This article was downloaded by: [Stech, Michael]
On: 24 June 2010
Access details: Access Details: [subscription number 923331511]
Publisher Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 3741 Mortimer Street, London W1T 3JH, UK
Systematics and Biodiversity
Publication details, including instructions for authors and subscription information:
http://www.informaworld.com/smpp/title~content=t913521959
Molecular, phytochemical and morphological characterization of the
liverwort genus Radula in Portugal (mainland, Madeira, Azores)
Michael Stecha; Manuela Sim-Simbc; M. Glória Esquíveld; Leena luísc; Susana Fontinhae; Carlos Lobof;
César Garciac; Soraia Martinsc; Cristiana Vieirag; José Barrosoh; Luis G. Pedroh; A.Cristina S.
Figueiredoh
a
Nationaal Herbarium Nederland, Universiteit Leiden branch, Leiden, the Netherlands b Faculdade de
Ciências de Lisboa, DBV, Centro de Biologia Ambiental, Universidade de Lisboa, Lisboa, Portugal c
Museu Nacional de História Natural, Lisboa, Portugal d Dep. Botânica e Engenharia Biológica, Centro
de Botânica Aplicada à Agricultura, Lisboa, Portugal e Centro de Estudos da Macaronésia,
Universidade da Madeira, Funchal, Madeira, Portugal f Jardim Botânico da Madeira, Funchal, Madeira,
Portugal g Faculdade de Ciências do Porto, Departamento de Botânica, CIBIO/Departamento de
Botânica - FCUP, Porto, Portugal h Faculdade de Ciências de Lisboa, DBV, Instituto de Biotecnologia e
Bioengenharia, Centro de Biotecnologia Vegetal, Universidade de Lisboa, Lisboa, Portugal
Online publication date: 23 June 2010
To cite this Article Stech, Michael , Sim-Sim, Manuela , Esquível, M. Glória , luís, Leena , Fontinha, Susana , Lobo, Carlos ,
Garcia, César , Martins, Soraia , Vieira, Cristiana , Barroso, José , Pedro, Luis G. and Figueiredo, A.Cristina S.(2010)
'Molecular, phytochemical and morphological characterization of the liverwort genus Radula in Portugal (mainland,
Madeira, Azores)', Systematics and Biodiversity, 8: 2, 257 — 268
To link to this Article: DOI: 10.1080/14772001003723579
URL: http://dx.doi.org/10.1080/14772001003723579
PLEASE SCROLL DOWN FOR ARTICLE
Full terms and conditions of use: http://www.informaworld.com/terms-and-conditions-of-access.pdf
This article may be used for research, teaching and private study purposes. Any substantial or
systematic reproduction, re-distribution, re-selling, loan or sub-licensing, systematic supply or
distribution in any form to anyone is expressly forbidden.
The publisher does not give any warranty express or implied or make any representation that the contents
will be complete or accurate or up to date. The accuracy of any instructions, formulae and drug doses
should be independently verified with primary sources. The publisher shall not be liable for any loss,
actions, claims, proceedings, demand or costs or damages whatsoever or howsoever caused arising directly
or indirectly in connection with or arising out of the use of this material.
Systematics and Biodiversity (2010), 8(2): 257–268
Research Article
Molecular, phytochemical and morphological characterization of the
liverwort genus Radula in Portugal (mainland, Madeira, Azores)
MICHAEL STECH1, MANUELA SIM-SIM2,3, M. GLÓRIA ESQUÍVEL4, LEENA LUÍS3, SUSANA FONTINHA5,
CARLOS LOBO6, CÉSAR GARCIA3, SORAIA MARTINS3, CRISTIANA VIEIRA7, JOSÉ BARROSO8,
LUIS G. PEDRO8 & A. CRISTINA S. FIGUEIREDO8
1
Nationaal Herbarium Nederland, Universiteit Leiden branch, PO Box 9514, 2300 RA Leiden, the Netherlands
Universidade de Lisboa, Faculdade de Ciências de Lisboa, DBV, Centro de Biologia Ambiental, C2, Campo Grande, 1749-016 Lisboa,
Portugal
3
Museu Nacional de História Natural, Jardim Botânico/CBA, Rua da Escola Politécnica, n◦ 58, 1250-102 Lisboa, Portugal
4
Centro de Botânica Aplicada à Agricultura, Dep. Botânica e Engenharia Biológica, ISA, UTL, 1399 Lisboa, Portugal
5
Centro de Estudos da Macaronésia, Universidade da Madeira, Campus da Penteada, 9000-390 Funchal, Madeira, Portugal
6
Jardim Botânico da Madeira, Caminho do Meio, Quinta do Bom Sucesso, 9050-251 Funchal, Madeira, Portugal
7
CIBIO/Departamento de Botânica – FCUP, Faculdade de Ciências do Porto, Departamento de Botânica, Edifı́cio FC4, Rua Do Campo
Alegre, s/n◦ , 4169-007 Porto, Portugal
8
Universidade de Lisboa, Faculdade de Ciências de Lisboa, DBV, Instituto de Biotecnologia e Bioengenharia, Centro de Biotecnologia
Vegetal, C2, Campo Grande, 1749-016 Lisboa, Portugal
Downloaded By: [Stech, Michael] At: 06:32 24 June 2010
2
(Received 4 June 2009; revised 16 November 2009; accepted 23 February 2010)
Relationships of the eight species of the liverwort genus Radula occurring in Portugal (mainland, the Madeira and Azores
archipelagos), including the Macaronesian endemics R. jonesii and R. wichurae, were evaluated based on molecular,
phytochemical and morphological–anatomical data. Maximum parsimony and Bayesian analyses were performed with
sequences from three plastid DNA markers (trnS GGA -rps4 spacer, rps4 gene, trnLUAA intron), volatile oil compounds, as
well as qualitative morphological–anatomical characters. In addition, the molecular data were subjected to maximum
likelihood analysis. The eight taxa, R. aquilegia, R. carringtonii, R. complanata, R. holtii, R. jonesii, R. lindenbergiana, R.
nudicaulis and R. wichurae, can be clearly distinguished from each other, either by molecular data alone or by combination
of characters from all three data sets. Radula aquilegia is monophyletic according to the molecular data, but shows
considerable, yet undescribed intraspecific morphological and phytochemical variability. Recognition of R. complanata and
R. lindenbergiana as separate species, previously based solely on the paroecious vs. dioecious sexual condition, is
moderately supported by the molecular phylogenetic analyses and strongly supported by the phytochemical data. The
Radula species, narrowly distributed in Macaronesia and Atlantic Europe, probably have two different origins. For Radula
holtii and R. nudicaulis, connections with Radula species from the Neotropics are indicated. The other species, among them
the two Macaronesian endemics, are closely related with the R. complanata/R. lindenbergiana complex, which is widely
distributed in the northern hemisphere.
Key words: Macaronesia, Portugal, Radula, species characterization, trnS-rps4 region, trnLUAA intron, volatile oils
Introduction
Radula Dumort., the only genus of the family Radulaceae (Porellanae, Jungermanniidae, Marchantiophyta),
comprises about 150 (–200) species (Frey & Stech,
2009). The genus is morphologically characterized by incubous, conduplicate-bilobed lateral leaves with the postical lobe (lobule) markedly smaller than the antical
lobe. Radula has traditionally been postulated to repreCorrespondence to: Michael Stech. E-mail: stech@nhn.
leidenuniv.nl
ISSN 1477-2000 print / 1478-0933 online
C 2010 The Natural History Museum
DOI: 10.1080/14772001003723579
sent an isolated lineage within leafy liverworts, based on
morphological–anatomical (e.g. nearly always Radula-type
branching, total absence of underleaves, lack of transfer
cells in the gametophyte-sporophyte junction), cytological
and phytochemical characters (Schuster, 1980a; Yamada,
1988).
Radula’s species distribution is mainly tropical, but
extends as far as southern Greenland in the North and
subantarctic Campbell Island in the South (Renner &
Braggins, 2004). In Europe, the highest diversity occurs in
Atlantic Europe and the Macaronesian archipelagos. Eight
Downloaded By: [Stech, Michael] At: 06:32 24 June 2010
258
M. Stech et al.
Radula species are known to occur in Portugal (Bouman &
Dirkse, 1990; Söderström et al., 2002; Luı́s et al., 2005):
R. aquilegia (Hook.f. & Taylor) Gottsche, Lindenb. & Nees,
R. carringtonii J.B. Jack, R. complanata (L.) Dumort., R.
holtii Spruce, R. lindenbergiana Gottsche ex C. Hartm.,
R. nudicaulis Steph., and the Macaronesian endemics R.
jonesii Bouman, Dirkse & Yamada and R. wichurae Steph.
