Cabrera Et Al. 2008 Araceae Phylogenetics

American Journal of Botany 95(9): 1153–1165. 2008.
PHYLOGENETIC RELATIONSHIPS OF AROIDS AND DUCKWEEDS

(ARACEAE) INFERRED FROM CODING AND NONCODING
PLASTID DNA1
Lidia I. Cabrera,2,6 Gerardo A. Salazar,2 Mark W. Chase,3 Simon J. Mayo,3 Josef

Bogner,4 and Patricia Dávila5
2Departamento de Botánica, Instituto de Biología, Universidad Nacional Autónoma de México, Apartado Postal 70-367, 04510
México, D.F., Mexico; 3Royal Botanic Gardens, Kew, Richmond, Surrey TW9 3DS, UK; 4Augsburger Str. 43a, D-86368,
Gersthofen, Germany; and 5Unidad de Biología, Tecnología y Prototipos (UBIPRO), Facultad de Estudios Superiores, Iztacala,
Universidad Nacional Autónoma de México, Avenida de los Barrios No. 1, Los Reyes Iztacala, 54090 Tlalnepantla, Estado de
México, Mexico
Familial, subfamilial, and tribal monophyly and relationships of aroids and duckweeds were assessed by parsimony and Bayes-
ian phylogenetic analyses of five regions of coding (rbcL, matK) and noncoding plastid DNA (partial trnK intron, trnL intron,
trnL–trnF spacer) for exemplars of nearly all aroid and duckweed genera. Our analyses confirm the position of Lemna and its allies
(formerly Lemnaceae) within Araceae as the well-supported sister group of all aroids except Gymnostachydoideae and Oron-
tioideae. The last two subfamilies form the sister clade of the rest of the family. Monophyly of subfamilies Orontioideae, Pothoid-
eae, Monsteroideae, and Lasioideae is supported, but Aroideae are paraphyletic if Calla is maintained in its own subfamily
(Calloideae). Our results suggest expansion of the recently proposed subfamily Zamioculcadoideae (Zamioculcas, Gonatopus) to
include Stylochaeton and identify problems in the current delimitation of tribes Anadendreae, Heteropsideae, and Monstereae
(Monsteroideae), Caladieae/Zomicarpeae, and Colocasieae (Aroideae). Canalization of traits of the spathe and spadix considered
typical of Araceae evolved after the split of Gymnostachydoideae, Orontioideae, and Lemnoideae. An association with aquatic
habitats is a plesiomorphic attribute in Araceae, occurring in the helophytic Orontioideae and free-floating Lemnoideae, but evolv-
ing independently in various derived aroid lineages including free-floating Pistia (Aroideae).
Key words: Araceae; Lemnaceae; molecular phylogenetics; plastid DNA; rbcL; subfamilial and tribal classification; trnK-
matK; trnL-trnF.
Over the last decade, DNA sequence data have contributed studies (e.g., Brooks and McLennan, 1991; Harvey and Pagel,
greatly to improve our understanding of the phylogenetic rela- 1991; Harvey et al., 1996; Givnish and Systma, 1997; Bateman,
tionships of flowering plants (reviews in Soltis and Soltis, 2004; 1999; Futuyma, 2004). In this work, we use DNA sequence
Soltis et al., 2005), permitting for the first time independent data to assess the phylogenetic relationships of Araceae (aroids)
testing of earlier classifications constructed on the basis of and Lemnaceae (duckweeds), which in spite of obvious mor-
similarities and differences in morphological attributes (e.g., phological differences have long been suspected to be closely
Cronquist, 1981; Takhtajan, 1997). DNA studies have also al- related, and to gain insights into evolution of the aquatic habit
lowed the systematic placement of groups problematic because in these groups.
of their highly modified vegetative or reproductive organs, such According to Mayo et al. (1997), Araceae include 105 genera
as aquatic family Podostemaceae (Soltis et al., 1999 and refer- and 3300 species occurring on all continents except Antarctica.
ences therein) and many parasitic plants (Nickrent et al., 1998). About 90% of genera and 95% of species are found in the trop-
Furthermore, molecular phylogenetic trees provide explicit ics. Aroids are one of the most ecologically and structurally di-
evolutionary frameworks for assessments of character evolu- verse groups of monocots. They occupy a wide variety of
tion, biogeography, and many other biological comparative habitats and display a notable diversity of life forms, including
geophytes, climbers, epiphytes, helophytes, and free-floating
1 Manuscript received 25 February 2008; revision accepted 11 June 2008.
aquatics (Croat, 1988; Grayum, 1990; Boyce, 1995; Mayo
The authors thank P. C. Boyce, M. Fay, F. Forest, E. Landolt, S. Renner, et al., 1997; Bown, 2000; Keating, 2002). Their vegetative parts
S.-M. Tam, T. Croat, and the staff of the botanical gardens of Leiden, are extremely varied; for instance, stems can be creeping or
Munich, and Singapore for providing material of various taxa; M. Hesse climbing and form rhizomes or distinct tubers; leaves range
for discussion and literature; the staff of the Molecular Systematics Section, from simple to complexly divided, and some such as those of
Jodrell Laboratory, Royal Botanic Gardens, Kew, and L. Márquez Amorphophallus titanum are among the largest produced by a
Valdelamar (Laboratorio de Biología Molecular, Instituto de Biología, herb (Bown, 2000). The most distinctive features of Araceae are
Universidad Nacional Autónoma de México) for assistance with DNA found in their inflorescences, which characteristically consist of
sequencing; and S. Magallón and R. Lira for useful criticisms of earlier a fleshy axis, the spadix, bearing small flowers usually arranged
drafts of the manuscript. This study was supported in part by the Posgrado
en Ciencias Biológicas, UNAM, a scholarship from the Consejo Nacional
in spirals and subtended by a conspicuous leaf-like or petal-like
de Ciencia y Tecnología (CONACyT) awarded to L.I.C. (No. 133137) and bract—the spathe. Flowers may be bisexual or unisexual, and a
the Royal Botanic Gardens, Kew. perigone is present in some groups. Unisexual flowers usually
6 Author for correspondence (e-mail: [email protected]) are borne in separate female and male zones of the spadix,
which often has a sterile apical appendix (Boyce, 1995; Mayo
doi:10.3732/ajb.0800073 et al., 1997, 1998; Judd et al., 2002; Soltis et al., 2005).
1153
1154 American Journal of Botany [Vol. 95
Historically, several major classifications of Araceae have In this study, we assess relationships of aroids and duck-
been proposed (reviews in Nicolson, 1960, 1987; Croat, 1990, weeds using DNA sequence and indel data from coding and
1998; Grayum, 1990; Mayo et al., 1995a, 1997). The earliest noncoding plastid DNA. The regions analyzed include the ex-
modern classification encompassing Araceae was that proposed ons of rbcL and matK plus the 3′ portion of the trnK intron
by Schott (1860), who based his groupings mainly on floral (downstream matK) and the trnL-trnF region, which consists
morphology (e.g., he divided the family into two major groups, mostly of the trnL intron and the IGS between trnL and trnF.
one with bisexual flowers and the other with unisexual flowers). These plastid regions have been broadly used for phylogenetic
On the other hand, Engler (e.g., 1876, 1920) relied on a broader estimation in angiosperms, including previous assessments of
spectrum of information sources, including vegetative mor- various aroid lineages (Barabé et al., 2002; Les et al., 2002;
phology and anatomy, in addition to floral morphology, and his Rothwell et al, 2004; Tam et al., 2004; Renner and Zhang, 2004;
system explicitly incorporated an evolutionary perspective Renner et al., 2004; Gonçalves et al., 2007). Our study is aimed
(Grayum, 1990; Mayo et al., 1997; Govaerts et al., 2002). at attaining a clearer picture of duckweed relationships to other
Hooker (1883) modified Schott’s classification and incorporated clades of Araceae and evaluating monophyly and relationships
many of Engler’s generic concepts, and subsequently Hutchinson for aroid subfamilies and tribes recognized in the classification
(1973) elaborated on Hooker’s system. Most contemporary of Mayo et al. (1997, 1998). We are also interested in the evolu-
aroid taxonomists (Bogner, 1979; Bogner and Nicolson, 1991; tion of the aquatic habit in Araceae. Throughout this paper,
Mayo et al., 1997, 1998) have been strongly influenced by the genera and suprageneric groups follow Mayo et al. (1997) un-
Englerian views, and some of them have discussed and modi- less otherwise specified.
fied previous systems by incorporating diverse types of infor-
mation into classifications on cladistic grounds (Grayum, 1990;
Mayo et al., 1997). Recently, several molecular phylogenetic MATERIALS AND METHODS
studies have been published that provide independent frame-
works for evaluating earlier proposals of relationship among Taxonomic sample—Exemplars representing 97 of 105 aroid genera ac-
cepted in Mayo et al. (1997) and all five genera of duckweeds recognized by
aroids (French et al., 1995; Renner and Zhang, 2004; Renner Les et al. (2002) were studied. Representatives of other families of Alismatales,
et al., 2004; Tam et al., 2004; Gonçalves et al., 2007) or evolu- including Alisma (Alismataceae), Tofieldia (Tofieldiaceae), Triglochin (Jun-
tion of specific traits (e.g., atypical bisexual flowers; Barabé caginaceae), as well as of Acorales (Acorus), Chloranthaceae (Hedyosmum),
et al., 2002, 2004). Magnoliales (Magnolia), and Piperales (Piper), were used as outgroups. A list
For over 250 years, botanists have been perplexed by the sys- of taxa with voucher information and GenBank accessions is given in Appendix
tematic position of duckweeds, a group of five genera and about 1; the aligned data matrix was deposited in TreeBase (http://TreeBASE.org;
matrix accession number M3914).
35 species of diminutive, specialized free-floating aquatics con-
sisting of minute fronds or thalli that bear only a few roots
DNA extraction, amplification, and sequencing—Genomic DNA was usu-
(Landoltia, Lemna, Spirodela) or none at all (Wolffia, Wolf- ally extracted from fresh or silica-gel-dried material, but in some instances leaf
fiella) and multiply predominantly by asexual means (Hillman, fragments from herbarium specimens were used. Genomic DNA was extracted
1961; Landolt, 1986, 1998; Les and Crawford, 1999). A link to using a modified 2× cetyltrimethylammonium bromide (CTAB) procedure
aroids has long been suspected (Engler, 1876; Beille, 1935, based on Doyle and Doyle (1987); DNA extracts were purified with QIAquick
cited in Lawalrée, 1945; Cronquist, 1981; Takhtajan, 1997), but minicolumns (Qiagen, Crawley, West Sussex, UK) following the manufactur-
the extreme morphological reduction in the duckweeds made it er’s protocol for cleaning PCR products or precipitated with 100% ethanol at
–20°C and purified on a cesium chloride/ethidium bromide density gradient
difficult to carry out meaningful comparisons for many struc- (1.55 g/mL).
tural traits routinely used in ascertaining taxonomic limits and Amplification of DNA was performed using commercial kits, including
relationships among aroids (e.g., Lawalrée, 1945; Landolt, PCR Master Mix (Advanced Biotechnologies, Epsom, Surrey, UK) and Taq
1986, 1998; Mayo et al., 1995b, 1997; Les et al., 1997, 2002). PCR Core Kit (Qiagen), according to the manufacturers’ protocols. To each
However, recently published molecular phylogenetic studies PCR reaction tube were added 1% of each primer (100 ng/μL) and 2–4% of a
0.4% aqueous solution of bovine serum albumin (BSA) to neutralize phenolic
have consistently placed the duckweeds among Araceae (French compounds and other potential inhibitors (Kreader, 1996).
et al., 1995; Barabé et al., 2002; Rothwell et al., 2004), and The rbcL exon was amplified with primers 1F and 1460R (Asmussen and
