Content uploaded by Tânia Wendt
Author content
All content in this area was uploaded by Tânia Wendt on Jul 16, 2014
Content may be subject to copyright.
Monophyly and Phylogenetic Relationships in Lymania (Bromeliaceae:
Bromelioideae) Based on Morphology and Chloroplast DNA Sequences
LEANDRO DE OLIVEIRA FURTADO DE SOUSA,
1
TA
ˆNIA WENDT,
1,4
GREGORY K. BROWN,
2
DOROTHY E. TUTHILL,
2
and TIMOTHY M. EVANS
3
1
Departamento de Bota
ˆnica, Universidade Federal do Rio de Janeiro, IB, CCS, Ilha do Funda
˜o, 21941-590,
Rio de Janeiro-RJ, Brazil;
2
Department of Botany, University of Wyoming, Laramie, Wyoming 82071 and Marie Selby Botanical
Gardens, Sarasota, Florida 34236 U.S.A.;
3
Biology Department, Hope College, 35 East 12
th
street, Holland, Michigan, 49423-9000 U.S.A.
4
Author for correspondence (twendt@biologia.ufrj.br)
Communicating Editor: Matt Lavin
ABSTRACT.A cladistic analysis of Lymania was conducted using morphology and sequences from three
chloroplast DNA regions: the matK coding region and the psbA-trnH and trnL-trnF intergenic spacers. The
monophyly of the genus and the phylogenetic relationships among related genera were examined. Of the nine
Lymania species, eight are endemic to southern Bahia, Brazil. Lymania is the first genus in Bromeliaceae subfamily
Bromelioideae to be subjected to a combined morphological and molecular analysis. The genera of Bromelioideae
have been particularly difficult to classify and there has been disagreement about their interrelationships and
monophyly. Morphological data show better resolution than molecular data alone. The partition homogeneity test
supported a combined analysis of the two data sets, yielding a single most parsimonious tree. In the combined
analysis, monophyly of Lymania is moderately supported, and the genus is closely related to species of Aechmea
subg. Lamprococcus and subg. subg. Ortigiesia. The morphological distinctiveness coupled with low molecular
divergence indicates relatively recent and rapid speciation within Lymania. The combined analysis of morphological
and molecular data as done in this study provides a framework for future research on other Bromelioideae genera
that could foster better taxonomic rearrangements.
KEYWORDS:Brazil, cladistics, combined analysis, matK,psbA-trnH,trnL-trnF.
Bromeliaceae is a well-defined monocot family, its
monophyly being supportedby several phylogenetic
and morphological characters (Gilmartin and Brown
1987; Gaut et al. 1992; Clark et al. 1993; Duvall et al.
1993; Givnish et al. 1999, 2004). Three subfamilies of
Bromeliaceae have been traditionally recognized,
Bromelioideae, Pitcairnioideae, and Tillandsioideae
(Mez 1934–1935; Smith and Downs 1974, 1977, 1979).
Several phylogenetic studies have been focused on
the relationships among the subfamilies (Ranker et
al. 1990; Terry et al. 1997a; Horres et al. 2000), and
genera (Gilmartin and Brown 1986; Varadarajan and
Gilmartin 1988; Terry et al. 1997b; Faria et al. 2004;
Barfuss et al. 2005). All these works reveal some
conflict within the generic concepts presented in the
last comprehensive monograph of the family by
Smith and Downs (1974, 1977, 1979)
Morphological characters traditionally used to
circumscribe genera within Bromeliaceae often fail
to delimit natural groups (Brown and Terry 1992;
Grant 1993), and the generic limits are frequently
undergoing change (e.g., Read 1984; Smith and
Spencer 1992; Brown and Leme 2005). The absence
of convincingly delimited monophyletic groups,
especially within the Bromelioideae, is well known
(Duval et al. 2003; Faria et al. 2004).
The circumscription of natural groups will be
greatly aided by phylogenetic studies using
morphological and molecular data. Combined
analyses of such data have not been undertaken in
Bromeliaceae even though they have proven more
robust than analyses of molecular or morphological
data alone (Donoghue and Sanderson 1992; Nandi
et al. 1998; Chase et al. 2000; Sytsma and Pires 2001).
