ArticlePDF Available

Phylogenetic Relationships and Classification in Marantaceae: Insights from Plastid DNA Sequence Data

May 2006
Taxon 55(2):281-296

Authors:

Ball Horticultural Company

The evolutionary distinctiveness and phylogenetic position of Marantaceae within the Zingiberales is strongly supported, yet relatively little phylogenetic research has been conducted on members of Marantaceae. Phylogenetic analyses of plastid DNA sequence data (matK and 3' intergenic spacer region, and the trnL-F intergenic spacer region) for 80 ingroup taxa representing 27 genera were conducted under parsimony criteria. Results identify five major clades and support significant realignments in tribal and "group" circumscriptions. Four non-monophyletic genera, Calathea, Marantochloa, Phrynium, and Schumannianthus are identified. Although significant work remains to identify relationships of the major lineages within Marantaceae and to redefine several problematic genera, we propose a working classification that better reflects the inferred evolutionary history of the family.

One of 48 shortest maximum parsimony trees for a combined analysis of the trnL-F intergenic spacer region, matK coding, and trnK 3' intergenic spacer region data for members of Marantaceae. Numbers above branches are maximum parsimony bootstrap percentages. Bold lines indicate branches with posterior probabilities of 0.95 or higher. Dashed lines indicate branches not found in the maximum parsimony strict consensus tree. Arrows indicate two (polyphyletic) Schumannianthus samples.

…

Figures - uploaded by Linda Marie (Prince) MacKechnie

Content may be subject to copyright.

Content uploaded by Linda Marie (Prince) MacKechnie

Content may be subject to copyright.

INTRODUCTION

The evolutionary distinctiveness and phylogenetic

position of Marantaceae within Zingiberales is strongly

supported by morphological and molecular data analyses

(Kress, 1990, 1995; Kress & al., 2001). Cannaceae, a

monogeneric family with about 10 species (Kubitzki,

1998; but see Tanaka, 2001), are the closest relative of

Marantaceae. These two families are sister to the family

pair Zingiberaceae and Costaceae. Marantaceae includes

31 genera and ~535 species distributed throughout warm

temperate and tropical parts of the world (Andersson,

1998) with 14 genera in the New World, 11 genera in

Africa (including Madagascar), and eight genera in Asia.

Only the genera Thalia and Halopegia occur on more

than one continent. The current geographic distribution

of the family is likely due to secondary intercontinental

dispersal events (Kress & Specht, in press; Prince &

Kress, in press). Halopegia has two representatives in

Africa (one on the continent plus one on Madagascar)

and 1 in Southeast Asia. Thalia is restricted to the New

World tropics however, T. geniculata L. is naturalized in

both Africa (Andersson, 1981a) and India (Sijimol & al.,

2000). Species distributions are significantly skewed

with the vast majority of the species (~450) restricted to

the New World, most belonging to the genus Calathea

(~300 species).

Andersson (1998) included a detailed discussion of

the vegetative and reproductive anatomy and morpholo-

gy of the family as well as embryology, ecology, phyto-

chemistry, and economic importance. An extensive liter-

ature review is beyond the scope of this paper and only a

few comments will be made here. There is a growing

demand for species of Calathea, Ctenanthe, Maranta,

and Stromanthe as ornamentals. These plants are easy to

grow in tropical and subtropical climates where they are

cultivated as landscape plants and as potted plants in

temperate climates. The family includes a number of edi-

ble species although only a single species, Maranta

arundinacea L., West Indian arrowroot, is economically

important. Fiber plants abound in the family. The fresh

leaves of Calathea allouya (Aubl.) Lindl., Ischnosiphon

spp., and Phrynium pubinerve Blume are used for wrap-

ping foods, to cover cargo, and as bottle stoppers (Duke,

1986). The split petioles of Donax ssp., Ischnosiphon

spp., and Schumannianthus dicotomus (Roxb.) Gagnep.

are used for making baskets, matting, and strings for a

musical instrument (Dodge, 1897; Watt, 1892; van den

Berg, 1984; W. J. Kress, pers. obs.).

Andersson (1998) described Marantaceae as prima-

rily “jungle weeds”, taking advantage of light gaps to

complete their life cycle. He noted two genera, Thalia

and Halopegia, that are “confined to open marshes”

although the latter often occurs in shady habitats as well

in tropical Asia. The growth habits of some African and

Asian Marantaceae are somewhat different from those

found in the Americas. “Marantaceae forests”, which are

large expanses of herbaceous vegetation dominated by

281

Prince & Kress • Phylogeny of Marantaceae55 (2) • May 2006: 281–296

Phylogenetic relationships and classification in Marantaceae: insights from

plastid DNA sequence data

Linda M. Prince1,2 & W. John Kress1

1Department of Botany, MRC-166, United States National Herbarium, National Museum of Natural History,

Smithsonian Institution, P. O. Box 37012, Washington, D. C. 20013, U.S.A. linda.prince@cgu.edu (author for

correspondence)

2Current address: Rancho Santa Ana Botanic Garden, 1500 North College Ave., Claremont, California 91711,

U.S.A.

The evolutionary distinctiveness and phylogenetic position of Marantaceae within the Zingiberales is strongly

supported, yet relatively little phylogenetic research has been conducted on members of Marantaceae.

Phylogenetic analyses of plastid DNA sequence data (matK and 3' intergenic spacer region, and the trnL-F

intergenic spacer region) for 80 ingroup taxa representing 27 genera were conducted under parsimony criteria.

Results identify five major clades and support significant realignments in tribal and “group” circumscriptions.

Four non-monophyletic genera, Calathea, Marantochloa, Phrynium, and Schumannianthus are identified.

Although significant work remains to identify relationships of the major lineages within Marantaceae and to

redefine several problematic genera, we propose a working classification that better reflects the inferred evo-

lutionary history of the family.

KEYWORDS:

classification, matK, Marantaceae, phylogeny, trnK intron, trnL-F intergenic spacer region.

members of Marantaceae, are a recognized community

type in Africa (Koechlin, 1965, 1968). These “forests”

are particularly important habitats providing important

food and nesting sites for several primates, including

lowland gorillas, chimpanzees, white-nosed monkeys,

moustached monkeys, and white-cheeked mangabeys

(Rogers & Williamson, 1987; Rogers & al., 1990;

Bermejo, 1999). They are also important food sources for

elephants (Hardin & al., 2002). Marantaceae play a sig-

nificant role in the ecology of natural and anthropologi-

cally influenced tropical communities around the world.

Relatively little systematic research has been con-

ducted on members of Marantaceae in dramatic contrast

with the largest family of the order Zingiberaceae (e.g.,

Burtt & Smith, 1972; Ngamriabsakul & al., 2000; Rang-

siruji, & al., 2000a, b; Searle & Hedderson, 2000; Kress

& al., 2002). Most of the research has been done by a

small group of dedicated scientists with interests in spe-

cific genera although Pischtschan (http://www.uni-mainz

.de/FB/Biologie/Botanikspeziell/home_e/pages/ag_class

enb_pischt_e.htm) is currently investigating evolution-

ary trends in floral morphology across the entire group.

New World taxa (14 genera/~440 species) have been

examined the most vigorously. Kennedy has described

some 75 new species, subspecies, and varieties of Cala-

thea over the last 25 years in addition to her extensive

studies on floral mechanisms in this large and complicat-

ed genus (e.g., Kennedy, 1975, 1978a, b, 1984, 1986,

1990, 1997). Ischnosiphon, Maranta subgen. Maranta,

and the Thalia geniculata complex were each revised by

Andersson (1977, 1981a, 1984, 1986), while Hagberg

(1986, 1990) investigated the genus Monotagma.

The African genera (Afrocalathea, Ataenidia, Halo-

pegia, Haumania, Hypselodelphys, Marantochloa, Me-

gaphrynium, Sarcophrynium, Thaumatococcus, and Tra-

chyphrynium) were revised in the 1950’s (Milne-Red-

head, 1950, 1952) and are currently under investigation

(see Lejoly, www.ulb.ac.be/rech/inventaire/projets/4/PR

1694.htm; Ley, http://www.uni-mainz.de/FB/Biologie/

Botanikspeziell/home_e/pages/ag_classenb_ley_e.htm).

The relatively small number of named species (~31) in

the African flora has remained stable over the past 50

years. Several of the African genera are reportedly mono-

typic including Afrocalathea, Ataenidia, Thaumatococ-

cus, and Trachyphrynium. It is unclear whether addition-

al species should be recognized. The most species-rich

genus in Africa is Marantochloa with ~15 described

species.

The Asian taxa (8 genera/~46 species) remain the

least understood. The genus Stachyphrynium has re-

ceived some recent attention (Li, 1985), but information

for most taxa in Southeast Asia is extremely limited. A

floristic project on Marantaceae of Sabah currently

underway has resulted in new data on the family

(Clausager & Borchsenius, 2003; www.sabah.edu.my/

sst/ums_danced_2.htm), whereas the species of Thailand

are being revised using molecular and morphological

methodologies (Suksathan & Borchsenius, 2003).

Taxonomic research has been hampered by the

small, delicate nature of the flowers, a relative rarity of

specimens in both herbaria and in the field, and the lack

of local experts to assist in identification. Classification

systems of the late 19

and early 20

century (Petersen,

1889; Schumann, 1902; Loesener, 1930) divided the

family into two tribes, Maranteae with one fertile locule

and Phrynieae with three fertile locules per ovary (see

Table 1). Andersson (1981b) criticized this division as

artificial when he considered several other characters. In

response, he created five informal groupings within

Marantaceae: a “Phrynium group”, a “Calathea group”,

a “Donax group”, a “Maranta group”, a “Myrosma

group”, and five genera of “uncertain affinity” (Table 1).

Andersson (1998) produced a familial treatment also rec-

ognizing five (slightly different) groups based on a suite

of vegetative and reproductive characters. Some of the

characters utilized in the delineation of groups include

inflorescence complexity, bracteole and interphyll num-

ber, corolla tube length, and staminode number and tex-

ture. Andersson’s (1998) construction of informal groups

rather than an explicit classification system emphasizes

the difficulty in interpreting morphological characters

and the paucity of information for the taxa. Represen-

tative floral and inflorescence types of Marantaceae are

shown in Figs. 1–12 including members of each of

Andersson’s groups.

Systematic evaluations of the family have generally

been framed in the context of the entire order Zingibera-

les. Taxon sampling of Marantaceae in many of these stu-

dies is too limited to allow application of the results to

specific questions within the family. This limited sam-

pling is true of the flavonoid study of Williams & Har-

borne (1977), the embryological study of Kamelina

(1990) and the developmental studies of Kirchoff

(1983a, b, 1986, 1988, 1991) and Kirchoff & Kennedy

(1985).

The only investigation of the family based on DNA

sequence data (rps16 intron) is that of Andersson &

Chase (2001). The rps16 intron trees resolved several

major clades, but the trees had poor bootstrap support for

many clades and representatives of several genera were

not included in the study. Our study uses different molec-

ular datasets (one protein-coding, the others non-coding)

with presumed different rates of evolution to generate

phylogenetic hypotheses of relationships for the family.

The trees are used to evaluate Andersson’s 1998 classifi-

cation, redefine generic limits if appropriate, and focus

efforts on new analyses of morphological characters.

Prince & Kress • Phylogeny of Marantaceae 55 (2) • May 2006: 281–296

282

MATERIALS AND METHODS

Taxa examined. —

Taxon sampling was predom-

inantly from the living collections of the Smithsonian

Institution Botany Research Greenhouses, many of

which were wild-collected by the second author

(Appendix 1). Sampling includes multiple representa-

tives of each of Andersson’s groups and representatives

of all of the taxa of “uncertain affinity.” A number of

species representatives were sampled for large genera,

such as Calathea, Ischnosiphon, Phacelophrynium, and

Phrynium. At least one specimen representing 27 of the

31 recognized genera in the family were sampled. In

addition, one representative for the sister family

Cannaceae (Canna paniculata) and the family Zingibe-

raceae (Siphonochilus kirkii) were included for outgroup

purposes. Species identification of a number of sterile

taxa was difficult due to the lack of monographic studies

and of recent floristic treatments.

DNA extraction, amplification, and se-

quencing. —

Total genomic DNA was extracted from

fresh, fast-frozen (-80ºC), or silica gel-dried leaf tissue

(Chase & Hills, 1991) following a minor modification of

the Doyle & Doyle (1987) CTAB method. Clean DNAs

were resuspended in Tris-EDTA (ethylenediaminete-

traacetic acid) buffer. Polymerase chain reactions were

carried out using Gibco BRL (Rockville, Maryland,

U.S.A.) Taq polymerase (following the manufacturer’s

instructions) with annealing temperatures of 50–54ºC.