Of the latter two species, Radula wichurae is classified
as ‘vulnerable’ (Sim-Sim et al., 2008) and R. jonesii as
‘endangered’ (IUCN, 2008). Three species are known to
mainland Portugal (R. complanata, R. holtii, R. lindenbergiana), seven to Madeira (R. aquilegia, R. carringtonii,
R. holtii, R. jonesii, R. lindenbergiana, R. nudicaulis,
R. wichurae), and six to the Azores (R. aquilegia, R.
carringtonii, R. holtii, R. lindenbergiana, R. nudicaulis, R.
wichurae). In the study area, the genus Radula is abundant
on tree trunks and branches (more rarely on leaves),
periodically wet rocks or shaded vertical rock walls, soil
slopes, and frequently growing on other bryophytes. Its
diversity and abundance becomes higher in mountainous
areas of the Macaronesian archipelagos (Luı́s et al., 2005).
Resolving intrageneric relationships within Radula
has long been problematic due to the uniformity and
phylogenetic uninformativeness of many morphological–
anatomical and karyological characters (e.g. Renner &
Braggins, 2004; Zheng & Zhu, 2009). Earlier classifications such as the distinction of 11 sections (Castle, 1937,
1969; Grolle, 1970) were considered artificial by Yamada
(1979) and Schuster (1980a, 1980b), who distinguished
three subgenera (Cladoradula, Odontoradula, Radula) or
four subgenera (Cladoradula, Metaradula, Odontoradula,
Radula), respectively. Both systems were based on revisionary work on species from different geographic regions
(Asian vs. mainly Laurasian taxa) and considered different morphological–anatomical characters (cf. the overview
in Renner & Braggins, 2004). In contrast, Jones (1977),
in his revision of the African Radula species, refrained
from using intrageneric ranks and only indicated informal
groups. Although revisionary work on the genus Radula is
going on (e.g. Yamada, 1999; So, 2005a, 2005b, 2006) and
new potentially informative characters are being employed
(Renner & Braggins, 2005), the intrageneric classification
remains problematic as long as no worldwide revision and
no molecular phylogenetic studies have been carried out.
Not only the intrageneric classification but also the delimitation of several species, among them European species
that are subject of the present study, is problematic. For
example, Radula lindenbergiana was considered a subspecies of R. complanata by Schuster (1980b), and several
authors believed that sterile plants cannot be referred to
either species based on morphology or habitat preferences
(Bouman & Dirkse, 1990; Paton, 1999; Luı́s et al., 2005;
Schumacker & Váňa, 2005). However, in European floras
and checklists R. lindenbergiana is still often treated as
a separate species (e.g. Schumacker & Váňa, 2005; Frey
et al., 2006). The distinction between R. aquilegia and R.
carringtonii is not completely clear either, which holds true
especially for the Madeiran and some Azorean populations,
as the diagnostic morphological characters used to identify
these two species are not always reliable (e.g. Bouman &
Dirkse, 1990; Paton, 1999).
Although systematic work on the Macaronesian Radula
species led to improved morphological species circumscriptions (Bouman & Dirkse, 1990), independent data sets are
needed to evaluate the significance of the morphological
characters and to assess the phylogenetic and biogeographic
relationships of the respective taxa. As the genus Radula is
well-known for its diversity of chemical compounds, we recently evaluated the chemotaxonomical value of volatile oil
characters to distinguish the Portuguese Radula species by
similarity analysis (Figueiredo et al., 2009). These data allow characterizing of all morphological species, except for
R. aquilegia, which appeared as separated into two groups,
and the closely related R. complanata and R. lindenbergiana.
In the present paper we provide the first molecular analysis, based on plastid trnS GGA -rps4 spacer, rps4 gene, and
trnLUAA intron sequences, of European Radula species,
focusing on Portugal (mainland, Madeira and Azores
archipelagos). Similar to recent molecular studies on the
liverwort genus Plagiochila (Sim-Sim et al., 2005) and the
moss genus Campylopus (Stech et al., 2007) in Madeira, we
aim at obtaining a molecular species inventory of the genus
Radula in Portugal. Additionally, we intend to provide
improved species circumscription of European Radula
species based on comparative phylogenetic analyses of
molecular, phytochemical, and morphological–anatomical
data as well as results from morphological revision of
herbarium material.
Materials and methods
Plant material
Material of Radula species from Portugal was collected during field trips by the authors to mainland Portugal (2001–2008), Madeira (2001–2008) and Azores
(2007–2008). For R. complanata, material from Switzerland (collected by one of the authors in 2008) was used in the
molecular and phytochemical analyses, as no suitable material from Portugal was available. In addition to the new collections, herbarium material from LISU, MADS, MADM,
MADJ, PO, COI, S, L and B was used for morphological
studies and partly for molecular analysis. In total, about 600
Radula specimens were determined and studied morphologically. Of these, 52 specimens were selected for detailed
morphological–anatomical analysis, because they were recently collected, well-developed and/or presenting fruiting structures, and phytochemical and/or molecular data
were available as well. Voucher information and geographic
origins of the specimens included in the phylogenetic
Characterization of the liverwort Radula in Portugal
259
Downloaded By: [Stech, Michael] At: 06:32 24 June 2010
Table 1. Geographic origin and voucher information of the Radula specimens and outgroup representatives analysed. Specimens included in the morphological
analysis are indicated by an asterisk. Specimens included in the phytochemical and molecular analyses are identified by their abbreviations used in Figueiredo et al.
(2009) or by their GenBank accession numbers, respectively.
Taxon
Abbr.
Geographic origin
Voucher no. (herb.)
Abbr. in
Figueiredo et al. Acc. no.
(2009)