Lymania Read comprises nine narrowly distrib-
uted species, which are known from a limited
number of collections (De Sousa 2004). Read (1984)
had described Lymania to accommodate all species
with furrowed- or winged-ovaries that were
segregated from Aechmea subgenus Lamprococcus,
Araeococcus, and Ronnbergia. New species were
added to Lymania (Leme 1987, 2006; Leme and
Forzza 2001), and recently De Sousa (2004) pro-
vided the first revision. However, the genus is
diagnosed with only a few characters, and as with
most genera in Bromelioideae, the monophyly is
questionable. The current study is designed to
examine the relationships among Lymania and its
relatives, which have not been evaluated within
a phylogenetic framework. Our specific aims are
to: 1) evaluate the utility of a combined phyloge-
netic analysis of morphological and molecular data
on a group of closely related Bromelioideae
species, 2) assess the monophyly of Lymania using
morphology and sequences from three chloroplast
DNA regions (matK,trnL-trnF, and psbA-trnH), and
Systematic Botany (2007), 32(2): pp. 264–270
#Copyright 2007 by the American Society of Plant Taxonomists
264
3) examine relationships among Lymania species,
and the distribution of specific morphological traits
within this genus.
MATERIALS AND METHODS
Taxon Sampling. Our data matrix comprises 23 species
representing five genera of subfamily Bromelioideae (Ap-
pendix 1). Almost all recognized species of Lymania were
included except for L. marantoides (L. B. Sm.) Read and L.
languida Leme, for which we were unable to obtain material.
Samples of the genera Aechmea (10 species), Araeococcus (2),
and Ronnbergia (2) were chosen because some Lymania
species were previously included in these genera. Two
species of Cryptanthus were designated as outgroups based
on Terry et al. (1997a) and Faria et al. (2004). Almost all plant
material was collected from naturally occurring populations
during two years of field studies conducted in the eastern
Brazilian Atlantic Rain Forest of Rio de Janeiro, Espı
´rito
Santo, and Bahia States. Material of Aechmea pedicellata and
Araeococcus goeldianus were obtained from The Marie Selby
Botanical Gardens (SEL).
Morphological Data. We scored 60 morphological char-
acters for this study (Appendix 2, 3). Meticulous annotation
concerning the habit and color of the plant structures, as well
as photo documentation, were made in the field or blooming
ex-situ. Flowers and fruits in different periods of maturation
were preserved in 70% ethanol prior to examination. The
remaining characters were determined after preparing the
herbarium vouchers. Continuous variables (e.g., mature leaf
length) were converted to discrete variables based on
observed gaps between species. As much as it was possible,
the same individual or clones were used for the morpholog-
ical and molecular data (Appendix 1).
DNA Extraction, Amplification and Sequencing. The
samples for DNA extraction were obtained from living
plants collected in the field or in cultivation and quickly
dried in silica gel. Total DNA was isolated using a modifica-
tion of the 23CTAB buffer method (Doyle and Doyle 1987;
Smith et al. 1991) followed by purification with the QIAamp
DNA Mini Kit (Qiagen). The protocols for PCR amplification
and sequencing of all three cpDNA regions follow Johnson
and Soltis (1995). For PCR amplification of matK, we used
primers two and five from Crayn et al. (2000). psbA-trnH and
trnL-trnF primer sequences were taken from Sang et al.
(1997). Almost all DNA sequences were originally produced
for this study, except for Cryptanthus beuckeri, for which the
matK sequence (GenBank AF 539965) was deposited earlier as
part of another study (Appendix 1). For the total combined
data matrix (morphology plus molecular), 3.7% of the cells
were scored as ‘‘missing.’’
Data Analysis. All matrices for this study were deposited
in TreeBASE (study number S1754). Cladistic analyses were
performed using maximum parsimony (PAUP version
4.0b.10, Swofford 2003). Heuristic searches were conducted
using the TBR branch swapping algorithm, MULTREES on,
and 10000 random addition replicates. All characters were
unordered and assigned equal weight. Relative levels of
homoplasy and synapomorphy in the data sets were
calculated using the consistency index (CI), the retention
index (RI), and the rescaled consistency index (RC) as
implemented in PAUP* 4.0b10. The cpDNA sequences were
aligned manually and combined into a single molecular
matrix. A matrix of morphological characters was created
using the computer program MacCLADE version 4.0
(Maddison and Maddison 2000). Morphological character
state changes were mapped using MacClade and AC-
CTRANS character state optimization. One thousand
bootstrap replicates with full heuristic search were conducted
to evaluate support for each clade.
