 9.393
* based on the complete alignment incl. outgroup
# based on charactersets for the mentioned members plus outgroup, based on substitutions only, if not
different indicated
$ based on charactersets for the mentioned members without outgroup, based on substitutions only
33
Microstructural variation in the trnK intron
A large number of length mutations has been identified and coded in a separate
matrix (395 indels in total). Among these, 217 indels are located within the trnK 5’
intron, 59 in the matK gene and 119 in the trnK 3’ intron. Most of these indel events
represent simple sequence repeats (SSR) of the flanking region (up to 25 bp in the
intron and 15 bp, 5 codons, in the gene). An AT-rich microsatellite-like stretch is found
in the domain I of the trnK intron (Fig 1, hotspot I). In Aristolochioideae (not present in
Thottea), this cryptic simple microsatellite ranges from 29 bp in Isotrema to 443 bp in
Endodeca. The internal structure of this repeat can be characterised as (AnTm)k. This
microsatellite region is absent in all other Piperales. A second poly-A/T microsatellite
is observed in Thottea (Fig. 2). Examination of the flanking regions and subsequent
structural analysis reveals a hairpin with the microsatellite forming a terminal loop,
which appears to have been inverted in the common ancestor of Thottea (Fig. 2).
Compared to expected triplet insertions and in frame deletions, “self repairing” out-of-
frame indels around 600 nt upstream of the matK ORF start (Fig. 3, Tab. 6),
apparently associated with a microsatellite, have been identified. Based on primary
homology assessment, length mutations involving one or two nucleotides must be
assumed, which are followed downstream by an additional length mutational event
involving one or two nucleotides respectively. Therefore, the frameshift affects only
five or six amino acids, the restoration of the original reading frame is observed in all
cases. However, an out-of-frame deletion of two codons is found in the two species of
Pararistolochia (around alignment position 2050). The matK gene of the Piperaceae
is highly variable in length on its 3’ end, this variation is probably due to point
mutations that result in early stop codons in several taxa (Fig. 4).
34
Figure 2. Position 356-397 of the alignment (trnK 5’ intron), showing a selection of taxa and the
potential inverted sequences (boldface) in Thottea. Below the potential part as “normal” and as reverse
complemented and aligned is given. Lowest shows the reverse complemented part aligned into the
original selection. The absolute number of inverted nucleotides could not be detected due to the
insertion of poly T’s. The secondary structure (∆G = -3.5) for this region is given as example from
Thottea corymbosa, demonstrating the perfect stem-loop region of the potential inverted region.
A. salvadorensis TTTCTTTGAACGGGACTCAAA-----------------AAAT-TAACCC----TTGGGTC
A. tomentosa TTTCTTGGAACGGGACTAAAA-----------------AAAT-TCACCC----TTGGGTC
A. manshurensis TTTCTTGGAACGGGACTAAAA-----------------AAAT-TCACCC----TTGGGTC
A. serpentaria TTTCTTGGAACGGGACTAAAA-----------------AAAT-GCACCC----TTGGGTC
Thottea dependens ATTCTTGGAACGGGATGCATTTTTTTTTTTTTTTTTTT-----TCATCT----TTAGGTC
Thottea corymbosa ATTCTTAGAACGGGATGAATTTTTTTTTTTT------------TCATCT----TTGGGTC
Thottea siliquosa ATTCTTGGAACGGGGTGCATTTTTTTTTT--------------TCATCT----TTGGGTC
Asarum caudatum ATTCTTGGAACGGGACCAAAT-----------------CAATATCACCATTGATTGGGTC
Saruma henryi ATTCTTGGAACGGGACCAAAT-----------------CAATATCACCATTGATTGGGTC
Anemopsis californica ATTCTTGGAATGGGACCAAAT-----------------CAAT-TCATCC----TTGGGTC
Thottea dependens GATGCATTTTTTTTTTTTTTTTTTT-----TCATC
Thottea corymbosa GATGAATTTTTTTTTTTT------------TCATC
Thottea siliquosa GGTGCATTTTTTTTTT--------------TCATC
Thottea dependens GATGAAAAAAAAAAAAAAAAA----AAAT-GCATC
Thottea corymbosa GATGAAAAAAAAAA-----------AAAT-TCATC
Thottea siliquosa GATGAAAAAAAA-------------AAAT-GCACC
A. salvadorensis TTTCTTTGAACGGGACTCAAA-----------------AAAT-TAACCC----TTGGGTC
A. tomentosa TTTCTTGGAACGGGACTAAAA-----------------AAAT-TCACCC----TTGGGTC
A. manshurensis TTTCTTGGAACGGGACTAAAA-----------------AAAT-TCACCC----TTGGGTC
A. serpentaria TTTCTTGGAACGGGACTAAAA-----------------AAAT-GCACCC----TTGGGTC
Thottea dependens TTTCTTGGAACGGGATGAAAAAAAAAAAAAAAAA----AAAT-GCATCC----TTGGGTC
Thottea corymbosa TTTCTTGGAACGGGATGAAAAAAAAAA-----------AAAT-TCATCC----TTGGGTC
Thottea siliquosa TTTCTTGGAACGGGATGAAAAAAAA-------------AAAT-GCACCC----TTGGGTC
Asarum caudatum ATTCTTGGAACGGGACCAAAT-----------------CAATATCACCATTGATTGGGTC
Saruma henryi ATTCTTGGAACGGGACCAAAT-----------------CAATATCACCATTGATTGGGTC
Anemopsis californica ATTCTTGGAATGGGACCAAAT-----------------CAAT-TCATCC----TTGGGTC
Thottea corymbosa ∆G = -3.5
G
A
T
G
A
A
T
T
T
T T T T T
T
T
T
T
T
C
A
T
C
35
Figure 3. Position 2020-2110 of the alignment. An example for the high amount of length mutational
events within the matK gene. Frame shift mutations with the respective “self repairing part” are
indicated in bold. Only parts of the complete matrix are shown. All effected AS are marked in bold and
arranged how they should be aligned on nucleotide level to demonstrate frame shift events.
