A MOLECULAR PHYLOGENY AND NEW INFR
ADANS. (LEGUMINOSAE-PAPILIONOID
MORPHOLOGY AND HYPOTHE
M.
# T
rden
oi-c
as,
ifica
Kew, Richmond, Surrey, TW9 3AB, United Kingdom; ∥Embrapa Recursos Genéticos e Biotecnologia, Parque
Estação Biológica, Avenue W5 Norte, 70770-917 Brasília, DF, Brazil; and #Instituto de Pesquisas
Jardim Botânico do Rio de Janeiro, Diretoria de Pesquisas, Rua Pacheco Leão, 915,
ical distribution through multiple long-distance dispersal events, which were facilitated by the occurrence of
cies are grown as ornamentals in botanical gardens and green-
houses (e.g.,M. bennettii F.Muell.).Mucuna pruriens is of wide
economic importance and is currently used in agriculture as for-
age and green manure, for biological control, and as a coffee
3 Author for correspondence; current address: Department of Bot-
any, National Museum of Natural History, MRC 166, Smithsonian
Institution, Washington DC 20013-7012 USA; e-mail: vatanparastm
@si.edu.
Int. J. Plant Sci. 177(1):76–89. 2016.
q 2015 by The University of Chicago. All rights reserved.
1058-5893/2016/17701-0007$15.00 DOI: 10.1086/684131substitute (Duke 1981; Garcia and Fragoso 2003; Ortiz-Ceballos
and Fragoso 2004; Ortiz-Ceballos et al. 2007a, 2007b). It also
Manuscript received May 2015; revised manuscript received August 2015;
electronically published November 24, 2015.seeds adapted to oceanic dispersal.
Keywords: Fabaceae, long-distance dispersal, sea-drifted seeds, systematics.
Introduction
Mucuna Adans. (Phaseoleae–Leguminosae) has a pantropi-
cal distribution and comprises approximately 105 species
(Lackey 1981; Schrire 2005). The highest diversity of the ge-
nus occurs in Asia (68 taxa), followed by Oceania (34 taxa),
the Americas (25 taxa), and Africa (19 taxa). Some species are
widely distributed, such as Mucuna sloanei Fawc. & Rendle,
occurring in the Americas, Hawaii, and Africa; M. gigantea
(Willd.) DC., in Africa, Asia, and the Pacific Islands; and M.
pruriens (L.) DC. across the entire tropical region. A number
of species are ecologically and economically important, and the
genus displays a high level of morphological variation, espe-
cially in its inflorescences, flowers, fruits, and seeds.
Most species of Mucuna are lianas (except the African en-
demic species M. stans Welw. ex Baker, which has a shrubby
habit), and they are often an important component of tropical
ecosystems. Because of their showy inflorescences, some spe-
1 The two lead authors contributed equally in developing this study
and should both be considered first author.
2 Author for correspondence; e-mail: tmariamoura@gmail.com.Jardim Botânico, 22460-030 Rio de Janeiro, RJ, Brazil
Editor: Patrick S. Herendeen
Premise of research. The genus Mucuna has a pantropical distribution and comprises approximately 105 spe-
cies, many of which show great economic value for forage, ornament, and medicine. To date, phylogenetic
relationships within Mucuna have not been investigated using molecular data. The aim of this study was to
build a phylogenetic framework forMucuna to address questions about its monophyly, infrageneric relationships,
divergence times, and biogeography.
Methodology. We sequenced plastid (trnL-F) and nuclear ribosomal (internal transcribed spacer) regions
and applied Bayesian and maximum likelihood analyses. An ancestral area reconstruction coupled with a di-
vergence time analysis was used to investigate the historical biogeography of the genus.
Pivotal results. Our results show that Mucuna is a monophyletic genus and that subgenus Stizolobium is a
monophyletic group within it. We present here the analyses and results that support the need to recircumscribe
subgenus Mucuna and to segregate a small group of species with large fruits into a newly proposed subgenus (to
be described formally elsewhere after additional investigations).
Conclusions. On the basis of ancestral area reconstruction and divergence time analyses, we conclude that
the genus Mucuna originated and first diversified in the Paleotropics around 29.2 Ma and achieved a pantrop-Tania M. Moura,1,2,* Mohammad Vatanparast,1,3,† Ana
Marcelo F. Simon,∥ Vidal F. Mansano,
*Comparative Plant and Fungal Biology Department, Royal Botanic Ga
of Biology, Graduate School of Science, Chiba University, 1-33 Yay
Vegetal, Instituto de Biologia, Universidade Estadual de Campin
Barão Geraldo, 13083-862 Campinas, SP, Brazil; §Ident76
This content downloaded from 160.11
All use subject to University of Chicago Press TermsSYMPOSIUM PAPER
AGENERIC CLASSIFICATION OF MUCUNA
EAE) INCLUDING INSIGHTS FROM
SES ABOUT BIOGEOGRAPHY
G. A. Tozzi,‡ Félix Forest,* C. Melanie Wilmot-Dear,§
adashi Kajita,† and Gwilym P. Lewis*
s, Kew, Richmond, Surrey, TW9 3AB, United Kingdom; †Department
ho, Inage-ku, Chiba 263-8-522, Japan; ‡Departamento de Biologia
Rua Monteiro Lobato, 255, Cidade Universitária Zeferino Vaz,
tion and Naming Department, Royal Botanic Gardens,1.254.017 on January 29, 2016 09:08:15 AM
 and Conditions (http://www.journals.uchicago.edu/t-and-c).
ENhas been used for the treatment of Parkinson’s disease (Naga-
shayana et al. 2000; Singhal et al. 2003).
Most species have pendant inflorescences; M. stanleyi C.T.
White, endemic to New Guinea, is the only lianescent Mucuna
with an erect inflorescence. Peduncles vary from a few centi-
meters long (e.g., 5–18 cm long inM. sloanei) to approximately
2 m long (e.g., M. globulifera T. M. Moura, N. Zamora &
A. M. G. Azevedo). The length of peduncle is partially associ-
ated with the pollination system. Species with short-pedunculate
inflorescences are mostly pollinated by birds; those with long-
pedunculate inflorescences (over 1 m long) are usually pollinated
by bats. The single lianescent species with an erect inflorescence
is visited by and may be pollinated by a possum (H. Fortune-
Hopkins, unpublished data). Inflorescences can be pseudorace-
mose (e.g.,M. flagellipesVogel, fromAfrica); pseudopaniculate
(e.g., M. paniculata Baker, from Madagascar); or umbelliform,
in which all flowers are closely clustered at the inflorescence
apex and where the floral internodes are so reduced as to not
be evident (e.g., M. elliptica Ruiz & Pavon, from South Amer-
ica). The species with pseudoracemose inflorescences are dis-
tributed across the entire geographical range of the genus; those
with pseudopanicles occur in the Paleotropics,whereas all the spe-
cies with umbelliform inflorescences are found in the Neotropics
(i.e.,M. argenteaT.M.Moura, G. P. Lewis&A.M. G. Azevedo,
M. cajamarca T. M. Moura, G. P. Lewis & A. M. G. Azevedo,
M. cuatrecasasii Hern. Cam. & C. Barbosa ex. L. K. Ruiz, M.
elliptica, M. klitgaardiae T. M. Moura, G. P. Lewis & A. M. A.
Azevedo, and M. pseudoelliptica T. M. Moura, G. P. Lewis &
A. M. G. Azevedo). Two widely distributed species (M. sloanei
and M. gigantea) have a reduced pseudoraceme in which the
brachyblasts and pedicels are progressively shorter toward the
inflorescence apex (rather than of uniform length). This has been
described as pseudoumbellate by some authors (e.g., Wilmot-
Dear 1990; Tozzi et al. 2005), although the internodes are clearly
visible on the inflorescence rachis.
Mucuna flowers show a remarkable variation in color of
the corolla, ranging from white (e.g., M. klitgaardiae), cream
(e.g.,Mucuna urens (L.) Medik.), or greenish (e.g.,M.monticola
Zamora, T. M. Moura & A. M. G. Azevedo) to yellow (e.g.,
M. japira A. M. G. A. Tozzi, Agostini & Sazima), orange (e.g.,
M. rostrata Benth.), red (e.g., M. bennetti), purple (e.g., M. pru-
riens), or almost black (e.g., M. hainanensis Hayata). The wing
petals of the corolla can be either longer than the standard (e.g.,
M. mutisiana (Kunth) DC.) or shorter (e.g., M. holtonii (Kuntze)
Moldenke). The flowers vary in size from 2.5 cm long (e.g.,
M. lane-poolei Summerh.) to 11 cm long (in M. cuatrecasasii).
The morphology of the fruits also presents an important
suite of taxonomic characters. The pod surface is sometimes or-
namented by lamellae (in a transversal, longitudinal, oblique, or
reticulate pattern) or ornamentation can be completely lacking.