For parsimony analysis, the morphological and cpDNA
data sets were each analyzed individually and in combina-
tion. Before combining the separate data sets, congruence
between them was examined using the incongruence length
differences (ILD) test (Farris et al. 1995), which is referred to
as the partition homogeneity test in PAUP. Cunningham
(1997) indicated that combining the data improved or did not
reduce the phylogenetic accuracy if a Pvalue of the ILD test
was greater than 0.01.
For Bayesian analyses of the combined morphological/
molecular data matrix, appropriate models of evolution were
selected for the matK region (HKY +G) and the non-coding
cpDNA regions (F81 +I+G) using a hierarchical ratio test as
implemented in Modeltest (Posada and Crandall 1998).
Bayesian analyses using MrBayes ver. 3.1.2 (Huelsenbeck
and Ronquist 2001; Ronquist and Huelsenbeck 2003) con-
sisted of 5,000,000 generations, with samples taken every
10,000 generations. Trees from the first 1.2 million genera-
tions were discarded (i.e. the first 25% of the generations
were the ‘‘burn-in’’ stage).
RESULTS
Morphological Analysis. The morphological
dataset produced nine most parsimonious trees of
219 steps (CI 50.45, RI 50.59; Table 1). The strict
consensus tree from morphological data showed
that the ingroup is well supported as monophyletic
with a bootstrap value of 100%, but there is little
resolution among genera (tree not shown). Lymania
is monophyletic, and forms a polytomy that is
sister to the clade containing species of Aechmea
subg. Lamprococcus and subg. Ortgiesia, though no
bootstrap support was found for this relationship.
Molecular Analysis. The aligned molecular
matrix had a total of 1858 characters with 1782
being constant, 45 parsimony uninformative, and
31 being parsimony informative. The cladistic
analysis of molecular data yielded 214 most
parsimonious trees of 83 steps (CI 50.86, RI 5
0.79; Table 1). The strict consensus (not shown) is
largely unresolved, with a large basal polytomy
and four main clades. The monophyly of Lymania
is not supported.
Combined Morphological and Molecular Analy-
sis. The partition homogeneity test produced a P
value of 0.15, indicating combinability for the two
data sets. The parsimony analysis of combined
data resulted in a single most parsimonious tree
(Fig. 1) of 310 steps (CI 50.56, RI 50.61; Table 1).
Bayesian analysis resulted in a tree that was similar
in overall topology to that produced by parsimony,
but with slightly poorer resolution and a paraphy-
letic (instead of polyphyletic) Ronnbergia (Fig. 2).
The monophyly of Lymania is moderately sup-
ported in these analyses. In each analysis, Lymania
is sister to Aechmea species from subg. Lamprococcus
and subg. Ortgiesia.Araeococcus is monophyletic
and sister to the clade formed by species of
2007] DE SOUSA ET AL.: PHYLOGENETIC OF LYMANIA 265
Lymania and Aechmea (subg. Lamprococcus and
subg. Ortgiesia).
DISCUSSION
Integration of Morphological and Molecular
Data in Lymania and Related Species. The
generic delimitation of Aechmea has not been
supported in previous studies (Faria et al. 2004)
and polyphyly of this genus has long been
recognized (Smith and Downs 1974, 1979). While
the present study includes seven of nine recog-
nized Lymania species, more extensive sampling of
Aechmea and other Bromelioideae genera will be
necessary to clarify relationships between the two
genera.
Scotland et al. (2003) noted that morphological
data are currently utilized less for phylogeny
reconstructions than are molecular sequence data.
This is true for bromeliad phylogenetic studies
(Ranker et al. 1990; Terry et al. 1997a, b; Crayn et al.