trnS-rps4
R. antilleana Castle
∗
R. aquilegia (Hook.f. & Taylor) Gottsche,
Lindenb. & Nees
∗
R. aquilegia
∗
R. aquilegia
∗
R. aquilegia
∗
R. aquilegia
∗
R. aquilegia
∗
R. aquilegia
∗
R. boryana (F. Weber) Nees
∗
R. carringtonii J.B.Jack
∗
R. carringtonii
∗
R. carringtonii
∗
R. carringtonii
∗
R. carringtonii
∗
R. carringtonii
∗
R. carringtonii
∗
R. carringtonii
∗
R. carringtonii
∗
R. carringtonii
∗
R. carringtonii
∗
R. carringtonii
∗
R. carringtonii
∗
R. complanata (L.) Dumort.
∗
R. complanata
R. complanata
R. episcia Spruce
R. grandis Steph.
R. gottscheana Taylor
∗
R. holtii Spruce
∗
R. holtii
∗
R. holtii
∗
R. holtii
∗
R. jonesii Bouman, Dirkse & Yamada
∗
R. lindenbergiana Gottsche ex C. Hartm.
∗
R. lindenbergiana
∗
R. lindenbergiana
∗
R. lindenbergiana
∗
R. lindenbergiana
∗
R. lindenbergiana
∗
R. lindenbergiana
∗
R. lindenbergiana
∗
R. nudicaulis Steph.
∗
R. nudicaulis
∗
R. nudicaulis
∗
R. nudicaulis
∗
R. nudicaulis
∗
R. wichurae Steph.
∗
R. wichurae
∗
R. wichurae
∗
R. wichurae
∗
R. wichurae
∗
R. wichurae
∗
R. wichurae
∗
R. wichurae
∗
R. wichurae
∗
R. wichurae
∗
R. wichurae
Outgroup
Leptoscyphus cleefii Fulford
–
a1
Peru
Madeira
282919 (B)
232860 (LISU), 3276 (MADJ)
–
M07
GU191807
GU191821
GU189044
GU189055
a2
a3
a4
a5
a6
a7
–
ca1
ca2
ca3
ca4
ca5
ca6
ca7
ca8
ca9
ca10
ca11
ca12
ca13
co1
co2
co3
–
–
–
h1
h2
h3
h4
j1
l1
l2
l3
l4
l5
l6
l7
l8
n1
n2
n3
n4
n5
w1
w2
w3
w4
w5
w6
w7
w8
w9
w10
w11
Azores (Flores)
Madeira
Madeira
Azores (Pico)
Azores (Pico)
Azores (Pico)
S. Tomé e Prı́ncipe
Madeira
Madeira
Madeira
Azores (Flores)
Madeira
Madeira
Azores (Pico)
Azores (Pico)
Azores (Pico)
Azores (Pico)
Azores (Flores)
Azores (Flores)
Azores (Flores)
Switzerland
Switzerland
Germany
Peru
New Zealand
Peru
Portugal mainland
Madeira
Madeira
Portugal mainland
Madeira
Portugal mainland
Portugal mainland
Madeira
Madeira
Madeira
Madeira
Portugal mainland
Portugal mainland
Madeira
Madeira
Madeira
Madeira
Madeira
Madeira
Madeira (Porto Santo)
Azores (Flores)
Azores (Flores)
Madeira
Azores (Graciosa)
Azores (Terceira)
Azores (Terceira)
Azores (Pico)
Azores (S. Jorge)
Azores (S. Jorge)
232879 (LISU)
232855 (LISU), 3084 (MADJ)
232859 (LISU), 3085 (MADJ)
232878 (LISU)
232876 (LISU)
232877 (LISU)
233167 (LISU)
232858 (LISU)
233270 (LISU)
233269 (LISU)
232882 (LISU)
232856 (LISU), 3083 (MADJ)
232857 (LISU)
232887 (LISU)
232886 (LISU)
232884 (LISU)
232885 (LISU)
232880 (LISU)
232881 (LISU)
232883 (LISU)
232898 (LISU)
232899 (LISU)
Frey 2–883 (herb. W. Frey)
256171 (B)
Frahm 28–8 (herb. W. Frey)
256159 (B)
232895 (LISU)
232861 (LISU), 8450 (MADJ)
232862 (LISU)
232894 (LISU)
232863 (LISU), 3277 (MADJ)
232896 (LISU)
233267 (LISU)
232867 (LISU), 3268 (MADJ)
232864 (LISU)
232866 (LISU), 3086 (MADJ)
232865 (LISU), 3087 (MADJ)
232897 (LISU)
152346 (LISU)
232872 (LISU), 3280 (MADJ)
232871 (LISU), 3275 (MADJ)
232870 (LISU), 3081 (MADJ)
232869 (LISU), 3082 (MADJ)
232868 (LISU)
232874 (LISU), 3278 (MADJ)
232875 (LISU), 3279 (MADJ)
232893 (LISU)
232888 (LISU)
232873 (LISU), 3088 (MADJ)
232889 (LISU)
232890 (LISU)
232891 (LISU)
232892 (LISU)
232901 (LISU)
232900 (LISU)
AF08
M06a
M06b
AP07a
AP07c
AP07b
–
M08
–
–
AF08c
M06a
M06b
AP07a
AP07b
AP07c
AP07d
AF08a
AF08b
AF08d
SW08a
SW08b
–
–
–
–
ML08b
M06a
M06b
ML08a
M07
ML08
–
M06c
M05
M06a
M06b
ML99
ML06
M08
M07
M06a
M06b
M06c
M07
PS08
AF08
AF99a
M06
AG99b
AT07a
AT07b
AP07
ASJ08a
ASJ08b
GU191822
–
–
–
–
–
GU191805
GU191816
GU191817
GU191819
GU191818
–
–
–
–
–
–
–
–
–
GU191827
GU191826
GU191825
GU191808
GU191811
GU191806
–
–
–
–
GU191820
GU191823
GU191824
–
–
–
–
–
–
GU191810
GU191809
–
–
–
GU191813
GU191812
GU191814
GU191815
–
–
–
–
–
–
–
GU189056
–
–
–
–
–
GU189042
GU189051
GU189052
GU189040
GU189053
–
–
–
–
–
–
–
–
–
GU189059
GU189058
AY007636
GU189045
AY007635
GU189043
GU189060
–
–
–
GU189054
GU189057
GU189041
–
–
–
–
–
–
GU189047
GU189046
–
–
–
GU189049
GU189048
GU189050
GU189039
–
–
–
–
–
–
–
–
see Vanderpoorten &
Long (2006)
Germany
see Vanderpoorten & Long
(2006)
Stech B950821.1 (L)
–
EU661835
DQ176709
–
GU191804
AY007634
Ptilidium ciliare (L.) Hampe
–
Acc. no.
trnL
260
M. Stech et al.
analyses of molecular, phytochemical and morphological
characters are summarized in Table 1. Detailed information on further morphologically studied material is available
from the authors.
DNA extraction, PCR and sequencing
Downloaded By: [Stech, Michael] At: 06:32 24 June 2010
Plant material was thoroughly cleaned with distilled water
and ultrasonic treatment. DNA extraction and PCR amplification of trnLUAA intron with primers CM and DM was
performed as described in Frey et al. (1999), amplification
of trnS GGA -rps4 spacer and rps4 gene with primers trnS-F
and rps5′ as described in Hernández-Maqueda et al. (2008).
R
PCR products were cleaned with the SureClean
solution
R
(Bioline) or the Wizard DNA Clean-up kit (Promega).
Manual or automated sequencing was performed as described in Frey et al. (1999) or by Macrogen Inc., South
Korea (www.macrogen.com), respectively. GenBank accession numbers are given in Table 1.
Alignment, sequence analysis and
phylogenetic reconstructions
Based on the criteria established in Kelchner (2000) and
Quandt & Stech (2005), DNA sequences were manually
R
v0.995 (Müller et al., 2006). Sealigned using PhyDE
lection of outgroup representatives was based on sequence
availability, alignability of the included non-coding regions,
and recent molecular phylogenetic reconstructions of liverworts. Ptilidium ciliare (L.) Hampe (Ptilidiaceae) was used
as close outgroup based on the sister-group relationship
of Ptilidium to the remaining Porellanae (e.g. He-Nygrén
et al., 2006; see also Ahonen, 2004), and Leptoscyphus
cleefii Fulford (Lophocoleaceae) as more remote outgroup
from Jungermannianae, the sister-group of Porellanae (e.g.
Forrest & Crandall-Stotler, 2004; He-Nygrén et al., 2006;
Heinrichs et al., 2007). Sequences of the latter species were
taken from Vanderpoorten & Long (2006).
Phylogenetic reconstructions according to the maximum
parsimony (MP) and maximum likelihood (ML) optimality criteria were performed using PAUP 4.0b10 (Swofford,
2002). Heuristic searches under parsimony were implemented using random sequence addition with 1000 replicates and employing the default settings otherwise. Gaps
were treated as missing data. Heuristic bootstrap searches
under parsimony were performed with 10 000 replicates and
100 random addition cycles per bootstrap replicate with the
same options in effect.
Maximum likelihood analyses were executed assuming
a general time reversible (GTR) model and a rate variation
among sites following a gamma distribution. GTR+Ŵ+I
was chosen as the model that best fits the combined data
according to the AIC criterion as evaluated by MrModeltest v2.3 (Nylander, 2004) employing MrMTgui (Nuin,
2005). The proposed settings by MrModeltest were exe-
cuted in PAUP: Basefreq = (0.3765 0.1238 0.1321), Nst =
6, Rmat = (0.8027 3.1192 0.2388 0.9415 3.3011), Shape =
3.8729, Pinvar = 0.4675. Model testing for the three partitions (trnS-rps4 spacer, rps4 gene, trnL intron) separately
indicated either GTR+Ŵ or GTR+I as best fitting models. Therefore, a Bayesian analysis (see below) with three
separate partitions was carried out to test for incongruence
with the analysis based on GTR+Ŵ+I as indicated by MrModeltest for the combined data set. Heuristic likelihood
bootstrap searches were performed with 100 replicates.
For further measurement of support, posterior probabilities were calculated using MrBayes v3.1 (Huelsenbeck &
Ronquist, 2001). As in the maximum likelihood analysis,
the GTR model of nucleotide substitution was employed,
assuming site-specific rate categories following a gamma
distribution. The a priori probabilities supplied were those
specified in the default settings of the program. Posterior
probability (PP) distributions of trees were created using
the Metropolis-coupled Markov chain Monte Carlo (MCMCMC) method. Four runs with four chains (106 generations
each) were run simultaneously, with the temperature of the
single heated chain set to 0.2. Chains were sampled every
10 generations and the respective trees written to a tree
file. Calculations of the consensus tree and the posterior
probability of clades were performed based upon the trees
sampled after the chains converged (we used 25% as default).
Analysis of phytochemical characters
The volatile oil compounds were isolated by distillationextraction and analysed by gas chromatography (GC) and
gas chromatography-mass spectrometry (GC-MS) as reported by Figueiredo et al. (2009). All 48 specimens
analysed in Figueiredo et al. (2009), plus R. boryana (F.