most contemporary aroid taxonomists now consider the duck- Chase, 2001) or, in the case of degraded DNA from herbarium specimens, us-
weeds to be members of the latter (e.g., Mayo et al., 1995b; ing as well internal primers 636F and 724R (Muasya et al., 1998). The PCR
Govaerts et al., 2002; Les et al., 2002; Keating, 2002; Renner program was: 2 min initial denaturation at 94°C; 28–30 cycles of 1 min at 94°C,
and Zhang, 2004; Bogner and Petersen, 2007). Phylogenetic 30 s at 48°C, and 1 min at 72°C; 7 min final extension at 72°C. When amplifica-
tion did not yield enough DNA for sequencing, 0.3–0.5 µL of PCR product was
classifications such as APG (1998, 2003) also do not recognize used directly as template for a second PCR with the same parameters as before
Lemnaceae as distinct from Araceae. Nevertheless, evidence but performing only 14 cycles.
for the precise phylogenetic position of the duckweeds within The trnK-matK region, including the matK gene and the 3′ portion of the
Araceae has remained inconclusive. For instance, in the plastid trnK intron, was usually amplified as a single segment with primers −19F
DNA restriction site analysis of French et al. (1995), Lemna is (Molvray et al., 2000) and trnK2R (Steele and Vilgalys, 1994), and the follow-
ing PCR parameters: 2 min 30 s at 94°C; 28–32 cycles of 1 min at 94°C, 45 s at
nested in subfamily Aroideae, but the trnL-trnF DNA sequence 52°C, and an initial 2.5-min extension at 72°C, increasing the time by 8 s on
analysis conducted by Barabé et al. (2002) placed Lemna as each consecutive cycle with a 7 min final extension at 72°C. However, de-
sister to all Araceae sampled except Lysichiton and Symplocar- graded DNA had to be amplified in smaller fragments using additional primers,
pus. More recently, Rothwell et al. (2004) obtained a similar including 390F (Sun et al., 2001), 731F (Molvray et al., 2000), 1309F (Civeyrel
result. They analyzed sequences of the intergenic spacer (IGS) and Rowe, 2001), and 1326R (Sun et al., 2001).
The trnL-trnF region, consisting of the intron in trnL and the trnL-trnF IGS,
between trnL and trnF of 22 exemplars of Araceae and six spe- was amplified either as a single piece with primers c and f or as two fragments
cies that represented all extant genera of Lemnaceae and found with primer combinations c–d and e–f (all from Taberlet et al., 1991). The PCR
that the duckweeds formed a polytomy with various groups of program was: 2 min at 94°C; 28–35 cycles of 30 s at 94°C; 30 s at 52°C; 2 min
“true Araceae.” at 72°C; final extension of 7 min at 72°C. In some instances, this region could
September 2008] Cabrera et al.—Aroid plastid DNA phylogenetics 1155
not be reliably amplified using primer c, and in such cases we used instead found 1105 shortest trees with a length of 7144 steps, an en-
primer c2 of Bellstedt et al. (2001). Some samples required reamplification, semble consistency index (CI, including parsimony uninforma-
which was conducted as in the first round of PCR but only for 16 cycles, using
0.5 μL of the product of the first PCR directly as template.
tive characters; Kluge and Farris, 1969) = 0.50, and an ensemble
PCR products were purified with QIAquick (Qiagen) or CONCERT (Life retention index (RI; Farris, 1989) = 0.71. Figure 1 shows the
Technologies, Paisley, UK) minicolumns following the manufacturers’ proto- strict consensus of the 1105 most parsimonious trees (MPTs),
cols. The cleaned products were used in cycle-sequencing reactions with the which is more resolved than any of the consensus trees from the
Big Dye Terminator Cycle Sequencing Ready Reaction kit version 3 or 3.1 three separate analyses (not shown). Likewise, overall clade
(Applied Biosystems, ABI, Warrington, Cheshire, UK). Cycle sequencing was support is higher than for any of the separate analyses, with 81
carried out in 5.25 μL reactions including 2 μL Big Dye, 0.25 μL primer at the
same concentration as for PCR, and 3 μL PCR product. Products of cycle-se-
clades receiving bootstrap support (BP) greater than 50, and
quencing were purified by ethanol precipitation or with Centri-Sep Sephadex 70.4% of them attaining a BP greater than 85. Acorus is sister
columns (Princeton Separations, Adelphia, New Jersey, USA). Both DNA (BP 100) to Alismatales (BP 100), the latter consisting of a
strands were sequenced in an ABI 377 automated sequencer or a 3100 Genetic clade with Tofieldia moderately supported (BP 82) as sister to
Analyzer. The chromatograms were edited and assembled with Sequencher Triglochin-Alisma (BP 100). Araceae are strongly supported
versions 3.1–4.1 (GeneCodes Corp., Ann Arbor, Michigan, USA). (BP 100) and within them diverge successively the following:
Sequences of both coding regions (rbcL and matK) were aligned visually
trying to maximize similarity (Simmons, 2004). The trnK intron and the trnL-F
Gymnostachydoideae/Orontioideae (BP 91), Lemnoideae (BP
region were initially aligned with Clustal W (Thompson et al., 1994) and sub- 100), Pothoideae/Monsteroideae (BP 100), Lasioideae (BP
sequently adjusted visually following the recommendations of Kelchner (2000). 100), and paraphyletic Aroideae including Calla (Calloideae)
A 233-bp segment of the trnL intron could not be aligned unambiguously and (BP 72). Pothoideae (BP 100) comprise Potheae, including Po-
was excluded from the analyses; the excluded portion amounts to about 4.3% of thos, Pothoidium, and Pedicellarum (BP 100), and monogene-
the data cells of the combined data set. ric Anthurieae. Monsteroideae (BP 100) include a trichotomy
with strongly supported Spathiphylleae (BP 100) and two fur-
Phylogenetic analyses—Previous phylogenetic analyses of aroids (Gonçalves ther clades in which monotypic tribes Heteropsideae (Heterop-
et al., 2007) and duckweeds (Les et al., 2002) based on plastid DNA sequences, sis) and Anadendreae (Anadendrum) are intermingled with
including those studied here, have shown that both resolution and overall boot-
strap support for clades improve when multiple DNA regions are analyzed in
genera of Monstereae. Internal relationships in Lasioideae (BP
combination. Parsimony analyses of the separate (rbcL, matK-trnK, trnL-trnF) and 100) obtained little support; Urospatha is sister to the rest, with
combined plastid data sets were performed using the program PAUP* version Dracontium, Dracontioides, and Anaphyllopsis forming a tri-
4.0b10 for Macintosh (Swofford, 2002) and consisted of heuristic searches with chotomy (BP 84) sister to a polytomy consisting of Lasimorpha
1000 replicates of random sequence addition with the MulTrees option (keeping through Lasia (BP 57). The group consisting of Stylochaeton
multiple trees) activated and tree-bisection-reconnection (TBR) branch swapping, sister to Zamioculcas/Gonatopus (BP 92) is sister to the rest of
saving up to 20 trees per replicate to reduce the time spent in swapping large is-
lands of trees (Maddison, 1991). All characters were unordered and equally
Aroideae (including Calla of Calloideae) (BP 84). The next
weighted. Individual gap positions were treated as missing data, but all nonauta- clade to diverge within Aroideae includes Calla (Calloideae) as
pomorphic indels were coded using the simple method of Simmons and Ochoter- sister (BP < 50) to Cryptocoryneae/Schismatoglottideae (BP
ena (2000) and appended to the sequence matrices as presence/absence characters. 100). Farther up the tree, Anubias and Callopsis diverge succes-
Internal support for clades was evaluated by nonparametric bootstrapping sively, followed by a clade in which Culcasieae are sister to
(Felsenstein, 1985), performing 500 bootstrap replicates, each with five heuristic Homolameneae/Philodendreae (BP < 50), and all these are sis-
replicates and TBR branch swapping, saving up to 20 shortest trees per replicate.
We also conducted a model-based analysis of the combined data set, exclud-
ter to another clade in which Aglaonemateae/Nephthytideae
ing the indels, using Bayesian Markov chain Monte Carlo (MCMC) inference (BP 96) are collective sisters of a weakly supported group (BP
(Yang and Rannala, 1997) as implemented in the computer program MrBayes < 50) consisting of Zantedeschia and a polytomy comprising
version 3.1.2 (Ronquist et al., 2005). A six-parameter model of molecular evo- Dieffenbachieae/Spathicarpeae (BP 100). Culcasieae through
lution with gamma distribution (Yang, 1993) and a proportion of invariant Spathicarpeae is sister to a clade in which Montrichardia is sis-
characters (Reeves, 1992) fit best each of the three separate data sets (rbcL, ter (BP < 50) to Thomsonieae (BP 89) plus Caladieae (includ-
trnK-matK, and trnL-trnF) in the program Modeltest 3.7 (Posada and Crandall,
1998) using both the likelihood ratio test (Goldman, 1993) and the Akaike in-
ing Zomicarpeae) and the “core aroid” clade (BP 100) within
formation criterion (Akaike, 1974). Accordingly, the GTR + I + G model was which the group comprising Ambrosineae through Arophyteae
set in MrBayes. Two independent analyses, each running four Markov chains (BP 97) is sister to a colocasioid grade (BP 100) that includes
and starting with a random tree, were run simultaneously for 3 200 000 genera- Pistieae and Arisaemateae/Areae.
tions, sampling trees every hundredth generation. The temperature of the heated
Markov chains was set to 0.2. Values for the rate matrix and proportion of each Bayesian analysis— The summary Bayesian tree is shown in
nucleotide were estimated from the data as part of the analyses. Stationarity of Fig. 2. In most respects, the relationships recovered by the
log likelihoods was reached around generation 100 000, and the first 800 000
generations (25% of the trees) of each analysis were discarded as the burn-in
Bayesian analysis mirror those depicted in the strict consensus
(Ronquist et al., 2005). A majority-rule consensus tree of the pooled 48 000 of the combined parsimony analysis (Fig. 1). The major differ-
remaining trees (24 000 from each run) was calculated using PAUP*, and inferences between them are in the positions of Stylochaeton/
ences about relationships and posterior probabilities (PP) were based on this Zamioculcadeae and Calla. Whereas in the parsimony tree
tree. Stylochaeton/Zamioculcadeae are moderately supported as sis-
ter to Aroideae (including Calla; BP 72), in the Bayesian results
Stylochaeton/Zamioculcadeae are sister to a weakly supported
RESULTS clade (PP 0.55) formed by Lasioideae and Aroideae. As in the
parsimony results, the Bayesian analysis has Calla nested
Maximum parsimony analyses— Data and statistics for each among members of Aroideae sensu Mayo et al. (1997). How-
region analyzed separately (Appendices S1–S5; see Supple- ever, in the parsimony analysis, Calla is recovered as sister to
mental Data with the online version of this article) and the com- Cryptocoryneae/Schismatoglottideae, whereas in the Bayesian
bined data set are summarized in Table 1. The combined matrix analysis, Calla is sister to the core Aroideae including
of all five regions consisted of 5188 characters, 1683 (32.4%) Pseudodracontium through Arum (PP 0.76; Fig. 2). Cryptoco-
of which were parsimony informative. The heuristic search ryneae/Schismatoglottideae in turn are sister to the Calla/core
Table 1. Summary of characteristics for the data sets analyzed in this study (separate analyses not discussed in the text are available as Appendices
S1–S5; see Supplemental Data with the online version of this article).
Data set No. of aligned characters No. (%) of variable /informative characters No. of shortest trees Tree length CI/RI
rbcL 1391 467 (33.6)/328 (23.6) 360 1465 0.42/0.69