Most species have dehiscent fruits, but two (M. poggei Taub.
and M. occidentalis T. M. Moura & G. P. Lewis, both endemic
toAfrica) have indehiscent fruits. Some species have fruits shorter
than 10 cm and contain approximately five seeds (e.g., M. pru-
riens); other species have fruits 10–30 cm long, but they again
contain approximately five seeds (e.g.,M.urens); in a third group
of species, the fruits can be over 50 cm long and have up to 18 seeds
MOURA ET AL.—MOLECULAR PHYLOG(e.g., M. macrocarpa Wall.). In addition, the seeds and hilum
provide taxonomically informative characters. Seeds can be re-
niform, discoid, or globose; the length of the hilum varies from
This content downloaded from 160.11
All use subject to University of Chicago Press Terms3–7 mm in length (circling less than 20% of the seed circumfer-
ence) to 8–9 cm in length (circling more than 50% of the seed
circumference).
On the basis of fruit and seed morphology, two subgenera
have been traditionally recognized in Mucuna (Wilmot-Dear
1984):M. subg.Mucuna andM. subg. Stizolobium (P. Browne)
Baker. Stizolobium P. Brownewas described by Browne (1756),
and De Candolle (1825) later down-ranked it to a section of
Mucuna, as M. sect. Stizolobium (P. Browne) DC. Currently,
the infrageneric classification of Mucuna recognizes two sub-
genera and no sections (Wilmot-Dear 1984, 1991). Neverthe-
less, due to differences in fruit and seed shape and hilum length,
some authors have treated Stizolobium as a distinct genus (e.g.,
Molina Rosito 1975; Stevens et al. 2001; Zamora 2010). Phylo-
genetic studies are necessary to clarify this issue.
Regional taxonomic studies have been published for the ge-
nus Mucuna across its pantropical distribution range (Verd-
court 1970, 1971, 1978, 1979a, 1979b, 1981; Wilmot-Dear
1984, 1987, 1990, 1991, 1992, 1993, 2008; Wiriadinata and
Ohashi 1990; Du Puy et al. 2002; Tozzi et al. 2005; Ren and
Wilmot-Dear 2010; Moura et al. 2012a, 2012b, 2013a,
2013b, 2013c, 2013d, 2013e, 2014, 2015; Moura and Lewis
2014; Zamora and Moura 2014), but there has been no global
taxonomic survey to date. Moreover, a comprehensive phylo-
genetic study of Mucuna has never been performed. A small
number of broader phylogenetic studies have included only
two or three species of Mucuna (e.g., Kajita et al. 2001; Ste-
fanovíc et al. 2009; Lima 2011) and thus have not adequately
covered the entire geographical range or morphological var-
iation of the genus. Although these studies have highlighted
the relationships between Mucuna and its closest allies, the
monophyly of the genus and its infrageneric groups remains
to be tested. In addition, the lack of a phylogenetic framework
for Mucuna precludes more precise inference about the area
of origin of the genus and possible dispersal routes across the
tropics.
In this study, we present a densely sampled phylogeny of
Mucuna and try to answer the following taxonomic and bio-
geographical questions: (1) Are the genus Mucuna and its pro-
posed subgenera monophyletic? (2) What are the infrageneric
relationships among Mucuna species? (3) Do pollination sys-
tems or geographical ranges correlate to clades identified by
the molecular phylogeny? (4) When and where did the genus
originate, and what processes may have produced its current
pantropical distribution?
Material and Methods
Taxon Sampling and DNA Extraction
Sixty-three taxa were sampled for this study, including 47 of
Mucuna and 16 representing outgroups. Due to the wide geo-
graphical distribution of the genus, most of the samples used
in this study came from herbarium collections. Material of a
few species was collected in the field and stored in silica gel.
Three sequences of Mucuna and nine outgroups were obtained
fromGenBank. The range of morphological variation and wide
Y AND CLASSIFICATION OF MUCUNA 77geographical distribution of Mucuna are represented in this
study. A list of the species and specimens sampled is presented
in the appendix.
1.254.017 on January 29, 2016 09:08:15 AM
 and Conditions (http://www.journals.uchicago.edu/t-and-c).
contigs and edited using the program Geneious, version 7.1.7
(Biomatters, Aukland, New Zealand), or SeqScape, version
NAFor internal transcribed spacer (ITS) analysis, 53 accessions
of Mucuna from 45 taxa were sequenced, and for the trnL-F
region, 48 accessions of 33 taxa were sequenced. Because it
was impossible to sequence all the samples for both markers
due to the high level of degradation of the DNA, we opted
for sequencing as much as we could for each locus and then
presenting the results separately. We also present a combined
analysis for the species sequenced for both markers (34 acces-
sions, 30 taxa).
The DNA extraction from leaf tissue was conducted in three
different laboratories: (1) the specimens from MO, UEC, and
CEN herbaria were extracted at theMissouri Botanical Garden
(St. Louis, MO) using MP FastDNA Green Spin Kit (MP Bio-
medicals). After extraction, the DNA was cleaned by DNA
Axigen AxyPrepPCR Clean-Up Kit; (2) the specimens from K,
some from L, and GH herbaria were extracted in the Jodrell
Laboratory, Royal Botanic Gardens, Kew, using the 2# cetyl-
trimethylammonium bromide method (Doyle and Doyle 1987),
and the DNA was cleaned by a cesium chloride-ethidium bro-
mide gradient (1.55 g/mL) and a dialysis procedure to yield
material suitable for long-term storage; (3) specimens from L
were extracted at Chiba University (Chiba, Japan), using the
DNeasy Plant Mini Kit (Qiagen) and following the man-
ufacturer’s instructions with a modified protocol for herbarium
materials. The concentration of genomic DNA was measured
with a GeneQuant 100 electrophotometer (GE Healthcare, Life
Sciences).
Polymerase Chain Reaction (PCR) and Sequencing
Two markers were used: the nuclear region ITS (White et al.
1990) and the plastid region trnL-F (Taberlet et al. 1991).When
amplification of the ITS region failed, internal primers ITS2 and
ITS3 (Baldwin 1992) were used to amplify the ITS region in two
fragments in association with primers ITS5 and ITS4, respec-
tively. Because the DNA obtained from the herbarium spec-
imens is generally degraded, both regions were amplified and
sequenced using internal primers for most of the samples. The
PCR and sequencing steps were conducted in two laboratories:
the Jodrell Laboratory, Royal Botanic Gardens, Kew, and the
Department of Biology at Chiba University.
For the analysis conducted in the Jodrell Laboratory, the PCR
was performed in 25-mL-volume reactions with the following
components: 1.0 mL template DNA; 22.5 mL of Reddy PCR
Master Mix (2.5 mM MgCl2; Thermo Scientific, Waltham,
MA); 0.5 mL of each primer (100 ng/mL); 5 mL of tricholse, bo-
vine serum albumin, and tween; 1.0 mL of dimethyl sulfoxide.
The same PCRmixwas used for both nuclear and plastid regions.
The PCR conditions for both regions were an initial denatura-
tion at 807C for 5 min, followed by 35 cycles of denaturation
at 957C for 1 min, primer annealing at 487–507C for 1 min, and
primer extension at 657C for 1 min; this was followed by a final
extension step of 7 min at 647C. For some samples that were dif-
ficult to amplify, the PCR conditions included a ramp of 0.37C/s,
as described by Shaw et al. (2007).
PCR products were checked on 1% agarose gel before being
cleaned with QIAquick PCR purification kit (Qiagen). Cycle se-
78 INTERNATIONAL JOURquencing reactions were performed in 5-mL-volume reactions,
using 0.3–1.0 mL of the PCR product, 0.25 mL BigDye, 1.5 mL
BigDye Buffer, 1.5 mL double distilled water (ddH2O), and
This content downloaded from 160.11
All use subject to University of Chicago Press Terms2.7 (Life Technologies, Applied Biosystems). A BLAST search
(http://blast.ncbi.nlm.nih.gov/) was conducted for all sequences
to check for possible contaminant DNA. Afterward, edited
alignments were performed using the Clustal W (Larkin et al.
2007) and MUSCLE (Edgar 2004) programs using default set-
tings with manual adjustments.
We performed phylogenetic analyses using two different
approaches: Bayesian inference (BI) and maximum likelihood
(ML). The Bayesian analysis was performed using a Markov
chain Monte Carlo (MCMC) method, as implemented in Mr-
Bayes, version 3.2.2 (Ronquist et al. 2012). The best-fit model
of DNA substitution for each molecular region was determined
using MrModeltest, version 2.2 (Nylander 2004), and the
Akaike information criterion (Akaike 1974). The GTR1 I1G
and GTR1G models were selected as the “best model” for
the ITS and trnL-F regions, respectively. For the combined anal-
ysis, two partitions were defined corresponding to the plastid
and nuclear regions. Two independent Metropolis-coupled
MCMCs with incremental heating temperature of 0.25 were
run for 50 million generations, with the parameters and the re-
sulting phylogenetic trees being sampled every ten-thousandth
generation. The analysis was repeated four times. The MCMC
sampling was considered sufficient when the effective sampling
size (ESS) for each parameter was higher than 200, as verified
with Tracer, version 1.6 (Rambaut et al. 2014). A burn-in period0.75 mL of the same primer as for PCR (diluted to 10%). The
cycle sequencing products were cleaned using Magnesil and
the automated workstation BiomeK NX58 (Beckman Coulter).