Weber) Nees as outgroup representative, were included
in the present study. In total, 101 chemical compounds
were considered, including those 83 reported by Figueiredo
et al. (2009), seven further distinct compounds identified
based on mass spectra only (putative sesquisabinene A,
isobazzanene, epi-β-santalene, γ -gurjunene, taynudol, cyclocolorenone and an unidentified component R), and 11
only present in the outgroup (α-terpinene, p-cymene, γ terpinene, 2,5-dimethyl styrene, terpinolene, p-cymen-8ol, thymol, carvacrol, α-gurjunene, and cis-calamenene). A
data matrix, which is available on request, was generated
based on the presence or absence of the respective compounds in the 49 samples (0 = absent, 1 = present). This
matrix was analysed with PAUP 4.0b10 according to the
maximum parsimony principle and subjected to a Bayesian
analysis with MrBayes v3.1 (data type = restriction site
for character states 0/1, coding = variable). Heuristic bootstrap searches under parsimony were performed with 1000
replicates and 10 random addition cycles per bootstrap
replicate.
Characterization of the liverwort Radula in Portugal
Downloaded By: [Stech, Michael] At: 06:32 24 June 2010
Analysis of morphological-anatomical
characters
For each of 51 Radula specimens (plus R. boryana as outgroup representative) for which phytochemical data and/or
DNA sequence data was generated, three stems from the
main axis were dissected off and each stem analysed at
5–10 mm from the apex. Both quantitative and qualitative
(discrete) characters were recorded, but in accordance with
the literature a preliminary analysis indicated that the qualitative characters were more useful as diagnostic characters
for distinguishing the Radula species than the quantitative
ones. Therefore, we used 13 qualitative characters (Table 2)
for phylogenetic analysis of the respective 52 specimens. Of
these, 11 characters were coded as binary characters (0/1),
whereas two characters were coded as multistate characters (0, 1 and 2). The respective character states for ‘perianth shape’ and ‘sexual condition’ could only be recorded
from fertile specimens; in sterile specimens the missing
data were indicated by ‘?’. The matrix was analysed with
PAUP 4.0b10 according to the maximum parsimony principle and subjected to a Bayesian analysis with MrBayes v3.1
(data type = standard for character states 0/1 and multistate
characters, coding = variable). Heuristic bootstrap searches
under parsimony were performed with 1000 replicates and
10 random addition cycles per bootstrap replicate.
Results
Sequence characterization and
alignment
Lengths of the trnS-rps4 spacer vary between 31 nucleotides (nt) (R. antilleana Castle) and 125 nt (R.
episcia Spruce) across the whole data set and within Radula.
Lengths of the trnL intron range from 303 to 372 nt across
the whole data set, within Radula from 341 (R. aquilegia)
to 372 nt (R. boryana). The rps4 gene is 608 nt in all sequenced samples. For R. holtii only the trnL intron could
be sequenced.
The combined alignment comprises 1263 positions
(trnS-rps4 positions 1–215, rps4 216–823, trnL 824–1263),
of which 68 were excluded from phylogenetic analysis due
to incomplete sequencing. Of the remaining 1195 included
characters, 333 (27.9%) are variable, and 217 (18.2%, or
65.2% of the variable positions) are parsimony-informative
(35 in trnS-rps4, 115 in rps4, 67 in trnL).
261
ML bootstrap values ≥ 60% and significant (≥ 94) posterior
probabilities (PP) indicated above the branches. Bayesian
analyses using the GTR+Ŵ+I model for the whole data
set versus individual models (GTR+Ŵ or GTR+I, respectively) for the three partitions resulted in the same topology
and (almost) equal PPs.
In all analyses, the basal split between the afroneotropical species Radula boryana and R. gottscheana
Taylor and a clade including all other Radula species receives maximum support. Within the latter clade, two of
the Portuguese species, R. holtii and R. nudicaulis, form a
maximally supported clade with the neotropical species R.
antilleana and R. episcia. The remaining European Radula
representatives form a clade, also with maximum support,
of which the sister-group relationship to R. grandis Steph.
from New Zealand is not well-supported. Radula aquilegia occupies a basal position within this second European
Radula clade, followed by R. jonesii. The sister-group relationships of R. complanata and R. lindenbergiana as well as
of R. carringtonii and R. wichurae are maximally supported.
Separate analyses of trnS-rps4 spacer, rps4 gene and trnL
intron, respectively, resulted in less well-resolved (consensus) trees, but did not reveal inconsistencies between the
different markers as inferred from visual comparison of the
respective tree topologies (data not shown).
Analysis of phytochemical characters
The maximum parsimony calculation of the phytochemical matrix (101 characters, of which 85 were parsimonyinformative) yielded two trees (lengths 167, CI excluding
uninformative characters 0.559, RI 0.935, RC 0.560). Both
trees differ in the branching order of the basal clades that
are thus resolved as a polytomy in the strict consensus tree
shown in Fig. 2. The specimens of R. carringtonii, R. complanata, R. holtii, R. nudicaulis and R. wichurae each form
clades with significant bootstrap support (99−100% BS)
and significant PP values of 96−100 (except R. wichurae).
Radula holtii is resolved at a basal position without support. The Madeiran and Azorean specimens of R. aquilegia
are separated on two clades with 99−100% BS and a PP
of 98 for the clade of the Azorean specimens. Concerning
species relationships, only the position of R. lindenbergiana
as sister to R. wichurae receives significant BS (99%) and
an almost significant PP of 93.
Molecular phylogenetic analysis
Analysis of morphological–anatomical
characters
The MP analysis retained a single most parsimonious tree
(length 496, CI excluding uninformative characters 0.754,
RI 0.889, RC 0.730). In the ML analysis a single optimal
tree with the same topology as the MP tree was found (lnL =
−3861.34269). The ML tree is shown in Fig. 1, with MP and
The eight Radula species referred to Portugal can be distinguished morphologically by a set of diagnostic characters
(Table 3), as inferred from the study of herbarium material.
Characters diagnostic for single species are the presence of
caducous leaves in Radula nudicaulis, the trumpet-shaped
262
M. Stech et al.
Downloaded By: [Stech, Michael] At: 06:32 24 June 2010
Table 2. Matrix of 13 morphological–anatomical characters of the 52 Radula specimens subjected to phylogenetic analysis.
Characters: 1. ramification: irregularly pinnate (0), regularly pinnate (1); 2. cell walls in the medullar cells: thick (0), thin (1); 3. leaf
lobe extension: lobes not covering stem in dorsal view (0), lobes covering stem in dorsal view (1); 4. overlapping of leaves: not
overlapping in dorsal view (0), overlapping in dorsal view (1); 5. apical margin of lobe and lobule: unbordered (0), bordered with
small hyaline cells (1); 6. lobule shape: subquadrate (0), subrectangular (1), ovate (2); 7. lobule inflation: flat (0), inflated (1); 8. lobule
apex: rounded (0), bluntly acute (1); 9. keel orientation: straight (0), convex (1), concave (2); 10. gemmae: absent (0), present (1); 11.
caducous leaves: absent (0), present (1); 12. perianth shape: subrectangular (0), trumpet-shaped (1); 13. sexual condition: dioecious
(0), monoecious (paroecious) (1).