Canella ATTACTCCAAAGAAA---------TCCATTTCCATT------TTT------TCA---------AAAGATAAT
Drimys ATTACTCCAAAGAAA---------TCCATTTCCCTT------TTT------TCA---------AAAAGGAAT
Pseudowintera ATTATTCCAAAGAAA---------TCCATTTCCATT------TTT------TCA---------AAAAGGAAT
A. pentandra ATTAGTTCAAAGAAA---------TCCATTTTTTTT------TTC------TCA---------AAAGAGAAT
A. bracteolata ATTAGTTCAAAGAAA---------TCCATTTCTTTT------TTC------TCA---------AAAGAGAAT
A. albida ATTAGTTCAAAGAAA---------TCCATTTATTTT------TTC------TCATTCTCA---AAAGAGAAT
A. pistolochia ATTAGTTCAAAGAAA---------TCCTTTTTTTTTT-----TTCTCAAA-TCA---------AAAGAGAAT
A. parvifolia ATTAGTTCAAAGAAA---------TCCTTTTTT---------TTC------TCA---------AAGGGGAAT
A. pichinch. ATTAATTCAAAGAAA---------TCCATTTTTATT------TTA------TCA---------AAAGGGAAT
P. triactina ATTAGTTCAAAGAAA---------TCCATTTTTTTT------TTC------TCA---------AAAGAGAAT
A. salvadoren. GTTAGTTCAAAGAAA---------TCCATTCTTTTTTTT---TTC------TCA---------AAAGAGAAT
A. californica GTTAGTTCAAAGAAA---------TCCATTCCTTTTTTTTTTTTC------TCA---------AAAGAGAAT
A. macrophylla GTTAGTTCAAAGAAA---------TCCATTCCTTTTTTT---TTC------TCA---------AAAGAGAAT
A. serpentaria GTTAGTTCAAAGAAA---------TCCATTCCTTTTTTT---TTC------TCA---------AAAGAAAAT
Th. siliquosa ATTAGTCCAAAGAAA---------TTCATTTCTTTT------TTC------TCA---------AAGGAGAAT
Asa. caudatum ATTAGTCTAAAGAAA---------TCTATCTCTTTCTTT---TTT------TCA---------AAAGGGAAT
Saruma henryi ATTAGTCCAAAGAAA---------TCTATCTCTTTCTTT---TTT------TCA---------AAAGGGAAT
Lactoris ATTAGTCCAAAAAAA---------TCGATTTCTTTT------TTT------TCA---------AATGGGAAT
Anemopsis ATTAGCCAAAAAAAA---------TCCATCTCT---------TTT------TCA------AAAGAAGAGAAT
Gymnotheca ATTAGCAAAAAAAAA---------TCCATCTCT---------TTT------TCA------AAAGAAGAGAAT
M. naranjoana ATTAGCCAAAAAAAA---------TTCTCTTCT---------TTT------TCA------AAAAAAGAGAAT
P. decuma ATTAGCAAAAAAAAA---------TTCTTTTCT---------TTT------TCA------AAAAAAGAAAAT
Pep. gracill. TTTAGCAAAAGA---------------------------------------------------AAAAAAAAT
Pep. fraseri TTTAGCAAAATA---------------------------------------------------AAAAAAAAT
Pep. trifolia TTTAGCAAAATAAAAAAAAA----TTCTTTTAT---------TTT------TCA--------AAAAAAAAAT
Pep. argyreia TTTAAAAAAAAAAAAAAAAA----TTCTTTTCT---------TTT------TCT--------AAAAAAAAAT
Pep. metallica TTTAGCAAAATAAAAAAAAA----TTCTTTTAT---------TTT------GAA--------AAATAAAAAT
Pep. marmorata TTTAGCAAAAGAAAAAAAAG----TTCTTTTAT---------TTT------TCA--------AAAAAAAAAT
Pep. fagerlin. TTTAGCAAAATAAAAAAAAAAA--TTCTTTTAT---------TTT------TCA-------AAAAAAAAAAT
Pep. clusiifo. TTTAGCAAAATAAAAAAAAA----TTCTTTTCT---------TTT------TCA--------AAAAAAAAAT
Pep. pernambu. TTTAGCAAAAGA---------------------------------------------------AAAAAAAAT
Canella IleThrProLysLys---------SerIleSerIle------Phe------Ser---------LysAspAsn
Drimys IleThrProLysLys---------SerIleSerIle------Phe------Ser---------LysArgAsn
Pseudowintera IleIleProLysLys---------SerIleSerIle------Phe------Ser---------LysArgAsn
A. pentandra IleSerSerLysLys---------SerIlePhePhe------Phe------Ser---------LysGluAsn
A. bracteolata IleSerSerLysLys---------SerIleSerPhe------Phe------Ser---------LysGluAsn
A. albida IleSerSerLysLys---------SerIleTyrPhe------Phe------SerPheSer---LysGluAsn
A. pistolochia IleSerSerLysLys---------SerPhePhePheP-----heLeuLys-Ser---------LysGluAsn
A. parvifolia IleSerSerLysLys---------SerPhePhe---------Phe------Ser---------LysGlyAsn
A. pichinch. IleAsnSerLysLys---------SerIlePheIle------Leu------Ser---------LysGlyAsn
P. triactina IleSerSerLysLys---------SerIlePhePhe------Phe------Ser---------LysGluAsn
A. salvadoren. ValSerSerLysLys---------SerIleLeuPhePhe---Phe------Ser---------LysGluAsn
A. californica ValSerSerLysLys---------SerIleProPhePhePhePhe------Ser---------LysGluAsn
A. macrophylla ValSerSerLysLys---------SerIleProPhePhe---Phe------Ser---------LysGluAsn
A. serpentaria ValSerSerLysLys---------SerIleProPhePhe---Phe------Ser---------LysGluAsn
Th. siliquosa IleSerProLysLys---------PheIleSerPhe------Phe------Ser---------LysGluAsn
Asa. caudatum IleSerLeuLysLys---------SerIleSerPhePhe---Phe------Ser---------LysGlyAsn
Saruma henryi IleSerProLysLys---------SerIleSerPhe------Phe------Ser---------LysGlyAsn
Lactoris IleSerProLysLys---------SerIleSerPhe------Phe------Ser---------AsnGlyAsn
Anemopsis IleSerGlnLysLys---------SerIleSer---------Phe------Ser------LysGluGluAsn
Gymnotheca IleSerLysLysLys---------SerIleSer---------Phe------Ser------LysGluGluAsn
M. naranjoana IleSerGlnLysLys---------PheSerSer---------Phe------Ser------LysLysGluAsn
P. decuma IleSerLysLysLys---------PhePheSer---------Phe------Ser------LysLysGluAsn
Pep. gracill. PheSerLysArg---------------------------------------------------LysLysAsn
Pep. fraseri PheSerLysIleLysLysAs----nSerPheIl---------ePh------eGl--------nLysLysAsn
Pep. trifolia PheLysLysLysLysLysAs----nSerPheLe---------uPh------eLe--------uLysLysAsn
Pep. argyreia PheSerLysIleLysLysAs----nSerPheIl---------eLe------uLy--------sAsnLysAsn
Pep. metallica PheSerLysArgLysLysSe----rSerPheIl---------ePh------eGl--------nLysLysAsn
Pep. marmorata PheSerLysIleLysLysLysI--leLeuLeuP---------heP------heL-------ysLysLysAsn
Pep. fagerlin. PheSerLysIleLysLysAs----nSerPheLe---------uPh------eGl--------nLysLysAsn
Pep. clusiifo. PheSerLysArg---------------------------------------------------LysLysAsn
Pep. pernambu. PheSerLysLys---------------------------------------------------LysLysAsn
36
Figure 4. The 3’ part of the matK gene (position 3517-3570), showing the different positions of stop
codons based on homology of the nucleotides a) nucleotide sequence b) amino acid sequence. Stop
codons are indicated in boldface or as star respectively. Only a representing sampling of the dataset is
indicated.