Complementary strands were sequenced on an ABI 3730 au-
tomated sequencer (Applied Biosystems) and then assembled;
software base-calling was verified using Sequencher 4.5 (Gene
Codes, Ann Arbor, MI).
At Chiba University, PCR reactions were performed in
volumes of 10 mL containing 0.2 units of ExTaq (TaKaRa) or
0.25 units of MightyAmp DNA Polymerase (TaKaRa) and
0.2 mM deoxynucleotide triphosphates, 10# PCR buffer con-
taining 1.5 mM magnesium chloride, 0.5–1 mM of each primer,
and 20 ng of genomic DNA. The PCR conditions were as
follows: 2 min for initial denaturation at 957C, followed by 35
amplification cycles of 45 s denaturation at 957C, 1 min anneal-
ing at 567C, 1 min extension at 727C, and a final 10 min exten-
sion at 727C. The PCR products were visualized on a 0.8% aga-
rose gel. PCR products were purified using illustra ExoStar
Enzymatic PCR and Sequencing Clean-Up Kit (GE Healthcare)
according to themanufacturer’s instructions. The cycle sequenc-
ing reactions were performed using the BigDye Terminator, ver-
sion 3.1, Cycle Sequencing Kit (Applied Biosystems), and cycle
sequencing products were purified using an ethanol precipita-
tion method. All base sequences were determined using an
ABI 3500 DNA sequencer (Applied Biosystems).
Phylogenetic Analyses
The sequences for all DNA regions were assembled into
L OF PLANT SCIENCESof onemillion generations per runwas applied, and the remaining
trees were used to reconstruct an “allcompat” consensus tree
with posterior probabilities (PP) for each node. Members of tribe
1.254.017 on January 29, 2016 09:08:15 AM
 and Conditions (http://www.journals.uchicago.edu/t-and-c).
erated Maximum Likelihood (RAxML), version 8.1.11 (Sta-
Evolutionary Analysis Sampling Trees (BEAST), version 1.8.1
seolus vulgaris L., published in Lavin et al. (2005). To sum-
marize plausible trees and to obtain a maximum clade credi-
whereas in the ITS trees (fig. 4) and combined trees (fig. 1), it
is sister to the Stizolobium clade. Given that some authors argue
ENbility tree, the Tree Annotator program implemented in the
BEAST package was used. Twenty-five percent of trees (i.e.,
7500 trees) were excluded as burn-in from the subsequent
calculations. Tracer, version 1.6 (Rambaut et al. 2014), was
used to check the ESSs, convergence, and confidence intervals
(CIs). The trees were visualized and edited using FigTree, ver-
sion 1.4.2.
Biogeographic Inferences
To investigate the historical biogeography of Mucuna, we
conducted ancestral state geographic distributions on phylo-(Drummond et al. 2012), using the combined matrix of ITS
and trnL-F, applying the same partition delimitation and evo-
lutionary models as those used for the MrBayes analysis. We
used an uncorrelated relaxed molecular clock with a lognor-
mal distribution of rates and a Yule speciation model (Yule
1925; Gernhard 2008). The analysis was run for 30 million
generations, sampling one tree every one-thousandth genera-
tion. As a calibration point, we applied a normal prior distri-
bution (mean 5 standard deviationp39.752.0 Ma) to the
root of the tree, based on the age estimate of the most recent
common ancestor of Platycyamus regnellii Benth. and Pha-matakis 2014), which implements a rapid hill-climbing algo-
rithm (Stamatakis 2006). Analyses were run for the best-scoring
ML tree inferences under the GTR-GAMMAmodel. Rapid boot-
strapping was performed with 1000 replications using the GTR-
CAT estimation to assess branch support (Stamatakis 2006). To
increase analysis speed, parallel versions of the RAXML, MPI/
Pthreads were used (Pfeiffer and Stamatakis 2010).
Partition Homogeneity Test
To assess congruence between the trnL-F and ITS data sets,
we performed the incongruence length difference (ILD) test (Far-
ris et al. 1994), implemented as a partition homogeneity test in
PAUP*, version 4.0b10 (Swofford 2002). The test was conducted
using a heuristic search with tree-bisection-reconnection branch-
swapping algorithm and with invariant characters excluded
(Cunningham 1997). Three random additions per replicate with
a time limit of 10 min were selected to run 1000 homogeneity
replicates.
Divergence Time Analysis
Estimates of divergence time were obtained using the Bayes-
ian inference approach implemented in the package BayesianDesmodieae, as well asApios americanaMedik., were selected as
outgroup taxa based on published phylogenies (Kajita et al. 2001;
Schrire 2005; Stefanovíc et al. 2009).
The ML analysis was performed using Randomized Axel-
MOURA ET AL.—MOLECULAR PHYLOGgenetic trees using the Bayesian Binary MCMC (BBM) method
implemented in the Reconstruct Ancestral State in Phylogenies
(RASP) program, version 3.2 (Yu et al. 2015). We divided the
This content downloaded from 160.11
All use subject to University of Chicago Press Termsthat combining different data sets (fig. 1) generally improves
phylogenetic accuracy, to increase the resolution of our trees re-
gardless of their incongruence (Cunningham 1997; Yoder et al.
2001), we decided to merge data sets and perform a combined
analysis.
The majority rule consensus tree resulting from the Bayes-
ian analysis of the combined data set revealed that the genus
Mucuna is monophyletic (fig. 1). Three main clades were re-
solved, here named the core Mucuna clade (which includes
the type species of Mucuna, M. urens), which thus represents
Mucuna subg. Mucuna; the Stizolobium clade (which includes
the species currently placed in M. subg. Stizolobium); and the
Macrocarpa clade. Themain diagnostic characteristic of species
in theMacrocarpa clade is the long pods, and therefore Macro-pantropical distribution of Mucuna into eight areas that were
based on the presence of endemic species: North America, in-
cluding Mexico (A), Asia (B), Central America (C), Papua
New Guinea (D), South America (E), Africa (F), Pacific (G),
andMadagascar (H). BBMcalculates the probabilities of ances-
tral ranges using the probabilities for each unit area. A con-
densed tree created by Tree Annotator from the output of the
BEAST analysis on the basis of a combined data set of trnL-F
and ITS was used. MCMC calculations were conducted with
2,000,000 generations and a sample frequency of 1000. We
used a FixedJC (Jukes-Cantor)model with the number of chains
equal to 10 and excluding 200 samples as burn-in using null
root distribution. The maximum number of areas selected was
six.
Results
Phylogenetic Analyses
The aligned trnL-F matrix consisted of 1203 characters for
58 samples, including 48 ofMucuna (33 taxa) and 10 outgroup
taxa; the aligned ITS matrix consisted of 902 characters for
62 samples, 53 of Mucuna (45 taxa), and nine outgroup taxa.
For both markers, Mucuna was supported as monophyletic
(figs. 4, 5). For the trnL-F marker, both subgenera tradition-
ally recognized within Mucuna were also monophyletic (fig. 4),
whereas Mucuna subg. Mucuna appeared as nonmonophyletic in
the analysis based on the ITS marker (fig. 5). A third clade, here
named the Macrocarpa clade, was revealed.
The ILD analysis suggested that the trnL-F and ITS data
sets are incongruent (P<0.003). This incongruence was also
detected from visual inspection of trees derived from individ-
ual independent analysis of the plastid and nuclear regions
(figs. 4, 5). The incongruence observed is mainly related to
the position of the Macrocarpa clade; otherwise, no incongru-
ence was found and the topologies of the main clades from
each region were congruent with support 10.7 posterior prob-
ability and 170%bootstrap. In the trnL-F tree, theMacrocarpa
clade (M. birdwoodiana, M. calophylla, M. macrocarpa, and
M. sempervirens) is sister to the core Mucuna clade (fig. 5),
Y AND CLASSIFICATION OF MUCUNA 79carpa is an appropriate name for the clade (T. M. Moura, un-
published data). The results obtained from the ML analysis
are in agreement with the Bayesian results (results not shown).
1.254.017 on January 29, 2016 09:08:15 AM
 and Conditions (http://www.journals.uchicago.edu/t-and-c).
NA80 INTERNATIONAL JOURThe BEAST analysis based on the combined data set (fig. 2)
estimated the stem age of Mucuna to be in the Oligocene to
early Miocene (29.2 Ma; 95% CI, 18.1–39.1). The subgenus
This content downloaded from 160.11
All use subject to University of Chicago Press TermsL OF PLANT SCIENCESEstimates of Divergence Times Mucuna (Mucuna clade) and the clade comprising the Stizo-
lobium and Macrocarpa subclades began diversifying during
Fig. 1 Fifty percent majority rule consensus tree resulting from Bayesian analysis of the combined data set of trnL-F and internal transcribed
spacer (ITS) sequences for Mucuna species. The numbers above and below branches are posterior probabilities and bootstrap supports, respec-
tively. Values !0.7 and !70% are not shown.the Miocene at 20.8 Ma (95% CI, 11.4–31.0) and 20.8 Ma
(95% CI, 10.5–32.1), respectively. Additionally, diversifica-
tion of subgenus Stizolobium (Stizolobium clade) and the Mac-
1.254.017 on January 29, 2016 09:08:15 AM
 and Conditions (http://www.journals.uchicago.edu/t-and-c).