Taxon
sample
1
2
3
4
5
6
7
8
9
10
11
12
13
R. boryana
R. aquilegia
R. aquilegia
R. aquilegia
R. aquilegia
R. aquilegia
R. aquilegia
R. aquilegia
R. carringtonii
R. carringtonii
R. carringtonii
R. carringtonii
R. carringtonii
R. carringtonii
R. carringtonii
R. carringtonii
R. carringtonii
R. carringtonii
R. carringtonii
R. carringtonii
R. carringtonii
R. complanata
R. complanata
R. holtii
R. holtii
R. holtii
R. holtii
R. jonesii
R. lindenbergiana
R. lindenbergiana
R. lindenbergiana
R. lindenbergiana
R. lindenbergiana
R. lindenbergiana
R. lindenbergiana
R. lindenbergiana
R. nudicaulis
R. nudicaulis
R. nudicaulis
R. nudicaulis
R. nudicaulis
R. wichurae
R. wichurae
R. wichurae
R. wichurae
R. wichurae
R. wichurae
R. wichurae
R. wichurae
R. wichurae
R. wichurae
R. wichurae
a1
a2
a3
a4
a5
a6
a7
ca1
ca2
ca3
ca4
ca5
ca6
ca7
ca8
ca9
ca10
ca11
ca12
ca13
co1
co2
h1
h2
h3
h4
j1
l1
l2
l3
l4
l5
l6
l7
l8
n1
n2
n3
n4
n5
w1
w2
w3
w4
w5
w6
w7
w8
w9
w10
w11
1
0
0
0
1
1
1
0
0
1
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
1
0
0
0
0
0
1
0
0
0
1
1
0
0
1
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
1
1
0
0
0
0
0
0
0
0
0
1
1
1
1
1
0
0
0
0
0
0
0
0
0
0
0
0
1
1
1
1
0
1
1
1
1
1
1
0
0
0
1
1
1
1
1
1
1
1
0
0
0
0
0
0
1
0
1
1
1
1
1
0
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
0
1
1
1
1
0
1
1
1
0
0
1
1
0
1
1
1
1
1
1
1
1
1
0
0
0
0
1
0
1
1
1
1
1
1
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
1
1
1
2
1
0
1
1
1
1
1
0
0
0
1
0
1
0
0
0
0
0
0
(01)
0
0
0
0
0
0
0
1
(01)
1
1
(01)
1
1
(01)
0
0
0
0
0
2
2
(02)
2
(02)
2
(02)
2
2
(02)
(02)
0
1
0
1
1
1
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
1
1
1
0
0
0
0
0
0
0
0
0
0
0
0
(01)
(01)
(01)
(01)
(01)
(01)
(01)
0
0
(01)
(01)
(01)
0
0
0
0
0
(01)
0
(01)
1
1
0
0
0
0
1
1
(01)
1
1
1
(01)
1
1
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
1
1
0
1
1
1
1
1
0
0
(01)
0
0
1
0
0
0
0
0
0
0
0
(01)
0
(01)
(01)
(01)
(02)
(01)
2
0
(01)
0
1
1
(01)
(01)
(01)
(01)
(01)
(01)
(02)
(02)
(02)
(02)
(02)
(02)
(02)
(02)
(02)
(02)
(02)
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
0
0
0
0
0
1
1
1
1
1
1
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
1
1
1
0
0
0
0
0
0
0
0
0
0
0
?
?
?
?
?
?
0
?
?
?
?
?
?
0
?
?
?
?
0
?
?
0
0
1
?
1
?
0
?
?
?
0
0
0
0
0
?
?
?
?
?
?
?
?
0
0
0
?
?
?
?
?
?
?
?
?
?
?
0
?
?
?
?
?
?
0
?
?
?
?
0
?
?
1
1
1
?
1
?
1
?
?
?
0
0
0
0
0
?
?
?
?
?
?
?
?
0
0
0
?
?
?
?
?
1
0
0
1
1
1
1
1
1
1
1
1
1
1
Characterization of the liverwort Radula in Portugal
263
R. wichurae w1 Madeira
95/97/100
R. wichurae w2 Madeira (Porto Santo)
R. wichurae w3 Azores (Flores)
100
R. wichurae w4 Azores (Flores)
R. carringtonii ca1 Madeira
R. carringtonii ca2 Madeira
95/93
/100
84/95/100
R. carringtonii ca3 Madeira
R. carringtonii ca4 Azores (Flores)
R. complanata co1 Switzerland
87/87
/100 R. complanata co2 Switzerland
Downloaded By: [Stech, Michael] At: 06:32 24 June 2010
79/86/100
100
R. complanata co3 Germany
R. lindenbergiana l1 Portugal
63/62
R. lindenbergiana l2 Portugal
/--
100
R. jonesii j1 Madeira
82/--/--
R. aquilegia a1 Madeira
86/79/96
R. aquilegia a2 Azores (Flores)
R. grandis New Zealand
100
75/91/-66/60/--
R. episcia Peru
R. holtii h1 Portugal
R. nudicaulis n1 Madeira
100
100
R. nudicaulis n2 Madeira
97/98/100
R. antilleana Peru
100
R. boryana São Tomé
R. gottscheana Peru
Leptoscyphus cleefii
Ptilidium ciliare
0.1
Fig. 1. Maximum likelihood tree based on plastid trnS-rps4 spacer, rps4 gene, and trnL intron sequences of 24 Radula specimens as
well as Leptoscyphus cleefii and Ptilidium ciliare as outgroup representatives. Bootstrap support values ≥ 60% from respective maximum
parsimony and maximum likelihood analyses as well as significant (≥ 94) Bayesian posterior probabilities are given at the branches
(maximum support in all analyses is indicated by ‘100’).
perianth in R. holtii and the lobe and lobule margins being
bordered with small hyaline cells in R. wichurae. Discoid
marginal gemmae on lobes and lobules are only found in
R. complanata and R. lindenbergiana, differentiation be-
tween thick-walled cortical and thin-walled medullary cells
in the stem occurs in R. holtii and R. nudicaulis. However,
species delimitations are not always sufficiently clear when
based solely on morphological–anatomical characters. For
264
M. Stech et al.
100/--
99/(93)
100/96
Downloaded By: [Stech, Michael] At: 06:32 24 June 2010
100/100
99/--
100/98
100/100
100/100
100/96
R. wichurae w1 Madeira
R. wichurae w2 Madeira (Porto Santo)
R. wichurae w3 Azores (Flores)
R. wichurae w4 Azores (Flores)
R. wichurae w5 Madeira
R. wichurae w6 Azores (Graciosa)
R. wichurae w7 Azores (Terceira)
R. wichurae w8 Azores (Terceira)
R. wichurae w9 Azores (Pico)
R. wichurae w10 Azores (São Jorge)
R. wichurae w11 Azores (São Jorge)
R. lindenbergiana l1 Portugal
R. lindenbergiana l3 Madeira
R. lindenbergiana l4 Madeira
R. lindenbergiana l5 Madeira
R. lindenbergiana l6 Madeira
R. lindenbergiana l7 Portugal
R. lindenbergiana l8 Portugal
R. carringtonii ca1 Madeira
R. carringtonii ca4 Azores (Flores)
R. carringtonii ca5 Madeira
R. carringtonii ca6 Madeira
R. carringtonii ca7 Azores (Pico)
R. carringtonii ca8 Azores (Pico)
R. carringtonii ca9 Azores (Pico)
R. carringtonii ca10 Azores (Pico)
R. carringtonii ca11 Azores (Flores)
R. carringtonii ca12 Azores (Flores)
R. carringtonii ca13 Azores (Flores)
R. aquilegia a1 Madeira
R. aquilegia a3 Madeira
R. aquilegia a4 Madeira
R. aquilegia a2 Azores (Flores)
R. aquilegia a5 Azores (Pico)
R. aquilegia a6 Azores (Pico)
R. aquilegia a7 Azores (Pico)
R. nudicaulis n1 Madeira
R. nudicaulis n2 Madeira
R. nudicaulis n3 Madeira
R. nudicaulis n4 Madeira
R. nudicaulis n5 Madeira
R. complanata co1 Switzerland
R. complanata co2 Switzerland
R. jonesii j1 Madeira
R. holtii h1 Portugal
R. holtii h4 Portugal
R. holtii h2 Madeira
R. holtii h3 Madeira
R. boryana São Tomé
Fig. 2. Strict consensus of two most parsimonious trees resulting from maximum parsimony analysis of volatile oil characters of 48
samples from nine Radula species, using R. boryana as outgroup representative. Bootstrap support values ≥ 60% and significant (≥ 94)
Bayesian posterior probabilities are given at the branches.
example, Radula aquilegia generally differs from R. carringtonii by inflated subrectangular rather than flat quadrate
lobules. Some populations of R. aquilegia, however, show
subquadrate and only slightly inflated lobules that may resemble those of R. carringtonii. Moreover, lobule morphology varies between populations of R. aquilegia collected
in distinct islands of the Azores archipelago. The populations collected in the islands of the central group (Pico I.)
present typical inflated subrectangular lobules, which contrasts with the populations from the western group (Flores
I.) whose lobules are less inflated and subquadrate. The
R. aquilegia populations from Madeira Island are less variable and usually have scarcely inflated subquadrate lobules,
which sometimes make their distinction from R. carringtonii difficult. In addition to the morphological differences,
such plants from Flores I. are more fragile than the typical
ones and always found in shady, humid rock crevices and
fissures and walls in the forest. Radula complanata and R.
lindenbergiana are highly similar in all vegetative characters and can only be distinguished in the presence of sexual
structures, since R. complanata is paroecious and R. lindenbergiana dioecious. Radula jonesii could be confused with
R. holtii when sterile, except for the absence of thin-walled
medullary cells of the stem.