Oligos for PCR were based on previously published

sequences for matK plus the flanking intergenic spacer

regions (Steele & Vilgalys, 1994; Manos & Steele, 1997)

and trnL-F intergenic spacer region (IGS; Taberlet & al.,

1992) while internal sequencing oligos were developed

specifically for taxa of Zingiberales. Information on all

primers is available in Table 2. PCR products were

cleaned via PEG8000/NaCl precipitation (Johnson &

Soltis, 1995) and sequenced using Applied Biosystems

(Foster City, California, U.S.A.) original (1/2 concentra-

tion) or Big-Dye I (1/4 concentration) chemistry

Terminator Cycle Sequencing Ready Reaction Kit fol-

lowing ABI protocol for a 377 Automated DNA

Sequencer. Only a portion of the matK region (last 1427

aligned base pairs of the coding region + 3' matK-trnK

IGS) was sequenced due to difficulty amplifying the 5'

Prince & Kress • Phylogeny of Marantaceae55 (2) • May 2006: 281–296

283

Table 1. Diversity, distribution, and classification of Marantaceae. Genera sampled are indicated by an *.

Genus Number of species/distribution Loesener (1930) Andersson (1998)

Afrocalathea *1 / Africa Phrynieae Maranta Group

Ataenidia *1 / Africa [Phrynieaea]Phrynium Group

Calathea *~300 / Americas Phrynieae Calathea Group

Cominsia *1 / Asia Phrynieae Uncertain Affinity

Ctenanthe *~10 / Americas Maranteae Myrosma Group

Donax *3–4 / Asia Phrynieae Donax Group

Halopegia *~3 / Africa, Madagascar & Asia Phrynieae Uncertain Affinity

Haumania *2 / Africa [Phrynieaea] Uncertain Affinity

Hylaeanthe *5–6 / Americas [Maranteaea]Myrosma Group

Hypselodelphys *4 / Africa [Phrynieaea]Donax Group

Ischnosiphon *~35 / Americas Maranteae Calathea Group

Koernickanthe *1 / Americas [Maranteaea]Maranta Group

Maranta *~25 / Americas Maranteae Maranta Group

Marantochloa *~15 / Africa & Madagascar Phrynieae Maranta Group

Megaphrynium *4 / Africa [Phrynieaea]Donax Group

Monophrynium 2–3 / Asia Phrynieae Phrynium Group

Monophyllanthe 2b/ Americas Maranteae Maranta Group

Monotagma *37 / Americas Maranteae Calathea Group

Myrosma 1 / Americas Maranteae Myrosma Group

Phacelophrynium *~6 / Asia Phrynieae Phrynium Group

Phrynium *~20 / Asia Phrynieae Phrynium Group

Pleiostachya *2 / Americas Maranteae Calathea Group

Sanblasia 1 / Americas [Phrynieaea]Calathea Group

Saranthe *5–10 / Americas Maranteae Myrosma Group

Sarcophrynium *~3 / Africa Phrynieae Donax Group

Schumannianthus *2 / Asia Phrynieae Donax Group

Stachyphrynium *~10 / Asia Phrynieae Phrynium Group

Stromanthe *10-15 / Americas Maranteae Myrosma Group

Thalia *5–7 / AmericascMaranteae Uncertain affinity

Thaumatococcus *1 / Africa Phrynieae Uncertain affinity

Trachyphrynium *1 / Africa Phrynieae Donax Group

aNot included in Loesener’s publication, classification based on number of fertile locules.

bAndersson (1998) lists 1–2 species, but a new species was recently published by Suárez & al. (2001).

cSee text for an explanation of the geographic distribution of Thalia.

Prince & Kress • Phylogeny of Marantaceae 55 (2) • May 2006: 281–296

Figs. 1–12. Representative floral and inflorescence types of Marantaceae. 1, Calathea crotalifera; 2, Calathea

warscewiczii; 3, Ischnosiphon helenae. 4, Stachyphrynium latifolium; 5, Ataenidia conferta; 6, Marantachloa purpurea;

7, Saranthe sp.; 8, Maranta bicolor, 9, Schumannianthus virgatus; 10, Donax canniformis; 11, Megaphrynium

macrostachyum; 12, Thaumatococcus daniellii.

284

IGS region. DNA fragments were compiled and edited in

Sequencher 3.1.1 (Gene Codes Corp., Ann Arbor,

Michigan, U.S.A.), aligned manually in Se-Al 2.0a11

(Rambaut, 1996), and imported into PAUP*4.0b10

(Swofford, 2002) for analysis. Indels were coded as

unordered multistate characters at the end of the data

matrix (Oxelman & al., 1997).

Phylogenetic analyses. —

Outgroup taxa were

included in the analyses of the trnL-trnF IGS and matK

coding regions (=matK). Results of analyses of the matK

and trnL-trnF IGS were used to identify appropriate taxa

within Marantaceae to serve as the functional outgroup

for all matK-trnK 3' IGS (=trnK IGS) and combined

analyses due to alignment difficulties with the trnK IGS

data partition for taxa outside the family. Data partition

homogeneity tests (Farris & al., 1994; Huelsenbeck &

Bull, 1996; Huelsenbeck & al., 1996) were not per-

formed since datasets are all from the plastid genome

which is not known to undergo recombination in

angiosperms (Doyle, 1992).

Separate and combined Fitch (1971) parsimony

analyses were conducted. Analyses varied in the taxa

included (sometimes excluding taxa with incomplete

sequences), the thoroughness of the search, and the data

partition examined. For each analysis a minimum of five

hundred random sequence addition replicates were con-

ducted with tree-bisection-reconnection (TBR) branch

swapping, holding 2 trees, saving no more than 10 trees

per replicate. Trees from the initial searches were then

subject to additional branch swapping and run to com-

pletion or until 50,000 trees were found. Alignment was

unambiguous for the matK region, but not so for the

trnL-F IGS and trnK IGS regions. Preliminary explorato-

ry analyses under the maximum parsimony (MP) criteri-

on (results not shown) indicated that the default PAUP*

coding of insertion/ deletion events (indels) as missing

data did not affect tree topology. Similarly, exclusion of

ambiguously aligned regions did not alter tree topology.

Based on these exploratory results, indels were coded as

multistate characters at the end of the matrix.

Maximum parsimony bootstrap (BS; Felsenstein,

1985) and Bayesian analyses were conducted to estimate

branch support. Bootstrap percentages were estimated

using 1000 replicates (10 random addition replicates,

hold 2 trees, subtree pruning-regrafting, saving a maxi-

mum of 10 trees per replicate) to maximize the accuracy

of the estimation while minimizing analysis time

(Salamin & al., 2003; Freudenstein & al., 2004).

Bayesian analyses were conducted in MrBayes

(Huelsenbeck & Ronquist, 2001) using 3 replicates of 5

million generations (sampling every 50 generations) for

each of the individual data partitions and one combined

dataset. Appropriate burn-in (number of generations dis-

carded prior to calculation of posterior probability) for

each analysis was determined by treating each million

generations as a data pool from which 40 samples were

drawn. The likelihood values were subjected to a

Bartlett’s test for homogeneity of variance (Bartlett,

1937a, b) to determine whether the data were het-

eroscedastic. This finding was expected due to the inclu-

sion of the first 40 data points in addition to 10 random-

ly drawn subsamples.

A Tukey-like multiple comparison test for differ-

ences among variances (Levy, 1975a, b) was used to

determine where the change from heteroscedasticity to

homoscedasticity occurred. Only the homoscedastic data

from the later portion of each MrBayes run were used to

calculate posterior probabilities. In all cases at least the

first 500,000 generations were discarded as burn-in. The

posterior probabilities (PP) provided on figures are based

on the pooled trees of the three independent MrBayes a-

nalyses from the last millions of generations with P ≥

0.05.

Character reconstruction. —

A summary tree

was constructed based on the results presented here and

the work of Andersson & Chase (2001). The taxon

Myrosma was added to the pruned tree to represent our

best estimate of relationships in the family. This was

accomplished using an indirect supertree approach

(Ponstein, 1966; Ragan, 1992; Bininda-Edmonds & al.,

2002). Summarized source trees from this study and

those of Andersson & Chase (2001) were redrawn in

MacClade 4.0 (Maddison & Maddison, 2000) then

imported into PAUP for matrix translation. All genera

were treated as monophyletic unless there was signifi-

cant (BS ≥50%, PP ≥0.95) support for paraphyly.

Phylogenetic analyses of the supertree matrix used 500

random addition replicates, saving all shortest trees. An

arbitrarily resolved tree was used to evaluate some of the

morphological characters investigated by Andersson &

Prince & Kress • Phylogeny of Marantaceae55 (2) • May 2006: 281–296

285

Table 2. Newly designed primers used in the amplification of matK to address phylogenetic relationships within

Marantaceae.

Primer Name Primer Sequence 5' to 3' Description

8R AGCACAAGAAAGTGCAAG complement of Steele & Vilgalys, 1994 - 8F

8Fa TACTTCGACTTTCCTGTGCC modification of Steele & Vilgalys, 1994 - 8F

5Fa CTCTATGGGTCTTCAAGGAT bp 806–825 of tobacco matK

5R AGGATCCTTGAAAATCCATAGA bp 807–828 of tobacco matK

5Ra TGATACCGAACATAATGCATG bp 831–851 of tobacco matK

mIF GTTCAGTACTTGTGAAACGTT bp 164–184 of tobacco matK

Chase (2001), namely their character numbers 12

(bracteoles), 15 (shape of corolla tube), 16 (staminodes

of outer staminal whorl), 17 (appendages of cucullate

staminode), and 18 (fruit). Character state reconstruc-

tions are shown under ACCTRAN optimization.

RESULTS

DNA sequence summary. —

The trnL-F IGS

region aligned matrix included 522 characters of which

95 characters were potentially parsimony informative.

Unaligned sequences ranged from 355 to 398 bp for

ingroup taxa. A total of 80 ingroup and 2 outgroup taxa

were sampled. Twenty-four indel events were inferred

and coded as multistate characters at the end of the data

matrix. These characters were included in some analyses.

(see Table 3 for summary). Three regions of alignment

ambiguity (characters 14–17, 74–85, and 277–298) were

also identified and excluded from some analyses.

The aligned matK region matrix included 1421

nucleotide characters for 77 ingroup and 2 outgroup taxa.

Inferred indel events were uncommon and generally

autapomorphic for specific families or genera. A total of

11 indel events were inferred and coded as multistate

characters at the end of the data matrix. These characters

were included in some analyses. One region of the matK

data matrix presented questionable homology (bases

134–139) particularly with reference to the Thalia genic-

ulata sequence.

The trnK IGS region was analyzed for ingroup taxa

only due to alignment difficulties with outgroup taxa.

Sarcophrynium was selected as the outgroup taxon based

on the results of rps16 intron analyses of Andersson &

Chase (2001), and results of our other datasets (matK and

trnL-F). Sequence data for outgroup taxa was coded as

missing. A total of 550 aligned nucleotide characters for

72 taxa yielded 84 potentially parsimony informative

characters (see Tables 3 for summary). In addition, eight-

een indels were coded as multistate characters at the end

of the matrix. A total of 129 characters were excluded

from some analyses due to alignment ambiguity. Those

characters are numbered 1994–2098 and 2343–2356 in

the data matrix.

Tree topologies. —

Maximum Parsimony and

Maximum Likelihood. A number of distinct MP analyses

were run to explore the distribution of the phylogenetic

signal in the different matrices. All analyses produced

trees of similar topology although the resolution and

branch support varied. Conflicts in tree topologies were

generally for unsupported (BS < 50%, PP < 0.95) branch-

es. Tree characteristics and indices are summarized in

Tables 3 for all analyses. Only the combined analysis tree

will be discussed in detail, although trees from the sepa-

rate analyses are available as Appendix 2 in the supple-

mental materials (online version of Taxon at http://www.

ingentaconnect.com/content/iapt/tax).

The trnL-F IGS matrix produced >50000 MP trees

(see Appendix 2, Fig. 1) when all characters and taxa

were included. Tree statistics (excluding parsimony

uninformative characters) were high. The analyses iden-

tified a monophyletic Marantaceae with several major

clades. A Calathea clade (united in the 50% majority-

rule consensus tree), a Donax clade, a Maranta clade,

Prince & Kress • Phylogeny of Marantaceae 55 (2) • May 2006: 281–296

286

Table 3. Summary statistics for plastid DNA data analyses of phylogenetic relationships in Marantaceae part 1 (A) and

part 2 (B).