ENMOURA ET AL.—MOLECULAR PHYLOGbiguous, and five ancestral areas are possible, B (MPp37.7%),
BF (MPp20%), AB (MPp7%), F (MPp6.3%), and BE
(MPp 5.7%), with 23% ambiguity (fig. 3, node V). Node VI
This content downloaded from 160.11
All use subject to University of Chicago Press TermsY AND CLASSIFICATION OF MUCUNA 81rocarpa clade occurred in the late tomiddleMiocene around 10.4
Ma (95% CI, 4.5–17.8) and 9.5 Ma (95% CI, 3.4–18.2),
respectively (fig. 2). Our results suggest that the main diversifi-
cation of the genus Mucuna occurred more recently, in the mid-
dle to late Miocene and Pliocene.
Biogeographic Inferences
The ancestral geographic ranges obtained by BBM analysis
(fig. 3) suggest that Mucuna originated in area B (Asia, node I)
with a marginal probability (MP) of 91.5%. Subclade Mucuna
(node II) and the subclade comprising the Macrocarpa and
Stizolobium clades (node III) are also postulated to have an
Asian origin with MPs of 95.9% and 85.4%, respectively. Al-
though the Macrocarpa clade has an exclusively Asian origin
(node IV, MPp97.6%), the Stizolobium clade (node V) is am-
was inferred to be of South American (area E) origin with
MPp73.5%, suggesting a dispersal event from Asia to the
Neotropics.
Discussion
Phylogenetic Analyses
Mucuna was recovered as monophyletic in all analyses and
comprises three main clades, two of them corresponding to
M. subg. Mucuna as traditionally circumscribed. However, the
positionof theMacrocarpa cladevaries between theanalyses, sug-
gesting that additional molecular data, coupled with a detailed
morphological analysis, are needed to clarify this relationship.
The Macrocarpa clade is characterized by species with long
fruits (often 150 cm in length, containing up to 18 seeds),
Fig. 2 Chronogram for the genus Mucuna, based on the concatenated matrices of trnL-F and internal transcribed spacer markers obtained
with a BEAST analysis. The numbers at nodes refer to mean age of the nodes (Ma). Gray bars represent 95% confidence intervals for the es-
timated mean dates. Qua: Quaternary, Plio: Pliocene.whereas the remaining species in the core Mucuna clade (i.e.,
excluding the Macrocarpa clade) have fruits up to 30 cm long
that usually contain no more than five seeds. Members of the
1.254.017 on January 29, 2016 09:08:15 AM
 and Conditions (http://www.journals.uchicago.edu/t-and-c).
NA82 INTERNATIONAL JOURseeds are reniform in outline (rather than round, as they are in
the other two clades), and the hilum is shorter. Although all the
taxa of subgenus Stizolobium cluster in a well-supported clade,
This content downloaded from 160.11
All use subject to University of Chicago Press TermsL OF PLANT SCIENCESStizolobium clade (M. subg. Stizolobium) also have fruits up to
10 cm long that usually contain nomore than five seeds, but the
the topology of the trees indicates that this group should be
treated as a subgenus within Mucuna (Wilmot-Dear 1984,
Fig. 3 Output of the Bayesian binary Markov chain Monte Carlo analysis from the RASP program, version 3.2 (Yu et al. 2015), showing
ancestral geographic reconstructions at each node of the phylogenetic tree of Mucuna. The proportion of colors in circles at each node represents
possible ancestral ranges, and letters in the centers of the circles show the most likely state of each node. Numbers on the horizontal axis cor-
respond to the estimated ages (Ma) obtained from the BEAST analysis. Ap North America, Bp Asia, Cp Central America, Dp Papua New
Guinea, E p South America, F p Africa, G p Pacific, and H p Madagascar. Black represents other ancestral ranges.1991; Moura et al. 2013b, 2014), instead of as an independent
genus, as suggested by some authors (Molina Rosito 1975;
Stevens et al. 2001; Zamora 2010).
1.254.017 on January 29, 2016 09:08:15 AM
 and Conditions (http://www.journals.uchicago.edu/t-and-c).
ENMOURA ET AL.—MOLECULAR PHYLOGysis of the ITS marker failed to include M. calophylla as a
member of the Macrocarpa clade, placing it instead as the sis-
ter taxon to subgenus Stizolobium, although with low support
This content downloaded from 160.11
All use subject to University of Chicago Press TermsY AND CLASSIFICATION OF MUCUNA 83As stated above, the Macrocarpa clade is morphologically
coherent with respect to fruit morphology. Only the ML anal-
(58%). Conversely, there is high support (99% bootstrap and
1 posterior probability support) in the combined ML and BI
Fig. 4 Fifty percent majority rule consensus tree resulting from Bayesian analysis of the trnL-F marker for Mucuna species. The numbers
above and below branches are posterior probabilities and bootstrap supports, respectively. Values !0.7 and !70% are not shown.analyses for a clade comprisingM.birdwoodiana,M. calophylla,
M.macrocarpa, andM. sempervirens. Incongruency between the
chloroplast DNA and ITS trees regarding the position of the
1.254.017 on January 29, 2016 09:08:15 AM
 and Conditions (http://www.journals.uchicago.edu/t-and-c).
Fig. 5 Fifty percent majority rule consensus tree resulting from Bayesian analysis of the internal transcribed spacer (ITS) marker for Mucuna
species. The numbers above and below branches are posterior probabilities and bootstrap supports, respectively. Values !0.7 and !70% are not
shown.This content downloaded from 160.111.254.017 on January 29, 2016 09:08:15 AM
All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
have almost pantropical distributions. The distribution of these
species is hypothesized to be a result of dispersal of seeds that
ENMacrocarpa clade (see above) and the observed variation in poly-
ploidy level among individuals of M. sempervirens might be the
result of incomplete lineage sorting and/or introgressive hybrid-
ization (Maddison and Knowles 2006; Petit and Excoffier
2009) among species of the Macrocarpa, core Mucuna, and
Stizolobium clades.
The corolla color, the pollination systems, and the lamellate
ornamentation of the pod surface were not found to be syna-
pomorphies for any of the three major clades in the analyses.
For example, M. nova-guineensis and M. bennettii, two red-
flowered species fromOceania, are placed in the same subclade,
whereas M. neocaledonica (from New Caledonia) and M. ros-
trata (Neotropics), which have red, purplish-red, or orange
flowers, are nested in different subclades. Similarly, M. japira
(endemic to Brazil) and M. sloanei (widely distributed) both
have yellow flowers but are not closely related. There is also
no evident clustering of species with similar pollination systems.
Species bearing long peduncles, which are characteristic of bat
pollination, appear in different subclades. For example, in the
clade formed byM. urens,M. flagellipes,M.mollis, M. rostrata,
and M. japira, the first three species have long peduncles and
are bat pollinated, whereas the other two species have short pe-
duncles and are pollinated by birds. Likewise, in the clade com-
prising M. argyrophylla, M. mutisiana, andM. sloanei, the first
two species have long peduncles and are probably bat polli-
nated, whereas M. sloanei has a short peduncle and is likely to
be bird pollinated. Our results suggest that the pollination syn-
drome and floral features have arisen in parallel multiple times
during the evolutionary history ofMucuna, a trend that has been
reported in other plant groups, such as Sinningeae (Gesne-
riaceae; Perret et al. 2003) and Bignonieae (Bignoniaceae; Al-
cantara and Lohmann 2010). The presence of lamellate orna-
mentation on the pod surface is apparently not synapomorphic
(e.g., M. argyrophylla, M. mollis, and M. gigantea, which lack
ornamented pods, group into different subclades, together with
species with ornamented fruit). No clustering of species on the
basis of inflorescence type, namely, pseudoracemose, pseudo-
paniculate, or pseudoumbellate, was observed, although the in-
clusion of severalNeotropical pseudoumbellate species (M.argen-
tea, M. cajamarca, M. cuatrecasasii, M. elliptica, M. klitgaardiae,
and M. pseudoelliptica), which were not included in our study,
would be needed to further support this finding. On the other
hand, the types of fruit, seed, and seed hilum do provide tax-
onomically informative characters that support our molecular
findings.