a)
Piper cinereum CATACAAATGACTTGACCAATCATTAA------TGA----TTGATCACAAG
Macropiper excelsum CATACAAATGACTTGACCAATCATGAA------TGA----TTGATCATAAG
Piper hispidum CATACAAATGACTTGACCAATCATGAA------TGA----TTGATCATAAG
Piper reticulatum CATACAAATGACTTAACCAATCATGAA------TGA----TTGGTCATAAG
Piper pulchrum CATACAAATGACTTGACCAATCATGAA------TGA----TTGGTCATAAG
Piper spoliatum CATACAAATGACTTGACCAATCATGAA------TGA----TTGGTCATAAG
Peperomia gracillima CATACAAATGACTTGACCAATCATGAA------TGA----TTGGTCATAAG
Peperomia fraseri CATACAAATGACTTGACCAATCAAGAA------TAA----TTGGTCATAAG
Peperomia ppucuppucu CATACAAATGACCTGACCAATCAATAATAATAATAA----TTGGTCATAAA
Peperomia trifolia CATACAAATGACCTGACCAATCAATAA------TAA----TTGGTCATAAG
Peperomia rhombea CGTACAAATTAACTCACCAATCAATAA------TAA----TTAGTCATAAG
Peperomia cuspidilimba CATACAAATTAACTGACCAATCAATAA------TAA----TTAGTCATAAG
Peperomia pereskiifolia CATACAAATGACCTGACCAATCAATAA------TAA----TTGGTCATAAG
Peperomia maypuensis CATACAAATGACTTGAACAATCAAGAA------TAA----TTGGTCATAAA
Peperomia argyreia CATACAAATGACTTGACCAATCAAGAA------TAA----TTGGTCATAAG
Peperomia marmorata CATACAAATGACTTGACCAATAAAGAA------TAA----TTAGTCATAAG
Peperomia vinasiana CAAACAAATGACTTGACCAATCAAGAA------TAA----TTAGTCATAAG
b)
Piper cinereum HisThrAsnAspLeuThrAsnHisEOF------EOF----LeuIleThrArg
Macropiper excelsum HisThrAsnAspLeuThrAsnHisGlu------EOF----LeuIleIleArg
Piper hispidum HisThrAsnAspLeuThrAsnHisGlu------EOF----LeuIleIleArg
Piper reticulatum HisThrAsnAspLeuThrAsnHisGlu------EOF----LeuValIleArg
Piper pulchrum HisThrAsnAspLeuThrAsnHisGlu------EOF----LeuValIleArg
Piper spoliatum HisThrAsnAspLeuThrAsnHisGlu------EOF----LeuValIleArg
Peperomia gracillima HisThrAsnAspLeuThrAsnHisGlu------EOF----LeuValIleArg
Peperomia fraseri HisThrAsnAspLeuThrAsnGlnGlu------EOF----LeuValIleArg
Peperomia ppucuppucu HisThrAsnAspLeuThrAsnGlnEOFEOFEOFEOF----LeuValIleArg
Peperomia trifolia HisThrAsnAspLeuThrAsnGlnEOF------EOF----LeuValIleArg
Peperomia rhombea HisThrAsnEOFLeuThrAsnGlnEOF------EOF----LeuValIleArg
Peperomia cuspidilimba HisThrAsnEOFLeuThrAsnGlnEOF------EOF----LeuValIleArg
Peperomia pereskiifolia HisThrAsnAspLeuThrAsnGlnEOF------EOF----LeuValIleArg
Peperomia maypuensis HisThrAsnAspLeuAsnAsnGlnGlu------EOF----LeuValIleLys
Peperomia argyreia HisThrAsnAspLeuThrAsnGlnGlu------EOF----LeuValIleArg
Peperomia marmorata HisThrAsnAspLeuThrAsnLysGlu------EOF----LeuValIleArg
Peperomia vinasiana GlnThrAsnAspLeuThrAsnGlnGlu------EOF----LeuValIleArg
Additional, non-functional matK copies in Zippelia.