A: Maximum Parsimony Analyses trnL-F IGS trnL-F IGS matK matK matK matK

(no. of taxa ××no. of characters) (82 ××522a) (82 ××266b) (79 ××1421a) (73 ××1421a) (79 ××1336b) (73 ××1336b)

Associated figures (Appendix 2, Fig. 1) (Appendix 2, Fig. 2)

Analysis typec500 ×10 500 ×all 500 ×10 500 x all 500 ×all 500 ×all

Number of shortest trees >50,000 234 >50,000 10803 28514 1296

Tree lengthd182 152 599 571 595 567

Consistency Indexd0.68 0.69 0.63 0.62 0.63 0.63

Retention Indexd0.89 0.91 0.86 0.86 0.86 0.86

Rescaled Consistency Indexd0.60 0.63 0.54 0.53 0.54 0.54

Number of tree islands n.a. 170 n.a. 15 17 6

B: Maximum parsimony analyses trnK 3' IGS trnK 3' IGS Combined Combined Combined Combined

(no. of taxa ××no. of characters) (73 ××550a)e(73 ××219b)e(83 ××2493a) (72 ××2493a) (83 ××1821b) (72 ××1821b)

(Appendix 2, Fig. 3) (Fig. 13)

Analysis typec500 ×10 500 ×all 500 ×10 500 ×all 500 ×10 500 ×all

Number of shortest trees >50,000 3690 >50 000 48 >50 000 648

Tree lengthd151 117 954 894 875 825

Consistency Indexd0.68 0.68 0.63 0.63 0.64 0.64

Retention Indexd0.89 0.90 0.86 0.86 0.87 0.87

Rescaled Consistency Indexd0.61 0.61 0.54 0.54 0.56 0.55

Number of tree islands n.a. 1 n.a. 1 n.a. 3

a Including all characters. bExcluding ambiguously aligned regions and indels; indels coded at end of matrix. c500 × 10 = 500 random addition replicates,

saving 10 shortest trees per replicate; 500 ×all = 500 random addition replicates, saving all shortest trees. dExcluding parsimony uninformative charac-

ters. eRooted using Sarcophrynium.

and a Stachyphrynium clade (united in 50% majority-rule

consensus tree) were indicated. Relationships among

these clades were unresolved and statistical support for

each of these clades was weak (BS values from less than

50% to 62%; PP all less than 0.95). A number of genera

do not form monophyletic groups including Calathea

and Schumannianthus.

Tree statistics for the corresponding matK results

were lower (Table 3A) due to the larger number of char-

acters in the dataset. When incomplete sequences were

removed there was a decrease in the number of shortest

MP trees found. Supported clades of the matK analyses

(see Appendix 2, Fig. 2) were similar to the trnL-F IGS

analysis. Some of the same clades were identified in

these analyses including the Donax clade and the

Maranta clade. The Stachyphrynium clade and the

Sarcophrynium clade were both identified in the matK

analyses. Tree topology was more highly resolved and

internal nodes had higher branch support than the trnL-F

IGS analysis.

In the analyses of the trnK IGS region (ingroup only;

Sarcophrynium the designated outgroup; see Appendix 2,

Fig. 3) resolution is less than for the matK coding region

(84 potentially parsimony informative characters, see

Table 3B). Significant topological differences had weak

BS (<50%) and PP (≤0.95) support. Two differences

from other analyses include the position of Calathea

utilis and Thalia although again there is no strong BS or

PP support for those placements.

Multiple combined MP data analyses were conduct-

ed. The first analysis utilized all ingroup taxa resulting in

a matrix of 469 potentially parsimony-informative char-

acters for 81 ingroup and 2 outgroup taxa (83 × 2493

analysis, Table 3B). The second analysis excluded

ingroup taxa with incomplete sequences, resulting in a

matrix of 442 potentially parsimony-informative charac-

ters (73 × 2493 analysis, Table 3B). The third and fourth

analyses excluded ambiguously aligned regions and

indels but coded gap characters separately (83×1821 and

73 × 1821 respectively). All analyses produced trees of

similar topology. Only the 73 × 2493 analysis will be dis-

cussed in detail (Fig. 13).

The combined analyses produced far fewer shortest

trees (48 versus >50,000) with index values (excluding

parsimony uninformative characters) of CI = 0.63, RI =

0.86, and RC = 0.54 (Table 3B). Five major clades were

identified. The largest lineage, the Calathea clade includ-

ed Calathea (in two parts), Haumania, Ischnosiphon,

Monotagma, and Pleiostachya. Branch support for the

clade was low (less than 50% BS and less than 0.95 PP),

but several internal branches were strongly supported.

Calathea I and II had strong BS (93–100%) and PP

(≥0.95) support. A close relationship between Ischnosi-

phon and Pleistachya was also strongly supported with

100% BS and ≥0.95 PP support, as was the monophyly

of Monotagma (100% BS, ≥0.95 PP).

The second lineage, the Donax clade, included

Donax, Phacelophrynium, Phrynium, Schumannianthus

(in part), and Thalia. There was moderate (68% BS) sup-

port for the clade, but high support for several internal

branches. A large Phrynium (including Phacelophryni-

um) lineage was supported by 100% BS and ≥0.95 PP

values. Schumannianthus dichotomus was also strongly

(99% BS, ≥0.95 PP) supported as sister to Donax rather

than sister to the other species of Schumannianthus.

The third major lineage was the Maranta clade that

included Ctenanthe, Hylaeanthe, Halopegia, Maranta,

Saranthe, Schumannianthus virgatus, and Stromanthe.

The Maranta clade had strong support (100% BS and

≥0.95 PP). Each genus was monophyletic (when more

than one species was sampled) except Schumannianthus

as discussed above, many with strong branch support.

The Stachyphrynium clade included Afrocalathea,

Ataenidia, Marantochloa, and Stachyphrynium. This

clade was strongly supported with a BS of 96% and PP

of ≥0.95, as were many of the internal branches. Almost

no monophyletic genera were present within this clade:

Ataenidia was nested within Marantochloa, and

Afrocalathea was embedded within Stachyphrynium.

The final group was the Sarcophrynium clade. This

clade included Hypselodelphys, Megaphrynium,

Sarcophrynium, Thaumatococcus, and Trachyphrynium.

Support was moderate for this clade with 69% BS and

≥0.95 PP, and many internal relationships were strongly

supported with 100% BS and ≥0.95 PP values. A clear

sister relationship was found for Hypselodelphys +

Trachyphrynium, and for Megaphrynium + Thauma-

tococcus.

DISCUSSION

General observations. —

All BS and PP values

in the following text refer to those of the combined

analysis as shown on Fig. 13. The likely first fork of the

family tree splits the Sarcophrynium clade (BS 81%, PP

≥0.95), including Hypselodelphys, Megaphrynium,

Sarcophrynium, Thaumatococcus, and Trachyphrynium,

from all other taxa in the family. These five genera share

the feature of a single, bilobed cucullate staminode

appendage.

Our limited sampling supports Sarcophrynium as

monophyletic. The results confirm a sister relationship

between Hypselodelphys and Trachyphrynium as has

been suggested by Andersson (1998) who stated, “The

only important difference [of Trachyphrynium] from

Hypselodelphys is in fruit structure, and the distinction is

not very convincing”. There are several other seed dif-

Prince & Kress • Phylogeny of Marantaceae55 (2) • May 2006: 281–296

287

Prince & Kress • Phylogeny of Marantaceae 55 (2) • May 2006: 281–296

Fig. 13. One of 48 shortest maximum parsimony trees for a combined analysis of the trnL-F intergenic spacer region,

matK coding, and trnK 3’ intergenic spacer region data for members of Marantaceae. Numbers above branches are

maximum parsimony bootstrap percentages. Bold lines indicate branches with posterior probabilities of 0.95 or higher.

Dashed lines indicate branches not found in the maximum parsimony strict consensus tree. Arrows indicate two (poly-

phyletic) Schumannianthus samples.

288

Stachyphrynium repens

Stachyphrynium spicatum1

Stachyphrynium spicatum2

Stachyphrynium latifolium

Stachyphrynium sp1

Afrocalathea rhizantha

Stachyphrynium sumatranum

Ataenidia conferta1

Ataenidia conferta2

Marantochloa purpurea

Marantochloa leucantha

Marantochloa sp1

Ctenanthe setosa

Ctenanthe villosa

Stromanthe stromanthoides

Stromanthe thalia

Saranthe sp1

Hylaeanthe hexantha

Maranta leuconeura

Maranta bicolor

Maranta sp1

Halopegia azurea

Halopegia blumei

Schumannianthus virgatus

Phrynium imbricatum1

Phrynium sp1

Phrynium imbricatum2

Phrynium imbricatum3

Phacelophrynium sp1

Phrynium sp3

Donax canniformis1

Schumannianthus dichotomus

Thalia geniculata

Calathea leopardina

Calathea rufibarba

Calathea aemulae

Calathea colorata

Calathea musiaca

Calathea mirabilis

Calathea comosae

Calathea ecuadoriana

Calathea pallidacosta

Calathea loeseneri

Calathea metallica

Calathea foliosa

Calathea warscewiczii

Calathea vinosa

Calathea micans

Calathea variegata

Calathea majestica

Calathea gymnocarpa

Ischnosiphon helenae

Ischnosiphon leucophoeusais

Ischnosiphon cerotus

Ischnosiphon puberulus

Ischnosiphon rotundifolius

Pleiostachya pruinosa

Calathea crotalifera

Calathea pluriplicata

Calathea utilis

Monotagma laxum

Monotagma papillosum

Monotagma parvulum

Monotagma smaragdinum

Haumania sp1

Hypselodelphys sp1

Trachyphrynium braunianum

Megaphrynium macrostachym

Thaumatococcus daniellii

Sarcophrynium brachystachys

Sarcophrynium sp1

Canna paniculata

Siphonochilus kirkii

CALATHEA II

100

DONAX

CLADE

SARCOPHRYNIUM CLADE

MARANTA

CLADE

STACHYPHRYNIUM

CLADE

CALATHEA CLADE

CALATHEA I

ferences that have been used in various taxonomic keys

such as an exarillate seed with an H-shaped perisperm

canal in Hypselodelphys versus an arillate seed with an

oval perisperm canal in Trachyphrynium (Koechlin,

1965, 1968). Additional molecular and morphological

studies are necessary to fully address the generic bound-

aries for these two taxa.

An unexpected sister relationship was identified for

Thaumatococcus (Fig. 12) and Megaphrynium (Fig. 11).

This relationship is supported by a striking vegetative

similarity between the two genera and the structural nec-

tary located at the tips of cataphylls in these two species

that has not yet been recorded for other members of the

family (Kirchoff & Kennedy, 1985).

The next node of the tree separates a Calathea clade

(Calathea, Haumania, Ischnosiphon, Pleiostachya,

Monotagma; <50% BS; <0.95 PP) that is sister to all

remaining taxa. Haumania is a member of the clade, but

the relationship is poorly supported. Ischnosiphon (Fig.

3), Monotagma, and Pleiostachya were historically all

included within Ischnosiphon. These genera were first

separated by Schumann in 1902. Species of Monotagma

were separated based on their one-flowered cymules, a

grouping clearly supported by results of molecular data

analyses. Taxa placed in Pleiostachya were segregated

based on their flattened inflorescences. Our results do not

provide sufficient resolution to determine whether

Pleiostachya is better maintained as a distinct genus or

returned to Ischnosiphon. It is also unclear whether the

small African genus Haumania belongs within the

Calathea clade. Andersson (1998) considered it a genus

of “uncertain affinity”. Our results provide little support

for an affiliation with any particular clade. With respect

to Calathea (Figs. 1 and 2), the paraphyly of the genus is

clear. Andersson & Chase (2001) first suggested that

Calathea might be a paraphyletic group based on a

shared character of distichous bract arrangement in

Ischnosiphon, Pleiostachya, and some Calathea species,

such as C. crotalifera (Fig. 1). Our results do not support

a basic dichotomy in the genus based on this character.

The other major clade of the family tree includes

three groups: the Donax clade, the Maranta clade, and

the Stachyphrynium clade. The Donax clade includes

Cominsia (see Appendix 2, Fig. 1), Donax,

Phacelophrynium, Phrynium, Schumannianthus dichoto-

mus, and Thalia. Thalia is another of Andersson’s (1998)

genera of “uncertain affinity”. Our results placed it sister

to the Donax clade with moderate support (68–100%

BS). Schumannianthus dichotomus was sister to Donax

(Fig. 10) in most analyses with strong support (99% BS,

≥0.95 PP). We were unable to obtain complete matK data

for Cominsia and several Phacelophrynium species

(except pseudogene copies), but the trnL-F IGS results

provide sufficient evidence to conclude monophyly of

Prince & Kress • Phylogeny of Marantaceae55 (2) • May 2006: 281–296

289

Table 4. A new informal working classification of Marantaceae with morphological characteristics of each clade. Bold

type face indicates key identification characteristics.