matK 1706 1021 (59.9)/717 (42.3) 11 460 3009 0.51/0.71
3′ portion of the trnK intron 296 173 (58.5)/115 (38.9) 8260 431 0.60/0.79
trnL intron 1006 371 (36.9)/219 (21.8) 17 760 869 0.61/0.73
trnL-trnF IGS 745 431 (57.9)/298 (40.0) 4180 1155 0.55/0.77
All data sets combined 5188 2480 (47.8)/1683 (32.4) 1105 7144 0.50/0.71
Aroideae clade with strong support (PP 0.95). Other differences corroborated by several molecular analyses (e.g., Duvall
with respect to the parsimony analysis include the positions of et al., 1993a, b; Chase et al., 1993, 1995a, b, 2000, 2006; this
Montrichardia and Zantedeschia, but these alternatives ob- study).
tained only weak support. In recent classifications, Gymnostachys has been included in
Araceae either as a member of subfamily Pothoideae (Grayum,
1990) or as a subfamily on its own (Bogner and Nicolson, 1991;
DISCUSSION Mayo et al., 1997, 1998). Our study strongly supports Gymnos-
tachys being sister to Orontioideae, in agreement with the plas-
Familial, subfamilial, and tribal limits and relation- tid DNA restriction site analysis of French et al. (1995), the
ships— Mayo et al. (1997) recognized seven subfamilies in Ar- morphology-based analysis of Mayo et al. (1997), and a cladis-
aceae, namely Gymnostachydoideae, Orontioideae, Pothoideae, tic analysis of DNA sequences of plastid trnL-trnF by Tam
Monsteroideae, Lasioideae, Aroideae, and Calloideae. Our re- et al. (2004). On morphological grounds, Gymnostachys differs
sults support the recognition of the first five subfamilies, but greatly from Orontioideae, and from all other Araceae, in hav-
Calla palustris, the sole member of Calloideae, is nested in ing linear, somewhat plicate leaves with serrate margins, undi-
Aroideae (Figs. 1, 2; see later), whereas the association of Sty- vided into blade and petiole, as well as an inconspicuous spathe
lochaeton and its sister group, Zamioculcadeae, to other Aroi- and a monopodial synflorescence consisting of several floral
deae is only weakly supported by parsimony and was not sympodia borne on an upright scape (Mayo et al., 1997; Buzgo,
recovered by the Bayesian analysis. On the other hand, this 2001). However, Buzgo (2001) noted several structural and de-
study provides evidence supporting inclusion of the duckweeds velopmental similarities between Gymnostachys and both Sym-
in Araceae, showing that they are sister to the “true Araceae” of plocarpus and Lysichiton (Orontioideae), including perigone
Mayo et al. (1997). In the following paragraphs, we discuss tetramery, unidirectional development of the inner whorls of
each of the major clades recovered by our analyses and argue tepals and stamens, and orthotropous, pendant ovules. There-
for recognition of eight subfamilies within Araceae. Our pro- fore, reproductive morphology mirrors our molecular data in
posal differs from the system of Mayo et al. (1997) in that the pointing to a close relationship between Gymnostachys and
duckweeds are included in Araceae as subfamily Lemnoideae, Orontioideae. Tam et al. (2004) suggested combining Gymnos-
Calla is included in Aroideae, and subfamily Zamioculca- tachys and the orontioids in a single subfamily, but in the view
doideae is accepted, but its original circumscription (Bogner of most of us (except MWC), the unique habit and inflorescence
and Hesse, 2005) is broadened to include also Stylochaeton. We morphology of Gymnostachys justifies its maintenance as a dis-
also consider issues related to monophyly and delimitation of tinct subfamily (Bogner and Nicolson, 1991; Mayo et al., 1997,
the tribes recognized by Mayo et al. (1997). 1998; Buzgo, 2001).
Placements of Orontium, Symplocarpus, and Lysichiton have
“Proto-Araceae”: Gymnostachydoideae and Orontioideae— varied among the various aroid classifications. For instance,
A close relationship between Gymnostachys, the sole member Engler (1876) placed all these genera in Symplocarpeae (=
of subfamily Gymnostachydoideae, and Orontioideae was un- Orontieae) within subfamily Pothoideae but later (1920) trans-
suspected until the 1990s. Instead, Gymnostachys was often as- ferred Symplocarpeae to subfamily Calloideae. Hutchinson
sociated with Acorus, either within Araceae (e.g., Schott, 1860; (1973), largely based on previous proposals by Schott (1860)
Engler, 1905, 1920; Hotta, 1970) or as a distinct family (Engler, and Hooker (1883), accommodated Orontium and Lysichiton in
1876). Early molecular studies based on plastid DNA restric- tribe Orontieae but included Symplocarpus in tribe Dracontieae.
tion site variation and sequence data yielded conflicting hy- Grayum (1990) considered all these genera as members of
potheses concerning Gymnostachys. On the one hand, Duvall Lasioideae, placing Symplocarpus and Lysichiton in tribe Sym-
et al.’s (1993a, b) analyses of rbcL sequences placed Gymnos- plocarpeae but Orontium in tribe Orontieae. Bogner and Nicolson
tachys as sister to the other aroids (with Lemna nested among (1991) resurrected Engler’s (1876) concept of Orontieae to
them) and identified Acorus as the sister of the rest of the mono- include Lysichiton, Orontium, and Symplocarpus but assigned
cots, whereas the plastid DNA restriction site analysis of Davis them to subfamily Lasioideae. Recently, Mayo et al. (1997)
(1995) grouped Gymnostachys with Acorus as sister to the rest raised Orontieae sensu Bogner and Nicolson (1991) to subfam-
of the monocot clade, with the two other aroids studied by him ilial rank (as Orontioideae). This group is strongly supported in
(Arisaema and Symplocarpus) either being sister to one another our analyses (Figs. 1, 2). All three genera consist of helophytes
and then to the alismatids or forming a polytomy with the alis- (Orontium often immersed), but there are no unambiguous
matids. Grayum (1987) and Rudall and Furness (1997) pro- morphological features that diagnose them as a group. Mayo
vided ample morphological evidence supporting the removal et al. (1997, 1998) listed a set of characters as distinctive of
of Acorus from Araceae, and their observations have been this group: a nonlinear, expanded leaf blade, anatropous or
Fig. 1. Strict consensus of 1105 trees from the combined parsimony analysis (length = 7144 steps, CI = 0.50, RI = 0.71). Bootstrap percentages >50
are indicated above the branches. Bars indicate the subfamilies and tribes recognized by Mayo et al. (1997). + = nonmonophyletic tribes in the classification
of Mayo et al. (1997).
hemianatropous ovules, and sparse or absent endosperm. How- of clear structural diagnostic features, Orontioideae are consis-
ever, an expanded leaf blade occurs throughout the aroids except tently recovered as monophyletic by plastid DNA sequences, in
in the highly modified Gymnostachydoideae and Lemnoideae agreement with their current subfamilial status.
and is also present in many alismatids. Buzgo (2001) found that
the ovule in Lysichiton is orthotropous, whereas that of Oron- Duckweeds (Lemnoideae)— The duckweeds have long been
tium is hemianatropous, and in addition he cited differences in believed to be closely related to Araceae, but their precise rela-
gynoecium development. Nevertheless, and in spite of the lack tionships have remained unclear (e.g., Mayo et al. 1997 and
references therein). Recently published molecular phylogenetic num, and Amydrium, all with a basic chromosome number of
studies have placed Lemna and its allies in Araceae, although x = 15 and an Old World tropical distribution, except for neo-
the phylogenetic position of the former varied among these tropical Monstera. These two monsteroid clades form a trichot-
studies or was unresolved. As noted earlier, French et al.’s omy with Spathiphylleae, which consist of Spathiphyllum and
(1995) plastid DNA restriction site analysis of Araceae placed Holochlamys. The last two genera share with Monstereae, as
Lemna among representatives of Aroideae, but in the trnL-trnF defined before, a basic chromosome number x = 15 (Mayo et
analysis of Barabé et al. (2002) Lemna is sister to the “true al., 1997; Bogner and Petersen, 2007), which may be an indica-
Araceae” of Mayo et al. (1997). However, in the analysis of the tion of a closer relation of Spathiphylleae with Monsterae than
trnL-trnF IGS by Rothwell et al. (2004), the duckweeds formed with Heteropsideae.
a polytomy with various groups of “true Araceae.”
Our data place the duckweeds as the strongly supported sister Lasioideae— Bogner and Nicolson (1991) proposed tribe
group of the “true Araceae,” in agreement with the findings of Lasieae to include 10 genera, and Mayo et al. (1997) recog-
Barabé et al. (2002) and with palynological evidence (Hesse, nized this group at subfamilial rank. We analyzed nine of the 10
2006). Therefore, the inclusion of Lemna and its allies in genera accepted by Mayo et al. (1997), and these form a strongly
Araceae, as suggested by several workers (e.g., Engler, 1876; supported group (BP 100, PP 1). Members of this group share
Mayo et al., 1995b; Govaerts et al., 2002; Keating, 2002; Hesse, some distinctive features such as no starch in the pollen grains,
2006; Bogner and Petersen, 2007), is fully justified from a phy- well-developed basal ribs of primary leaf veins, dracontioid
logenetic standpoint. Maintaining Lemnaceae as a distinct fam- leaf-margin development, basipetal flowering sequence, anthers
ily results in paraphyly of Araceae. Araceae sensu Mayo et al. dehiscing by oblique pore-like slits, and a basic chromosome
(1997) could be split to maintain Lemnaceae, creating a new number of x = 13. Relationships within this group are only
family for the “proto-aroids” or a new family each for Gymnos- weakly supported, except for a clade formed by the neotropical
tachys and the orontioids. In this way, Araceae would contain genera Dracontium, Dracontioides, and Anaphyllopsis (BP 84,
only the “true Araceae” of Mayo et al. (1997). However, we see PP 1.00). Another clade consists of the Old World tropical gen-
no advantage in dismembering this natural group, with the en- era (Lasimorpha, Podolasia, Pycnospatha, Cyrtosperma, and