2000; Horres et al. 2000; Duval et al. 2003; Barfuss
TABLE 1. Summary statistics for parsimony analyses of morphological, molecular, and combined data sets for Lymania and
related genera.
psbA-trnH trnL-trnF matK Total Molecular Morphology Morphology +Molecular
No. of characters in matrix 602 389 867 1858 60 1918
No. of constant characters 577 375 830 1782 0 1782
No. of uninformative
characters 16 8 21 45 4 49
No. of informative
characters 9 6 16 31 56 87
No. of trees - - - 214 9 1
No. of steps - - - 83 219 310
CI - - - 0.86 0.45 0.56
RI - - - 0.79 0.59 0.61
RC - - - 0.67 0.27 0.34
Figure - - - 2 1 3
FIG. 1. Single most parsimonious tree from the simulta-
neous analysis of morphological and three chloroplast DNA
sequence data of Lymania and related species of Bromelioi-
deae with Cryptanthus as the outgroup. Bootstrap percentages
greater than 50% are given above the branches. Selected
characters mapped onto the parsimonious tree are indicated
with black bars. Numbers refer to characters in the list of
morphological characters. Characters and states discussed
in text.
FIG. 2. Consensus tree from Bayesian analysis of mor-
phological and cpDNA sequence data. Posterior probabilities
greater than 50% are given above the branches.
266 SYSTEMATIC BOTANY [Volume 32
et al. 2005). There is a general assumption that
morphological data should show higher levels of
homoplasy, lower levels of resolution, and lower
levels of support (e.g., bootstrap and posterior
probability values) than molecular data (Scotland
et al. 2003). An increased number of characters in
an analysis generally increases accuracy, and this is
seen as one of the primary advantages of molecular
data. However DNA studies undertaken for
Bromeliaceae, and specifically for subfamily Bro-
melioideae, have suffered from an unexpectedly
low level of phylogenetic resolution due to an
unusually low substitution rate (Terry et al. 1997a,
b; Horres et al. 2000; Barfuss et al. 2005).
In our study, morphological data produced
a more highly resolved phylogeny than molecular
data alone. The combined analyses resulted in
complete resolution among species within Lymania
and higher support values across the tree (Figs. 1
and 2). This is the first study in Bromeliaceae in
which morphological and molecular data have
been combined for a phylogenetic analysis, and
this approach appears to hold great promise for
systematic studies in the family. While combining
data from different data sets still remains some-
what controversial (see reviews by de De Queiroz
et al. 1995; Huelsenbeck et al. 1996), there are
reasonable arguments in favor of combined anal-
ysis of data sets. In Bromeliaceae, a family that
exhibits a great deal of convergent and parallel
morphological evolution coupled with a low nu-
cleotide substitution rate, integration of multiple
data sets provides perhaps the greatest hope for
yielding a stable, phylogenetically-based classifi-
cation.
Monophyly of Lymania.The morphological
and combined data sets each support the mono-
phyly of Lymania, albeit with relatively low
support values (Figs. 1, 2). While there is support
for a monophyletic genus, no single morphological
character unambiguously unites the clade. The
presence of a furrowed or winged ovary, which
has been used to define the genus, is also present in
Aechmea carvalhoi (Fig. 1). Likewise, a bottleform
rosette is found in all Lymania species, but it has
been derived in Araeococcus parviflorus as well
(Fig. 1). This high degree of convergent and
parallel evolution illustrates a difficulty that
pervades Bromelioideae systematics and has con-
sistently hampered attempts to establish solid
generic circumscriptions.
Although we have not yet found a single di-
agnostic morphological character for Lymania, the
current data support maintenance of generic
status, as advocated by De Sousa (2004). Our study
included only six (of ca. 14) species from A. subg.
Lamprococcus, however, and the relationships be-
tween this subgenus and Lymania remain unclear.
It is possible that future studies that incorporate
additional species of A. subg. Lamprococcus could
result in modification of the circumscription of
Lymania.
Relationships among Lymania Species and
Distribution of Morphological Traits. The com-
bined analyses resulted in complete resolution
among Lymania species, with members of the
genus falling into two relatively well supported
clades (Figs. 1, 2). One clade (L. alvimii, L. spiculata,
L. azurea, and L. smithii) is united by the presence of
scape bracts that soon disintegrate, and a furrowed
(as opposed to winged) ovary (Fig. 1). The second
clade within the genus (L. brachycaulis, L. corallina,
and L. globosa) is united by flower length (ranging
from 2-3.5 cm), sepal apex acute to attenuate,
sepals carinate, and ovary winged. While each of
these characters is homoplasious within Brome-
lioideae as a whole, they are unambiguous
synapomorphies within Lymania.