Morphological characteristics

Clade name Petaloid staminodes Cucullate stami- Fruit

and taxa Bracteoles Corolla tube in outer Whorl node appendage

Calathea: Calathea (in usually present generally >5××1(2 in Haumania, solitary, simple dehiscent capsule

two parts), Haumania,longer than absent in some (indehiscent in

Monotagma, Pleiostachya wide Calathea spp.) Haumania)

Donax: Cominsia, Donax, present and generally 1–5×2 (1 in Thalia and solitary, simple dehiscent capsule

Phrynium, Phacelophry- glandlike, or long as wide Phacelophryni- (2 in Thalia) (indehiscent in

nium, Schumannianthus absent um)Thalia)

dichotomus, Thalia

Maranta: Ctenanthe, Ha- usually absent length variable usually 2 solitary, simple dehiscent capsule

lopegia, Hylaenanthe, Koer- (indehiscent in

nickanthe, Maranta, Saran- Halopegia)

the, Schumannianthus virga-

tus, Stromanthe, Myrosma

Stachyphrynium: Afro- absent length variable 2 (except in Ataeni- usually solitary and dehiscent capsule

calathea, Ataenidia, Maran- dia) simple (2 in Afroca-

tochloa, Stachyphrynium lathea and Ataenidia)

Sarcophrynium: Hypse- present and usually 1–5×(usually) 2 (absent in 2 divergent, unequal caryopsis-like

lodelphys, Megaphrynium, gland-like longer than Thaumatococcus)

Sarcophrynium, Thaumato- wide

coccus, Trachyphrynium

Taxa of unknown affinity

Monophrynium absent 3.5 mm long 2 solitary, simple unknown

Monophyllanthe small and scale- 1–5×longer 1, rudimentary solitary, simple dehiscent capsule

like than wide

Sanblasia sheath-like ~30××longer 1 small and simple unknown

than wide

Cominsia and to place these two genera within the

Donax clade. Phacelophrynium is unresolved in all

analyses that include more than one sample (trnL-F as

shown in Appendix 2, Supplemental Materials, and other

analyses not shown). Phrynium representatives (exclud-

ing Phrynium sp. #3) form a monophyletic group with

strong support (99% BS, ≥ 0.95 PP). Phacelophrynium

and Phrynium are currently being reviewed (Clausager &

Borchsenius, 2003).

The Maranta clade includes Ctenanthe, Halopegia,

Hylaeanthe, Maranta, Saranthe (Fig. 7), Schuman-

nianthus virgatus, and Stromanthe. All genera (except

Schumannianthus) appeared monophyletic based on our

limited sampling. Ctenanthe and Stromanthe are sister

groups. This relationship was also found by Andersson &

Chase (2001), but they could not show support for

monophyly of the two genera. Our results provide sup-

port for each genus (69–99% BS, ≥0.95 PP for

Ctenanthe) and for their sister relationship (78% BS,

≥0.95 PP). According to Kirchoff & Kennedy (1985)

these two genera share a tendency toward foliar non-

structural nectaries. Our results confirm (75% BS) the

close relationship between Hylaeanthe and Maranta

(Fig. 8) first proposed by Andersson & Chase (2001).

Although the congeneric status of the two species placed

in Schumannianthus has been suspect (H. Kennedy, pers.

comm.), the sister relationship of Halopegia and Schu-

mannianthus virgatus (Fig. 9) was somewhat surprising.

Schumannianthus dichotomus is native to Southeast Asia

(Vietnam, Thailand, Malaysia, etc.) and S. virgatus to Sri

Lanka and Malabar. Because S. dichotomus is the type of

the genus the generic status of S. virgatus will need to be

reconsidered.

The final clade is the Stachyphrynium clade, which

includes Afrocalathea, Ataenidia, Marantochloa, and

Stachyphrynium. Members of the group share a number

of morphological features, including presence of inter-

phylls (in only some species of Marantochloa), absence

of bracteoles, 2 outer staminodes, an ovary of three

uniovulate locules, dehiscent fruits, and arillate seeds

(unknown in Afrocalathea). Our results place the distinc-

tive monotypic genus Ataenidia within the

Stachyphrynium clade. The results also suggest that

Ataenidia (Fig. 5) might be better interpreted as a highly

modified member of Marantochloa (Fig. 6) with a great-

ly congested inflorescence. In addition to the characters

listed above, Ataenidia and Marantochloa also share a

caulescent habit, and the presence of interphylls.

Afrocalathea and Stachyphrynium (Fig. 4) form the other

group in the Stachyphrynium clade. This result was unex-

pected given the geographic distribution of the two gen-

era, however, they do share a rosulate habit. We hope to

confirmation of these findings with the addition of a sec-

ond accession of Afrocalathea. Stachyphrynium is cur-

rently under review (Clausager & Borchsenius, 2003;

Suksathan & Borchsenius, 2004).

Previous classifications. —

As suggested by

Andersson (1981b) the number of fertile locules is a poor

indicator of relationships within Marantaceae. When the

number of fertile locules is compared to the molecular

tree (Fig. 13), the reduction of fertile locules from three

to one appears to be independently derived as many as

three times, once each in the Calathea clade, Maranta

clade, and Thalia. The only apparent reversal is in the

small Calathea II lineage. Although this character does

not divide the family into two distinct lineages, it is a

useful character for grouping some genera. Within the

Maranta clade, those taxa with one fertile locule were

strongly supported (100% BS, ≥0.95 PP). Support for

uniting those genera with one fertile locule in the

Calathea clade was moderate (75% BS, ≥0.95 PP).

The five clades identified in this study share some

similarities with the groups proposed by Andersson

(Table 1). Our Calathea clade (excluding Haumania)

was the only one of Andersson’s groups demonstrated to

be monophyletic in our analyses. This result agrees with

the earlier rps16 intron study by Andersson & Chase

(2001). However, the group had only weak to moderate

support (60% BS; ≥0.95 PP when Haumania is exclud-

ed). The Calathea clade can be characterized by the fol-

lowing suite of characters (Andersson, 1998): brachy-

blastic cymules, interphylls usually or occasionally pres-

ent, bracteoles usually present; a long or very long corol-

la tube, and a single outer staminode. Our analyses found

weak support (<50% BS) for the inclusion of Haumania

in this clade. Andersson considered it a genus of uncer-

tain affinity. Haumania differs in all the major characters

with its dolichoblastic cymules, lack of interphylls and

bracteoles, short corolla tube, and two outer staminodes

rather than one. Morphological evidence does not sup-

port the inclusion of Haumania in the Calathea clade and

our results do not refute this proposal.

Our Maranta clade is equivalent to Andersson’s

Myrosma group plus Maranta, Koernickanthe,

Halopegia, and Schumannianthus virgatus. The clade

had high (100% BS, ≥0.95 PP) support. Andersson char-

acterized his Myrosma group by the subbrachyblastic to

dolichoblastic cymules, interphylls lacking, bracteoles

present or lacking, a short to very short corolla tube, two

outer staminodes, and a distally branched perisperm

canal in the seed. Maranta shares all of the above char-

acters except that they possess a longer corolla tube (1/2

to twice the length of the lobes).

Genera included in Andersson’s three other groups

(Donax, Maranta, and Phrynium groups) are scattered

around the tree in four different clades. Members of the

Donax group were split into three clades here, the Donax

clade (Donax, Schumannianthus dichotomus), the

Prince & Kress • Phylogeny of Marantaceae 55 (2) • May 2006: 281–296

290

Maranta clade (Schumannianthus virgatus), and the

Sarcophrynium clade (Hypselodelphys, Megaphrynium,

Sarcophrynium, Trachyphrynium). Members of the

Maranta group were split between the Maranta clade

(Koernickanthe, Maranta) and the Stachyphrynium clade

(Afrocalathea, Marantochloa), and members of the

Phrynium group were split between the Donax clade

(Phacelophrynium, Phrynium) and the Stachyphrynium

clade (Ataenidia, Stachyphrynium).

We were able to resolve with varying degrees of sup-

port the placement of the five genera of uncertain affini-

ty in Andersson’s classification. Haumania is sister to all

remaining members of our Calathea clade, but with

weak (<50–60% BS) support. Similarly, Thalia is sister

to all remaining members of the Donax clade with

68–100% BS support. Cominsia is also placed within the

Donax clade, but with moderate support (77% BS; see

Appendix 2). Halopegia is strongly supported as sister to

Schumannianthus virgatus and as a member of the

Maranta clade (100% BS, ≥0.95 PP). Thaumatococcus is

strongly supported as sister to Megaphrynium (100% BS,

≥0.95 PP) and as part of the Sarcophrynium clade (69%

BS, ≥0.95 PP).

Andersson (1998) provided the first modern evalua-

tion of the family, but his taxonomic groups do not

entirely agree with our current estimates of relationships

in Marantaceae based on these molecular data. We

acknowledge that one of the groups identified in our

study (the Calathea clade) had weak support and that

some future rearrangements may be necessary. Four of

our clades (the Donax, Maranta, Stachyphrynium, and

Sarcophrynium clades) were moderately to strongly sup-

ported. Based on the information currently available,

including prior work by Andersson & Chase (2001), we

propose a new informal working classification that

reflects current knowledge of this difficult family (Table

4).

Generic limits. —

Our study identified at least

four potentially non-monophyletic genera: Calathea,

Marantochloa, Phrynium, and Schumannianthus.

Calathea is the largest genus in the family with ~300

species. Andersson & Chase (2001) suspected that

Calathea might be paraphyletic and suggested that the

genus might be divided based on inflorescence compres-

sion (see Figs. 1–2). Our results confirm at least two lin-

eages of Calathea, but do not support a division of the

genus based on an inflorescence compression character.

Until the type for the genus (C. discolor G. Mey.) is sam-

pled, it is difficult to say which clade, Calathea I or Ca-

lathea II, should be retained as Calathea sensu stricto.

One sample of Marantochloa was sister to Ataenidia

(Fig. 5), rather than grouped with the other Marantochlo-

a(see Fig. 6) sequences, and the support for this associ-

ation was strong (100% BS, ≥0.95 PP). However, the

current circumscription of Marantochloa is wide with

much intrageneric morphological variation. Additional

sampling is needed to evaluate the paraphyly of the lar-

gest African genus and current generic circumscription.

Phrynium sequences form a monophyletic lineage

with the exception of the inclusion of a single Phacelo-

phrynium sample. Difficulties encountered in obtaining

clean matK and matK 3' IGS sequence data prevent any

firm conclusions at this time but suggest Phrynium and

Phacelophrynium may be paraphyletic.

As discussed above, Schumannianthus dichotomus

and S. virgatus are not closely related with S. dichotomus

sister to Donax in the Donax clade and S. virgatus sister

to Halopegia in the Maranta clade. Nomenclatural

rearrangements are needed.

Morphological observations. —

Interpretation

of homologies in morphological characters has proved

difficult in Marantaceae. Andersson (1981b, 1998) dra-

matically improved our understanding with his thorough

evaluation of available information on the family. He

subjected a data matrix of 22 morphological characters

(variable in ingroup taxa) to cladistic analyses (Anders-

son & Chase, 2001). Most of the 22 characters are homo-

plasious when mapped onto our estimated phylogeny

(Supertree; Fig. 14A). A small number of those charac-

ters were mapped onto our estimated phylogeny (Fig. 14:

B, bracteoles; C, shape of corolla tube; D, staminodes of

outer staminal whorl; E, appendages of cucullate stamin-

ode; F, fruit). We did not attempt to recode or reanalyze

the published characters except fruit type for Cannaceae

which was mis-coded in the original matrix.

The bracteole character coding (Fig. 14B) is com-

plex, including both presence/absence information and

texture/shape information. Ambiguity with reconstruc-

tion at the base of the tree and in the Calathea clade

makes interpretation difficult. The most common charac-

ter state is an absence of bracteoles (black branches, Fig.

14B). This is unlikely to be the plesiomorphic character

state, thus loss of bracteoles must have occurred numer-

ous times.

Corolla tube shape (Fig. 14C) is another potentially

continuously distributed character. Andersson & Chase

(2001) recognized 4 distinct states within the family.

Moderately long corolla tube (length 1–5×the width)

appears to be the pleisomorphic character state in the

family. Extension in tube length arose at least 4 times, in

the Calathea clade (excluding Haumania), in Cominsia,

Koernickanthe, and Afrocalathea. Reductions also

occurred multiple times, in Thalia, Ctenanthe +

Stromanthe, Halopegia + Schumannianthus virgatus, and

in Stachyphrynium.

Outer whorl staminodes (Fig. 14D) are generally

petaloid in showy. Most members of the Calathea clade

have reduced the number of staminodes from 2 to 1, as

Prince & Kress • Phylogeny of Marantaceae55 (2) • May 2006: 281–296

291

Prince & Kress • Phylogeny of Marantaceae 55 (2) • May 2006: 281–296

Fig. 14. Hypothesized relationships in Marantaceae based on all available molecular data with character state distribu-

tions indicated. ? indicate missing data. A, supertree; B, bracteoles; C, shape of corolla tube; D, staminodes of outer

staminal whorl; E, appendages of the cucullate staminode; F, fruit type.

292

have Phacelophrynium and Thalia. A total loss of outer

staminodes has occurred independently at least 3 times,

in Stromanthe, Ataenidia, and Thaumatococcus.