In the MP analysis of the 13 diagnostic morphological–
anatomical characters 8000 most parsimonious trees were
recovered (length 38, CI 0.395, RI 0.807, RC 0.318). In the
strict consensus tree (Fig. 3) the samples of R. nudicaulis,
all but one R. aquilegia, and one R. carringtonii form a
basal polytomy together with the outgroup R. boryana.
The specimens of R. holtii and R. wichurae each form
one clade. Their relationships with the specimens of R.
complanata, R. jonesii, R. lindenbergiana, all but one
R. carringtonii, and one R. aquilegia are unresolved. No
Characterization of the liverwort Radula in Portugal
265
Table 3. Morphological–anatomical diagnostic characters of the eight Radula species occurring in Portugal.
Cauloid (stem)
Leaf lobes
Leaf lobules
R. aquilegia
Irregularly to regularly
pinnately branched;
medullary cells
thick-walled
R. carringtonii
Irregularly to (rarely)
regularly pinnately
branched; medullary
cells thick-walled
Irregularly pinnately
branched; medullary
cells thick-walled
Imbricate or rarely
contiguous; generally
covering the stem in
dorsal view; margins
unbordered
Imbricate or rarely
contiguous; generally
covering the stem;
margins unbordered
Imbricate; generally
covering the stem;
margins unbordered
Subrectangular or rarely
subquadrate; inflated;
apex rounded; keel
convex; margins
unbordered
Subquadrate or rarely
subrectangular; flat; apex
rounded; keel straight;
margins unbordered
Subquadrate to
subrectangular; flat; apex
rounded; keel straight to
slightly convex; margins
unbordered
Subquadrate; flat; apex
rounded; keel straight to
convex; margins
unbordered
Subquadrate; flat; apex
from strictly rounded to
bluntly acute; keel
straight to concave;
margins unbordered
Subrectangular to
subquadrate; flat; apex
rounded; keel straight to
slightly convex, rarely
concave; margins
unbordered
Subquadrate; inflated; apex
bluntly acute to rounded;
keel straight to convex;
margins unbordered
Ovate, sometimes almost
quadrate; flat; apex
rounded; keel straight to
slightly concave; margins
bordered with small
hyaline cells
R. complanata
Irregularly pinnately
branched; medullary
cells thin-walled and
hyaline
Regularly pinnately
branched; medullary
cells thick-walled
Contiguous; not covering
the stem; margins
unbordered
R. lindenbergiana
Irregularly to regularly
pinnately branched;
medullary cells
thick-walled
Imbricate to contiguous;
generally covering the
stem; margins
unbordered
R. nudicaulis
Irregularly to regularly
pinnately branched;
medullary cells
thin-walled and hyaline
Irregularly to regularly
pinnately branched;
medullary cells
thick-walled
Imbricate to contiguous;
generally covering the
stem; margins
unbordered
Imbricate; generally
covering the stem;
margins bordered with
small hyaline cells
R. holtii
Downloaded By: [Stech, Michael] At: 06:32 24 June 2010
R. jonesii
R. wichurae
Imbricate to contiguous;
not covering the stem;
margins unbordered
significant bootstrap support and no posterior probabilities
in the respective Bayesian analysis were achieved.
Discussion
All three data sets employed in the present study
(DNA sequence data, volatile oil composition, and
morphological–anatomical characters) contribute to inferring species circumscriptions and phylogenetic relationships of the Radula species referred to Portugal, albeit to different degrees. The molecular phylogenetic reconstruction (Fig. 1) provides the highest resolution in
terms of species relationships, although this may partly
be due to the different taxon sampling with further Radula
species included from other geographic regions. Radula
aquilegia, R. carringtonii, R. holtii, R. jonesii, R. nudicaulis, R. wichurae and the R. complanata/R. lindenbergiana complex can be clearly distinguished from each other
based on the molecular data. The phytochemical data (Fig.
Sexual
reproduction
Asexual reproduction
Dioecious;
perianth
subrectangular
Gemmae and
caducous leaves
absent
Dioecious;
perianth
subrectangular
Gemmae and
caducous leaves
absent
Monoecious
(paroecious);
perianth
subrectangular
Discoid marginal
gemmae present at
lobe and lobule
margins; caducous
leaves absent
Gemmae and
caducous leaves
absent
Monoecious
(paroecious);
perianth
trumpet-shaped
Monoecious
(paroecious);
perianth
subrectangular
Gemmae and
caducous leaves
absent
Dioecious;
perianth
subrectangular
Discoid marginal
gemmae present at
lobe and lobule
margins; caducous
leaves absent
Reproductive
structures not
observed
Caducous leaves
present, especially
in adult plants;
gemmae absent
Gemmae and
caducous leaves
absent
Dioecious;
perianth
subrectangular
2) support these species circumscriptions except for R.
aquilegia, of which the Azorean and Madeiran samples
form separate clades (see discussion below). Support for
the clades of R. aquilegia, R. carringtonii, R. complanata
and R. wichurae is comparable to that obtained in the molecular analyses, and higher for R. lindenbergiana. The latter
is clearly separated from R. complanata when the phytochemical data are analysed in a phylogenetic framework, in
contrast to the similarity analysis performed in Figueiredo
et al. (2009). The morphological–anatomical characters are
less informative in the phylogenetic approach (Fig. 3), with
only the specimens of R. holtii and R. wichurae being resolved as distinct clades, although several characters are
diagnostic for single species or species pairs (cf. results).
The low resolution and absence of statistical support may
result from the low number of characters, but also from a
considerably higher homoplasy and lower synapomorphy
content in the morphological data set, as inferred from the
low rescaled consistency (RC) value of the resulting trees
Downloaded By: [Stech, Michael] At: 06:32 24 June 2010
266
M. Stech et al.
R. wichurae w1 Madeira
R. wichurae w2 Madeira (Porto Santo)
R. wichurae w3 Azores (Flores)
R. wichurae w4 Azores (Flores)
R. wichurae w5 Madeira
R. wichurae w6 Azores (Graciosa)
R. wichurae w7 Azores (Terceira)
R. wichurae w8 Azores (Terceira)
R. wichurae w9 Azores (Pico)
R. wichurae w10 Azores (São Jorge)
R. wichurae w11 Azores (São Jorge)
R. holtii h1 Portugal
R. holtii h4 Portugal
R. holtii h2 Madeira
R. holtii h3 Madeira
R. lindenbergiana l1 Portugal
R. lindenbergiana l3 Madeira
R. lindenbergiana l2 Portugal
R. lindenbergiana l4 Madeira
R. lindenbergiana l5 Madeira
R. lindenbergiana l6 Madeira
R. lindenbergiana l7 Portugal
R. lindenbergiana l8 Portugal
R. complanata co1 Switzerland
R. complanata co2 Switzerland
R. jonesii j1 Madeira
R. carringtonii ca1 Madeira
R. carringtonii ca2 Madeira
R. carringtonii ca3 Madeira
R. carringtonii ca4 Azores (Flores)
R. carringtonii ca5 Madeira
R. carringtonii ca7 Azores (Pico)
R. carringtonii ca8 Azores (Pico)
R. carringtonii ca9 Azores (Pico)
R. carringtonii ca10 Azores (Pico)
R. carringtonii ca11 Azores (Flores)
R. carringtonii ca12 Azores (Flores)
R. carringtonii ca13 Azores (Flores)
R. aquilegia a2 Azores (Flores)
R. nudicaulis n3 Madeira
R. nudicaulis n4 Madeira
R. nudicaulis n5 Madeira
R. nudicaulis n1 Madeira
R. nudicaulis n2 Madeira
R. carringtonii ca6 Madeira
R. aquilegia a1 Madeira
R. aquilegia a3 Madeira
R. aquilegia a4 Madeira
R. aquilegia a5 Azores (Pico)
R. aquilegia a6 Azores (Pico)
R. aquilegia a7 Azores (Pico)
R. boryana São Tomé
Fig. 3. Strict consensus of 8000 most parsimonious trees resulting
from maximum parsimony analysis of morphological-anatomical
characters of 52 samples from nine Radula species, using R.
boryana as outgroup representative.
(0.318 versus 0.560 in the phytochemical and 0.730 in the
molecular analysis). In fact, only three out of the 13 characters show individual RC values of 1.00 (characters 5, 12
and 13, cf. Table 2), while the RC values of the remaining
characters range between 0.08–0.59.