While editing electropherograms of Zippelia, conspicuous overlapping peaks were
observed from positions 3264 to 3271 in the alignment on the forward strand or
positions 4047 to 4084 on the reverse strand. Different PCR products obtained with
different primer combinations and sequencing of both strands showed the same
pattern, thus exluding contamination or Taq errors. Subsequent to cloning and
generating independent sequences per colony, five different products, differing by a
37
large number of point mutations and 9 length mutations, were obtained. All
sequences were highly similar to matK sequences in Piperaceae, but most
sequences exhibited a large number of internal stop codons. The sequence obtained
from clone 2 showed a 7 bp gap within domain X, that was not corrected, thus clearly
pointing to a non-functional copy of matK. The conserved domain X (135 nt in length),
associated with the maturase activity of the protein, can be found from pos. 3237 to
3371 in the Piperales matK alignment (Fig. 1). This variability pattern in Zippelia could
only be explained by assuming the presence of at least five additional copies of matK.
The only copy without internal stop codons and out-of-frame mutations, and thus
unambiguously recognized as functional was used for the phylogenetic analyses.
Evidence for polymorphic sites or additional matK copies was not found in any other
investigated taxon or generated sequence for this study.
Rate heterogeneity between lineages
Between the seven main Piperales clades we observed highly significant (P<0.0001)
rate heterogeneity, with the Piperaceae having pronounced rates compared to the
remaining Piperales. Within Piperaceae, Peperomia displays the fastest evolutionary
rate, followed by Piper and Sarcorhachis/Manekia. For the remaining clades,
Asaroideae have the lowest and Lactoris the fastest rates.
Trees resulting from maximum parsimony, likelihood and Bayesian inference
The different methods employed in this study, maximum parsimony (Fig. 5), likelihood
and Bayesian inference (Fig. 6) resulted in a nearly identical tree topology for the
major groups. The present data set (2066 constant and 1186 parsimony informative
(PI) characters) results in 440 most parsimonious trees (MPT) of 4303 steps; the strict
consensus tree is depicted in Fig. 5 with the bootstrap and decay values depicted
aling the branches. The coding of length mutational events has added 300 PI sites to
the data matrix. The combined analysis has resulted in 3434 MPTs (5065 steps)
(Table 5). The likelihood phylogram (-ln 30144.36066 ) is shown in Fig. 6 with the
significantly supported branches (prosterior probabilities above 90) being emphasized
by thick lines.
38
The relationship between Aristolochiaceae and Lactoris remains uncertain, as the
analyses do not resolve the tree with high support. MP analyses show Asaroideae
sister to a Lactoris + Aristolochioideae clade (no support), whereas the Bayesian and
the likelihood analysis show Asaroideae sister to Piperaceae + Saururaceae (no
support; PP 64). The Aristolochiaceae are paraphyletic with respect to Lactoris and
split into two well- supported clades (Asaroideae and Aristolochioideae). The sister
group relationship of Aristolochioideae and Lactoridaceae receive BS of 82% / 83%
as well as PP of 100. Within subfamily Aristolochioideae, Thottea branches first
followed by the Endodeca and Isotrema clade (maximal support) which is sister to the
remaining Aristolochioideae. Among the remaining Aristolochioideae, Pararistolochia
as well as segregates recognised by e.g. Huber (1985, 1993) like Einomeia are
monophyletic. Hubers’s segregate “Howardia” is found to be polyphyletic, as the
Aristolochia grandiflora complex has been treated as part of “Howardia”.
Relationships among these segregates are generally not well supported in parsimony
analyses.
Saururaceae are monophyletic and sister to Piperaceae. Within Saururaceae, results
obtained with the different search methods, are congruent, resolving Houttuynia and
Anemopsis together and this clade sister to the remaining genera Saururus and
Gymnotheca. Piperaceae are subdivided into two well-supported clades, the first one
consisting of the small genera Manekia and Zippelia, and the two large genera Piper
s. l. and Peperomia forming the second clade. Several infrageneric clades, such as
Pothomorphe, Macrostachys and Macropiper, can be recognised with moderate to
high support. In contrast to this, species relationships in Peperomia are fully resolved
with generally high support, although the infrageneric clades are often polyphyletic
(e.g. Tildenia, Rhynchophorum).
Figure 5. Strict consensus tree of 440 most parsimonious trees (MPT’s; length 4303, CI = 0.559, RI =
0.913, RC = 0.511). Bootstrap support values as well as decay indices are depicted along the
branches; support values derived with an appended indel matrix are shown below the branches,
whereas regular support values are depicted above. Indels were coded with SeqState (Müller, 2005a)
using the SIC-approach (Simmons and Ochoterena, 2000). Members of “Howardia” are indicated with
an asterisk (*).The Asian tropics and South Pacific clade is highlighted by a grey box. (see next page)
39
100/46
100/49
100/39
100/43
-/1
-/-
100/69
100/76 100/11
100/11
82/6
83/5
100/37
100/43
100/86
100/91 99/7
98/8
100/40
100/42
100/28
100/32
100/23
100/23
92/5
89/5
100/23
100/23
67/1
78/2
90/4
89/4
93/5
94/5 100/10
100/10
100/26
100/28
100/25
100/28
94/6
95/7
100/14
100/13 51/2
-/1
-/1
-/1 78/3
63/2
-/1
-/- 93/7
93/6
56/1
-/1
100/39
100/39 69/1
66/1
66/1
59/1
100/49
100/52
72/2
79/3
100/27
100/29 97/4
99/5
100/15
100/18 100/6
100/6
100/91
100/100
100/22
100/23
83/3
78/3
93/4
92/4 100/29
100/29
100/79
100/88
100/15
100/14 99/15
100/17
100/40
100/42
100/19
100/23
100/7
100/7
100/11
100/12
92/1
93/1
63/1
74/1 99/5
99/7
98/6
99/8 99/6
100/7
-/1
-/- 50/1
-/- 66/2
75/2 66/1
-/-
-/1
-/- 100/18
100/18
-/1
-/- -/1
-/- 97/6
97/5
100/49
100/51
100/42
100/40
100/16
100/21
55/1
75/3
86/3
100/6 100/22
100/23
59/1
-/1
96/7
100/8
95/5
93/5
100/12
100/12
98/6
97/7
96/5
94/7
100/13
100/15 99/4
98/4
90/4
97/7
97/7
98/9 100/12
100/13
92/3
73/3 98/9
98/9 64/1
66/1 66/2
-/3
-/1
-/1
54/1
-/-
Saruma
Asarum
Lactoris
Asaroideae
Lactoridaceae
Aristolochioideae
Thottea
Endodeca
Isotrema
Pararistolochia
Howardia*
Einomeia
A. grandiflora complex*
Aristolochia s. str.