Staminode appendages (Fig. 14E) were the least ple-

siomorphic character examined. Appendages of the

cucullate staminode are generally solitary and simple,

but it is unclear if this is the ancestral condition. Absence

of staminode appendages appears to be a derived state in

the family, occurring in Ataenidia. Another potentially

derived state includes appendage proliferation as found

in Thalia, Afrocalathea, and the Sarcophrynium clade.

The final character examined is fruit type (Fig. 14F).

The predominant fruit type is a dehiscent capsule, a fea-

ture shared with Cannaceae. Derived fruit types include

an indehiscent fruit as found in the Sarcophrynium clade

(except Megaphrynium, Andersson 1998) and Donax and

Haumannia, and a caryopsis-like fruit in Thalia and

Halopegia. An evaluation of fresh material would likely

yield a more refined set of character states.

Cominisia, Halopegia, Haumania, Thalia, and

Thaumatococcus, Andersson’s taxa of uncertain affinity,

can be evaluated based on their estimated phylogenetic

placement and morphological characters. Thalia and

Haumania each have a different character state than all

other members of their clade for 3 of the 5 characters

examined. This character conflict and weak branch sup-

port emphasize the need for additional data. Conversely,

most of the morphological characters examined support

the position of Cominsia, Halopegia, and Thaumatococ-

cus in the molecular tree.

Much work remains to be done on the morphology

of Marantaceae. The application of additional molecular

data has strengthened several hypotheses set forth by

Andersson & Chase (2001). Our study has identified five

primary clades that may eventually warrant formal

recognition. In addition a number of “problem” taxa, i.e.,

Calathea, Marantochloa, Phrynium, and Schumanni-

anthus, in need of additional study have been identified.

We hope this study provides impetus for further research

in Marantaceae, especially on character evolution and

biogeographic patterns, using both molecular and mor-

phological tools.

ACKNOWLEDGEMENTS

The authors dedicate this paper to the memory of Lennart

Andersson and his significant contributions to systematic botany,

especially in Marantaceae. We also thank Ray Baker, Finn

Borchsenius, Mike Bordelon, Alan Carle, Mark Collins, Helen

Kennedy, Qing-Jun Li, Ida Lopez, John Mood, David Orr,

Piyakaset Suksathan, and Yong-Mei Xia for discussion, assistance,

and tissue samples that made this investigation possible. Thanks to

Charlie Butterworth for helpful comments on earlier drafts of the

manuscript. Final thanks go to Mark Chase and Finn Borchsenius

for their critical and helpful comments on the later versions of the

manuscript. Their comments greatly improved both the content

and structure of the paper. This work was funded by the

Smithsonian Scholarly Studies Program, the Biotic Surveys and

Inventories Program of the National Museum of Natural History,

and the National Geographic Society.

LITERATURE CITED

Andersson, L. 1977. The genus Ischnosiphon (Marantaceae).

Opera Bot. 43: 1–113.

Andersson, L. 1981a. Revision of the Thalia geniculata com-

plex (Marantaceae). Nordic J. Bot. 1: 48–56.

Andersson, L. 1981b. The neotropical genera of Marantaceae:

circumscription and relationships. Nordic J. Bot. 1:

218–245.

Andersson, L. 1984. Notes on Ischnosiphon (Marantaceae).

Nordic J. Bot. 4: 25–32.

Andersson, L. 1986. Revision of Maranta Subgen. Maranta

(Marantaceae). Nordic J. Bot. 6: 729–756.

Andersson, L. 1998. Marantaceae. Pp. 278–293 in: Kubitzki,

K. (ed.), Families and Genera of Vascular Plants: 4.

Flowering Plants: Monocotyledons Alismatanae and

Commelinanae (Except Gramineae). Springer-Verlag,

Berlin.

Andersson, L. & Chase, M. W. 2001. Phylogeny and classifi-

cation of Marantaceae. Bot. J. Linn. Soc. 135: 275–287.

Bartlett, M. S. 1937a. Some examples of statistical methods of

research in agriculture and applied biology. J. Royal

Statistical Soc. Suppl. 4: 137–170.

Bartlett, M. S. 1937b. Properties of sufficiency and statistical

tests. Proc. Royal Statististical Soc. Ser. A. 160: 268–282.

Bermejo, M. 1999. Status and conservation of primates in

Odzala National Park, Republic of the Congo. Oryx 33:

323–331.

Bininda-Edmonds, O. R. P., Gittleman, J. L. & Steel, M. A.

2002. The (super)tree of life: procedures, problems, and

prospects. Annual Rev. Ecol. Syst. 33: 265–289.

Burtt, B. L. & Smith, R. M. 1972.Tentative keys to the sub-

families, tribes and genera of the Zingiberales. Notes Roy.

Bot. Gard. Edinburgh 31: 171–76.

Chase, M. W. & Hills, H. H. 1991. Silica gel: an ideal materi-

al for field preservation of leaf samples for DNA studies.

Taxon 40: 215–220.

Clausager, K. & Borchsenius, F. 2003. The Marantaceae of

Sabah, northern Borneo. Kew Bull. 58: 647–678.

Dodge, C. R. 1897. A Descriptive Catalogue of Useful Fiber

Plants of the World. United States Department of

Agriculture, Fiber Investigations Report No. 9.

Government Printing Office, Washington.

Doyle, J. J. 1992. Gene trees and species trees: molecular sys-

tematics as one-character taxonomy. Syst. Bot. 17:

144–163.

Doyle, J. J. & Doyle, J. L. 1987. A rapid DNA isolation pro-

cedure for small quantities of fresh leaf tissue. Phytochem.

Bull. Bot. Soc. Amer. 19: 11–15.

Duke, J. A. 1986. Isthmian Ethnobotanical Dictionary, ed. 3.

Scientific Publications, Jodhpur.

Prince & Kress • Phylogeny of Marantaceae55 (2) • May 2006: 281–296

293

Farris, J. S., Källersjö, M., Kluge, A. G. & Bult, C. 1994.

Testing significance of congruence. Cladistics 10:

315–319.

Felsenstein, J. 1985. Confidence intervals on phylogenies: an

approach using the bootstrap. Evolution 39: 783–791.

Fitch, W. M. 1971. Toward defining the course of evolution:

minimum change for a specific tree topology. Syst. Zool.

20: 406–416.

Freudenstein, J. V., van den Berg, C., Goldman, D. H.,

Kores, P. J., Molvray, M. & Chase, M. W. 2004. An

expanded plastid DNA phylogeny of Orchidaceae and

analysis of jackknife branch support strategy. Amer. J. Bot.

91: 149–157.

Hagberg, M. 1986. A revision of Monotagma (Marantaceae).

Rep. Bot. Inst., Aarhus Univ. 15: 42–44.

Hagberg, M. 1990. Two new species of Monotagma from the

Venezuelan Guayana. Ann. Missouri Bot. Gard. 77:

421–424.

Hardin, R., Rupp, S. & Eves, H. E. 2002. Sangha River

Region. Yale School of Forestry & Environmental Studies

Bulletin 102: 8–28.

Huelsenbeck, J. P. & Bull, J. J. 1996. A likelihood ratio test

to detect conflicting phylogenetic signal. Syst. Biol. 45:

92–98.

Huelsenbeck, J. P & Ronquist, F. 2001. MrBayes: Bayesian

inference of phylogeny. Bioinformatics 17: 754–755.

Huelsenbeck, J. P., Bull, J. J. & Cunningham, W. C. 1996.

Combining data in phylogenetic analysis. Trends Ecol.

Evol. 11: 152–158.

Johnson, L. A. & Soltis, D. E. 1995. Phylogenetic inference in

Saxifragaceae sensu stricto and Gilia (Polemoniaceae)

using matK sequence. Ann. Missouri Bot. Gard. 82:

149–175.

Kamelina, O. P. 1990. The development of male and female

embryonic structures in the Marantaceae family. Bot.

Zhurn. (Leningrad) 75: 480–493.

Kennedy, H. 1975. Notes on Central American Marantaceae II.

New species and records from Panama and Costa Rica.

Bot. Not. 128: 312–322.

Kennedy, H. 1978a. Systematics and Pollination of the

“Closed-Flowered” Species of Calathea (Marantaceae).

Univ. Calif. Publ. Bot. 71, Berkeley.

Kennedy, H. 1978b. Notes on Central American Marantaceae:

3. New species of Calathea from Costa Rica and Panama.

Brenesia 14/15: 349–356.

Kennedy, H. 1984. Calathea ecuadoriana and C. contrafenes-

tra: new Ecudarian species in Calathea Series Comosae

(Marantaceae). Canad. J. Bot. 62: 15–19.

Kennedy, H. 1986. Additional Ecuadorian species in Calathea

series Comosae (Marantaceae). Nordic J. Bot. 6: 143–150.

Kennedy, H. 1990. Taxonomic notes on Calathea (Maranta-

ceae) from the Venezuelan Guayana: a new species and a

new combination. Phytologia 69: 373–377.

Kennedy, H. 1997. New species of Calathea (Marantaceae)

endemic to Costa Rica. Canad. J. Bot. 75: 1356–1362.

Kirchoff, B. K. 1983a. Allometric growth of the flowers in five

genera of the Marantaceae and in Canna (Cannaceae). Bot.

Gaz. 144: 110–118.

Kirchoff, B. K. 1983b. Floral organogenesis in five genera of

Marantaceae and in Canna (Cannaceae). Amer. J. Bot. 70:

508–523.

Kirchoff, B. K. 1986. Inflorescence structure and development

in the Zingiberales: Thalia geniculata (Marantaceae).

Canad. J. Bot. 64: 859–864.

Kirchoff, B. K. 1988. Floral ontogeny and evolution in the gin-

ger group of the Zingiberales. Pp. 45–56 in: Leins, P.,

Tucker, S. C., Endress, P. K. & Erbar, C. (eds.), Aspects of

Floral Development: Proceedings of the Double

Symposium “Floral Development: Evolutionary Aspects

and Special Topics”. J. Cramer, Berlin.

Kirchoff, B. K. 1991. Homeosis in the flowers of the Zingibe-

rales. Amer. J. Bot. 78: 833–837.

Kirchoff, B. K. & Kennedy, H. 1985. Foliar, nonstructural

nectaries in the Marantaceae. Canad. J. Bot. 63:

1785–1788.

Koechlin, J. 1965. Scitaminales: Marantaceae. Pp. 99–160 in:

A. Aubréville (ed.), Flore du Cameroun. Muséum

National d’Histoire Naturelle, Paris.

Koechlin, J. 1968. Scitaminales: Marantaceae. Pp. 91–170 in:

Aubréville, A. (ed.), Flore du Cameroun. Muséum

National d’Histoire Naturelle, Paris.

Kress, W. J. 1990. The phylogeny and classification of the

Zingiberales. Ann. Missouri Bot. Gard. 77: 698–721.

Kress, W. J. 1995. Phylogeny of the Zingiberanae: morpholo-

gy and Molecules. Pp. 443–460 in: Ruddall, P. J., Cribb, P.

J., Cutler, D. F. & Humphries, C. J. (eds.), Monocotyle-

dons: Systematics and Evolution, vol. 2. Royal Botanic

Gardens, Kew.

Kress, W. J., Prince, L. M., Hahn, W. J. & Zimmer, E. A.

2001. Unraveling the evolutionary radiation of the fami-

lies of the Zingiberales using morphological and molecu-

lar evidence. Syst. Biol. 50: 926–944.

Kress, W. J., Prince, L. M. & Williams, K. J. 2002. The phy-

logeny and classification of the gingers (Zingiberaceae):

evidence from molecular data. Amer. J. Bot. 89:

1682–1696.

Kress, W. J. & Specht, C. D. 2006. Using local clocks and dis-

persal-vicariance analysis to investigate the origin and

diversification of the major lineages of the tropical mono-

cot Order Zingiberales. Pp.621–632. in: Columbus, J. T.,

Friar, E. A., W., Porter, J. M., Prince, L. M. & Simpson,

M. G. (eds.). Monocots: Comparative Biology and

Evolution. Rancho Santa Ana Botanic Garden, Claremont.

Kubitzki, K. 1998. Cannaceae. Pp. 103–106 in: Kubitzki, K.

(ed.), Families and Genera of Vascular Plants: 4.

Flowering Plants: Monocotyledons Alismatanae and

Commelinanae (except Gramineae). Springer-Verlag,

Berlin.

Levy, K. J. 1975a. Some multiple range tests for variances.

Educational and Psychological Measurement 35:

599–604.

Levy, K. J. 1975b. An empirical comparison of several multi-

ple range tests for variances. Journal of the American

Statistical Association 70: 180–183.

Li, H. 1985. Notes on the genus Stachyphrynium K. Schumann

(Marantaceae). Acta Phytotax. Sin. 23: 144–148.

Loesener, T. 1930. Marantaceae. Pp. 654–693 in: Engler, A. &

Prantl, K. (eds.), Die natürlichen Pflanzenfamilien, ed. 2,

15a, Engelmann, Leipzig.