suing loss of phylogenetic information and inflation of Lasia), but this obtained only weak support (BP 57). On the
nomenclature. other hand, Urospatha is placed by parsimony as sister to a
clade encompassing the two aforementioned monophyletic
Pothoideae and Monsteroideae— Both these subfamilies are groups, but in the Bayesian analysis it is sister to the Dracon-
strongly supported as monophyletic by our data, as is their sis- tium clade (Fig. 2).
ter-group relationship (Figs. 1, 2). Within Pothoideae, monoge-
neric tribe Anthurieae (Anthurium) is sister to Potheae. All three Expanded Zamioculcadoideae, including Stylochaeton—
genera of Potheae (Pothos, Pedicellarum, and Pothoidium) Recently, Bogner and Hesse (2005) proposed a new subfam-
possess main axes with monopodial growth (a feature only oth- ily, Zamioculcadoideae, to include the distinctive genera
erwise found among the aroids in the monsteroid genus Het- Zamioculcas and Gonatopus. These have been placed in recent
eropsis) and flowering shoots always with axillary inflorescences. classifications either in Pothoideae (Grayum, 1990), Lasioideae
Potheae are restricted to Southeast Asia, Madagascar, and east- (Bogner and Nicolson, 1991), or Aroideae (Mayo et al., 1997,
ern Australia, whereas Anthurium, with sympodial growth, is 1998). The new subfamily was diagnosed by the possession of
restricted to the Neotropics (Mayo et al., 1997). In all these gen- a unique type of pinnatisect or trisect leaves, absence of latic-
era, the secondary and tertiary veins are reticulate, and the ifers and biforines, perigone with four free tepals, zona-aperturate
spathe does not enclose the spadix. Although Pothoideae have pollen grains with a double tectum, and the capacity to propa-
long been recognized as a subfamily (e.g., Engler, 1876, 1920; gate from fallen leaflets (Hesse et al., 2001; Bogner and Hesse,
Grayum, 1990), some authors have included Anthurium in 2005). Bogner and Hesse (2005) also discussed the systematic
Lasioideae (Bogner and Nicolson, 1991). position of Stylochaeton, the only other aroid genus with uni-
Within Monsteroideae, tribe Monstereae is paraphyletic with sexual flowers that also has perigoniate flowers, but because it
monogeneric tribe Heteropsideae nested in a strongly supported lacks the distinctive leaf and pollen features of Zamioculcas
clade consisting of Stenospermation, Heteropsis, Alloschemone, and Gonatopus, those authors concluded that it does not fit into
and Rhodospatha (all these belonging in Monstereae sensu their concept of Zamioculcadoideae. Grayum (1990) consid-
Mayo et al., 1997). This result is in agreement with the analysis ered Stylochaeton an isolated and primitive genus without any
of Tam et al. (2004) based on sequences of trnL-trnF. Anaden- clear link to Zamioculcadeae (including Zamioculcas and
drum (Anadendreae) is placed in another strongly supported Gonatopus), placing it in subfamily Lasioideae. On the other
clade in which it forms a trichotomy with Rhaphidophora hand, in the morphological cladogram of Araceae in Mayo et al.
and (Scindapsus-(Monstera-(Epipremnum-Amydrium))). Tribal (1997), Zamioculcadeae and monotypic Stylochaetonae branch
monophyly would be achieved by expanding Heteropsideae to off successively at the basal nodes of Aroideae. In our study,
also include Stenospermation, Alloschemone, and Rhodospatha, Stylochaeton is strongly supported (BP 92, PP 1.00) as sister
as suggested by Tam et al. (2004). All these genera have a basic to Zamioculcas-Gonatopus. Recognizing Zamioculcadoideae
chromosome number x = 14 and are exclusively neotropical. but leaving out Stylochaeton breaks up this relationship and re-
On the other hand, Monstereae might be redefined to include sults in paraphyly of Aroideae, unless a further subfamily is
Anadendrum, Rhaphidophora, Scindapsus, Monstera, Epiprem- created for Stylochaeton. We find this last option unappealing
←
Fig. 2. Bayesian summary tree from analysis of all regions combined. Numbers above branches are posterior probabilities. Bars indicate subfamilies
recognized by Mayo et al. (1997).
and argue here for the expansion of Zamioculcadoideae to in- (Culcasia and Cercestis). The last relationship is weakly sup-
clude also Stylochaeton. Thus redelimited, Zamioculcadoideae ported by the bootstrap (BP < 50) but obtained a high posterior
consist of geophytic plants restricted to sub-Saharan Africa, probability (PP 0.97). These three groups—Philodendreae,
lacking laticifers and having perigoniate, unisexual flowers. Homalomeneae, and Culcasieae—consist mainly of climbing
hemiepiphytes that possess resin canals in roots, stems, and
Aroideae, including Calla (formerly Calloideae)—Calla leaves and sclerotic hypodermis in the roots. Given their close
palustris, the only member of subfamily Calloideae sensu relationship and great morphological similarity, these three
Bogner and Nicolson (1991; Mayo et al., 1997, 1998), has been tribes might be merged in a more inclusive concept of Philo-
referred to as a highly autapomorphic taxon of obscure affinities dendreae. Tribe Aglaonemateae, consisting of Aglaonema and
(Mayo et al., 1997). Indeed, Calla is puzzling because of its pe- Aglaodorum, are sister to Nephthytideae (Nephthytis, Ancho-
culiar combination of features, including bisexual flowers with manes, and Pseudohydrosme). Both of these strongly supported
some staminate flowers at the apex of the spadix (a condition tribes share adjacent male and female flower zones, free sta-
reminiscent of Orontium, Orontioideae) and absence of a perig- mens, and collenchyma arranged in threads peripheral to the
one—which seems to link it to Aroideae but is also typical of vascular strands of leaf blades and petioles (with the exception
many Monsteroideae. In the cladogram of Mayo et al. (1997), of Nephthytis, in which collenchyma can form interrupted
Calla is the earliest member of the “true Araceae” to diverge, bands; Keating, 2002). Aglaonemateae and Nephthytideae
being sister to all the other Araceae except the protoaroids, but jointly are sisters to a clade in which Zantedeschia is weakly
the plastid DNA restriction site analysis of French et al. (1995) supported as sister to a strongly supported group containing all
placed Calla as sister to a clade matching subfamily Aroideae members of Dieffenbachieae and Spathicarpeae.
sensu Mayo et al. (1997). Our combined parsimony analysis Our analyses failed to provide evidence for the monophyly of
(Fig. 1) places Calla within the clade that is sister to the rest of Dieffenbachieae and Spathicarpeae as classified by Mayo et al.
subfamily Aroideae (excluding Stylochaeton, Zamioculcas, and (1997). Both the parsimony and the Bayesian analyses recov-
Gonatopus, here placed in an expanded subfamily Zamioculca- ered two clades of genera of Spathicarpeae, namely Synandros-
doideae; discussed before), with bootstrap support <50; its sister padix, Spathicarpa, and Taccarum (BP 64, PP 1.00) and
clade is strongly supported and composed of Cryptocoryneae/ Mangonia, Asterostigma, Spathantheum, and Gorgonidium (BP
Schismatoglottideae. Our Bayesian analysis (Fig. 2) also places < 50, PP 0.83). In the parsimony consensus tree these two clades
Calla within Aroideae but in a strongly supported clade (PP form a polytomy with Bognera, Gearum, and Dieffenbachia,
0.95) containing Cryptocoryneae and Schismatoglottideae as which does not exclude the possibility of monophyly. However,
collective sisters to a weakly supported (PP 0.76) group in which in the Bayesian tree Dieffenbachieae are paraphyletic, with
Calla is in turn sister to a clade containing various tribes of de- Bognera occupying an isolated position and Dieffenbachia em-
rived Aroideae (Pseudodracontium through Arum in Fig. 2; PP bedded in Spathicarpeae as sister of the Mangonia clade (PP
1.00). It is worth noting that in both analyses, Calla, a helophyte, 0.95). Thus, all these closely related genera might best be merged
is located near tribes Cryptocoryneae and Schismatoglottideae, in a single tribe pending further study, with Spathicarpeae hav-
which consist predominantly of helophytic and rheophytic spe- ing nomenclatural priority. These results are in agreement with
cies. Given its embedded position within Aroideae, Calla must a recently published phylogenetic analysis of the genera of
be included in that subfamily, and tribal status (as Calleae Schott) Spathicarpeae and Dieffenbachieae by Gonçalves et al. (2007)
seems advisable in view of its peculiar morphology. based on plastid DNA sequences and morphology and a broader
Aroideae as understood here (i.e., including Calla but ex- taxonomic sample, in which they expand Spathicarpeae to in-
cluding Stylochaeton, Zamioculcas, and Gonatopus) obtained clude also Bognera and Dieffenbachia. The group is nearly ex-
strong support (BP 84, PP 1.00) and morphologically is distin- clusively South American, with only Dieffenbachia spreading
guished by several features, such as the aperigoniate, unisexual through Central America north to southeastern Mexico.
flowers (except Calla) and laticifers (except Pistia). Both The last major clade of Aroideae recovered by our analyses
aperigoniate flowers and laticifers represent putative synapo- is strongly supported (BP 100, PP 1.00) and consists of two
mophies of Aroideae, but unisexual flowers are symplesiomor- groups. In the first group, tribe Thomsonieae (Amorphophallus
phic (they are also present in Zamiculcadoideae). The clade and Pseudodracontium) are strongly supported as sister of Ca-
including Schismatoglottideae-Cryptocoryneae (and Calla, in ladieae, in which paraphyletic Zomicarpeae are nested. Of the
the combined parsimony analysis) is sister to the rest. Monoge- genera included in Zomicarpeae by Mayo et al. (1997), Ule-
neric Anubiadeae (Anubias) and Callopsideae (Callopsis) di- arum and Filarum are sisters to each other, whereas Zomi-
verge successively. Callopsis is weakly supported as sister to carpella is strongly supported as sister to Scaphispatha (BP