Biogeographic Inferences
Present-day global legume distributions are most likely a
combination of long-distance dispersal (LDD) and vicariance,
based on the evidence from phylogenetic studies and fossil
records (Schrire et al. 2005; Bessega et al. 2006). More impor-
tantly, the transoceanic distribution of various crown clades of
the legumes is considered to be the result of LDD, because the
young ages estimated for most legume groups preclude vicari-
ance as an explanation for their disjunct distributions (Lavin
et al. 2004). According to our divergence time analysis using
MOURA ET AL.—MOLECULAR PHYLOGBEAST, the genus Mucuna evolved sometime in the Oligocene
to early Miocene (39.1–18.1 Ma), whereas the core Mucuna,
Macrocarpa, and Stizolobium clades diversified in the Miocene
This content downloaded from 160.11
All use subject to University of Chicago Press Termshave drifted by ocean currents, because Mucuna seeds have
been found even along the coast of New Zealand (Mason
1961) and on beaches in Europe (Nelson et al. 2000). LDD
plays an important role in the biodiversity and biogeography
of a number of legume genera, including Apios (Li et al. 2014),
Canavalia and Dalbergia (Vatanparast et al. 2011; Vatanparast
et al. 2013), Lonchocarpus (Silva et al. 2012), Zornia (Fortuna-
Perez et al. 2013), and members of tribe Fabeae (Schaefer et al.
2012) and indeed of legumes in general (Lavin et al. 2004; Pen-
nington et al. 2006).
Although our taxon sampling could be increased, our anal-
yses showed that at least two dispersal events from the Paleo-
tropics (Asia) to the Neotropics can be inferred. One dispersal
event to Central America has led to a clade comprising
M. sloanei, M. mutisiana, and M. argyrophylla, whereas another
inferred dispersal event to South America has resulted in a clade
comprisingM. urens, M. mollis, M. rostrata, andM. japira. For
Africa, four dispersal events can be inferred. Although the Af-
rican species of M. subg. Stizolobium (M. coriacea, M. poggei,
M. pruriens, and M. stans) could have resulted from a single
colonization from Asia, M. subg. Mucuna arrived on the Afri-
can continentmultiple times:M. flagellipes fromSouthAmerica,
M. gigantea from Asia, and M. sloanei from Central America
(all by LDD). Mucuna manogarivensis, which is endemic to
Madagascar, is sister to M. atropurpurea from Asia, but the
ancestor of these two species appears to be Central Ameri-
can (although this relationship received low support of PP;
0.53). The species from Oceania, M. nova-guineensis, M. ben-
nettii, and M. mollissima, also have an origin in Asia (fig. 3),
whereas the origin of M. diabolica (an Australian species) is
uncertain.
Conclusions
Our results confirm that the genus Mucuna is monophyletic,
as is the previously described subgenus Stizolobium. Mucuna
subg. Mucuna contains two main clades, the core Mucuna clade
and Macrocarpa clade. Although the position of the Macro-
carpa clade is variable in our analyses, it is consistent in species
content for both studied markers. The species of this clade are
morphologically distinctive with respect to fruit type, but we
are carrying out additional investigations before formally de-
scribing a new subgenus. On the basis of the divergence time(fig. 2). Our results suggest that the genus Mucuna originated
in Asia and has since undergone multiple colonization events
into Africa and North, Central, and South America. Although
the role of long-distance dispersal in the biogeography of land
plants has generally been underestimated (Cain et al. 2000;
Vatanparast 2010), when one considers the topology of our
phylogenetic trees and ancestral area reconstruction, it is plau-
sible that several LDD events underpin the pantropical distribu-
tion of Mucuna across Asia, Africa, and the Americas. The ma-
jority of extantMucuna species (180%) are restricted to a single
continent, whereas some species, M. gigantea and M. sloanei,
Y AND CLASSIFICATION OF MUCUNA 85analysis, we conclude that the genus Mucuna originated in the
Paleotropics, in theOligocene to earlyMiocene, and gradually ex-
panded its geographic range via a number of dispersal events into
1.254.017 on January 29, 2016 09:08:15 AM
 and Conditions (http://www.journals.uchicago.edu/t-and-c).
GQ413945.1, trnL-F: JN402868.1. Desmodium uncinatum (Jacq.) DC., ITS: GQ413950.1. Erythrina fusca Lour., Brazil: cul-
; tr
en
ed
h &
.1.
404
: KT
cya
lvar
i H
cu
Mucuna argyrophylla Standl., Mexico, M. Sousa 11399 (MO), IT
ul
KT
603
&
607G.P. Lewis & A.M.G. Azevedo., Colombia, B. Kats & A. van D
DC., Ceylon, A. G. Robyns 7327 (K), ITS: KT696030, trnL-F:
Janeiro Botanical Garden, T. M. Moura 996 (UEC), ITS: KT69
L.G. Saw (L 0462392), trnL-F: KT696088. M. biplicata Teijsm
KT696089. M. birdwoodiana Tutcher, China, W. T. Tang 20tivated in Rio de Janeiro Botanical Gardens, ITS: KT729507
Moura 1004 (UEC), ITS: KT729508; trnL-F: KT729513. K
(MO), trnL-F: KT696082. Kennedia nigricans Lindl., cultivat
KT696083. Kennedia prostrata R. Br., Australia, R. J. Smit
(Dum. Cours.) G. Don, ITS: GU572175.1, trnL-F: JN402793
archboldianus Merr. & L. M. Perry, New Guinea, M. Fallen
Brazil: cultivated in Rio de Janeiro Botanical Gardens, trnL-F
Platy
Platycyamus regnellii Benth., Brazil, B. A. S. Pereira & D. A
Brazil, E. Ule 9496 (K), trnL-F: KT696080. Platycyamus ule
KT696081.
MuRech., Fiji, W. Greenwood 1109 (K), ITS: KT696034. M. bracteat
KT696090. M. bracteata DC. ex Kurz 02, Burma, G. B. Vogt 49
R. C. Ching 21690 (GH), ITS: KT696035, trnL-F: KT696092. M
This content downloaded from 160.11
All use subject to University of Chicago Press TermsnL-F: KT729512. Erythrina speciosa Andrews, Brazil, T. M.
nedia coccinea (Curtis) Vent., Australia, T. R. Lally 1568
in the United States, H. Van der Werff 8254 (MO), trnL-F:
A. Shade 39 (K), trnL-F: KT696084. Lespedeza cuneata
Lespedeza maritima Nakai, ITS: GU572190.1. Strongylodon
(MO), ITS: KT696078. Strongylodon macrobotrys A. Gray,
696131.
mus
enga 2474 (K), KT696079. Platycyamus ulei Harms 44721,
arms 45190, Peru, E. Meneces s.n. (MO 4268531), trnL-F:
na
S: KT696029, trnL-F: KT696085. M. argentea T. M. Moura,
men AVD 265 (K), ITS: KT729509. M. atropurpurea (Roxb.)
696086. M. bennettii F. Muell., Brazil, cultivated in the Rio de
1, trnL-F: KT696087. M. biplicata Teijsm & Binn., Malaysia,
Binn., Malaysia, C. Hansen 136 (K), ITS: KT696032, trnL-F:
(MO), ITS: KT696033, trnL-F: KT729514. M. brachycarpaAfrica and via at least two dispersal events into the Neotropics.
These findings suggest that long-distance oceanic dispersal of
Mucuna seeds has had a central role in forming the present-
day pantropical distribution of and diversification within the
genus.
Acknowledgments
We thank Conselho Nacional de Desenvolvimento Científico
e Tecnológico (CNPq), Brazil, and the Science without Borders
Program for the postdoctoral scholarship to T.M.Moura (pro-
cess 245590/2012-9); Fundação de Amparo à Pesquisa do
Estado de São Paulo (FAPESP), Brazil (process 2012/04635-8),
and Fundação de Amparo à Pesquisa do Estado do Rio de
Janeiro (FAPERJ), Brazil (process E-26/110.331/2012), for fi-
nancial support; Coordenação de Aperfeiçoamento de Pessoal
de Nível Superior (CAPES), Brazil, for T. M. Moura’s PhD
scholarship; the ShirleyA.GrahamFellowship,Missouri Botan-
ical Garden, to T.M.Moura; the Biologia Vegetal postgraduate
program at Universidade Estadual de Campinas (UNICAMP),
Brazil; the Royal Botanic Gardens, Kew; and the Missouri Bo-
tanical Garden for support. We also thank Dra. Anete Pereira
de Souza and Centro de Biologia Molecular e Engenharia
Genética (CBMEG) Laboratory, Brazil, for support at the be-
ginning of this study; the Jodrell Laboratory staff (Royal Bo-
tanic Gardens, Kew), especially Edith Kapinos, Dion Devey,
László Csiba, JimClarkson, and Tim Fulcher, for valuable tech-
nical support; we thank the curators of the herbaria at K, L,
and MO for providing DNA material; all researchers and
friends that helped with laboratory activities, including DNA
extraction and data analysis, particularly Ana Paula Fortuna
Perez, André Simões, Anne Bruneau, David Bogler, Domingos
Cardoso, Eve Lucas, Fiorella Fernanda Mazine Capelo, Helen
Fortune-Hopkins, Laura Lima, Marcos José da Silva, Mariana
de Oliveria Bünger, and Valentina D’Amico. M. Vatanparast
thanks Tetsukazu Yahara and Shuichiro Tagane from Kyushu
University for providing fresh material of Mucuna. M. Va-
tanparast and T. Kajita were supported by a grant from the
Environment Research and Technology Development Fund
(S-9) of the Ministry of the Environment, Japan. We thank
Dr. Benjamin Torke and one anonymous reviewer for con-
structive comments on the first submitted draft of the paper
and also Patrick Herendeen for his valuable editorial input.