Maddison, D. R. & Maddison, W. P. 2000. MacClade 4.0:

Analysis of Phylogeny and Character Evolution. Sinauer

Associates, Inc., Sunderland, Massachusetts.

Milne-Redhead, E. 1950. Notes on African Marantaceae: I.

Kew Bull. 2: 157–163.

Prince & Kress • Phylogeny of Marantaceae 55 (2) • May 2006: 281–296

294

Milne-Redhead, E. 1952. Notes on African Marantaceae: II.

Phrynium Willd. emend. K. Schum. in Africa. Kew Bull.

1952: 167–170.

Ngamriabsakul, C., Newman, M. F. & Cronk, Q. C. B.

2000. Phylogeny and disjunction in Roscoea

(Zingiberaceae). Edinburgh J. Bot. 57: 39–61.

Oxelman, B., Lidén, M. & Berglund, D. 1997. Chloroplast

rps16 intron phylogeny of the tribe Sileneae

(Caryophyllaceae). Pl. Syst. Evol. 206: 393–410.

Petersen, O. G. 1889. Marantaceae. Pp. 31–43 in: Engler, A. &

Prantl, K. (eds.), Die natürlichen Pflanzenfamilien, vol.

2(6). Engelmann, Leipzig.

Ponstein, J. 1966. Matrices in Graph and Network Theory.

Van Gorcum, Assen, The Netherlands.

Prince, L. M. & Kress, W. J. 2006. Biogeography of the

prayer plant family: getting to the root problem in

Marantaceae. Pp. 643–657 in: Columbus, J. T., Friar, E.

A., Porter, J. M., Prince, L. M. & Simpson, M. G. (eds.),

Monocots: Comparative Biology and Evolution. Rancho

Santa Ana Botanic Garden, Claremont.

Ragan, M. A. 1992. Phylogenetic inference based on matrix

representation of trees. Molec. Phylog. Evol. 1: 53–58.

Rambaut, A. 1996. Se-Al (v2.0a11) Sequence Alignment Edi-

tor. Available at http://evolve.zoo.ox.ac.uk/. Univ. Oxford,

Oxford.

Rangsiruji, A., Newman, M. F. & Cronk, Q. C. B. 2000a.

Origin and relationships of Alpinia galanga (Zingibera-

ceae) based on molecular data. Edinburgh J. Bot. 57:

9–37.

Rangsiruji, A., Newman, M. F. & Cronk, Q. C. B. 2000b. A

study of the infrageneric classification of Alpina (Zingibe-

raceae) based on the ITS region of nuclear rDNA and the

trnL-F spacer of chloroplast DNA. Pp. 695–709 in: Wil-

son, K. L. & Morrison, D. A. (eds.), Monocots - Syste-

matics and Evolution. CSIRO Publishing, Collingwood.

Rogers, M. E. & Williamson, E. A. 1987. Density of herba-

ceous plants eaten by gorillas in Gabon: some preliminary

data. Biotropica 19: 278–281.

Rogers, M. E., Maisels, F., Williamson, E. A., Fernandez,

M. & Tutin, C. E. G. 1990. Gorilla diet in the Pope

Reserve, Gabon: a nutritional analysis. Oecologia 84:

326–339.

Salamin, N., Chase, M. W., Hodkinson, T. R. & Savolainen,

V. 2003. Assessing internal support with large phylogenet-

ic DNA matrices. Molec. Phylogen. Evol. 27: 528–539.

Schumann, K. M. 1902. Marantaceae. Pp. 1–184 in: Engler,

A. (ed.), Das Pflanzenreich IV, 48. W. Engelmann,

Leipzig.

Searle, R. J. & Hedderson, T. A. J. 2000. A preliminary phy-

logeny of the Hedychieae tribe (Zingiberaceae) based on

ITS sequences of the nuclear rRNA cistron. Pp. 710–718

in: Wilson, K. L. & Morrison, D. A. (eds.), Monocots:

Systematics and Evolution. CSIRO Publishing, Colling-

wood.

Sijimol, P. S., Rajesh, K. P. & Madhusoodanan, P. V. 2000.

On the occurrence of Thalia geniculata L. (Marantaceae)

in India. J. Econ. Taxon. Bot. 24: 727–730.

Steele, K. P. & Vilgalys, R. 1994. Phylogenetic analyses of

Polemoniaceae using nucleotide sequences of the plastid

gene matK. Syst. Bot. 19: 126–142.

Suárez, S., Galeano, G. & Kennedy, H. 2001. Una neuva

especie del género Monophyllanthe (Marantaceae) de la

Cuenca Amazónica. Novon 11: 356–359.

Suksathan, P. & Borchsenius, F. 2003. Two new species of

Stachyphrynium (Marantaceae) from SE Asia. Willdeno-

wia 33: 403–408.

Swofford, D. L. 2002. PAUP*. Phylogenetic Analysis Using

Parsimony (*and Other Methods). Version 4. Sinauer

Associates, Sunderland, Massachusetts.

Taberlet, P., Gielly, L., Pautou, G. & Bouvet, J. 1991.

Universal primers for amplification of three non-coding

regions of chloroplast DNA. Pl. Molec. Biol. 17:

1105–1109.

Tanaka, N. 2001. Taxonomic revision of the family Cannaceae

in the New World and Asia. MAKINOA 1: 1–74.

van den Berg, M. E. 1984. Ver-o-Peso: the ethnobotany of an

Amazonian market. Advances Econ. Bot. 1: 140–149.

Watt, G. 1892. A Dictionary of the Economic Products of

India, vol. 6. W. H. Allen and Co., London.

Williams, C. & Harborne, J. B. 1977. The leaf flavonoides of

the Zingiberales. Biochem. Syst. Ecol. 5: 221–229.

Prince & Kress • Phylogeny of Marantaceae55 (2) • May 2006: 281–296

295

Supplementary materials (Appendix 2) are available in the online version of Taxon at http://www.ingentaconnect.com/content/iapt/tax

Prince & Kress • Phylogeny of Marantaceae 55 (2) • May 2006: 281–296

296

Appendix 1. Source and voucher information for taxa sampled in a phylogenetic study of Marantaceae based on chloro-

plast DNA sequence data.

Species, Voucher, trnK (including matK) GenBank Number, trnL-F GenBank Number.

INGROUP TAXA: Afrocalathea rhizantha (K. Schumann) K. Schum., ex hort., A. Carle s.n. no voucher, AY140262, AY140341; Ataeni-

dia conferta (Benth.) Milne-Redh. #1, Kress 99-6572 US, AY140263, AY140342; Ataenidia conferta (Benth.) Milne-Redh. #2, DJ Harris

6671 (RBGE), AY140264, AY140343; Calathea aemula Körn., Kress & Kennedy 99-6390 US, AY140265, AY140344; Calathea colorata

Benth. & Hook. f., Kress & Kennedy 99-6396 US, AY140266, AY140345; Calathea comosa (L. f.) Lindl., ex hort., Lyon Arboretum no

voucher, AY140267, AY140346; Calathea crotalifera S. Watson, Kress 78-0899 DUKE, AY140268, AY140347; Calathea ecuadoriana

H. Kennedy, Kress 01-6966 US, AY140269, AY140348; Calathea foliosa Rowlee ex Woodson & Schery, Hammel 11993 DUKE,

AY140270, AY140349; Calathea gymnocarpa H. Kennedy, Kress & Kennedy 99-6402 US, AY140271, AY140350; Calathea leopardina

Regel, Kress & Kennedy 99-6393 US, AY140272, AY140351; Calathea loeseneri J. F. MacBr., Kress 99-6594 US, AY140273, AY140352;

Calathea majestica (Linden) H. Kennedy, Kress & Kennedy 99-6401 US, AY140274, AY140353; Calathea metallica Planchon & Linden,

Kress 99-6586 US, AY140275, AY140354; Calathea micans (L. Mathieu) Körn., Madison & al. 5486 SEL, AY140276, AY140355;

Calathea mirabilis E. Morren, Kress & Kennedy 99-6392 US, AY140277, AY140356; Calathea musaica (Bull.) L. H. Bailey, Kress 02-

7178 US, AY140278, AY140357; Calathea pallidicosta H. Kennedy, Kress & Kennedy 99-6395 US, AY140279, AY140358; Calathea

pluriplicata H. Kennedy, Kress & Kennedy 99-6399 US, AY140280, AY140359; Calathea rufibarba Fenzl, Kress 01-6856 US, AY140281,

AY140360; Calathea utilis H. Kennedy, Kress & Kennedy 99-6389 US, AY140282, AY140361; Calathea variegata Linden ex Körn.,

Kress 94-3740 no voucher, AY140283, AY140362; Calathea vinosa H. Kennedy, Kress 77-0879 DUKE, AY140284, AY140363; Calathea

warscewiczii (L. Mathieu ex Planch.) Planch. & Linden, L. Conde 8 DUKE, AY140285, AY140364; Cominsia gigantea K. Schum., ex

hort., Lyon Arboretum no voucher, AY140286, AY140365; Cominsia sp. #1, Kress 95-5555 US, AY140287, AY140366; Ctenanthe burle-

marxii H. Kennedy, Albers & Merkel s.n. no voucher, none, AY140367; Ctenanthe setosa Eichl., Kress 94-3684 no voucher, AY140288,

AY140368; Ctenanthe villosa H. Kennedy, Kress 01-6961 US, AY140289, AY140369; Donax canniformis #1 (G. Forst.) K. Schum., Kress

99-6527 US, AY140290, AY140371; Donax canniformis #2 (G. Forst.) K. Schum., Kress 99-6318 US, none, AY140370; Halopegia azurea

K. Schum., Kress & Bordelon 01-6830 US, AY140291, AY140372; Halopegia blumei (Körn.) K. Schum., Kress & al. 98-6204 US,

AY140292, AY140373; Haumania sp. #1, DJ Harris 6672 RBGE, AY140293, AY140374; Hylaeanthe hexantha (Poepp. & Endl.) Jonker-

Verhoel & Jonker, Kress & al. 96-5691 US, AY140294, AY140375; Hypselodelphys sp. #1, DJ Harris 6670 RBGE, AY140295,

AY140376; Ischnosiphon cerotus Loes., Kress & Kennedy 99-6376 US, AY140297, AY140378; Ischnosiphon helenae L. Andersson, Kress

& Bordelon 95-5285 US, AY140298, AY140379; Ischnosiphon leucophaeus (Poepp. & Endl.) Körn., Kress 99-6377 US, AY140299,

AY140380; Ischnosiphon puberulus Loes., Kress & Kennedy 99-6383 US, AY140300, AY140381; Ischnosiphon rotundifolius (Poepp. &

Endl.) Körn., Kress & Kennedy 99-6379 US, AY140301, AY140382; Koernickanthe orbiculata (Körn.) L. Andersson, Silva & Santos 717

1983 US, none, AY140383; Maranta arundinacea L., Kress 01-6853 US, none, AY140384; Maranta bicolor Ker-Gawl., Kress 01-6851

US, AY140302, AY140385; Maranta leuconeura E. Morren, Wurdack s.n. NCU, AY140303, AY140386; Maranta sp. #1, Kress 94-3730

no voucher, AY140304, AY140387; Marantochloa leucantha (K. Schum.) Milne-Redh., Kress 99-6582 US, AY140305, AY140388;

Marantochloa purpurea (Ridley) Milne-Redh., Kress 99-6591 US, AY140306, AY140389; Marantochloa sp. #1, Kress 99-6547 US,

AY140307, AY140390; Megaphrynium macrostachyum (Benth.) Milne-Redh., Kress 98-6290 US, AY140308, AY140391; Monotagma

laxum (Poepp. & Endl.) K. Schum., Kress & Kennedy 99-6381 US, AY140309, AY140392; Monotagma papillosum Hagberg ined., Kress

& Kennedy 94-6411 US, AY140310, AY140393; Monotagma parvulum Loes., Kress & Kennedy 99-6382 US, AY140311, AY140394;

Monotagma smaragdinum (Linden & André) K. Schum., Kress & Kennedy 99-6380 US, AY140312, AY140395; Phacelophrynium sp. #1,

ex hort., Lyon Arboretum no voucher, AY140313, AY140396; Phacelophrynium maximum #1 (Blume) K. Schum., Kress & Kennedy 99-

6384 US, AY140314, AY140397; Phacelophrynium maximum #2 (Blume) K. Schum., Kress & Kennedy 99-6409 US, AY140315,

AY140398; Phrynium imbricatum #1 Roxb., Kress 95-5531 US, AY140317, AY140399; Phrynium imbricatum #2 Roxb. , Kress 99-6387

US, AY140319, AY140401; Phrynium imbricatum #3 Roxb., Kress & al. 00-6798 US, AY140320, AY140402; Phrynium sp. #1, Kress 00-

6616 US, AY140322, AY140405; Phrynium sp. #2, Kress 00-6800 US, AY140316, none; Phrynium sp. #3, Kress 00-6799 US, AY140333,