the rest of Aroideae by parsimony (BP < 50), but strongly so by 100, PP 1.00) within a weakly supported clade of Caladieae that
the Bayesian analysis (PP 1.00). In the Bayesian analysis, also includes Chlorospatha and Xanthosoma. Mayo et al. (1997)
Montrichardia (Montrichardieae) formed a polytomy with grouped tribes Caladieae and Zomicarpeae in their “Caladium
two strongly supported clades of Aroideae, but in the parsi- alliance,” and the mingling of their constituent genera in the
mony strict consensus it was recovered as sister to one of them molecular trees supports the inclusion of all of them in a more
(BP < 50). broadly defined Caladieae, as in Keating (2002). Thus rede-
The remaining Aroideae form a weakly supported major fined, Caladieae are supported by two putative morphological
clade that in turn contains two groups. The first of them in- synapomorphies, i.e., anastomosing laticifers and at least a par-
cludes, on the one hand, members of Culcasieae, Philodendreae, tial adnation of the female part of the spadix to the spathe (Mayo
Homalomeneae, Aglaonemateae, and Nephthytideae, and on et al., 1997; Keating, 2002). Such adnation is more evident in
the other Zantedeschieae, Dieffenbachieae, and Spathicar- the genera previously assigned to Zomicarpeae but only be-
peae. Philodendreae (Philodendron) and Homalomeneae cause of the smaller number of female flowers they produce.
(Homalomena) form a strongly supported sister to Culcasieae Except for Southeast Asian geophytes of the genus Hapaline,
the tribe is restricted to the neotropics. In our analyses, Hapa- vasions of the aquatic habitat have occurred in subfamilies
line is the weakly supported sister of Jasarum, a distinctive ge- Lasioideae (nearly all genera) and Aroideae (mainly Calla,
nus consisting of a single species that is a submerged aquatic in Cryptocoryneae, Schismatoglottideae, Peltandreae, Anubias,
oligotrophic upland streams (Mayo et al., 1997). Montrichardia, Jasarum, and Pistia). All these are secondarily
The second group (BP 100, PP 1.00) encompasses a clade with aquatic in a morphologically canalized aroid clade and have the
monotypic tribes Ambrosinae (Ambrosina) and Arisareae (Ari- characteristic aroid features. However, both orontioids and
sarum) as collective sisters of another clade formed by Peltan- duckweeds are much simpler and diverged from the main aroid
dreae and Arophyteae. Both Ambrosina and Arisarum have line before morphology became canalized in the manner now
reticulate higher order venation, basal placentation, and copious considered to be the hallmark of the family. Thus, the early
endosperm. The two genera of Peltandreae, Peltandra and Typho- evolution of the aroid lineage, estimated to have occurred more
nodorum, do not group with each other in the strict consensus of than 120 million years ago as indicated by early Cretaceous
the parsimony analysis, but they form a clade in the Bayesian tree pollen assigned to tribe Spathyphylleae (Friis et al., 2004), ap-
(PP 0.81). Both genera consist of helophytes inhabiting fresh and pears to have involved “experimentation” with a wide range of
brackish water, but they have a bizarre disjunction, with Pelt- habits and inflorescence architectures, which were later cana-
andra restricted to eastern North America and Typhonodorum lized in the main aroid clade into the typical highly stereotyped
found in the Comores, Madagascar, Mauritius, and some islands syndrome of spathe and spadix. This group then underwent a
off the coast of Tanzania (Mayo et al., 1997). The Madagascan major radiation in the tropics, more recently evolving again into
endemic tribe Arophyteae obtained moderate and strong support tropical aquatic habitats (e.g., Pistia, Cryptocoryne) and tem-
in the parsimony and Bayesian analyses, respectively. The clade perate environments (Areae). Both protoaroids and duckweeds,
formed by Ambrosinae, Arisareae, and Peltandreae is sister to the lacking the canalized development typical of the family, have
“Pistia clade” as defined by of Renner and Zhang (2004), which not undergone a comparable radiation of taxa. This same pat-
consists of genera placed by Mayo et al. (1997) in tribes Pistieae, tern—a few, relatively species-poor, morphologically atypical
Colocasieae, Arisaemateae, and Areae. lineages diverging prior to a major morphological canaliza-
Our results do not support monophyly of Colocasieae, in tion—is found in most other large plant families as well, includ-
which a clade formed by Arisaemateae (Arisaema, Pinellia) ing orchids, grasses, sedges, legumes, and composites (cf.
and Areae (Typhonium, Eminium, Helicodiceros, Biarum, Dra- Chase, 2004). Many of these atypical lineages have also been
cunculus, and Arum) is nested. Alocasia is weakly supported as proposed as separate families, as in the case of the duckweeds.
sister to the Arisaemateae/Areae clade in the parsimony analy- Several earlier workers have pointed to the morphological
sis (BP 57) but strongly so in the Bayesian analysis (PP 1.00). and ecological similarities between the duckweeds and Pistia
The recently published study of the “Pistia clade” by Renner as evidence of a close relationship (reviewed in Landolt, 1986,
and Zhang (2004), based on noncoding plastid and mitochon- 1998; Mayo et al., 1995b, 1997; Stockey et al., 1997; Lemon
drial DNA, recovered Colocasieae as paraphyletic to Arisae- and Posluszny, 2000b; Les et al., 2002), but both DNA sequence
mateae and Areae, but in their tree the species of Colocasia data (Barabé et al., 2002; Rothwell et al., 2004; this paper) and
sampled (C. gigantea) grouped with two species of Alocasia as structural studies (Mayo et al., 1997 and references therein;
collective sisters of the Arisaemateae-Areae clade. In our Hesse, 2006) have shown that these two groups are only dis-
Bayesian tree, Protarum (Colocasieae sensu Mayo et al., 1997) tantly related. The duckweeds are an early offshoot of the aroid
is sister to a weakly supported clade (PP 0.66) consisting of lineage and are represented in the fossil record since the late
Pistia (Pistieae) as sister of the remaining Colocasieae plus Ari- Cretaceous by the genus Limnobiophyllum (Kvaček, 1995;
saemateae and Areae, but in the parsimony strict consensus tree Stockey et al., 1997), whereas the oldest fossils attributable to
Protarum and Pistia are sister taxa with BP < 50 (Figs. 1, 2). In Pistia date back only to late Oligocene/early Miocene (Renner
the study of Renner and Zhang (2004: Fig. 3), Protarum and and Zhang, 2004; Wilde et al., 2005). Their similarities are best
Pistia form a polytomy with a clade that includes other Colo- interpreted as independent evolutionary acquisitions resulting
casieae, Areae, and Arisemateae. These results suggest the need from adaptation to the aquatic environment (Landolt, 1986,
to revise the current tribal limits of Colocasieae. Distinctive 1998; Rothwell et al., 2004; cf. Grace, 1993). However, the no-
Protarum should be removed from Colocasieae to monotypic table likeness of Pistia in habit and development to the duck-
Protareae, as proposed by Engler (1920), but resolution of rela- weeds (Lemon and Posluszny, 2000a, b) makes Pistia a useful
tionships among Colocasia, Alocasia, and the other Colocasieae living model of the sort of modifications that may have led to
requires further study, including more thorough sampling. the extreme reduction and specialization of the duckweeds.
Arisaemateae, consisting of Pinellia and Arisaema, obtained
only BP < 50 and PP 0.57, and there are no obvious morphologi- Concluding remarks— This is the first molecular phyloge-
cal features permitting diagnosis. On the other hand, Typhonium, netic study that includes a nearly complete generic sample of
Eminium, Helicodiceros, Biarum, Dracunculus, and Arum all aroids and all extant genera of duckweeds, as well as represen-
belong in the strongly supported (BP 99, PP 1.00) tribe Areae of tatives of other alismatids and nonmonocot outgroups, to assess
Mayo et al. (1997). This tribe is predominantly Mediterranean monophyly and relationships of the major lineages of Araceae.
and Asian (with Typhonium reaching Australia) and includes The position of the duckweeds (subfamily Lemnoideae) as sis-
plants from temperate, seasonally dry, cold environments. ter of all other Araceae except the “proto-aroids” is strongly
supported. The delimitation of most subfamilies, tribes, and in-
Evolution of the aquatic habit in Araceae— With the excep- formal alliances proposed by Mayo et al. (1997) is consistent
tion of the highly autapomorphic Gymnostachys, all of the small with the phylogenetic trees, with the notable exception of Aroi-
clades diverging at the basalmost nodes consist of aquatic deae, which is paraphyletic unless Calla, placed in its own sub-
plants: Orontioideae are helophytes, and Lemnoideae include family (Calloideae), is included. The recently proposed subfamily
only free-floating aquatics. Several additional, independent in- Zamioculcadoideae is supported by our data but should also
include Stylochaeton because retaining it in Aroideae, as pro- Chase, M. W., M. F. Fay, D. S. Devey, O. Maurin, N. Rønsted, J.
posed by Bogner and Hesse (2005), renders the latter nonmono- Davies, Y. Pillon, et al. 2006. Multi-gene analyses of monocot
phyletic. The following eight subfamilies might then be relationships: A summary. Aliso 22: 63–75.
recognized in Araceae (in ascending order of their branching in Chase, M. W., D. E. Soltis, R. G. Olmstead, D. Morgan, D. H. Les,
B. D. Mishler, M. R. Duvall, et al. 1993. Phylogenetics of seed
the phylogenetic trees): Gymnostachydoideae, Orontioideae, plants: An analysis of nucleotide sequences from the plastid gene
Lemnoideae, Pothoideae, Monsteroideae, Lasioideae, Zamio- rbcL. Annals of the Missouri Botanical Garden 80: 528–580.
culcadoideae, and Aroideae. Our phylogenetic hypotheses are Chase, M. W., D. E. Soltis, P. S. Soltis, P. J. Rudall, M. F. Fay, W.
derived from DNA sequence data from a single compartment of J. Hahn, S. Sullivan, et al. 2000. Higher–level systematics of