While material of L. marantoides and the newly-
described L. languida (Leme 2006) was not available
for this study, the distribution of characters within
the Lymania clade may provide some insight
regarding the placement of these two species. It
has been hypothesized that each of these two
species is morphologically closely related to L.
corallina (Leme 2006; De Sousa 2004). The presence
of the alate-carinate ovary may unite them with the
winged ovary clade, which would support a rela-
tively close affinity to L. corallina. Other morpho-
logical characters, however, make the placement
within that clade more problematic. Members of
the winged ovary clade also share an acute sepal
apex. Lymania languida, however, is described as
having an obtuse sepal apex. Likewise, the scape
bract in L. languida is described as soon disinte-
grating, a character state that unites members of
the furrowed ovary clade. The presence of char-
acteristics from each of the two clades may suggest
a relatively basal position for L. languida within the
Lymania clade or hybrid origin. It is not possible to
hypothesize more about L. marantoides relation-
ships because information regarding several mor-
phological characters is not available.
ACKNOWLEDGEMENTS. We thank the Brazilian Forestry
Service (IBAMA) and the Brazilian Research Council (CNPq)
for the field collection permits; the numerous colleagues who
assisted us with plant collection; H. Luther, who provided
some of the plant material; A. P. G. Faria for assistance with
data analysis and for constructive comments; Matt Lavin and
three anonymous reviewers for helpful comments on the
manuscript; the United States National Science Foundation
(DEB-0129446, DEB-0129414) for funding; the Brazilian
Council for Graduate Studies (CAPES) for a research grant
2007] DE SOUSA ET AL.: PHYLOGENETIC OF LYMANIA 267
to L. O. F. De Sousa; and CNPq for a productivity grant to T.
Wendt. This paper is part of a Master’s thesis undertaken at
the Post-Graduate Program in Botany of the Universidade
Federal do Rio de Janeiro by the first author.
LITERATURE CITED
BARFUSS, M. H. J., R. SAMUEL,W.TILL, and T. F. STUESSY. 2005.
Phylogenetic relationships in subfamily Tillandsioideae
(Bromeliaceae) based on DNA sequence data from seven
plastid regions. American Journal of Botany 92: 337–351.
BROWN, G. K. and E. M. C. LEME. 2005. The re-establishment
of Andrea (Bromeliaceae: Bromelioideae), a monotypic
genus from Southeastern Brazil threatened with extinc-
tion. Taxon 54: 63–70.
——— and R. G. TERRY. 1992. Petal appendages in Brome-
liaceae. American Journal of Botany 79: 1051–1071.
CHASE, M. W., M. F. FAY, and V. SAVOLAINEN. 2000. Higher-
level classification in the Angiosperms: new insights
from the perspective of DNA sequence data. Taxon 49:
685–704.
CLARK, W. D., B. S. GAUT,M.R.DUVALL, and M. T. CLEGG.
1993. Phylogenetic relationships of the Bromeliiflorae -
Commeliniflorae - Zingiberiflorae complex of monocots
based on rbcL sequence comparisons. Annals of the
Missouri Botanical Garden 80: 987–998.
CRAYN, D. M., R. G. TERRY,J.A.C.SMITH, and K. WINTER.
2000. Molecular systematic investigations in Pitcairnioi-
deae (Bromeliaceae) as a basis for understanding the
evolution of crassulacean acid metabolism (CAM).
Pp. 569–579 in Monocots: Systematics and Evolution, eds.
K. L. Wilson and D. A. Morrison. Collingwood: CSIRO
Publishing.
CUNNINGHAM, C. W. 1997. Can three incongruence tests
predict when data should be combined? Molecular
Biology and Evolution 14: 733–740.
DEQUEIROZ, A. L., M. J. DONOGHUE, and J. KIM. 1995.
Separate versus combined analysis of phylogenetic
evidence. Annual Review of Ecology and Systematics 26:
657–681.
DESOUSA, L. O. F. 2004. Revisa
˜o taxono
ˆmica e filogenia do ge
ˆnero
Lymania Read (Bromeliaceae: Bromelioideae). M.S. thesis.
Rio de Janeiro: Universidade Federal do Rio de Janeiro.
DONOGHUE, M. J. and M. J. SANDERSON. 1992. The suitability
of molecular and morphological evidence in reconstruct-
ing plant phylogeny. Pp. 340–368 in Molecular systematics
of plants, eds. P. S. Soltis, D. E. Soltis, and J. J. Doyle. New
York: Chapman and Hall.