Aristolochia
Saururaceae
Piperaceae
Piper s.l.
Peperomia
Manekia
Zippelia
Geophila
Micropiper
Panicularia
Anemopsis
Houttuyni a
Gymnotheca
Saururus
Tildenia
Rhynchophorum
Rhynchophorum
Tildenia
a
Sphaerocarpidium
Piper s. str.
Piper s. str.
Macropiper
Macrostachys
Potomorphe
Enckea
40
Figure 6. Likelihood phylogram. Significantly supported branches are indicated by thick lines (Posterior
Probability ≥ 90%). The Asian tropics and South Pacific clade is highlighted by a grey box.
0.1
Saruma
Asarum
Lactoris
Thottea
Aristolochia
Piper s.l.
Peperomia
Manekia
Zippelia
Anemopsis
Houttuynia
Gymnotheca
Saururus
41
Discussion
Microstructural changes in the matK CDS
The matK ORF is maintained in all taxa of Piperales analysed in this study. Length
variability, resulting from stop codons, of the ORF is confined to the 3' end, where
minimal impact on the protein structure is expected (Hilu and Liang, 1997). This
parallels results in other groups of angiosperms with highly variable matK sequences
such as Lentibulariaceae (Müller and Borsch, 2005a), and the general situation in
angiosperms as derived from the molecular evolutionary analysis of a 550-taxon data
set (Borsch et al., pers. comm.). Microstructural changes in Piperales mostly involve
one to three codons, rarely up to five codons. Since non-trimeric structural mutations
are generally suppressed in coding regions (Metzgar et al., 2000), the matK gene
therefore appears as a functional gene. The only situation where microstructural
changes are not in triplets is associated with an internal microsatellite (Fig. 2). SSR
expansion and contraction does not seem to be a rare process in coding regions, but
often may result in changes of gene function (Li et al., 2004).
Microstructural changes in the trnK group II intron
The perhaps most striking microstructural variation is the extensive insertion of an
AT-rich, microsatellite-like sequence in the domain I of the trnK intron. For the overall
Piperales analysis, this portion had to be excluded (hotspot I), but at the species and
even population level it has been found to be highly informative (Wanke et al.,
2006a). As in other non-coding chloroplast DNA, most microstructural changes are
SSRs of four to nine nucleotides in length, whereas length mutations involving one to
two nucleotides are particularly rare. An overall higher frequency of microstructural
changes in the trnK intron as compared to the matK gene also conforms to a general
trend and has been found in other studies with complete trnK/matK sequences in
Amaranthaceae (Müller and Borsch, 2005b), Lamiales (Müller et al., 2004;
Rahmanzadeh et al., 2005) and Magnoliales (Sauquet et al., 2003).
42
Presence of matK pseudogenes in Piperaceae
The highly deviant copies of matK found in Zippelia can only be explained by the
occurrence of several additional non-functional copies (paralogous pseudogenes) of
matK. The presence in some copies of non-triplet microstructural changes in domain
X which are not corrected shortly downstream, as well as the frequent occurrence of
stop codons, are evidence for a pseudogenic nature of these additional matK copies.
Given that all matK sequences from Zippelia group together in a parsimony analysis
(trees not shown), it can be hypothesized that either gene duplication events occurred
rather recently in an ancestor of Zippelia begoniifolia or that plastids might harbour
different copies of the genome. Additional pseudogenic copies of matK have already
been reported for other land plant lineages such as Bryophyta and Marchantiophyta
(Jankowiak, 2004), Antocerotophyta and Lycophyta (Quandt et al., unpublished) and
were also found in Valerianaceae (Hidalgo et al., 2004; Bell et al., 2001), and
Nepenthaceae (Meimberg et al., 2006). Similar to Piperaceae, pseudogenes in
Valerianaceae could be clearly distinguished from the functional matK CDS, and
excluded from the actual phylogenetic analysis. There has been some discussion on
a possible pseudogenic nature of matK in Orchidaceae, based on low transition-
transversion ratios and the presence of internal stop codons (Kores et al., 2000).
However, this has not been confirmed by other recent studies (Kocyan et al., 2004;
Van den Berg et al., 2005). Recently, the non-functional trnK/matK copies in
Nepenthaceae were found to be translocated to the mitochondrial genome (Meimberg
et al. 2006). Since multiple PCR amplification products of matK in Zippelia were
obtained with primers annealing to internal parts of the ORF (at least one primer),
there is currently no information on where these matK copies may be located.
Although duplication of chloroplast genes resulting in paralogous copies is rare, there
are reports of extensive trnF duplications in Asteraceae (Vijverberg and Bachmann,
1999) and Brassicaceae (Koch et al., 2005). For trnF, Koch et al., (2005) clearly
showed the tandem arrangement of copies in the plastome, whereas the situation is
less clear for rbcL paralogues which occur in several angiosperm lineages.
Cummings et al. (2003) provided evidence in Brassicaceae, Solanaceae and
monocots for rbcL transfers into the mitochondrial genome, although there may also
43
be paralogous rbcL copies in the plastome of Orobanchaceae (Wolfe and Randle,
2004).
Figure 7. Relative substitutional rates in Piperales with reference to Peperomia (outgroup: Canella,
Drimys, and Pseudowintera). X axis: taxa, y axis: d = K(Peperomia, outgroup) – K(taxa, outgroup); with
K(i,j) = maximum likelihood estimate (GTR +G +I model) of substitutions per site between taxa.
Significant rate differences (<0.0001) compared to Peperomia are indicated with an asterisk.
Rate heterogeneity and lineage specific resolution contrast
Relative rates of Peperomia and Piper are much higher as compared to
Aristolochiaceae (Fig. 7) which is also paralleled by branchlength in Fig. 6.