Despite the low resolution of the phylogenetic reconstruction based on morphology, the present study shows
that all eight Radula taxa referred to Portugal can be clearly
distinguished from each other, either by molecular data
alone or by combination of molecular, phytochemical, and
morphological–anatomical characters. Radula carringtonii
is clearly separated from the morphologically similar R.
aquilegia according to the molecular and phytochemical
data, and is resolved with maximum support as sister to the
Macaronesian endemic R. wichurae in the molecular tree
(Fig. 1). Although phylogenetic relationships of the second
Macaronesian endemic R. jonesii are less well-supported
(Fig. 1), this species can be clearly distinguished from the
other Radula species, especially from the most similar R.
holtii, by DNA sequences, volatile oil composition (Fig. 2;
Figueiredo et al., 2009), and the different stem anatomy.
The more complicated cases concern the incongruence
of the three data sets in R. aquilegia and the differentiation
between R. complanata and R. lindenbergiana. The separation of the Azorean and Madeiran samples of R. aquilegia
in the volatile oil analyses (Fig. 2; Figueiredo et al., 2009)
may indicate that the morphological species R. aquilegia
comprises different cryptic taxa, which may have originated
from independent evolution on both archipelagos. Our morphological studies revealed a larger intraspecific variability
in R. aquilegia than previously known, which, however,
shows a different geographic pattern than the phytochemical variation. The Azorean flora comprises both the typical
R. aquilegia and a morphological–ecological form closer
to R. carringtonii, the latter occurring in the western group
of Azores islands but also in Madeira. As the molecular
data clearly support the morphological species R. aquilegia
as a monophyletic group, the observed morphological and
phytochemical variation may thus be influenced by different ecological conditions on the archipelagos, but does not
justify recognizing formal (intra-)specific taxa.
Radula complanata and R. lindenbergiana have always been considered closely related, or even conspecific
(Schuster, 1980a). Several authors emphasize that separation is only possible in the presence of reproductive structures (Schuster, 1980a; Paton, 1999; Luı́s et al., 2005;
Schumacker & Váňa, 2005). The molecular analyses support a close relationship of both species (Fig. 1), but statistical support is rather low especially for the clade of R.
lindenbergiana, since the sequences of the included specimens of both taxa differ by only three substitutions (two in
rps4 gene, one in trnL intron). In addition, different types
of loops of a putative hairpin secondary structure in the
most variable stem-loop region P8 of the trnL intron (cf.
Quandt & Stech, 2005), an A-rich type in R. complanata
and a T-rich type in R. lindenbergiana (Fig. 4), may serve
as a molecular character. Both loop types originate from inversion of the loop sequence. However, such inversions in
non-coding plastid DNA regions of bryophytes have been
shown to be homoplastic at different taxonomic levels (e.g.
Quandt et al., 2003; Stech, 2004; Quandt & Stech, 2005)
and can also occur within one species (own unpublished
results). In contrast, partial sequences of the more variable
internal transcribed spacers (ITS) of the nuclear ribosomal
DNA clearly allow distinguishing R. complanata and R. lindenbergiana (own unpublished results). Unfortunately, tree
construction was not possible as no reliable ITS sequences
of the other Radula species could be obtained, mainly due to
fungal contamination. Taking the molecular data together
with the difference in volatile oil composition and sexual
condition, we consider a separation of R. complanata and
R. lindenbergiana at species level reasonable. However,
Characterization of the liverwort Radula in Portugal
A
A
T
T
T
G
A
C
G
A
A
T
A
A
A
A
A
A
C
T
G
C
T
T
A
T
T
T
T
T
G
A
C
G
A
A
T
T
T
A
A
A
G
C
T
G
C
T
T
A
Downloaded By: [Stech, Michael] At: 06:32 24 June 2010
Fig. 4. Putative hairpin secondary structure in stem-loop region
P8 of the trnL intron. A. A-rich type of loop in Radula complanata
(G = −10.8). B. T-rich type of loop (or U in the respective RNA)
in R. lindenbergiana (G = −10.6).
an extended sampling, including Portuguese material of R.
complanata, and analysis of further molecular markers is
necessary to arrive at a final conclusion.
Despite the limited taxon sampling across the genus
Radula, the molecular phylogenetic reconstructions allow
first conclusions concerning intrageneric and biogeographic relationships of the species under study. The basal
split between R. boryana/R. gottscheana and the remaining
species corresponds to the distinction of subgenera Cladoradula and Radula, respectively, by Schuster (1980a).
The inclusion of R. aquilegia and R. carringtonii in subg.
Cladoradula by Castle (1969) is rejected. Within subgenus
Radula, relationships are difficult to analyse as not all
European species have been decidedly assigned to intrasubgeneric taxa. At this taxonomic level, at least the hitherto
recognised taxa of Castle (1969) may not represent monophyletic groups, as indicated by the positions of R. wichurae
(section Marginata), close to R. complanata/R. lindenbergiana and R. holtii (section Radula), and of R. nudicaulis
(also section Radula) being separated from this clade.
The Radula species narrowly distributed in Macaronesia
and Atlantic Europe probably have two different origins.
For Radula holtii and R. nudicaulis connections with the
Neotropics are indicated by the molecular data, which corresponds to the pronounced neotropical influence on the
Macaronesian bryoflora recently indicated by a number of
molecular studies of mosses (e.g. Stech et al., 2007) and especially liverworts (e.g. Feldberg & Heinrichs, 2005; SimSim et al., 2005; Vanderpoorten & Long, 2006). The other
species, among them the Macaronesian endemics R. jonesii
and R. wichurae, are closely related with, and may be derived from, the R. complanata/R. lindenbergiana complex,
which is widely distributed in the northern hemisphere.
Relationships with northern-hemisphere species have also
been revealed in previous studies of Macaronesian mosses
(e.g. Stech et al., 2007, 2008) and liverworts (e.g. Sim-Sim
et al., 2005; Hentschel et al., 2007). Biogeographic patterns
of the European Radula species thus seem to be rather com-
267
plex, similar to other diverse and species-rich genera such as
Plagiochila (Sim-Sim et al., 2005) and Campylopus (Stech
et al., 2007).
Acknowledgements
This study was partially funded by Fundação para a
Ciência e a Tecnologia (Lisbon) under research contract
POCI/AGR/57487/2004. Sincere thanks are due to B.
Giesicke (Berlin) and M.C.M. Eurlings (Leiden) for technical assistance.
References
AHONEN, I. 2004. Molecular phylogeny of the liverwort order Porellales (Marchantiophyta, Jungermanniopsida). In:
GOFFINET, B., HOLLOWELL, V. & MAGILL, R., Eds., Molecular
systematics of bryophytes. Monographs in Systematic Botany
from the Missouri Botanical Garden 98, 168–188.
BOUMAN, A.C. & DIRKSE, G.M. 1990. The genus Radula in Macaronesia. Lindbergia 16, 119–127.
CASTLE, H. 1937. A revision of the genus Radula. Introduction and
Part I. Subgenus Cladoradula. Annales Bryologici 9, 13–56.
CASTLE, H. 1969. Radula (L.) Dumortier – A synopsis of the taxonomic revision of the genus. Revue Bryologie et Lichénologie
36, 5–44.
FELDBERG, K. & HEINRICHS, J. 2005. On the identity of Herbertus
borealis (Jungermanniopsida: Herbertaceae), with notes on
the possible origin of H. sendtneri. Journal of Bryology 27,
343–350.
FIGUEIREDO, A.C., SIM-SIM, M., BARROSO, J.G., PEDRO, L.G.,
ESQUÍVEL, M.G., FONTINHA, S., LUÍS, L., MARTINS, S. &
STECH, M. 2009. Liverwort Radula species from Portugal: chemotaxonomical evaluation of volatiles composition.
Flavour and Fragrance Journal 24, 316–325.
FORREST, L.L. & CRANDALL-STOTLER, B.J. 2004. A phylogeny of
the simple thalloid liverworts (Jungermanniopsida, Metzgeriidae) as inferred from five chloroplast genes. In: GOFFINET,
B., HOLLOWELL, V. & MAGILL, R., Eds., Molecular Systematics of Bryophytes. Monographs in Systematic Botany from the
Missouri Botanical Garden 98, 119–140.
FREY, W., FRAHM, J.-P., FISCHER, E. & LOBIN, W. 2006. The Liverworts, Mosses and Ferns of Europe (English edition revised
and edited by BLOCKEEL, T.L.). Harley Books, Colchester.