DOYLE, J. J. and J. L. DOYLE. 1987. A rapid DNA isolation
procedure for small quantities of fresh leaf tissue.
Phytochemical Bulletin 19: 11–15.
DUVALL, M. R., M. T. CLEGG,M.W.CHASE,W.D.CLARK,W.J.
KRESS,H.G.HILLS,L.E.EGUIARTE,J.F.SMITH,B.S.
GAUT,E.A.ZIMMER, and G. H. LEARN. 1993. Phyloge-
netic hypotheses for the monocotyledons constructed
from rbcL sequence data. Annals of the Missouri Botanical
Garden 80: 607–619.
DUVAL, M. F., G. S. C. BUSO,F.R.FERREIRA,J.L.NOYER,G.
COPPENS D’EECKENBRUGGE,P.HAMON, and M. E. FER-
REIRA. 2003. Relationships in Ananas and other related
genera using chloroplast DNA restriction site variation.
Genome 46: 990–1004.
FARIA, A. P. G., T. WENDT, and G. K. BROWN. 2004. Cladistic
relationships of Aechmea (Bromeliaceae: Bromelioideae)
and allied genera. Annals of the Missouri Botanical Garden
91: 303–319.
FARRIS, J. S., M. KA
¨LLERSJO
¨,A.G.KLUGE, and C. BULT. 1995.
Testing significance of incongruence. Cladistics 10:
315–319.
GAUT, B. S., S. V. MUSE,W.D.CLARK, and M. T. CLEGG. 1992.
Relative rates of nucleotide substitution at the rbcL locus
of monocotyledonous plants. Journal of Molecular Evolu-
tion 35: 292–303.
GILMARTIN, A. J. and G. K. BROWN. 1986. Cladistic tests of
hypotheses concerning evolution of xerophytes and
mesophytes within Tillandsia subgenus Phytarrhiza (Bro-
meliaceae). American Journal of Botany 73: 387–397.
——— and ———. 1987. Bromeliales, related monocots, and
resolution of relationships among Bromeliaceae subfa-
milies. Systematic Botany 12: 493–500.
GIVNISH, T. J., T. M. EVANS,J.C.PIRES, and K. J. SYTSMA. 1999.
Polyphyly and convergent morphological evolution in
Commelinales and Commelinidae: evidence from rbcL
sequence data. Molecular Phylogenetics and Evolution 12:
360–385.
———, K. C. MILLAM,T.M.EVANS,J.C.HALL,J.C.PIRES,P.E.
BERRY, and K. J. SYTSMA. 2004. Ancient vicariance or
recent long-distance dispersal? Inferences about phy-
logeny and south American-African disjunctions in
Rapateaceae and Bromeliaceae based on ndhF sequence
data. International Journal of Plant Sciences 165(4 Suppl.):
S35–S54.
GRANT, J. R. 1993. True tillandsias misplaced in Vriesea
(Bromeliaceae: Tillandsioideae). Phytologia 75: 170–175.
HORRES, R., G. ZIZKA,G.KAHL, and K. WEISING. 2000.
Molecular phylogenetics of Bromeliaceae: evidence from
trnL(UAA) intron sequences of chloroplast genome.
Plant Biology 2: 306–315.
HUELSENBECK, J. P., J. J. BULL, and C. W. CUNNINGHAM. 1996.
Combining data in phylogenetic analysis. Trends in
Ecology and Evolution 11: 152–158.
——— and F. RONQUIST. 2001. MRBAYES: Bayesian inference
of phylogeny. Bioinformatics 17: 754–755.
JOHNSON, L. A. and D. E. SOLTIS. 1995. Phylogenetic inference
in Saxifragaceae sensu stricto and Gilia (Polemoniaceae)
using matK sequences. Annals of the Missouri Botanical
Garden 82: 149–175.
LEME, E. M. C. 1987. Novas Bromelia
´ceas Nativas do Brasil.
Bradea 4: 392–404.
———. 2006. A new Lymania species from Bahia, Brazil.
Journal of the Bromeliad Society 56: 42–45.
——— and R. C. FORZZA. 2001. A new Lymania species from
Bahia, Brazil. Journal of the Bromeliad Society 51: 195–198.