Nevertheless, internal resolution within the Piper clade is significantly less as
compared to the Peperomia clade. Further comparison between those two lineages
shows that internally the Peperomia clade exhibits about twice the amout of sequence
variation in trnK/matK as compared to Piper (Table 5). The difference between
genetic variation or parsimony informative sites within Piper, versus Piper compared
to the outgroup, and the high relative rate indicates, that most of the parsimony
Asaroideae *
Aristolochioideae (excl. Thottea)
*
Saururaceae *
Thottea *
Lactoridaceae *
Manekia/Zippelia *
Piper s.l. *
Peperomia*
-0,18
-0,16
-0,14
-0,12
-0,1
-0,08
-0,06
-0,04
-0,02
0
Piperales groups
d
44
informative sites have been accumulated before the radiation of present Piper
species but after the split of the Piper-Peperomia lineage. There are at least two
possible explanations for the resolution contrast between Piper and Peperomia, two
groups that show fairly similar global distribution patterns of extant species (although
there are more species of Piper in tropical Asia). One is that rates continued to
accelerate during the crown group diversification of Peperomia, thus leading to the
accumulation of more variability among species. It seems that Peperomia species are
more often narrow endemics, occupying specialized niches as epiphytes or
succulents, what may lead to small effective population sizes. More work on these
aspects is certainly needed. The other is that rates have slowed down in Piper,
thereby hindering the accumulation of a fairly good amount of historic information. It
may be assumed that the crowngroup diversification in both genera started at about
the same time, considering their comparable global distribution. Unfortunately, in the
absence of reliable fossil material, no molecular dating approach is available for
Piperales. The situation described here parallels findings in Lentibulariaceae, where
trnK/matK sequences with accelerated rates fully resolve the Genlisea-Utricularia
clade (Müller and Borsch 2005) in contrast to its sister clade Pinguicula (Cieslack et
al. 2005).
Relationships of Lactoris
This study provides strong evidence for the sistergroup relationship of Lactoris and
Aristolochioideae and therewith the paraphyly of the family Aristolochiaceae as
currently circumscribed. Depending on the combination of markers between different
studies, or even within the same study, several molecular phylogenetic analyses have
found Aristolochiaceae to be either para- or monophyletic (Duvall, 2000; Qiu et al.,
2000; Doyle and Endress, 2000; Soltis et al., 2000; Savolainen et al., 2000; Hilu et al.,
2003; Borsch et al., 2003; Borsch et al., 2005). However, several of these studies
have not sampled both Asarum and Saruma, which could influence the branching
pattern, and the supports for the relationships were often low. Based on
morphological characters, the position of Lactoris was either hypothesized as close to
Saururaceae (tenuinucellate ovules, development and morphology of stipules;
45
Igersheim and Endress, 1998; González and Rudall, 2001) or to Aristolochiaceae.
The latter hypothesis was favoured by Doyle and Endress (2000) who listed several
characters as potential synapomorphies for these taxa, e.g. presence of a perianth,
presence of tepals, nearly sessile anthers that are strongly extrorse with a broad
connective, and stamens basally fused with the gynoecium. But as already cited by
González and Rudall (2001), most of these characters are symplesiomorphies of the
whole order Piperales or even magnoliids. The basal fusion of stamens with the
gynoecium as in Lactoris is present in Aristolochia but not in Asarum, Saruma or even
Thottea. In addition, the only synapomorphy for Asaroideae and Lactoris cited by
Doyle and Endress (2000) is the extended anther connective. From a molecular point
of view, combining three fast evolving regions (Borsch et al., 2005) or a combination
of nine genes from all three genomes (Qiu et al., 2005) increased the support for the
most frequently found molecular tree (Lactoris sister to Aristolochioideae) to BS 89
(PP 100) in Borsch et al. (2005) or 78 (BS) for the protein coding genes in Qiu et al.
(2005). In addition, the inclusion of Hydnoraceae could only enhance the poly-or
paraphyly of Aristolochiaceae, as the datasets of Nickrent et al. (2002) already
suggests.
Phylogeny of Aristolochiaceae
The two subfamilies of Aristolochiaceae, Aristolochioideae and Asaroideae, are each
well supported in the present analysis. In most analyses, the monophyly of
Aristolochiaceae was only well supported (molecular and morphology), if Lactoris was
not included. Under inclusion of Lactoris the monophyly of Aristolochiaceae was often
only poorly supported, but another position of Asaroideae within Piperales was not
favored, thus the Asaroideae were unresolved and a branch of its own. The inclusion
of this subfamily into Aristolochiaceae is a consequence of the historical treatment.
Asaroideae and Aristolochioideae have never been seriously discussed as two
independent lineages.
The subfamily Asaroideae consisting of the genera Asarum and Saruma, the sister
relationship of these two genera, has been supported by other molecular (Qiu et al.,
2000; Soltis et al., 2000; Neinhuis et al., 2005) and morphological analyses (Kelly and
46
González, 2003). The latter mentioned the PII sieve tube plastid inclusions,
pluricellular stigmatic papillae and seeds with elaiosomes as synapomorphies.
Aristolochioideae include the two major lineages Thottea and Aristolochia s.l..
Additionally, these two lineages are well circumscribed based on molecular (for an
overview, see Neinhuis et al. 2005) and morphological data (for an overview, see
Kelly and González (2003)). The Aristolochia s.l. clade comprises two lineages: the
first containing Isotrema and Endodeca, the second containing Pararistolochia and
Aristolochia. These results are congruent with former studies (González and
Stevenson, 2002; Kelly and González, 2003; Neinhuis et al., 2005). Within
Aristolochia the relationships of the informal group "Howardia" + segregates,
Einomeia and Aristolochia s. str., all accepted by Huber (1985), remain unclear.
These taxa have always been treated as Aristolochia except by Huber (1985), who
further subdivided this clade into what he informally calls ''Howardia'' (hexandrous
Central and South American species), Einomeia (pentandrous Central American
species) and Aristolochia s.str. (Mediterranean and Paleotropical species). He also
cited strong synapomorphies for the monophyly of these groups (e.g. the
pentamerous organisation of the gynostemium in Einomeia (Huber, 1985; González
and Stevenson, 2002, Kelly and González, 2003). The relationship of the segregates
Aristolochia grandiflora complex, Einomeia, “Howardia” p.p. and Aristolochia s. str.
are incongruent among the present and the analysis of Neinhuis et al. (2005), but the
different branching patterns are in both poorly supported.