Appendix
Voucher Information and GenBank Numbers
Shown below are voucher information and GenBank numbers (trnL-F and ITS) for all specimens used in this study.
Outgroups
Apios americana Medik., ITS: AF467019.1, trnL-F: EU717312.1. Craspedolobium schochii Harms, China, A. Henry 9241-A
(K), ITS: KT696028. Desmodium barbatum (L.) Benth., trnL-F: EU717290.1. Desmodium microphyllum (Thunb.) DC., ITS:
86 INTERNATIONAL JOURNAL OF PLANT SCIENCESa DC. ex Kurz, Thailand, E. F. Anderson 4108 (MO), trnL-F:
5 (K), trnL-F: KT696091. M. calophylla W. W. Sm., China,
. championii Benth., Hong Kong, Hu & But 20316 (MO),
1.254.017 on January 29, 2016 09:08:15 AM
 and Conditions (http://www.journals.uchicago.edu/t-and-c).
.n.
trn
96
nth
ilip
s.n
ert
erd
(MO), ITS: KT696044. M. interrupta Gagnep., ITS: AB775135.1. M. irritans Burtt Davy, Malawi, E. A. Banda et al. 3573
(MO), ITS: KT696045. M. japira A. M. G. Azevedo, K. Agostini & M. Sazima, Brazil, T. M. Moura 630 (UEC), ITS:
ENKT696046, trnL-F: KT696101. M. klitgaardiae T. M. Moura, G.P. Lewis & A.M.G. Azevedo, Ecuador, B. Klitgaard 99502
(MO), ITS: KT696047. M. lamellata Wilmot-Dear, China, C. Ford 64 (K), ITS: KT696048, trnL-F: KT696102. M. lamellata
Wilmot-Dear (as M. paohwashanica T. Tang & F. T. Wang), China, X. Q. Wang & Y. N. Xiong 001 (MO), ITS:
KT696049, trnL-F: KT696103. M. macrocarpa Wall., China, G. Z. Li s.n. (MO), ITS: KT696050, trnL-F: KT696104. M.
macrophylla Miq., Indonesia, F. J. A. J. Verheijen 1330/31 (L), ITS: KT696051. M. macropoda Baker f., Papua New Guinea,
H. Hopkins s.n. (L), ITS: KT696052. M. manongarivensis Du Puy & Labat, Madagascar, L. Gautier et al. 3785 (MO), ITS:
KT696053, trnL-F: KT696105. M. melanocarpa Hochst. ex A. Rich., Ethiopia, I. Friis, W. Abebe & E. Getachew 13442
(K), trnL-F: KT696106. M. mollis (Kunth) DC., Colombia, H. Murphy & E. Parra 684 (MO), ITS: KT696054, trnL-F:
KT696107. M. mollissima Teijsm. & Binn. ex Kurz, Papua New Guinea, J. H. Waterhouse s.n. (K), ITS: KT696055, trnL-F:
KT696108. M. mollissima Teijsm. & Binn. ex Kurz [as M. cyanosperma K. Schum.], Papua New Guinea, B. Eddie s.n. (K),
ITS: KT696056, trnL-F: KT696109. M. monosperma (Roxb.) DC., ITS: AB775136.1. M. mutisiana (Kunth) DC., Colombia,
J.H. Kirkbride Jr. 2555 (MO), ITS: KT696057, trnL-F: KT696110. M. neocaledonica Baker f., New Caledonia, G. McPherson
5261 (K), ITS: KT696058, trnL-F: KT696111. M. nova-guineensis Scheff., Papua New Guinea, W.N., Takeuchi (L 0254121),
ITS: KT696059, trnL-F: KT696112. M. poggei Taub. 01, Malawi, R-Smith, Pope & Goyder 5836 (K), ITS: KT696060, trnL-F:
KT696113. M. poggei Taub. 02, Zambia, D. K. Harder et al. 3073 (MO), trnL-F: KT696114. M. pruriens var. utilis (Wall. ex
Wight) Baker ex Burck, cultivated in Brazil, T. M. Moura 994 (UEC), ITS: KT696063, trnL-F: KT696118. M. pruriens var. utilis
(Wall. ex Wight) Baker ex Burck (as M. hassjoo (Piper & Tracy) Mansf.), Taiwan, M.T. Kao 10159 (MO), ITS: KT696062,
trnL-F: KT729516. M. pruriens var. utilis (Wall. ex Wight) Baker ex Burck 31, Ecuador, Z. T. Almeida 002 (MO), ITS:
KT696064. M. pruriens var. utilis (Wall. ex Wight) Baker ex Burck 32, Ecuador, Z. T. Almeida 002 (MO), ITS: KT696065.
M. pruriens (L.) DC. var. pruriens, Venezuela, R. Ramirez 26 (MO), ITS: KT696061, trnL-F: KT729515. M. pruriens (L.)
DC. 038, French Guiana, J. J. de Granville (U 013577), trnL-F: KT696115. M. pruriens (L.) DC. M7, Mexico, R. Duno de
Stefano1830 (MO), trnL-F: KT696116. M. rostrata Benth., Brazil, M. Simon 1639 (CEN), ITS: KT696066, trnL-F:
KT696120. M. rostrata Benth. 36, Peru, F. Woytkowski 5378 (MO), trnL-F: KT729517. M. samarensis Merr. 027, Philippines,
C. E. Ridsdale (L 0396765), ITS: KT696067, trnL-F: KT696121. M. schlechteri Harms, New Guinea, R. D. Hoogland 4241
(GH), ITS: KT696068. M. sempervirens Hemsl., China, D.E. Boufford (L 0254117), ITS: KT696069, trnL-F: KT696122. M.
sempervirens Hemsl., Asia, Hemsl s.n. (MO), trnL-F: KT696123. M. sloanei Fawc. & Rendle, Brazil, T. M. Moura 1005
(UEC), ITS: KT696070, trnL-F: KT696124. M. stanleyi C. T. White, Papua New Guinea, L. J. Brass 24277 (L), ITS:
KT729511. M. stanleyi C. T. White 2, Papua New Guinea, Hopkins & Hopkins 1018 (K), ITS: KT696071. M. stans Welw.
ex Baker, Tanzania, F. Furuya 95 (MO), ITS: KT696072, trnL-F: KT696125. M. stenoplax Wilmot-Dear, Thailand, S.
Phusomsaeng & S. Pinnin 49 (K), trnL-F: KT696126. M. stenoplax Wilmot-Dear, Asia, T. Kajita s./n. (FU), ITS: KT696073,
trnL-F: KT696127. M. stenoplax Wilmot-Dear 22, Thailand, J. F. Maxwell (L 0401600), trnL-F: KT696128. M. urens (L.)
Medik., Brazil, T. M. Moura 629 (UEC), ITS: KT696074, trnL-F: KT696130. M. urens (L.) Medik. 39, Guyana, T. R. van Andel
(U0085459), trnL-F: KT696129. M. warburgii K. Schum. & Lauterb., Papua New Guinea, S. Lenean 1443 (K), ITS: KT696075.
Literature Cited
Akaike H 1974 A new look at the statistical model identification.
IEEE Trans Autom Control 19:716–723.
Alcantara S, LG Lohmann 2010 Evolution of floral morphology and
pollination system in Bignonieae (Bignoniaceae). Am J Bot 97:782–
796.
Baldwin B 1992 Phylogenetic utility of the internal transcribed
spacers of nuclear ribosomal DNA in plants: an example from
the Compositae. Mol Phylogenet Evol 1:3–16.
Bessega C, JC Vilardi, O Saidman 2006 Genetic relationships among
Browne P 1756 The civil and natural history of Jamaica in three
parts. Osborne & Shipton, London.
Cain ML, BG Milligan, AE Strand 2000 Long-distance seed dispersal
in plant populations. Am J Bot 87:1217–1227.
Cunningham CW 1997 Can three incongruence tests predict when
data should be combined? Mol Biol Evol 14:733–740.
De Candolle ACP 1825 Prodromus systematis naturalis regni vegeta-
bilis 2. Treuttel & Würtz, Paris.
Doyle JJ, JL Doyle 1987 A rapid DNA isolation procedure for smallKT696042,trnL-F: KT696100. M. hainanensis Hayata, ITS: GU217596.1. M. hainanensis Hayata (as M. nigricans (Lour.)