AY140416; Pleiostachya pruinosa (Regel) K. Schum., Prince s.n. NCU, AY140323, AY140406; Saranthe sp. #1, Kress 96-5737 US,

AY140324, AY140407; Sarcophrynium brachystachys (Benth.) K. Schum., Kress & Bordelon 01-7007 US, AY140325, AY140408;

Sarcophrynium sp. #1, DJ Harris 6668 RBGE, AY140326, AY140409; Schumannianthus dichotomus Gagnep., ex hort., Lyon Arboretum

no voucher, AY140327, AY140410; Schumannianthus virgatus Rolfe, Kress 99-6563 US, AY140328, AY140411; Stachyphrynium lati-

folium (Blume) K. Schum., Kress 99-6343 US, AY140329, AY140412; Stachyphrynium repens (Körn.) Suksathan & Borchs., Kress 99-

6319 US, AY140321, AY140403; Stachyphrynium sumatranum (Miq.) K. Schum., Kress & Kennedy 99-6410 US, AY140318, AY140400;

Stachyphrynium sp. #1, Kress 94-5281 no voucher, AY140330, AY140413; Stachyphrynium spicatum #1 (Roxb.) K. Schum., Kress 00-

6713 US, AY140331, AY140414; Stachyphrynium spicatum #2 (Roxb.) K. Schum., Kress 99-6587 US, AY140332, AY140415; Stromanthe

stromanthoides (J. F. Macbr.) L. Andersson, Kress 94-3676 no voucher, AY140334, AY140417; Stromanthe thalia (Vell.) J. M. A. Braga,

Kress 99-6355 US, AY140335, AY140418; Thalia dealbata Fraser ex Roscoe, Kress 99-6569 US, AY140336, AY140419; Thalia genicu-

lata L., Wurdack 261 NCU, AY140337, AY140420; Thaumatococcus daniellii (Benn.) Benth. & Hook. f., Kress 98-6288 US, AY140338,

AY140421; Trachyphrynium braunianum (K. Schum.) Baker, Kress 01-6954 US, AY140339, AY140422.

OUTGROUP TAXA: Canna paniculata Ruiz & Pav., Prince 95-211 NCU, AY140340, AY140423; Siphonochilus kirkii (Hook. f.) B.L.

Burtt, Kress 94-3692 US, AF478895, AY140429.

55 (2) • May 2006: Appendix 2 Prince & Kress • Phylogeny of Marantaceae

Fig. 1. One of >50,000 shortest maximum parsimony trees resulting from an analysis of the trnL-F intergenic spacer

region for members of Marantaceae. Numbers above branches are maximum parsimony jackknife values based on

100,000 fastswap replicates. Bold lines indicated branches with posterior probabilities of 0.95 or higher. Dashed lines

indicate branches not found in the maximum parsimony strict consensus tree.

Ctenanthe setosa

Ctenanthe villosa

Hylaeanthe hexantha

Maranta leuconeura

Maranta arundinacea

Stromanthe thalia

Koernickanthe orbiculata

Maranta bicolor

Maranta sp. #1

Stromanthe stromanthoides

Saranthe sp. #1

Ctenanthe burle marxii

Halopegia azurea

Halopegia macrostachya

Schumannianthus virgatus

Ataenidia conferta #2

Marantochloa purpurea

Marantochloa leucantha

Marantochloa sp. #1

Stachyphrynium jagorianum

Stachyphrynium latifolium

Stachyphrynium sp. #1

Stachyphrynium sp. #2

Stachyphrynium spicatum

Afrocalathea rhizantha

Stachyphrynium parvum

Sarcophrynium brachystachys

Sarcophrynium sp. #1

Donax canniformis #1

Donax canniformis #2

Phrynium sp. #3

Schumannianthus dichotomous

Cominsia gigantea

Cominsia sp. #1

Phacelophrynium sp. #1

Phacelophrynium sp. #2

Phacelophrynium sp. #3

Phrynium oliganthum

Phrynium philippinense

Phrynium rheedei

Phrynium sp. #1

Thalia dealbata

Thalia geniculata

Ischnosiphon cerotus

Ischnosiphon puberulus

Pleiostachys pruinosa

Ischnosiphon leucophoeusais

Ischnosiphon rotundifolius

Ischnosiphon helenae

Calathea crotalifera

Calathea pluriplicata

Calathea utilis

Monotagma laxum

Monotagma papillosum

Monotagma parvulum

Monotagma smaragdinum

Calathea micans

Calathea variegata

Calathea majestica

Calathea gymnocarpa

Haumania sp. #1

Calathea comosae

Calathea ecuadoriana

Calathea pallidicosta

Calathea loeseneri

Calathea metallica

Calathea foliosa

Calathea vinosa

Calathea warscewiczii

Calathea aemulae

Calathea mirabilis

Calathea colorata

Calathea musaica

Calathea rufibarba

Calathea leopardina

Hypselodelphys sp. #1

Trachyphrynium braunianum

Megaphrynium macrostachyum

Thaumatococcus daniellii

Canna paniculata

Siphonochilus kirkii

Ataenidia conferta #1

100

57 85

MARANTA

CLADE

STACHYPHRYNIUM

CLADE

DONAX

CLADE

CALATHEA

CLADE

TRACHYPHRYNIUM

CLADE, PRO PARTE

TRACHYPHRYNIUM

CLADE, PRO PARTE

OUTGROUP TAXA

Prince & Kress • Phylogeny of Marantaceae 55 (2) • May 2006: Appendix 2

Fig. 2. One of 4319 shortest maximum parsimony trees resulting from an analysis of the matK coding region for mem-

bers of Marantaceae. Numbers above branches are maximum parsimony jackknife values based on 100,000 fastswap

replicates. Bold lines indicated branches with posterior probabilities of 0.95 or higher. Dashed lines indicate branches

not found in the maximum parsimony strict consensus tree.

Calathea leopardina

Calathea rufibarba

Calathea aemula

Calathea colorata

Calathea musaica

Calathea mirabilis

Calathea comosa

Calathea ecuadoriana

Calathea loeseneri

Calathea metallica

Calathea pallidicosta

Calathea foliosa

Calathea warscewiczii

Calathea vinosa

Calathea micans

Calathea variegata

Calathea majestica

Calathea gymnocarpa

Ischnosiphon helenae

Ischnosiphon leucophaeus

Ischnosiphon cerotus

Ischnosiphon puberulus

Ischnosiphon rotundifolius

Pleiostachys pruinosa

Calathea crotalifera

Calathea pluriplicata

Calathea utilis

Monotagma laxum

Monotagma papillosum

Monotagma parvulum

Monotagma smaragdinum

Phrynium oliganthum

Phrynium sp. #1

Phrynium rheedei

Phrynium philippinense

Phrynium sp. #2

Phrynium sp. #3

Phacelophrynium sp. #1

Donax canniformis #1

Schumannianthus dichotomous

Thalia geniculata

Stachyphrynium jagorianum

Stachyphrynium sp. #2

Stachyphrynium spicatum

Stachyphrynium latifolium

Stachyphrynium sp. #1

Afrocalathea rhizantha

Stachyphrynium parvum

Ataenidia conferta #1

Ataenidia conferta #2

Marantochloa purpurea

Marantochloa leucantha

Marantochloa sp. #1

Ctenanthe setosa

Ctenanthe villosa

Stromanthe stromanthoides

Stromanthe thalia

Saranthe sp. #1

Maranta bicolor

Maranta sp. #1

Maranta leuconeura

Hylaeanthe hexantha

Halopegia azurea

Halopegia macrostachya

Schumannianthus virgatus

Haumania sp. #1

Hypselodelphys sp. #1

Trachyphrynium braunianum

Megaphrynium macrostachyum

Thaumatococcus daniellii

Sarcophrynium brachystachys

Sarcophrynium sp. #1

Canna paniculata

Siphonochilus kirkii

100

67 64

100 59

99 67 99

100

100 96

100

100 89 67

100 99 100

100

76 95

82 68 100

100 100

100

MARANTA

CLADE

STACHY-

PHRYNIUM

CLADE

DONAX

CLADE

CALATHEA

CLADE

OUTGROUP TAXA

TRACHY-

PHRYNIUM

CLADE

Prince & Kress • Phylogeny of Marantaceae55 (2) • May 2006: Appendix 2

Fig. 3. One of >50,000 shortest maximum parsimony trees resulting from an analysis of the trnK 3' intergenic spacer

region for members of Marantaceae. Numbers above branches are maximum parsimony jackknife values based on

100000 fastswap replicates. Bold lines indicated branches with posterior probabilities of 0.95 or higher. Dashed lines

indicate branches not found in the maximum parsimony strict consensus tree. Arrows indicate taxa found placed in

unexpected positions on the phylogeny.

Calathea colorata

Calathea leopardina

Calathea mirabilis

Calathea musaica

Calathea rufibarba

Calathea aemula

Calathea comosa

Calathea ecuadoriana

Calathea loeseneri

Calathea metallica

Calathea pallidicosta

Monotagma laxum

Monotagma papillosum

Monotagma parvulum

Monotagma smaragdinum

Calathea foliosa

Calathea vinosa

Calathea warscewiczii

Calathea gymnocarpa

Calathea micans

Calathea variegata

Calathea majestica

Calathea crotalifera

Calathea pluriplicata

Stachyphrynium jagorianum

Stachyphrynium sp. #2

Stachyphrynium parvum

Stachyphrynium latifolium

Stachyphrynium sp. #1

Stachyphrynium spicatum

Afrocalathea rhizantha

Thalia geniculata

Megaphrynium macrostachyum

Thaumatococcus daniellii

Ctenanthe setosa

Ctenanthe villosa

Saranthe sp. #1

Stromanthe stromanthoides

Stromanthe thalia

Maranta bicolor

Maranta sp. #1

Maranta leuconeura

Halopegia azurea

Halopegia macrostachya

Schumannianthus virgatus

Hylaeanthe hexantha

Phrynium oligantha

Phrynium philippinense

Phrynium rheedei

Phrynium sp. #1

Donax canniformis #1

Phrynium sp. #2

Schumannianthus dichotomous

Phrynium sp. #3

Phacelophrynium sp. #1

Calathea utilis

Ischnosiphon cerotus

Ischnosiphon leucophaeus

Ischnosiphon helenae

Ischnosiphon puberulus

Ischnosiphon rotundifolius

Pleiostachya pruinosa

Haumania sp. #1

Sarcophrynium brachystachys

Sarcophrynium sp. #1

Ataenidia conferta #1

Ataenidia conferta #2

Marantochloa leucantha

Marantochloa purpurea

Marantochloa sp. #1

Hypselodelphys sp. #1

Trachyphrynium braunianum

STACHYPHRYNIUM

CLADE PRO PARTE

DONAX CLADE

CALATHEA CLADE

PRO PARTE

CALATHEA CLADE

PRO PARTE

TRACHYPHRYNIUM

CLADE PRO PARTE

MARANTA CLADE

TRACHYPHRYNIUM

CLADE PRO PARTE

STACHYPHRYNIUM

CLADE PRO PARTE

TRACHYPHRYNIUM

CLADE PRO PARTE

Molecular phylogenetics of Maranta (Marantaceae: Zingiberales): non-monophyly and support for a wider circumscription

Article

Feb 2023

Neotropical Maranta has repeatedly emerged as non-monophyletic in molecular phylogenetic studies, but no taxonomic changes have been proposed due to previous weak support for the main clades and overall sparse taxonomic sampling. Our study includes a phylogenetic hypothesis strictly based on molecular evidence, using nuclear ribosomal (nrITS) and plastid (rps16, trnL-F and rpl32-trnL) markers for Maranta and allied genera. Thirty-two species from eight genera were sampled, and maximum likelihood (ML) and Bayesian inference (BI) analyses were carried out. Non-ambiguous indels were scored in both analyses to test their contribution to internal support. Our results confirm that Maranta, as previously delimited, is non-monophyletic, with species of Hylaeanthe, Myrosma and Koernickanthe nested among clades of Maranta. The combined BI analysis without indels was the best resolved, and inclusion of indels increased support only for terminal clades. The sampled species comprise two sister clades, one with Maranta + Hylaeanthe + Myrosma + Koernickanthe and the other with Ctenanthe + Saranthe + Stromanthe. The infrageneric classification proposed by Schumann for Maranta (M. subgenera Maranta and Calateastrum) is partially corroborated by our results, but the remaining subgenera need to be further studied. Based on our strongly supported phylogenetic results, we revise the taxonomy of these genera, expanding the limits of Maranta. We propose two new combinations and a new name for Hylaeanthe in Maranta.