Phylogeny of Piperaceae
This study confirms the monophyly of the Piperaceae and its subdivision into two
major clades: one including the large genera Piper and Peperomia, and another the
small genera Zippelia and Manekia. The same topology was recovered by Jaramillo
and collaborators (2004) based on the analysis of the slowly evolving genes 18S
rDNA, atpB, and rbcL sequence data for a reduced sampling within Piper and
Peperomia.
The first clade consists of the core Piperaceae with the large pantropical genera Piper
and Peperomia. It is suggested that several outstanding characters in Peperomia
47
(e.g. paniculate inflorescences, peltate leaves) on which Dahlstedt's classification
(Dahlstedt, 1900) has been based have evolved several times independently (Wanke
et al., 2006d). Two subgenera from Dahlstedt (1900), Micropiper and
Sphaerocarpidium (including Erasmia) and the three sections of subgenus
Rhynchophorum could be regarded as monophyletic (Wanke et al., 2006d). This is
also supported by morphological data. A very clear morphological synapomorphy for
the subgenus Micropiper is the so-called pseudocupula at the base of the fruit
(Dahlstedt, 1900). The fruits of all species of the subgenus Sphaerocarpidium are
characterized by a large amount of sticky papillae distributed on the surface. The
three sections of Rhynchophorum are each characterized by a typical fruit shape and
fruit apex. The subgenus Panicularia was described on the basis of paniculate
inflorescences but it is shown that this unusual character has evolved several times
independently. The same accounts for the subgenus Tildenia in which the species
with peltate leaves and shortened internodes are classified. The tuberous species
belonging to the latter subgenus have been classified in a segregate section (Hill,
1907) which forms the basalmost clade in the genus Peperomia.
Piper (including Macropiper, Pothomorphe) is very diverse, varying in inflorescence
position (terminal/axillary) and structure (solitary or clustered spikes or racemes),
sexuality (bisexual or unisexual and then dioecious) and stamen number. The
phylogeny presented here provides little resolution within Piper s.l. compared to
earlier studies using the ITS region (Jaramillo and Manos, 2001). But the current
analysis provides support for the monophyly of a Paleotropical clade, including taxa
from both the Asian tropics and the South Pacific Islands. Piper species in the
Paleotropics differ from their congeners in Tropical America in being dioecious plants
with a climbing growing habit, while their Neotropical counterparts have bisexual
flowers and several growing habits, i.e. shrubs, herbs treelets, but they are never real
climbers. The monophyly of Paleotropical taxa had been suggested before (Callejas,
1986; Jaramillo and Manos, 2001) but it was never well supported. The segregates
Macrostachys, Enckea and Macropiper are also highly supported (BS 100/100, 98/98
and 100/100), however, the present analysis does not provide support for many
48
segregates that had been supported in previous analyses using ITS sequence data
(Jaramillo and Manos, 2001; Jaramillo et al., 2004; Jaramillo and Callejas, 2004).
The second clade consists of Zippelia (monotypic) and Manekia (5 to 6 species). Both
taxa have been associated with Piper (de Candolle, 1866, 1923; Callejas, 1986). The
herbaceous, Asiatic genus Zippelia, with a floral structure similar to Saururus (Omori,
1982), has been placed either in Saururaceae (Blume, 1830; Wu and Wang, 1957;
Heywood, 1993) or in Piperaceae (Engler, 1893; Willis, 1957; Wu and Wang, 1958).
In a cladistic analysis of taxa in Saururaceae and Piperaceae, mostly based on
ontogenetic characters, the similarities between Zippelia and Saururus are identified
as plesiomorphies and Zippelia appears as the basal taxon in Piperaceae (Tucker et
al., 1993). Manekia was not included in this analysis. Zippelia shares several
synapomorphies with Piperaceae, which indicate a close phylogenetic relationship
with other taxa of Piperaceae, e.g. a double vascular cylinder in the stem, lack of
discrete style, single ovule, basal placentation and fusion of two ventral bundles into
one in each carpel (Liang and Tucker, 1995). Zippelia appears to represent a
morphologically transitional genus between Saururaceae and Piperaceae, although
indisputably belonging to the latter (Tucker et al., 1993). Several characters suggest
that Zippelia is a more isolated evolutionary line in Piperaceae, as expressed by
floral development, which is different from the other piperaceous taxa (Liang and
Tucker, 1995), as well as unique glochidiate fruits and a Drusa type of embryo sac
(Lei et al., 2002).
The little studied genus Manekia from Central America, northern South America and
the Atlantic Forest of Brazil is a liana with fleshy, axillary inflorescences, similar to
those of Peperomia.
Phylogeny of Saururaceae
The close relationship of Saururaceae and Piperaceae and the monophyly of
Saururaceae are unquestionable. Comprehensive studies support the relationship
between the two families based on morphology (Doyle and Endress, 2000; Tucker et
al., 1993) and the monophyly of Saururaceae based on molecular data (e.g. Neinhuis
et al., 2005; Jaramillo et al., 2004; Meng et al., 2002, 2003). This is also supported by
49
our results. The relationships found based on trnK/matK support the relationships
found from most other molecular studies ((Saururus + Gymnotheca) and (Anemopsis
+ Houttuynia)). Discrepancy occurs between morphological studies and molecular
results within the family, and even among molecular results from different genomes.
Most of the morphological and especially ontogenetic studies have considered
Saururus to be the most basal branch, sister to all remaining Saururaceae (e.g.
Tucker et al., 1993; Liang, 1995, Lei et al., 1991; Okada, 1986). The morphological
characters, especially those of the flower within the perianthless Piperales have to be
treated with caution. This was already mentioned by Jaramillo et al. (2004) and
Neinhuis et al. (2005), as these plants show a high degree of reduction and fusion,
which makes the detection of reversals or parallelisms more complicated.