Steud.), Taiwan, W. L. Wagner 6735 (K), ITS: KT696043. M. holtonii (Kuntze) Moldenke, Colombia, W. Devia et al. 2302ITS: KT696036. M. coriacea Baker, Zimbabwe, R. D. Barnes s
ex Keuch., Australia, K. Kenneally 6391 (K), ITS: KT696038,
A.M.G. Azevedo., Ecuador, M. Blanco 2532 (MO), ITS: KT6
ITS: KT696040, trnL-F: KT696095. M. flagellipes Vogel ex Be
KT696041, trnL-F: KT696096. M. gigantea (Willd.) DC., Ph
KT696097. M. gigantea (Willd.) DC. 19, Hawaii, G. Spence
gigantea subsp. quadrialata (Baker) Verdc.], Kenya, S. H. Rob
(Willd.) DC. [as M. gigantea subsp. quadrialata (Baker) V
MOURA ET AL.—MOLECULAR PHYLOGAmerican species of the genus Prosopis (Mimosoideae, Leguminosae)
inferred from ITS sequences: evidence for long-distance dispersal.
J Biogeogr 33:1905–1915.
This content downloaded from 160.11
All use subject to University of Chicago Press Terms(K), ITS: KT696037, trnL-F: KT696093. M. diabolica Backer
L-F: KT696094. M. ecuatoriana T. M. Moura, G.P. Lewis &
039. M. elmeri Merr., Indonesia, Ambriansyah (L 0501638),
., Ghana, C. C. H. Jonkind & D. K. Abbin 1588 (MO), ITS:
pines, C. E. Ridsdale (L 0460611), ITS: KT729510, trnL-F:
. (MO), trnL-F: KT696098. M. gigantea (Willd.) DC. [as M.
son & Q. Luke 5438 (MO), trnL-F: KT696099. M. gigantea
c.], Kenya, S.A. Robertson & Q. Luke (72675MO), ITS:
Y AND CLASSIFICATION OF MUCUNA 87quantities of fresh leaf tissue. Phytochem Bull 19:11–15.
Duke JA 1981 Handbook of legumes of world economic importance.
Plenum, New York.
1.254.017 on January 29, 2016 09:08:15 AM
 and Conditions (http://www.journals.uchicago.edu/t-and-c).
NADu Puy DJ, J-N Labat, R Rabevohitra, J-F Villiers, J Bosser, J Moat
2002 The Leguminosae of Madagascar. Royal Botanic Gardens,
Kew.
Drummond AJ, MA Suchard, D Xie, A Rambaut 2012 Bayesian
phylogenetics with BEAUti and the BEAST 1.7. Mol Biol Evol 29:
1969–1973.
Edgar RC 2004 MUSCLE: multiple sequence alignment with high
accuracy and high throughput. Nucleic Acids Res 32:1792–1797.
Farris J, M Källersjö, A Kluge, C Bult 1994 Testing significance of
incongruence. Cladistics 10:315–319.
Fortuna-Perez A, M Silva, L Queiroz, G Lewis, A Simões, A Tozzi,
T Sarkinen, A Souza 2013 Phylogeny and biogeography of the ge-
nus Zornia (Leguminosae: Papilionoideae: Dalbergieae). Taxon 62:
723–732.
Garcia JA, C Fragoso 2003 Influence of different food substrate
on growth and reproduction of two tropical earthworm species
(Pontoscolex corethrurus and Amynthas corticis). Pedobiologia 47:
754–763.
Gernhard T 2008 The conditioned reconstructed process. J Theor
Biol 253:769–778.
Kajita T, H Ohashi, Y Tateishi, CD Bailey, JJ Doyle 2001 rbcL and
legume phylogeny, with particular reference to Phaseoleae, Millet-
tieae, and allies. Syst Bot 26:515–536.
Lackey JA 1981 Tribe 10. Phaseoleae DC. Pages 301–327 in RM
Polhill, PH Raven, eds. Advances in legume systematics, part 1.
Royal Botanic Gardens, Kew.
Larkin MA, G Blackshields, NP Brown, R Chenna, PA McGettigan,
H McWilliam, F Valentin, IM Wallace, A Wilm, R Lopez 2007
Clustal W and Clustal X, version 2.0. Bioinformatics 23:2947–
2948.
Lavin M, PS Herendeen, MF Wojciechowski 2005 Evolutionary
rates analysis of Leguminosae implicates a rapid diversification of
lineages during the Tertiary. Syst Biol 54:575–594.
Lavin M, BD Schrire, GP Lewis, RT Pennington, A Delgado-Salinas,
M Thulin, CE Hughes, AB Matos, MF Wojciechowski 2004 Meta-
community process rather than continental tectonic history better
explains geographically structured phylogenies in legumes. Philos
Trans R Soc B 359:1509–1522.
Li J, J Jiang, H Stel, A Homkes, J Corajod, K Brown, Z Chen 2014
Phylogenetics and biogeography of Apios (Fabaceae) inferred from
sequences of nuclear and plastid genes. Int J Plant Sci 175:764–
780.
Lima LCP 2011 Estudos filogenéticos em Desmodium Desv.
(Leguminosae–Papilionoideae) e revisão taxonômica das espécies
ocorrentes no Brasil. Tese, Universidade Estadual de Feria de Santana,
Feira de Santana.
Maddison W, L Knowles 2006 Inferring phylogeny despite incom-
plete lineage sorting. Syst Biol 55:21–30.
Mason R 1961 Dispersal of tropical seeds by ocean currents. Nature
191:408–409.
MolinaRositoA 1975 Enumeración de las plantas deHonduras. Ceiba
19:1–118.
Moura TM, R Gereau, GP Lewis 2015 Mucuna mooneyi (Legumino-
sae: Papilionoideae), a new species from Ethiopia. Kew Bull 70:32.
doi:10.1007/s12225-015-9581-x.
Moura TM, GP Lewis 2014 Taxonomic studies in theMucuna poggei
complex (Leguminosae: Papilionoideae). Kew Bull 69:9544. doi:10
.1007/s12225-014-9544-7.
Moura TM, GP Lewis, VF Mansano, AMGA Tozzi 2013c Three
new species of Mucuna (Leguminosae: Papilionoideae: Phaseoleae)
from South America. Kew Bull 68:143–150.
——— 2014 Taxonomic Studies in Mucuna Adans. (Leguminosae –
Papilionoideae) from Peru. Syst Bot 39:884–896.
88 INTERNATIONAL JOURMoura TM, VF Mansano, R Gereau, AMGA Tozzi 2013e Mucuna
jarocha (Leguminosae-Papilionoideae-Phaseoleae), a new species
from Mexico. Phytotaxa 89:43–46.
This content downloaded from 160.11
All use subject to University of Chicago Press TermsMoura TM, VF Mansano, BM Torke, GP Lewis, AMGA Tozzi 2013b
A taxonomic revision of Mucuna (Fabaceae: Papilionoideae: Pha-
seoleae) in Brazil. Syst Bot 38:631–637.
Moura TM, BM Torke, VF Mansano, AMGA Tozzi 2012a A new
combination for an endemic Hawaiian species of Mucuna (Legu-
minosae: Papilionoideae), with a key to the Hawaiian taxa of the
genus. Kew Bull 67:1–5.
Moura TM, AMGA Tozzi, VF Mansano, GP Lewis 2013a Lecto-
typification of names of Neotropical species of Mucuna (Legumi-
nosae: Papilionoideae: Phaseoleae). Taxon 62:391–393.
Moura TM, NA Zamora, GP Lewis, VF Mansano, AMGA Tozzi
2013d Mucuna globulifera (Leguminosae: Papilionoideae), a new
species from Costa Rica, Panama and Colombia. Kew Bull 68:151–
155.
Moura TM, NA Zamora, BM Torke, VF Mansano, AMGA Tozzi
2012b A new species of Mucuna (Leguminosae-Papilionoideae-
Phaseoleae) from Costa Rica and Panama. Phytotaxa 60:1–8.
Nagashayana N, P Sankarankutty, MR Nampoothiri, PK Mohan, KP
Mohanakumar 2000 Association of L-DOPA with recovery fol-
lowing Ayurveda medication in Parkinson’s disease. J Neurol Sci
176:124–127.
Nelson CE, W Walsh, A Hathway, PH Oswald 2000 Sea beans and
nickar nuts: a handbook of exotic seeds and fruits stranded on
beaches in north-western Europe. BSBI Handbooks for Field Iden-
tification. Botanical Society of the British Isles, London.
Nylander JAA 2004 MrModeltest version 2. Evolutionary Biology
Centre, Uppsala University, Uppsala, Sweden.
Ortiz-Ceballos AI, C Fragoso 2004 Earthworm populations under
tropical maize cultivation: the effect of mulching with velvet bean.
Biol Fert Soils 39:438–445.
Ortiz-Ceballos AI, C Fragoso, GG Brown 2007b Synergistic effect of
a tropical earthwormBalanteodrilus pearsei and velvet beanMucuna
pruriens var. utilis on maize growth and crop production. Appl Soil
Ecol 35:356–362.
Ortiz-Ceballos AI, JJ Peña-Cabriales, C Fragoso, GG Brown 2007a
Mycorrhizal colonization and nitrogen uptake by maize: combined
effect of tropical earthworms and velvet bean mulch. Biol Fert Soils
44:181–186.