FREY, W. & STECH, M. 2009. Marchantiophyta, Bryophyta, Anthocerotophyta. In: FREY, W., Ed., Syllabus of Plant Families. A.
Engler′ s Syllabus der Pflanzenfamilien, 13th edition, Part 3.
Bryophytes and Seedless Vascular Plants. Gebr. Borntraeger
Verlagsbuchhandlung, Stuttgart, pp. 13–263.
FREY, W., STECH, M. & MEISSNER, K. 1999. Chloroplast DNArelationship in palaeoaustral Lopidium concinnum (Hypopterygiaceae, Musci). An example of stenoevolution in
mosses. Studies in austral temperate rain forest bryophytes
2. Plant Systematics and Evolution 218, 67–75.
GROLLE, R. 1970. Radula casteli sp. nov. und Anmerkungen zur
Gattung Radula. The Bryologist 73, 662–668.
HEINRICHS, J., HENTSCHEL, J., WILSON, R., FELDBERG, K. & GRADSTEIN, S.R. 2007. Evolution of leafy liverworts (Jungermanniidae, Marchantiophyta): estimating divergence times from
chloroplast DNA sequences using penalized likelihood with
integrated fossil evidence. Taxon 56, 31–44.
HENTSCHEL, J., ZHU, R.-L., LONG, D.G., DAVISON, P.G., SCHNEIDER,
H., GRADSTEIN, S.R. & HEINRICHS, J. 2007. A phylogeny of
Porella (Porellaceae, Jungermanniopsida) based on nuclear
Downloaded By: [Stech, Michael] At: 06:32 24 June 2010
268
M. Stech et al.
and chloroplast DNA sequences. Molecular Phylogenetics
and Evolution 45, 693–705.
HE-NYGRÉN, X., JUSLÉN, A., AHONEN, I., GLENNY, D. & PIIPPO,
S. 2006. Illuminating the evolutionary history of liverworts
(Marchantiophyta)—towards a natural classification. Cladistics 22, 1–31.
HERNÁNDEZ-MAQUEDA, R., QUANDT, D., WERNER, O. & MUÑOZ,
J. 2008. Phylogeny and classification of the Grimmiaceae/Ptychomitriaceae complex (Bryophyta) inferred
from cpDNA. Molecular Phylogenetics and Evolution 46,
863–877.
HUELSENBECK, J.P. & RONQUIST, F. 2001. MrBayes: Bayesian inference of phylogenetic trees. Bioinformatics 17, 754–755.
IUCN. 2008. Bryophyte Specialist Group 2000. Radula jonesii. In: 2008 IUCN Red List of Threatened Species.
http://www.iucnredlist.org (accessed 5 May 2009).
JONES, E.W. 1977. African Hepatics. XXX. The genus Radula
Dumortier. Journal of Bryology 9, 461–504.
KELCHNER, S.A. 2000. The evolution of non-coding chloroplast
DNA and its application in plant systematics. Annals of the
Missouri Botanical Garden 87, 482–498.
LUÍS, L., DRAPER, D. & SIM-SIM, M. 2005. The distribution of
the genus Radula in mainland Portugal and the Madeira
Archipelago. Lindbergia 30, 3–10.
MÜLLER, K., MÜLLER, J., NEINHUIS, C. & QUANDT, D. 2006.
PhyDE – Phylogenetic Data Editor, v0.995. Program distributed by the authors. http://www.phyde.de (accessed 23
February 2010).
NUIN, P.A.S. 2005. MTgui—a simple interface to ModelTest.
Program distributed by the author. University of Toronto,
http://www.genedrift.org/mtgui.php (accessed 23 February
2010).
NYLANDER, J.A.A. 2004. MrModeltest v2. Program distributed by
the authors, Evolutionary Biology Centre, Uppsala University.
PATON, J.A. 1999. The Liverwort Flora of the British Isles. Harley
Books, Colchester.
QUANDT, D., MÜLLER, K. & HUTTUNEN, S. 2003. Characterisation
of the chloroplast DNA psbT-H region and the influence of
dyad symmetrical elements on phylogenetic reconstructions.
Plant Biology 5, 400–410.
QUANDT, D. & STECH, M. 2005. Molecular evolution and secondary structure of the chloroplast trnL intron in bryophytes.
Molecular Phylogenetics and Evolution 36, 429–443.
RENNER, M.A.M. & BRAGGINS, J.E. 2004. The sterile gametophyte as a source of informative characters in the
genus Radula (Radulaceae: Hepaticae). Nova Hedwigia 78,
243–268.
RENNER, M.A.M. & BRAGGINS, J.E. 2005. Systematically relevant
characters of the Radula sporophyte (Radulaceae: Hepaticae).
Nova Hedwigia 81, 271–300.
SCHUMACKER, R. & VÁŇA, J. 2005. Identification Keys to the Liverworts and Hornworts of Europe and Macaronesia (Distribution and Status). 2nd edition. Sorus, Poznań.
SCHUSTER, R.M. 1980a. Phylogenetic studies on Jungermanniidae
II. Radulineae (Part I). Nova Hedwigia 32, 637–693.
SCHUSTER, R.M. 1980b. The Hepaticae and Anthocerotae of North
America, East of the Hundredth Meridian, Vol. 3. Columbia
University Press, New York.
SIM-SIM, M., ESQUÍVEL, M.G., FONTINHA, S. & STECH, M.
2005. The genus Plagiochila (Plagiochilaceae, Hepaticophytina) in Madeira Island—Molecular relationships, ecology, and biogeographic affinities. Nova Hedwigia 81, 449–
462.
SIM-SIM, M., LUÍS, L., GARCIA, C., FONTINHA, S., LOBO, C.,
MARTINS, S. & STECH, M. 2008. New data on the status
of bryophytes from Madeira Island with reference to new
species. Journal of Bryology 30, 226–228.
SO, M.L. 2005a. A synopsis of Radula (Radulaceae, Marchantiophyta) in New Zealand and Tasmania. Journal of the Hattori
Botanical Laboratory 98, 149–174.
SO, M.L. 2005b. Radula (Radulaceae, Marchantiophyta) in
Hawaii. Journal of the Hattori Botanical Laboratory 98,
175–191.
SO, M.L. 2006. Radula (Radulaceae, Marchantiophyta) in the
South Pacific. Journal of the Hattori Botanical Laboratory
99, 207–232.
SÖDERSTRÖM, L, URMI, E. & VÁŇA, J. 2002. Distribution of Hepaticae and Anthocerotae in Europe and Macaronesia. Lindbergia
27, 21–36.
STECH, M. 2004. Supraspecific circumscription and classification
of Campylopus (Dicranaceae, Bryopsida) based on inferences
from sequence data. Systematic Botany 29, 817–824.
STECH, M., SIM-SIM, M., ESQUÍVEL, M.G., FONTINHA, S.,
TANGNEY, R., LOBO, C., GABRIEL, R. & QUANDT, D.
2008. Explaining the ‘anomalous’ distribution of Echinodium Jur. (Bryopsida): independent evolution in Macaronesia and Australasia. Organisms Diversity and Evolution 8,
282–292.
STECH, M., SIM-SIM, M. & FRAHM, J.-P. 2007. Campylopus
(Leucobryaceae, Bryopsida) on Madeira Island—molecular
relationships and biogeographic affinities. Nova Hedwigia
Beiheft 131, 91–100.
∗
SWOFFORD, D.L. 2002. PAUP : Phylogenetic Analysis Using Par∗
simony ( and Other Methods) Version 4.0b10. Sinauer Associates Inc., Sunderland, MA.
VANDERPOORTEN, A. & LONG, D.G. 2006. Budding speciation
and neotropical origin of the Azorean endemic liverwort,
Leptoscyphus azoricus. Molecular Phylogenetics and Evolution 40, 73–83.
YAMADA, K. 1979. A revision of Asian taxa of Radula, Hepaticae.
Journal of the Hattori Botanical Laboratory 45, 201–322.
YAMADA, K. 1988. Studies on Taiwan Hepaticae VII. Radulaceae.
Bulletin of the National Science Museum, Tokyo; Série B 14,
41–51.
YAMADA, K. 1999. Taxonomy of the section Amentulosae of the
genus Radula (Radulaceae, Hepaticae). Haussknechtia Beiheft 9, 391–397.
ZHENG, M. & ZHU, R.-L. 2009. Karyological studies on
some species of Radula (Radulaceae, Jungermanniopsida,
Marchantiophyta). Nova Hedwigia 88, 229–244.