Concluding the present study, the evolution of trnK/matK in Piperales presents a case
of striking rate heterogeneity of this gene in flowering plants. Particularly high rates
are present in Peperomia, leading to an internally well resolved gene tree for this
clade. High rates are further reflected in the accumulation of numerous specific length
mutations, including self repairing frame shifts. Nevertheless, a complete reading
frame of the matK gene is maintained, with unrepaired frame shift mutations being
restricted to the downstream end of the gene. Further work will be necessary to
explore possible causes that lead to the observed rate heterogeneity, including
sequence comparisons of other genomic regions, analysis of speciation patterns,
population structures and effective population sizes.
50
Acknowledgements
Wilhelm Barthlott (Bonn) for the opportunity to carry out parts of the lab work in his
institute.
Thanks to Nadin Fliegner (Dresden) for lab assistance. Financial support for this
study came from the Department of Biology, Ghent University and the Friends of the
Botanical Garden, Gent (travel grant to MS). NSF-USA grant DEB-9972600 to MAJ.
TB thanks the Deutsche Forschungsgemeinschaft for a Heisenberg-Scholarship.
Supply of material from Favio Gonzalez, Mario Blanco, Burfard Westlund, Shao-Wu
Meng, Daniel Crawford, Tod Stuessy, R.O. Gardner, O. Vargas and the Botanical
Gardens Dresden, Berlin and Bonn is gratefully acknowledged.
51
Chapter 3 Systematics of the genus Aristolochia
(Aristolochiaceae)
3.1 Systematics of pipevines – Combining morphological
and fast-evolving molecular characters to investigate
the relationships within subfamily Aristolochioideae
(Aristolochiaceae)
This study is in press as:
Wanke, S., Gonzalez, F., Neinhuis, C., 2006. International Journal of Plant
Sciences.
52
Abstract
A combined phylogenetic analysis of the Aristolochioideae was conducted based on
72 morphological characters and molecular datasets (matK gene, trnK intron, trnL
intron, trnL-trnF spacer). The analysis sampled 33 species as the ingroup, including
two species of Thottea and 30 species of Aristolochia and the monotypic genus
Euglypha, which represent all the infrageneric taxa formally described; Saruma henryi
and Asarum caudatum were used as the outgroup. The results corroborate a sister-
group relationship between Thottea and Aristolochia, and the paraphyly of
Aristolochia with respect to Euglypha that consequently should be included into
Aristolochia. Two of the three subgenera within Aristolochia (Isotrema and
Pararistolochia) are shown to be monophyletic, whereas the signal obtained from the
different datasets about the relationships within subg. Aristolochia is low and
conflicting, resulting in collapsed or unsupported branches. The relationship between
the New World and the Old World species of subgenus Aristolochia is conflictive
because morphological data support these two groups as monophyletic, whereas
molecular data show the monophyletic Old World species of Aristolochia nested
within the New World species. A sister group relationship is proposed between A.
lindneri and pentandrous species, which suggests that a group of five species from
central and southern South America (including A. lindneri) could be monophyletic and
sister to Aristolochia subsection Pentandrae, a monophyletic taxon consisting of
about 35 species from southern USA, Mesoamerica, and the West Indies.
53
Introduction
Aristolochiaceae, a member of the Piperales (Borsch et al. 2005, Qiu et al. 2005),
consists of approximately 550 species, most of which are tropical, and subtropical.
Although generic circumscription within the family has been in dispute for about two
centuries (cf. González and Stevenson 2002, Neinhuis et al. 2005), recent authors
recognize four genera in two subfamilies. The subfamily Asaroideae, characterized by
an actinomorphic perianth, consists of two genera: the monotypic Saruma, endemic
from central China, and Asarum with about 86 species from temperate areas of North
America, Europe, and Asia.
The subfamily Aristolochioideae includes Thottea, with less than 30 species with an
actinomorphic perianth, which are restricted to tropical Asia; and Aristolochia
(including the monotypic South American generic segregates Holostylis, and
Euglypha), which is by far the largest genus of the family. Aristolochia has a
monosymmetric perianth and is primarily pantropical but with some offshoots in
subtropical, and temperate areas. The most consistent synapomorphies that relate
these genera are found on the seed coat (González and Stevenson 2002, González
and Rudall 2003), which consists of a two cell-layered testa, and a three cell-layered
tegmen. The cells of the inner layer of the testa have crystals and thickened inner
walls; the three layers of the tegmen are tangentially elongated and fibrous; fibers of
the outer, and inner layers are parallel to the longitudinal axis of the seed, whereas
those of the middle layer are perpendicular to them. In addition, the following unique
combination of characters strongly suggests that these genera form a monophyletic
family: alternate, distichous leaves with palmate, reticulate venation, adaxial
prophylls, oil cells, trimerous perianth (double in Saruma), two (Saruma, Asarum, and
some Thottea spp.) or one (Aristolochia, and the remaining Thottea spp.) whorl(s) of
six stamens (five in Aristolochia subsection Pentandrae Duchartre, and more than six
in most Aristolochia subgenus Pararistolochia Schmidt), six carpels, and pollen in
monosulcate or inaperturate monads.
54
Throughout the taxonomic history of the family, every possible combination of generic
relationships can be found (summarized in González and Stevenson 2002). However,
the most accepted, but yet contradictory classifications at a subfamily level are those
by Schmidt (1935), and Huber (1960, 1985, 1993). Schmidt (1935), following the
classic treatment by Klotzsch (1859), proposed that Asaroideae consists of Asarum
plus Saruma plus Thottea. In contrast, Huber (1985, 1993) transferred Thottea to the
subfamily Aristolochioideae along with Aristolochia.
The genus Aristolochia has been treated in its broad sense by many authors
(Duchartre 1854a, 1864, Hoehne 1942, Pfeifer 1966, 1970, Hou 1984, Nardi 1984,
1991, Ma 1989, among others). However, as many as 15 segregates have been
proposed (for a detailed revision see González and Stevenson 2002), of which
Einomeia, Endodeca, “Howardia”, Isotrema, and Pararistolochia have recently been
used at the generic level especially by Huber (1985, 1993). The splitting of
Aristolochia is based primarily on floral, and fruit characters such as the morphology
of the gynostemium, the gross shape of the perianth, the dehiscence of fruits, and the
morphology of the seeds.
Furthermore, Huber (1985, 1993) recognized two tribes, Isotrematinae (with
Endodeca Raf., and