Myanmaranthus roseiflorus, a New Genus and Species of Marantaceae from Myanmar

Article

Full-text available

Sep 2022

Myanmaranthus Nob.Tanaka, Suksathan & K.Armstr., a new genus of Marantaceae from northern Myanmar, is described with a single species, M. roseiflorus Nob.Tanaka & K.Armstr. Its relationship to all other genera in Asian Marantaceae is investigated through morphological examination and molecular phylogenetic analyses based upon chloroplast (rps16 intron, trnL�trnF) and nuclear (ITS and ETS) sequences. Myanmaranthus differs morphologically from the most closely related genus, Phrynium Willd., in having a combination of the following characters: a rosulate habit, a loose paniculate inflorescence arising from the rhizome, the absence of interphylls and bracteoles, and fertile bracts each holding a single pink flower. Thus far, this new taxon is known only from the type locality in Kachin State, Myanmar. A key to the genera of Asian Marantaceae is provided.

Linking high diversification rates of rapidly growing Amazonian plants to geophysical landscape transformations promoted by Andean uplift

Article

Jan 2022

Amazonia is extremely biodiverse, but the mechanisms for the origin of this diversity are still under debate. We propose a diversification model for Amazonia based on the interplay of intrinsic clade functional traits, habitat associations and past geological events, using as a model group the species-rich Neotropical family Marantaceae. Our results show that the species richness of the lineage is predicted by functional strategy, rather than clade age, and thus the fast vs. slow growth functional trade-off is a major determinant of clade diversification in Marantaceae. Rapidly growing clades were mostly associated with highly productive habitats, and their origin and diversification dynamics matched the expansion of fertile soils mediated by Andean uplift c. 23 Mya. Fast-growth strategies probably led to fast molecular evolution, speeding up speciation rates and species accumulation, resulting in higher numbers of extant species. Our results indicate that pure allopatric-dispersal models disconnected from past geological and ecological forces may be inadequate for explaining the evolutionary and diversity patterns in Amazonian lowlands. We suggest that a coupling of the functional trait-niche framework with diversification dynamics provides insights into the evolutionary history of tropical forests and helps elucidate the mechanisms underlying the origin and evolution of its spectacular biodiversity.

Linking high diversification rates of rapidly growing Amazonian plants to geophysical landscape transformations promoted by Andean uplift

Article

Jan 2022

Ethnobotanical study of the family Marantaceae R. Br in Bangladesh Agricultural University Botanical Garden

Article

Full-text available

Jun 2023

The Marantaceae family is a diverse group of plants that has drawn the interest of scientists and researchers worldwide due to their distinctive morphological characteristics, ecological and economic relevance. The Bangladesh Agricultural University Botanical Garden is home to an abundance of Marantaceae species, making it a useful resource for examining the diversity and significance of this plant family. This present study was designed to survey and document the family Marantaceae with an overview of the family emphasizing its morphological, economic, and ethnobotanical relevance based on a literature review. During the study, we found 25 species (two of which have two varieties each) belonging to 8 genera of which Goeppertia contributed the most species, with 16, followed by Maranta with 4 (including varieties) and Thalia with 2; the remaining 5 genera each contributed one species. Our findings reveal the remarkable diversity and significance of Marantaceae plants in this region, highlighting the necessity for their conservation and protection.

Accommodating trait overlap and individual variability in species diagnosis of Ischnosiphon (Marantaceae)

Article

Jul 2020

Recognition and delimitation of taxonomic categories of biological organisms are still challenging and full of controversy. We used Ischnosiphon as a model to unravel the importance of morphometrics as individual-based variables to disentangle the morphological variability of plant species. Ischnosiphon spp. continue to be problematic for users, taxonomists and ecologists, due mainly to the huge morphological variability, the species criteria and circumscription proposed for many taxa and the many habitat and vegetative macro-morphological characters lacking in most currently available exsiccates. Twenty-three morphometric variables were sampled from 228 individuals, belonging to 22 Ischnosiphon spp. Principal components and discriminant multivariate analyses were used to describe and identify patterns of morphological variation in Ischnosiphon. Individual-landmark assessment analysed with multivariate methods captured morphometric intraspecific diversity and morphological variability in Ischnosiphon spp., along with the continuous variation of important morphological traits. By examining the morphology of Ischnosiphon spp. through individual-landmark assessment, we demonstrate that different morphological species concepts used today in the identification of the species are difficult to apply. We propose a replicable and analytical framework to accommodate individual variability in species diagnosis in morphologically diverse plant groups.

Vascular bundle modifications in nodes and internodes of climbing Marantaceae

Article

Oct 2020

Nodes are interfaces between stems and leaves. Vascular bundles originate here and elongate into leaves and internodes. In Marantaceae, internodal bundles are highly diverse, including inverted bundles in the climbing genus Haumania. The objective of this paper is to characterize bundle forms, their position across the stem and their connection to leaves and short shoots in Haumania spp. and other unrelated African branch-angle climbers in the family (Hypselodelphys, Trachyphrynium). We question whether bundle inversion is a genus-specific trait in Haumania or related to the climbing growth form. Vascular bundles in internodes are scattered across the stem diameter in a characteristic pattern. Four (to five) bundle types follow each other in a centripetal order from highly sclerenchymatic 'a'-bundles close to the epidermis to 'd'-bundles in the centre with a low sclerenchyma proportion. Inverted bundles only appear in internodes of Haumania, making this trait a synapomorphy for the genus. The nodes show stem, leaf and short shoot bundles in a remarkably diverse pattern with partitioned phloem clusters and apparently augmented xylem elements. Our preliminary conclusion is that the inversion of bundles happens when leaf and short shoot traces join the main axis bundle layers.

Anatomy of leaf edges in Marantaceae in the Neotropics: the relationship between vernation and leaf asymmetry and contributions to the systematics of the family

Article

Nov 2019

Marantaceae consist of species with asymmetric leaves of two types: those with either a wider left or right half; this asymmetry is related, respectively, to clockwise or counterclockwise convolute vernation. In this study, we analysed whether anatomical differences in the leaf edges, i.e. the anatomical asymmetry, were related to the orientation of the convolute vernation and to the asymmetry of leaf morphology, and whether these differences supported the organization of the clades in the family. Transverse sections of the mid third of the leaf buds expanded to the height of the right and left edges of the blades were prepared for 19 species belonging to 11 genera, using cyto-histological techniques. Anatomical analyses of the blade edges revealed that there is a relationship between morphological asymmetry and anatomical asymmetry that has never before been ascribed to the family. The anatomical data support differences between the arrangements in two of the three Neotropical informal groups. In the Calathea clade, Calathea showed much more similarity with Goeppertia than with Ischnosiphon and Monotagma, since they are the only genera that do not present with anatomical asymmetry. In the Maranta clade, Ctenanthe, Saranthe and Stromanthe appear to be related to one another, as they share strong anatomical asymmetry and fibrous edges. These characteristics, however, are not observed in Myrosma, which in turn is more anatomically similar to Maranta.

Accommodating trait overlap and individual variability in species diagnosis of Ischnosiphon (Marantaceae)

Article

Jul 2020

Taxonomic Study of the Species of Maranta Plum ex. L. (Marantaceae) from Northeastern Brazil: A Neglected Diversity Center for the Genus with Five New Species

Article

Oct 2021

This study summarizes the taxonomy of the species of Maranta from northeastern Brazil. While 20 names have been proposed in the region, only 13 are accepted. We provide two new records, M. rugosa and M. polystachya , and five new species are described, of which M. bahiensis and M. villosovaginata are endemic to Bahia, M. chrysogina to Ceará, M. vieirae to Maranhão, and M. lorifolia occurs in Bahia and Espírito Santo. Maranta noctiflora , M. parvifolia , M. phrynioides , and M. rupicola L. have distributions outside of the study area, while M. anderssoniana and M. hatschbachiana are synonymized. Our final checklist includes 21 species, which represent more than half of the species in the genus, highlighting northeastern Brazil as a center of diversity for Maranta . Among the species listed, M. lorifolia , M. tuberculata , and M. zingiberina are thought to be endangered (EN), while M. gigantea is likely critically endangered (CR). Finally, this study provides informal conservation statuses, descriptions, distribution maps, and an identification key to the species. Typifications for M. subterranea and M. polystachya are also provided.

HOMEOSIS IN THE FLOWERS OF THE ZINGIBERALES

Article

Jun 1991

Bruce Kirchoff

Homeosis has played an important role in the evolution of the flowers of the Zingiberales, especially those of the Ginger Group. In the Zingiberaceae, two members of the outer androecial whorl are replaced by a lip, and two members of the inner androecial whorl are replaced by petaloid staminodes. Most of the androecium of the Costaceae has also been replaced by petaloid structures, and the single fertile stamen is often attached to an enlarged petaloid “filament.” The Cannaceae and Marantaceae have one-half of one fertile anther and three to four variously modified staminodes. In contrast, homeosis has played a minor role in floral evolution of the Banana Group. Only in the Heliconiaceae has a stamen been replaced by a staminode. In none of the families of the Zingiberales do the staminodes assume the total “form or character” of any perianth members. Because of this, it is reasonable to extend the definition of homeosis to include replacement by an organ like, but not identical to, some other part of the plant.

FLORAL ORGANOGENESIS IN FIVE GENERA OF THE MARANTACEAE AND IN CANNA (CANNACEAE)

Article

Apr 1983

Bruce Kirchoff

The paired flowers of all species of the Marantaceae studied, except Monotagma plurispicatum, are produced through the division of an apical meristem with a tunica-corpus structure. The solitary flowers of M. plurispicatum develop from a similar meristem which does not bifurcate. The paired flowers of Canna indica are produced in the axil of a florescence bract through the formation of a bract and an axillary flower on the side of the primordium which gives rise to the largest flower of the pair. The sequence of organ initiation for both families is: calyx, corolla and inner androecial whorl, outer androecial whorl, gynoecium. The sequence of sepal formation is opposite in the two families. In the Cannaceae it leads directly into the spiral created by the formation of the other organs, while in the Marantaceae the sequence of sepal formation follows a spiral opposite to that of the other floral organs. The members of the corolla and inner androecial whorl separate from common primordia. In general these common primordia separate into a petal and an inner androecial member through the initiation of two growth centers, at the same level, in the dorsal and ventral flanks of the primordium. In Ischnosiphon elegans and Pleiostachya pruinosa the stamen is initiated at a lower position than the petal in the ventral flank of the common primordium. A similar pattern of initiation is described for the callose staminode in Marantochloa purpurea and Canna indica. This pattern is interpreted as a variation on the more generalized pattern of inner androecial formation found in the other genera.

Flowering Plants · Monocotyledons: Alismatanae and Commelinanae (except Gramineae)

Book

Jan 1998

Klaus Kubitzki

Volumes III and IV of this encyclopaedia provide a novel classification of the monocotyledons, a group encompassing plants of most diverse life-forms such as aquatics, terrestrial and epiphytic herbs, and tall trees. Of the 106 families now recognized 104 are treated in the two volumes, while the economically or horticulturally important grass and orchid families are relegated to two subsequent volumes. The classification followed here is based on recent molecular studies as well as on the vast body of information available on this plant group. The wealth and precision of information, but also the keys for the identification of genera and details on their properties, including distribution and diversification, make this work an important source for both the scholar and the practitioner in the fields of pure and applied plant sciences.

CONFIDENCE LIMITS ON PHYLOGENIES: AN APPROACH USING THE BOOTSTRAP

Article

Jul 1985
EVOLUTION

Joseph Felsenstein

The recently-developed statistical method known as the "bootstrap" can be used to place confidence intervals on phylogenies. It involves resampling points from one's own data, with replacement, to create a series of bootstrap samples of the same size as the original data. Each of these is analyzed, and the variation among the resulting estimates taken to indicate the size of the error involved in making estimates from the original data. In the case of phylogenies, it is argued that the proper method of resampling is to keep all of the original species while sampling characters with replacement, under the assumption that the characters have been independently drawn by the systematist and have evolved independently. Majority-rule consensus trees can be used to construct a phylogeny showing all of the inferred monophyletic groups that occurred in a majority of the bootstrap samples. If a group shows up 95% of the time or more, the evidence for it is taken to be statistically significant. Existing computer programs can be used to analyze different bootstrap samples by using weights on the characters, the weight of a character being how many times it was drawn in bootstrap sampling. When all characters are perfectly compatible, as envisioned by Hennig, bootstrap sampling becomes unnecessary; the bootstrap method would show significant evidence for a group if it is defined by three or more characters.

Tentative keys to the subfamilies, tribes and genera of the Zingiberales

Article

Jan 1972

Ver-o-peso: The ethnobotany of an Amazonian market

Article

Jan 1984

M.A. Van Den Berg

A new species of the genus Monophyllanthe (Marantaceae) from the Amazonian Basin

Article

Jan 2001

A new species of Monophyllanthe, M. araracuarensis S. Suáez, Galeano & H. Kennedy (Marantaceae) from Amazonian Colombia and adjacent Brazil, is described and illustrated. This is the second species described in the genus, which was previously known only from French Guiana and Suriname. The new species is characterized by its sparsely branched inflorescence, with 3 to 7 bracts per branch, borne on a separate, leafless shoot directly from the rhizome.

MRBAYES: Bayesian inference of phylogeny

Article