Pennington RT, JE Richardson, M Lavin 2006 Insights into the his-
torical construction of species-rich biomes from dated plant phylog-
enies, neutral ecological theory and phylogenetic community struc-
ture. New Phytol 172:605–616.
Perret M, A Chautems, R Spichiger, G Kite, V Savolainen 2003 Sys-
tematics and evolution of tribe Sinningieae (Gesneriaceae): evidence
from phylogenetic analyses of six plastid DNA regions and nuclear
ncpGS. Am J Bot 90:445–460.
Petit RJ, L Excoffier 2009 Gene flow and species delimitation. Trends
Ecol Evol 24:386–393.
Pfeiffer W, A Stamatakis 2010 Hybrid MPI/Pthreads parallelization
of the RAxML phylogenetics code. Paper presented at IEEE Inter-
national Symposium on Parallel and Distributed Processing, Atlanta,
April 19–23.
Rambaut A, MA Suchard, D Xie, AJ Drummond 2014 Tracer, ver-
sion 1.6. http://beast.bio.ed.ac.uk/Tracer.
Ren S, CM Wilmot-Dear 2010 Mucuna. Pages 207–218 in Z Wu,
PH Raven, eds. Flora of China: Fabaceae. Missouri Botanical Gar-
den, St. Louis.
Ronquist F, M Teslenko, P van der Mark, DL Ayres, A Darling, S
Höhna, BLarget, L Liu,MASuchard, JPHuelsenbeck 2012 MrBayes
3.2: efficient Bayesian phylogenetic inference and model choice across
a large model space. Syst Biol 61:539–542.
Schaefer H, P Hechenleitner, A Santos-Guerra, M Menezes de
Sequeira, RT Pennington, G Kenicer, MA Carine 2012 Systemat-
L OF PLANT SCIENCESics, biogeography, and character evolution of the legume tribe Fabeae
with special focus on the middle-Atlantic island lineages. BMC Evol
Biol 12:250.
1.254.017 on January 29, 2016 09:08:15 AM
 and Conditions (http://www.journals.uchicago.edu/t-and-c).
Schrire BD 2005 Phaseoleae. Pages 393–431 in G Lewis, B Schrire, B
Mackinder, M Lock, eds. Legumes of the world. Royal Botanic
Gardens, Kew.
Schrire BD, M Lavin, GP Lewis 2005 Global distribution patterns
of the Leguminosae: insights from recent phylogenies. Biol Skr 55:
375–422.
Shaw J, EB Lickey, EE Schilling, RL Small 2007 Comparison of
whole chloroplast genome sequences to choose noncoding regions
for phylogenetic studies in Angiosperms: the tortoise and the hare
III. Am J Bot 94:275–288.
Silva MJ, LP de Queiroz, AMGA Tozzi, GP Lewis 2012 Phylogeny
and biogeography of Lonchocarpus sensu lato and its allies in the
tribe Millettieae (Leguminosae, Papilionoideae). Taxon 61:93–108.
Singhal B, J Lalkaka, C Sankhla 2003 Epidemiology and treatment of
Parkinson’s disease in India. Parkinsonism Relat Disord 9:105–109.
Stamatakis A 2006 RAxML-VI-HPC: maximum likelihood-based phy-
——— 1971 Mucuna. Pages 561–571 in Flora of tropical east Africa,
Leguminosae (part 4) subfamily Papilionoideae. Royal Botanic
Gardens, Kew.
——— 1978 New taxa of Leguminosae from Papua New Guinea.
Kew Bull 32:455–473.
——— 1979a A manual of New Guinea legumes. Botany Bulletin
no. 11. Office of Forests, Lae, Papua New Guinea.
——— 1979b A note on Mucuna schlechteri Harms (Leguminosae-
Papilionoideae-Phaseoleae). Kew Bull 34:521–525.
——— 1981 New taxa of Mucuna (Leguminosae-Phaseoleae) from
East Africa and Australia. Kew Bull 35:743–752.
White TJ, T Bruns, S Lee, J Taylor 1990 Amplification and direct se-
quencing of fungal ribosomal RNA genes for phylogenetics. Pages
315–322 in MA Innis, DH Gelfand, JJ Sninsky, TJ White, eds. PCR
protocols: a guide to methods and applications. Academic Press,
MOURA ET AL.—MOLECULAR PHYLOGENY AND CLASSIFICATION OF MUCUNA 89logenetic analyses with thousands of taxa andmixedmodels. Bioinfor-
matics 22:2688–2690.
——— 2014 RAxML version 8: a tool for phylogenetic analysis and
post-analysis of large phylogenies. Bioinformatics 30:1312–1313.
Stefanovíc S, BE Pfeil, JD Palmer, JJ Doyle 2009 Relationships among
phaseoloid legumes based on sequences from eight chloroplast re-
gions. Syst Bot 34:115–128.
Stevens WD, C Ulloa, A Pool, OM Montiel Jarquin 2001 Flora de
Nicaragua. Monogr Syst Bot Mo Bot Gard 85:i–xlii.
Swofford D 2002 PAUP*: phylogenetic analysis using parsimony
(*and other methods). http://paup.csit.fsu.edu/.
Taberlet P, L Gielly, G Pautou, J Bouvet 1991 Universal primers for
amplification of three non-coding regions of chloroplast DNA.
Plant Mol Biol 17:1105–1109.
Tozzi AMGA, K Agostini, M Sazima 2005 A new species of Mucuna
Adans. (Leguminosae, Papilionoideae, Phaseoleae) from southeast-
ern Brazil, with a key to Brazilian species. Taxon 54:451–455.
Vatanparast M 2010 Phylogeography of a pantropical plant with
sea-drifted seeds; Canavalia rosea (Sw.) DC., (Fabaceae). PhD the-
sis. Chiba University, Japan.
Vatanparast M, BB Klitgård, FACB Adema, RT Pennington, T
Yahara, T Kajita 2013 First molecular phylogeny of the pantrop-
ical genus Dalbergia: implications for infrageneric circumscription
and biogeography. S Afr J Bot 89:143–149.
Vatanparast M, K Takayama, MS Sousa, Y Teteishi, T Kajita 2011
Origin of Hawaiian endemic species of Canavalia (Fabaceae) from
sea-dispersed species revealed by chloroplast and nuclear DNA se-
quences. J Jpn Bot 86:15–25.
Verdcourt B 1970 Studies in the Leguminosae-Papilionoideae for the
“Flora of Tropical East Africa.” Kew Bull 24:235–307.This content downloaded from 160.11
All use subject to University of Chicago Press TermsLondon.
Wilmot-Dear CM 1984 A revision of Mucuna (Leguminosae-Phaseoleae)
in China and Japan. Kew Bull 39:23–65.
——— 1987 A revision of Mucuna (Leguminosae-Phaseoleae) in the
Indian subcontinent and Burma. Kew Bull 42:23–46.
——— 1990 A revision of Mucuna (Leguminosae-Phaseoleae) in the
Pacific. Kew Bull 45:1–35.
——— 1991 A revision of Mucuna (Leguminosae-Phaseoleae) in the
Philippines. Kew Bull 46:213–251.
——— 1992 A revision ofMucuna (Leguminosae-Phaseoleae) in Thai-
land, Indochina and the Malay Peninsula. Kew Bull 47:203–245.
——— 1993 Anew species ofMucuna (Leguminosae-Phaseoleae) from
Thailand, and a revised key to the species in Thailand, Indochina and
the Malay Peninsula. Kew Bull 48:29–35.
——— 2008 Mucuna (Leguminosae) in Thailand. Thai Forest Bull
(Bot) 36:114–139.
Wiriadinata H, H Ohashi 1990 Four new species of Mucuna (Legu-
minosae) of the Lesser Sunda Islands. J Jap Bot 65:97–108.
Yoder A, J Irwin, B Payseur 2001 Failure of the ILD to determine
data combinability for slow loris phylogeny. Syst Biol 50:408–424.
Yu Y, AJ Harris, C Blair, X He 2015 RASP (Reconstruct Ancestral
State in Phylogenies): a tool for historical biogeography. Mol Phy-
logenet Evol 87:46–49.
Yule UG 1925 A mathematical theory of evolution, based on the
conclusions of Dr. JCWillis, FRS. Philos Trans R Soc B 88:433–436.
Zamora N 2010 Fabaceae. Pages 395–775 in BE Hammel, MH
Grayum, C Herrera, N Zamora, eds. Manual de Plantas de Costa
Rica. Vol 5. Monogr Syst Bot Mo Bot Gard 119:395–775.
Zamora NA, TMMoura 2014 Mucuna tapantiana (Fabaceae: Faboi-
deae: Phaseoleae), a new species from Costa Rica. Kew Bull 69:9490.
doi:10.1007/S12225-014-9490-4.1.254.017 on January 29, 2016 09:08:15 AM
 and Conditions (http://www.journals.uchicago.edu/t-and-c).