the plant genome (plastid DNA), and it would be desirable to the monocotyledons: an assessment of current knowledge and a new
compare them with results from other sources of evidence, in- classification. In K. L. Wilson, and D. A. Morrison [eds.], Monocots:
cluding, for instance, nuclear markers and structural characters. Systematics and evolution, 3–16, CSIRO, Collinwood, Australia.
We hope that the ideas presented here stimulate further research Chase, M. W., D. W. Stevenson, P. Wilkin, and P. J. Rudall.
focused on producing a better understanding of the evolution of 1995b. Monocot systematics: A combined analysis. In P. J. Rudall,
this highly diverse group of monocots. P. J. Cribb, D. F. Cutler, and C. J. Humphries [eds.]. Monocotyledons:
Systematics and evolution, vol. 2, 685–730. Royal Botanic Gardens,
Kew, UK.
LITERATURE CITED Civeyrel, L., and N. Rowe. 2001. Relationships of Secamonoideae
Akaike, H. 1974. A new look at the statistical model identification. IEEE based on the plastid gene matK, morphology and biomechanics.
Transactions on Automatic Control 19: 716–723. Annals of the Missouri Botanical Garden 88: 583–602.
APG [Angiosperm Phylogeny Group]. 1998. An ordinal classification Croat, T. B. 1988. Ecology and life-forms of Araceae. Aroideana 11:
for the families of flowering plants. Annals of the Missouri Botanical 4–56.
Garden 85: 531–553. Croat, T. B. 1990. A comparison of aroid classification systems.
APG [Angiosperm Phylogeny Group]. 2003. An update of the Aroideana 13: 44–63.
Angiosperm Phylogeny Group classification for the orders and families Croat, T. B. 1998. History and current status on systematic research with
of flowering plants: APG II. Botanical Journal of the Linnean Society Araceae. Aroideana 21: 26–145.
141: 399–436. Cronquist, A. 1981. An integrated system of classification of flowering
Asmussen, C. B., and M. W. Chase. 2001. Coding and noncoding plants. Columbia University Press, New York, New York, USA.
plastid DNA in palm systematics. American Journal of Botany 88: Davis, J. I. 1995. A phylogenetic structure for monocotyledons, as in-
1103–1117. ferred from chloroplast restriction site variation, and a comparison of
Barabé, D., A. Bruneau, F. Forest, and C. Lacroix. 2002. The correla- measures of clade support. Systematic Botany 20: 503–527.
tion between development of atypical bisexual flowers and phylogeny Doyle, J. J., and J. L. Doyle. 1987. A rapid DNA isolation procedure from
in the Aroideae (Araceae). Plant Systematics and Evolution 232: 1–19. small quantities of fresh leaf tissue. Phytochemical Bulletin 19: 11–15.
Barabé, D., C. Lacroix, A. Bruneau, A. Archambault, and M. Duvall, M. R., M. T. Clegg, M. W. Chase, W. D. Clark, W. J. Kress,
Gibernau. 2004. Floral development and phylogenetic position of H. G. Hills, L. E. Eguiarte, et al. 1993a. Phylogenetic hypotheses
Schismatoglottis (Araceae). International Journal of Plant Sciences for the monocotyledons constructed from rbcL sequence data. Annals
165: 173–189. of the Missouri Botanical Garden 80: 607–619.
Bateman, R. M. 1999. Integrating molecular and morphological evidence Duvall, M. R., G. H. Learn, L. E. Eguiarte, and M. T. Clegg.
of evolutionary radiations. In P. M. Hollingsworth, R. M. Bateman, 1993b. Phylogenetic analysis of rbcL sequences identifies Acorus
and R. J. Gornall [eds.], Molecular systematics and plant evolution, calamus as the primal extant monocotyledon. Proceedings of the
432–471. Taylor & Francis, London, UK. National Academy of Sciences, USA 90: 4641–4644.
Beille, L. 1935. Précis de botanique pharmaceutique, éd. 2, II: 193–194. Engler, A. 1876. Vergleichende Untersuchungen über die morphologis-
Bellstedt, D. U., H. P. Linder, and E. Harley. 2001. Phylogenetic chen Verhältnisse der Araceae. I. Natürliches System der Araceae.
relationships in Disa based on non-coding trnL–trnF chloroplast se- Nova Acta Academiae Caesareae Leopino–Carolinae Germanicae
quences: evidence of numerous repeat regions. American Journal of Naturae Curiosorum 39: 133–155.
Botany 88: 2088–2100. Engler, A. 1905. Araceae–Pothoideae. In A. Engler [ed.], Das
Bogner, J. 1979. A critical list of the aroid genera. Aroideana 1: 63–73. Pflanzenreich 21 (IV.23B), 1–330.
Bogner, J., and M. Hesse. 2005. Zamioculcadoideae, a new subfamily of Engler, A. 1920. Araceae—Pars generalis et index familiae generalis. In
Araceae. Aroideana 28: 3–20. A. Engler [ed.], Das Pflanzenreich 74 (IV.23A), 1–71.
Bogner, J., and D. H. Nicolson. 1991. A revised classification of Araceae Farris, J. S. 1989. The retention index and the rescaled consistency index.
with dichotomous keys. Willdenowia 21: 35–50. Cladistics 5: 417–419.
Bogner, J., and G. Petersen. 2007. The chromosome numbers of the Felsenstein, J. 1985. Confidence limits on phylogenies: An approach
aroid genera. Aroideana 30: 82–90. using the bootstrap. Evolution; International Journal of Organic
Bown, D. 2000. Aroids: Plants of the Arum family, 2nd ed. Timber Press, Evolution 39: 783–791.
Portland, Oregon, USA. French, J. C., M. G. Chung, And Y. K. Hur. 1995. Chloroplast DNA
Boyce, P. C. 1995. Introduction to the family Araceae. Curtis’s Botanical phylogeny of the Ariflorae. In P. J. Rudall, P. J. Cribb, D. F. Cutler,
Magazine 12: 122–125. and C. J. Humphries [eds.], Monocotyledons: Systematics and evolu-
Brooks, D. R., and D. A. McLennan. 1991. Phylogeny, ecology, and tion, vol. 1, 255–275. Royal Botanic Gardens, Kew, UK.
behavior. University of Chicago Press, Chicago, Illinois, USA. Friis, E. M., K. R. Pedersen, and P. R. Crane. 2004. Araceae from the
Buzgo, M. 2001. Flower structure and development of Araceae compared early Cretaceous of Portugal: Evidence on the emergence of mono-
with alismatids and Acoraceae. Botanical Journal of the Linnean cotyledons. Proceedings of the National Academy of Sciences, USA
Society 136: 393–425. 101: 16565–16570.
Chase, M. W. 2004. Monocot relationships: An overview. American Futuyma, D. J. 2004. The fruit of the tree of life. In J. Cracraft and
Journal of Botany 91: 1645–1655. M. J. Donoghue [eds.], Assembling the tree of life, 25–39. Oxford
Chase, M. W., M. R. Duvall, H. G. Hillis, J. G. Conran, A. V. Cox, University Press, Oxford, UK.
L. E. Eguiarte, H. Hartwell, et al. 1995a. Molecular phyloge- Givnish, T. J., and K. J. Systma. [eds.]. 1997. Molecular evolution and
netics of Lilianae. In P. J. Rudall, P. J. Cribb, D. F. Cutler, and C. J. adaptive radiation. Cambridge University Press, Cambridge, UK.
Humphries [eds.]. Monocotyledons: Systematics and evolution, vol. Goldman, N. 1993. Statistical tests of models of DNA substitution.
1, 109–137. Royal Botanic Gardens, Kew, UK. Journal of Molecular Evolution 36: 182–198.
Gonçalves, E. G., S. J. Mayo, M.-A. Van Sluys, and A. Salatino. Les, D. H., E. Landolt, and D. J. Crawford. 1997. Systematics of the
2007. Combined genotypic–phenotypic phylogeny of the tribe Lemnaceae (duckweeds): Inferences from micromolecular and mor-
Spathicarpeae (Araceae) with reference to independent events of in- phological data. Plant Systematics and Evolution 204: 161–177.
vasion to Andean regions. Molecular Phylogenetics and Evolution 43: Maddison, R. D. 1991. The discovery and importance of multiple islands
1023–1039. of most-parsimonious trees. Systematic Zoology 40: 315–328.
Govaerts, R., D. G. Frodin, J. Bogner, J. Boos, P. Boyce, B. Cosgriff, Mayo, S. J., J. Bogner, and P. C. Boyce. 1995a. The acolytes of the
T. B. Croat, et al. 2002. World checklist and bibliography of Araceae. Curtis’s Botanical Magazine 12: 153–168.
Araceae and Acoraceae. Royal Botanic Gardens, Kew, UK. Mayo, S. J., J. Bogner, and P. C. Boyce. 1995b. The Arales. In
Grace, J. B. 1993. The adaptive significance of clonal reproduction in P. J. Rudall, P. J. Cribb, D. F. Cutler, and C. J. Humphries [eds.],
angiosperms: An aquatic perspective. Aquatic Botany 44: 159–180. Monocotyledons: Systematics and evolution, vol. 1, 277–286. Royal
Grayum, M. H. 1987. A summary of evidence and arguments supporting Botanic Gardens, Kew, London, UK.
the removal of Acorus from Araceae. Taxon 36: 723–729. Mayo, S. J., J. Bogner, and P. C. Boyce. 1997. The genera of Araceae.
Grayum, M. H. 1990. Evolution and phylogeny of the Araceae. Annals of Royal Botanic Gardens, Kew, UK.
the Missouri Botanical Garden 77: 628–697. Mayo, S. J., J. Bogner, and P. C. Boyce. 1998. Araceae. In K. Kubitzki
Harvey, P. H., A. J. Leigh Brown, J. Maynard Smith, and S. Nee. [ed.], The families and genera of vascular plants, vol. 4, 26–74.
1996. New uses for new phylogenies. Oxford University Press, Springer–Verlag, Berlin, Germany.
Oxford, UK. Molvray, M. P., P. J. Kores, and M. W. Chase. 2000. Polyphyly
Harvey, P. H., and M. D. Pagel. 1991. The comparative method in evo- of mycoheterotrophic orchids and functional influences on flo-
lutionary biology. Oxford University Press, Oxford, UK. ral and molecular characters. In K. L. Wilson and D. A. Morrison
Hesse, M. 2006. Pollen wall ultrastructure of Araceae and Lemnaceae in [eds.], Monocots: Systematics and evolution, 441–448. CSIRO,
relation to molecular classifications. Aliso 22: 204–208. Collingwood, Australia.
Hesse, M., J. Bogner, H. Halbritter, and M. Weber. 2001. Palynology Muasya, A. M., D. A. Simpson, M. W. Chase, and A. Culham. 1998. An
of the perigoniate Aroideae: Zamioculcas, Gonatopus and Stylochaeton assessment of suprageneric phylogeny in Cyperaceae using rbcL DNA
(Araceae). Grana 40: 26–34. sequences. Plant Systematics and Evolution 211: 257–271.
Hillman, W. S. 1961. The Lemnaceae, or duckweeds: A review of the de- Nickrent, D. L., R. J. Duff, A. E. Colwell, A. D. Wolfe, N. D.
scriptive and experimental literature. Botanical Review 27: 221–287. Young, K. E. Steiner, and C. W. Depamphilis. 1998. Molecular
Hooker, J. F. 1883. Aroideae. In G. Bentham and J. D. Hooker. Genera phylogenetic and evolutionary studies of parasitic plants. In D.
Plantarum, vol. 3, 995–1000. Reeve, London, UK. E. Soltis, P. S. Soltis, and J. J. Doyle [eds.], Molecular system-
Hotta, M. 1970. A system of the family Araceae in Japan and adjacent ar- atics of plants II: DNA sequencing, 211–241. Kluwer, Boston,
eas. Memoirs of the Faculty of Science of Kyoto Imperial University. Massachusetts, USA.
Series for Biology 4: 72–96. Nicolson, D. H. 1960. A brief review of classifications in the Araceae.
Hutchinson, J. 1973. The families of flowering plants. Clarendon Press, Baileya 8: 62–67.
Oxford. UK. Nicolson, D. H. 1987. History of aroid systematics. Aroideana 10:
Judd, W. S., C. S. Cambell, E. A. Kellogg, P. F. Stevens, and M. 23–30.
J. Donoghue. 2002. Plant systematics: A phylogenetic approach. Posada, D., and K. A. Crandall. 1998. Modeltest: Testing the model
Sinauer, Sunderland, Massachusetts, USA. of DNA substitution. Bioinformatics (Oxford, England) 14: 817–818.
Keating, R. C. 2002. Acoraceae and Araceae. In M. Gregory and D. F. Reeves, J. H. 1992. Heterogeneity in the substitution process of amino
Cutler [eds.], Anatomy of the monocotyledons, vol. 9, 1–327. Oxford acid sites of proteins coded for by mitochondrial DNA. Journal of
University Press, Oxford, UK. Molecular Evolution 35: 17–31.
Kelchner, S. A. 2000. The evolution of non–coding chloroplast DNA Renner, S. S., and L.-B. Zhang. 2004. Biogeography of the Pistia clade
and its application in plant systematics. Annals of the Missouri (Araceae): Based on chloroplast and mitochondrial DNA sequences
Botanical Garden 87: 482–498. and Bayesian divergence time inference. Systematic Biology 53:
Kluge, A. G., and J. S. Farris. 1969. Quantitative phyletics and the evo- 422–432.
lution of anurans. Systematic Zoology 18: 1–32. Renner, S. S., L.-B. Zhang, and J. Murata. 2004. A chloroplast phy-
Kreader, C. A. 1996. Relief of amplification inhibition in PCR with bo- logeny of Arisaema (Araceae) illustrates Tertiary floristic links be-
vine serum albumin or T4 gene 32 protein. Applied and Environmental tween Asia, North America, and East Africa. American Journal of
Microbiology 62: 1102–1106. Botany 91: 881–888.
KvaČek, Z. 1995. Limnobiophyllum Krassilov—A fossil link between the Ronquist, F., J. P. Huelsenbeck, and P. Van Der Mark. 2005. MrBayes
Araceae and the Lemnaceae. Aquatic Botany 50: 49–61. 3.1 manual, draft 5/17/2005. Program documentation and manual.
Landolt, E. 1986. Biosystematic investigations in the family of duck- Website http://morphbank.ebc.uu.se/mrbayes/ [accessed 17 May 2005].
weeds (Lemnaceae), vol. 2, The family of Lemnaceae—A monographic Rothwell, G. W., M. R. Van Atta, H. W. Ballard Jr., and R. A.
study, vol. 1. Veröffentlichungen des Geobotanischen Institutes der Stockey. 2004. Molecular phylogenetic relationships among
ETH, Stiftung, Rübel, Zürich, Heft 71. Zürich, Switzerland. Lemnaceae and Araceae using the chloroplast trnL–trnF spacer.
Landolt, E. 1998. Lemnaceae. In K. Kubitzki [ed.], The families and genera Molecular Phylogenetics and Evolution 30: 378–385.
of vascular plants, vol. 4, 264–269. Springer–Verlag, Berlin, Germany. Rudall, P. J., and C. A. Furness. 1997. Systematics of Acorus: Ovule
Lawalrée, A. 1945. La position systématique des Lemnaceae et leur and anther. International Journal of Plant Sciences 158: 640–651.
classification. Bulletin de la Société Royale de Botanique de Belgique Schott, H. W. 1860. Prodromus systematis aroidearum. Congretationis
77: 27–38. Mechitharisticae, Vienna, Austria.
Lemon, G. D., and U. Posluszny. 2000a. Shoot development in Pistia Simmons, M. P. 2004. Independence of alignment and tree search.
stratiotes (Araceae). International Journal of Plant Sciences 161: Molecular Phylogenetics and Evolution 31: 874–879.
721–732. Simmons, M. P., and H. Ochoterena. 2000. Gaps as characters in
Lemon, G. D., and U. Posluszny. 2000b. Comparative shoot develop- sequence–based phylogenetic analyses. Systematic Biology 49:
ment and evolution in the Lemnaceae. International Journal of Plant 369–381.
Sciences 161: 733–748. Soltis, D. E., M. W. Mort, P. S. Soltis, C. Hibsch-Jetter, E. A. Zimmer,
Les, D. H., and D. J. Crawford. 1999. Landoltia (Lemnaceae), a new and D. Morgan. 1999. Phylogenetic relationships of the enigmatic
genus of duckweeds. Novon 9: 530–533. angiosperm family Podostemaceae inferred from 18S rDNA and rbcL
Les, D. H., D. J. Crawford, E. Landolt, J. D. Gabel, and R. T. sequence data. Molecular Phylogenetics and Evolution 11: 261–272.
Kimball. 2002. Phylogeny and systematics of Lemnaceae, the duck- Soltis, D. E., and P. S. Soltis. 2004. The origin and diversification of
weed family. Systematic Botany 27: 221–240. angiosperms. American Journal of Botany 91: 1614–1626.
Soltis, D. E., P. S. Soltis, M. W. Chase, M. W. Mort, D. C. Albach, Takhtajan, A. 1997. Diversity and classification of flowering plants.
M. Zanis, V. Savolainen, et al. 2000. Angiosperm phylogeny in- Columbia University Press, New York, USA.
ferred from 18S rDNA, rbcL, and atpB sequences. Botanical Journal Tam, S.-M., P. C. Boyce, T. M. Upson, D. Barabé, A. Bruneau, F.
of the Linnean Society 133: 381–461. Forest, and J. S. Parker. 2004. Intergeneric and infrafamil-
Soltis, D. E., P. S. Soltis, P. K. Endress, and M. W. Chase. ial phylogeny of subfamily Monsteroideae (Araceae) revealed by
2005. Phylogeny and evolution of angiosperms. Sinauer, Sunderland, chloroplast trnL–F sequences. American Journal of Botany 91:
Massachusetts, USA. 490–498.
Steele, K. P., and R. Vilgalys. 1994. Phylogenetic analysis of Thompson, J. D., D. G. Higgins, and R. J. Gibson. 1994. Clustal W:
Polemoniaceae using nucleotide sequences of the plastid gene matK. Improving the sensitivity of progressive multiple sequence align-
Systematic Botany 19: 126–142. ment through sequence weighting, position-specific gap pen-
Stockey, R. A., G. L. Hoffman, and G. W. Rothwell. 1997. The fos- alties and weight matrix choice. Nucleic Acids Research 22:
sil monocot Limnobiophyllum scutatum: Resolving the phylogeny of 4673–4680.
Lemnaceae. American Journal of Botany 84: 355–368. Wilde, V., Z. KvaČek, and J. Bogner. 2005. Fossil leaves of the
Sun, H., W. McLewin, and M. F. Fay. 2001. Molecular phylogeny of Araceae from the European Eocene and notes on other aroid fossils.
Helleborus (Ranunculaceae), with an emphasis on the east Asian– International Journal of Plant Sciences 166: 157–183.
Mediterranean disjunction. Taxon 50: 1001–1018. Yang, Z. 1993. Maximum-likelihood estimation of phylogeny form DNA
Swofford, D. L. 2002. PAUP*: Phylogenetic analysis using parsimony (*and sequences when substitution rates differ over sites. Molecular Biology
other methods), version 4.0b. Sinauer, Sunderland, Massachusetts, USA. and Evolution 10: 1396–1401.
Taberlet, P., L. Gielly, G. Pautou, and J. Bouvet. 1991. Universal Yang, Z., and B. Rannala. 1997. Bayesian pylogenetic inference using
primers for amplification of three non-coding regions of chloroplast DNA sequences: A Markov chain Monte Carlo method. Molecular
DNA. Plant Molecular Biology 17: 1105–1109. Biology and Evolution 14: 717–724.
Appendix 1. Voucher information and GenBank accession numbers for taxa used in this study. Herbarium acronyms or botanical gardens: Bot. Gard. Osaka City
Univ. = Botanical Garden Osaka City University, Osaka, Japan; CHRB = Rutgers University-Cook College, New Brunswick, New Jersey, USA; CONN
= University of Connecticut, Storrs, Connecticut, USA; K = Royal Botanic Gardens, Kew, UK; L = Nationaal Herbarium Nederland, Leiden University
branch, Leiden, Netherlands; M = Botanische Staatssammlung München, Munich, Germany; MEXU = Herbario Nacional, Universidad Nacional Autónoma
de México, Mexico City, Distrito Federal, Mexico; MO = Missouri Botanical Garden, Saint Louis, Missouri, USA; MT = Université de Montréal, Montréal,
Québec, Canada; NCU = University of North Carolina, Chapel Hill, North Carolina, USA; SING = Singapore Botanic Gardens, Singapore, Singapore; UPS =
Uppsala University, Uppsala, Sweden; — = not available.
FAMILY. Species, Voucher (Herbarium), GenBank accessions (+ = sequenced (K), AM905797, AM920619, AM932347 + AM933343. Arophyton
as two separate fragments): rbcL, matK-trnK, trnL-F. buchetii Bogner, Bogner 207 (M), AM905820, AM920642, AM932367
+ AF521870. Arum hygrophilum Boiss., Chase 10990 (K), AM905809,
ACORACEAE. Acorus calamus L., French 232 (CHRB), M91625, —, —; A.
AM920631, AM932296 + AM933353. Asterostigma pavonii Schott,
calamus L., Tamura & Yamshita 6008 (Bot. Gard. Osaka City Univ.), —,
Sizemore 95-062B (L), AM905768, AM920590, AM932325 + AM933321.
AB040154, —; A. calamus L., Joly 226 (MT), —, —, AY054741.
Biarum tenuifolium (L.) Schott, Chase 282 (K), AM905810, AM920632,
ALISMATACEAE. Alisma canaliculatum A.Braun & C.D.Bouché, Tamura
AM932357 + AM933354. Bognera recondita (Madison) Mayo &
& Fuse 10018 (Bot. Gard. Osaka City Univ.), —, AB040179, —; A.
Nicolson, Bogner 1995 (M), AM905765, AM920587, AM932322
plantago-aquatica L., Les s.n. (CONN), L08759, —, —; A. plantago
+ AM933318. Bucephalandra motleyana Schott, Tomey s.n. (M),
aquatica L, Chase 11275 (K), —, —, AM932372 + AM933368,
AM905822, AM920644, AM932369 + AM933365.
ARACEAE. Aglaodorum griffithii (Schott) Schott, Bogner 1767 (M),
AM905758, AM920580, AM932318 + AM933314. Aglaonema modestum Caladium bicolor (Aiton) Vent., Barabé & Chantha 96 (MT), —, —, AY054708;
Schott ex Engl., Chase 10671 (K), AM905757, AM920579, —; A. C. lindenii (André) Madison, Chase 10670 (K), AM905788, AM920610,
modestum Schott ex Engl., Barabé & Chantha 86 (MT), —, —, AY054700. —. Calla palustris L., Chase 11802 (K), AM905819, AM920641,
Alloschemone occidentalis (Poepp.) Engl. & K.Krause, Chase 9996 (K), AM932366 + AM933363. Callopsis volkensii Engl., Chase 10668
AM905744, AM920566, AM932310 + AM933306. Alocasia odora (Roxb.) (K), AM905773, AM920595, AM932330 + AM933325. Carlephyton
K.Koch., Chase 10674 (K), AM905802, AM920624, —; A. odora (Roxb.) glaucophyllum Bogner, Mangelsdorff 124 (M), AM905821, AM920643,
K.Koch., Barabé & Chantha 93 (MT), —, —, AY054705. Ambrosina basii AM932368 + AM933364. Cercestis mirabilis (N.E.Br.) Bogner,
L., Chase 12339 (K), AM905798, AM920620, AM932348 + AM933344. Chase 11772 (K), AM905817, AM920639, AM932364 + AM933361.
Amorphophallus hottae Bogner & Hett., Lam s.n. (L), AM905785, Chlorospatha sp., Chase 11912 (K), AM905791, AM920613, AM932341
AM920607, —; A. paeoniifolius (Dennst.) Nicolson, Barabé & Chantha + AM933339. Colletogyne perrieri Buchet, Pronk s.n. (M), AM905823,
98 (MT), —, —, AY054703. Amydrium humile Schott, Chase 9974 (K), AM920645, AM932370 + AM933366. Colocasia esculenta (L.) Schott,
AM905745, AM920567, —; A. zippelianum (Schott) Nicolson, Barabé Chase 10669 (K), AM905800, AM920622, AM932349 + AM933345.
& Chantha 99 (MT), —, —, AY054735. Anadendrum sp., Chase 9985 Cryptocoryne lingua Becc. ex Engl., Chase 10998 (K), AM905779,
(K), AM905740, AM920547, AM932308 + AM933304. Anaphyllopsis AM920601, — + AM933329. Culcasia liberica N.E.Br., Chase 11777
americana (Engl.) A.Hay, Chase 11914 (K), AM905753, AM920575, (K), AM905816, AM920638, AM932363 + AM933360. Cyrtosperma
—; A. americana (Engl.) A.Hay, Barabé 83 (MT), —, —, AY054726. macrotum Engl., Chase 11771 (K), AM905750, AM920572, AM932313
Anchomanes difformis (Blume) Engl., Chase 10687 (K), AM905761, + AM933309.
AM920583, —; A. difformis (Blume) Engl., Barabé 155 (MT), —, —, Dieffenbachia aglaonemifolia Engl., Chase 10678 (K), AM905764, AM920586,
AY054711. Anthurium acaule (Jacq.) Shott, Chase10884 (K), AM905735, —; D. pittieri Engl. & K.Krause, Barabé & Chantha 88 (MT), —, —,
AM920557, —; A. jenmanii Engl., Barabé & Chantha 92 (MT), —, AY054714. Dracontium polyphyllum L., Chase 10688 (K), AM905747,
—, AY054730. Anubias barteri Schott, Chase 10997 (K), AM905756, AM920569, —; D. polyphyllum L., Barabé 50 (MT), —, —, AY054727.
AM920578, —; A. barteri Schott, Barabé & Chantha 90 (MT), —, —, Dracontioides desciscens Engl., Chase 11916 (K), AM905754, AM920576,
AY054710. Aridarum nicholsonii Bogner, Bogner 2835 (M), AM905784, AM932316 + AM933312. Dracunulus vulgaris Schott, Chase 11760 (K),
AM920606, AM932337 + AM933334. Ariopsis peltata J.Grah., Chase AM905812, AM920634, AM932359 + AM933356.
11913 (K), AM905804, AM920626, AM932352 + AM933348. Arisaema
franchetianum Engl., Chase 10478 (K), AM905806, AM920628, Epipremnum falcifolium Engl., Barabé & Turcotte 100 (MT), —, —,
AM932354 + AM933350. Arisarum vulgare O.Targ-Tozz, Chase 10992 AY054732; E. pinnatum (L.) Engl., Chase 9977 (K), AM905746,
AM920568, —. Eminium spiculatum (Blume) Kuntze, Chase 11806 AM920563, —. Remusatia vivipara (Roxb.) Schott, Chase 11775
(K), AM905813, AM920635, AM932360 + AM933357. (K), AM905803, AM920625, AM932351 + AM933347. Rhodospatha
oblongata Poepp. & Endl., Croat 53888 (MO), AM905739, AM920562,
Filarum manserichense Nicolson, Sizemore 1996-001 (L), AM905795,
AM932307 + AM933303.
AM920617, AM932345 + AM933342.
Scaphispatha gracilis Brongn. ex Shott, Gonçalves 172 (MO), AM905793,
Gearum brasiliense N.E.Br., Chase 10693 (K), AM905763, AM920585,
AM920615, AM932343 + AM933340. Schismatoglottis trifasiata Engl.,
AM932321 + AM933317. Gonatopus angustus N.E.Br., Chase 10675
Chase 10692 (K), AM905782, AM920604, AM932335 + AM933332.
(K), AM905777, AM920599, AM932333 + AM933328. Gorgonidium
Scindapsus heredaceus Schott, Chase 9986 (K), AM905742, AM920564,
sp., Cultivated (L), AM905767, AM920589, AM932324 + AM933320.
AM932309 + AM933305. Spathantheum intermediaum Bogner,
Gymnostachys anceps R.Brown, Chase 9473 (K), AM905727,
Chase 11776 (K), AM905769, AM920591, AM932326 + AM933322.
AM920548, AM932297 + AM933293.
Spathicarpa hastifolia Hook, Chase 10995 (K), AM905772, AM920594,
Hapaline benthamiana Schott, Chase 10676 (K), AM905787, AM920609, AM932329 + AM933324. Spathiphyllum wallisii Hort, Chase 210 (NCU),
AM932339 + AM933336. Helicodiceros muscivorus (L. f.) Engl., AJ235807, AM920559, —; S. wallisii Hort, Barabé & Turcotte 105 (MT),
Chase 11759 (K), AM905811, AM920633, AM932358 + AM933355. —, —, AY054738. Spirodela polirhiza (L.) Schleid., Chase 11096 (K),
Heteropsis oblongifolia Kunth, Ramírez 11848 (L), AM905737, AM905731, AM920553, AM932300 + AM933296. Stenospermation
AM920560, —; Heteropsis sp., Barabé, Forest & Gibernau 147 (MT), popayanense Schott, Barabé & Lavoie 159 (MT), —, —, AY054737; S.
—, —, AY054739. Homalomena magna A. Hay, Chase 10691 (K), ulei K.Krause, Chase 9987 (K), AM905738, AM920561, —. Steudnera
AM905774, AM920596, —; Homalomena sp., Barabé 151 (MT), —, colocasiifolia K.Koch, Chase 10682 (K), AM905801, AM920623,
—, AY054724. Holochlamys beccarii (Engl.) Engl., Chase 10677 (K), AM932350 + AM933346. Stylochaeton bogneri Mayo, Chase 10685
AM905736, AM920558, AM932306 + AM933302. (K), AM905776, AM920598, AM932332 + AM933327. Symplocarpus
foetidus (L.) Nuttall, French 219 (CHRB), L10247, —, —; S. foetidus
Jasarum steyermarkii G.S.Buting, Berry 5531 (MO), AM905792, AM920614,
(L.) Nuttall, Chase 11749 (K), —, AM920551, —; S. foetidus (L.) Nuttall,
AM932342 + AM933339.
Barabé 154 (MT), —, —, AY054741. Synandrospadix vermitoxicus
Lagenandra ovata Thwaites, Chase 10991 (K), AM905780, AM920602, (Griseb.) Engl., Chase 11774 (K), AM905771, AM920593, AM932328
— + AM933330. Landoltia punctata (G. Mey) Les & D.J.Crawford, + AM933292. Syngonium auritum (L.) Schott, Chase 10994 (K),
Landolt 7248 (—), AY034223, AY034185 + AY034301, —; L. punctata AM905789, AM920611, AM932340 + AM933337.
(G. Mey) Les & D.J.Crawford, Chase 14451 (K), —, —, AM932301 +
Taccarum weddelianum Brongn. ex Schott, Hennipman 8315 (L), AM905770,
AM933297. Lasia spinosa (L.) Thwaites, Chase 11779 (K), AM905749,
AM920592, AM932327 + AM933323. Thyphonium blumei Nicolson &
AM920571, AM932312 + AM933308. Lasimorpha senegalensis Shott,
Sivad., Chase 10694 (K), AM905808, AM920630, —; T. giganteum Engl.,
Bogner s.n. (M), AM905755, AM920577, AM932317 + AM933313.
Chase 11803 (K), —, —, AM932356 + AM933352. Thyphonodorum
Lemna minor L., Chase 11761 (K), AM905730, AM920552, AM932299
lindleyanum Schott, Chase 11780 (K), AM905814, AM920636,
+ AM933295. Lysichiton americanus Hultén & H.St.John, Chase 11748
AM932361 + AM933358.
(K), AM905728, AM920549, —; L. camtschatcense (L.) Schoott, Barabé
153 (MT), —, —, AY054740. Ulearum sagittatum Engl., Chase 10695 (K), AM905794, AM920616,
Mangonia tweediana Schott, Bogner 2376 (L), AM905766, AM920588, AM932344 + AM933341. Urospatha saggitifolia (Rudge) Schott, Chase
AM932323 + AM933319. Monstera adansonii Schott, Chase 9980 (K), 11773 (K), AM905748, AM920570, AM932311 + AM933307.
AM905743, AM920565, —; M. adansonii Schott, Barabé & Chantha Wolffia columbiana H.Karts, Landolt 7467 (—), AY034255, AY034217
94 (MT), —, —, AY054734. Montrichardia arborescens (L.), Schott, + AY034333, —; W. columbiana H.Karts, Chase 14447 (K), —, —,
Cultivated (SING), AM905818, AM920640, AM932365 + AM933362. AM932303 + AM933299. Wolffiella oblonga Hegelm., Landolt 8984
Nephthytis afzellii Schott, Chase10689 (K), AM905759, AM920581, —; N. (—), AY034242, AY034204 + AY034320, —; W. oblonga Hegelm.,
afzellii Schott, Barabé & Chantha 95 (MT), —, —, AY054702. Chase 14359 (K), —, —, AM932302 + AM933298.
Orontium aquaticum L., Qui 97112 (NCU), AM905729, AM920550, Xanthosoma helleborifolium (Jacq.) Schott, Chase 10683 (K), AM905790,
AM932298 + AM933294. AM920612, —; Xanthosoma sp., Barabé & Turcotte 107 (MT), —, —,
AY054709.
Pedicellarum paiei M.Hotta, Bogner 2196 (M), AM905733, AM920555,
AM932304 + AM933300. Peltandra virginica (L.) Raf., Chase 11770 Zamioculcas zamiifolia (Lodd.) Engl., Chase 10686 (K), AM905778,
(K), AM905815, AM920637, AM932362 + AM933359. Philodendron AM920600, —; Z. zamiifolia (Lodd.) Engl., Barabé & Chantha 84
deltoideum Poepp. & Endl., Chase 10891 (K), AM905775, AM920597, (MT), —, —, AY054725. Zanthedeschia albomaculata (Hook. f.) Bail,
AM932331 + AM933326. Phymatarum borneense M.Hotta, Chase Chase 11758 (K), AM905762, AM920584, AM932320 + AM933316.
10979 (K), AM905783, AM920605, AM932336 + AM933333. Pinellia Zomicarpella amazonica Bogner, Bogner 1985 (M), AM905796,
pedatisecta Schott, Chase 11752 (K), AM905807, AM920629, AM932355 AM920618, AM932346 + —.
+ AM933351. Piptospatha ridleyi N.E.Br., Chase 10680 (K), AM905781, CHLORANTHACEAE. Hedyosmum mexicanum C.Cordem., Salazar s.n.
AM920603, AM932334 + AM933331. Pistia stratiotes L., Chase 10996 (MEXU), AM905824, AM920646, AM932371 + AM933367.
(K), AM905799, AM920621, —; P. stratiotes L., Barabé 153 (MT), —,
—, AY054706. Podolasia stipitata N.E.Br., Chase 11915 (K), AM905752, JUNCAGINACEAE. Triglochin maritima L., Les s.n. (CONN), U80714,
AM920574, AM932315 + AM933311. Pothoidium lobbianum Schott, —, —; T. maritima L, Chase 8279 (K), —, AM920647, AM932373 +
Bogner 1272 (M), AM905734, AM920556, AM932305 + AM933301. AM933369.
Pothos scandens D.Don, Chase 9989 (K), AM905732, AM920554, —; MAGNOLIACEAE. Magnolia macrophylla Michx., BG 790346 (MO), —,
P. scandens D.Don, Barabé & Lavoie 157, —, —, AY054731. Protarum —, AF040680; M. pseudokobus Abe & Akasawa, Tamura 10015 (Bot.
sechellarum Engl., Bogner s.n. (M), AM905805, AM920627, AM932353 Gard. Osaka City Univ.), —, AB040152, —; M. umbrella L., — (—),
+ AM933349. Pseudodracontium lacourii (Linden & Andre) N.E.Br., AF206791, —, —.
Chase 10681 (K), AM905786, AM920608, AM932338 + AM933335.
Pseudohydrosme gabunensis Engl., Wieringa 3308 (L), AM905760, PIPERACEAE. Piper mullesua Buch.-Ham., — s.n. (—),—,—, AY032651;
AM920582, AM932319 + AM933315. Pycnospatha arietina Thorel P. nigrum L., Tamura & Fuse s.n. (Bot. Gard. Osaka City Univ.), —,
ex Gagnep., Sizemore s n. (L), AM905751, AM920573, AM932314 + AB040153, —; P. betle L., Qiu 91048 (NCU), L12660, —, —.
AM933310.
TOFIELDIACEAE. Tofieldia pusilla Pers., Lundqvist 12935 (UPS),
Raphidophora africana N.E.Br., Barabé & Turcotte 110 (MT), —, —, AJ286562, —, —; T. pusilla Pers., Chase 1851 (K), —, AM920648,
AY054736; R. crassifolia Aldewer., Chase 7398 (K), AM905741, AM932374 + AM933370

Cabrera Et Al. 2008 Araceae Phylogenetics

Uploaded by

Copyright:

Available Formats

Cabrera Et Al. 2008 Araceae Phylogenetics

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Cabrera Et Al. 2008 Araceae Phylogenetics

Uploaded by

Copyright:

Available Formats

American Journal of Botany 95(9): 1153–1165. 2008.

PHYLOGENETIC RELATIONSHIPS OF AROIDS AND DUCKWEEDS

Lidia I. Cabrera,2,6 Gerardo A. Salazar,2 Mark W. Chase,3 Simon J. Mayo,3 Josef

rbcL 1391 467 (33.6)/328 (23.6) 360 1465 0.42/0.69

You might also like