MADDISON, D. R. and W. P. MADDISON. 2000. MacClade 4:
Analyses of Phylogeny and Character evolution. Version
4.0. Sunderland: Sinauer Associates.
MEZ, C. 1934–1935. Bromeliaceae. Pp. 1–667 in Das Pflanzen-
reich IV.32, ed. A. Engler. Leipzig: Wilhelm Engelmann.
NANDI, O. I., M. W. CHASE, and P. K. ENDRESS. 1998. A
combined cladistic analysis of angiosperms using rbcL
and nonmolecular data sets. Annals of the Missouri
Botanical Garden 85: 137–212.
POSADA, D. and K. A. CRANDALL. 1998. Modeltest: testing the
model of DNA substitution. Bioinformatics 14: 817–818.
RANKER, T. A., D. E. SOLTIS,P.S.SOLTIS, and A. J. GILMARTIN.
1990. Subfamilial phylogenetic relationships of the
Bromeliaceae: evidence from chloroplast DNA restric-
tion site variation. Systematic Botany 15: 425–434.
RONQUIST, F. and J. P. HUELSENBECK. 2003. MRBAYES 3:
Bayesian inference of phylogeny. Bioinformatics 19:
1572–1574.
READ, R. W. 1984. The ‘‘ Evolution’’ of a new genus Lymania
gen. nov. Journal of the Bromeliad Society 34: 199–201,
212–216.
SANG, T., D. J. CRAWFORD, and T. F. STUESSY. 1997. Chloroplast
DNA phylogeny, reticulate evolution, and biogeography
of Paeonia (Paeoniaceae). American Journal of Botany 84:
1120–1136.
268 SYSTEMATIC BOTANY [Volume 32
SCOTLAND, R. W., R. G. OLMSTEAD, and J. R. BENNETT. 2003.
Phylogeny reconstruction: the role of morphology.
Systematic Biology 52: 539–548.
SMITH, J. F., K. S. SYTSMA,J.S.SHOEMAKER, and R. L. SMITH.
1991. A qualitative comparison of total cellular DNA
extraction protocols. Phytochemical Bulletin 23: 2–9.
SMITH, L. B. and R. J. DOWNS. 1974. Pitcairnioideae (Brome-
liaceae). Pp. 1–658 in Flora Neotropica. Mon. 14, part 1.
New York: Hafner Press.
——— and ———. 1977. Tillandsioideae (Bromeliaceae).
Pp. 663–1492 in Flora Neotropica. Mon. 14, part 2. New
York: Hafner Press.
——— and ———. 1979. Bromelioideae (Bromeliaceae).
Pp. 1493–2142 in Flora Neotropica. Mon. 14. part 3. New
York: The New York Botanical Garden.
——— and M. A. SPENCER. 1992. Reduction of Streptocalyx
(Bromeliaceae: Bromelioideae). Phytologia 72: 96–98.
SWOFFORD, D. L. 2003. PAUP* Phylogenetic Analysis Using
Parsimony (* and other methods). Version 4. Sunder-
land: Sinauer Associates.
SYTSMA, K. J. and J. C. PIRES. 2001. Plant systematics in the
next 50 years: Re-mapping the new frontier. Taxon 50:
713–732.
TERRY, G. R., G. K. BROWN, and R. G. OLMSTEAD. 1997a.
Examination of subfamilial phylogeny in Bromeliaceae
using comparative sequencing of the plastid locus ndhF.
American Journal of Botany 84: 664–670.
———, ———, and ———. 1997b. Phylogenetic relationships
in subfamily Tillandsioideae (Bromeliaceae) using ndhF
sequences. Systematic Botany 22: 333–345.
VARADARAJAN, G. S. and A. J. GILMARTIN. 1988. Phylogenetic
relationships of groups of genera within the subfamily
Pitcairnioideae (Bromeliaceae). Systematic Botany 13:
283–293.
APPENDIX 1. Taxonomic sampling for combined morpho-
logical and molecular phylogenetic analysis of Lymania and
related species of Bromelioideae with Cryptanthus as the
outgroup. Subgenera assignments correspond to those
traditionally recognized (i.e., Smith and Downs 1979).
Voucher details are listed in the following sequences:
collection locality, collector name and number, and herbaria
at which the voucher is deposited (parentheses). GenBank
accession numbers are in the order matK,trnL-trnF,psbA-
trnH.
Aechmea subg. Aechmea Ruiz & Pav. 2A. lingulata (L.)
Baker: Brazil, Espı
´rito Santo, Faria 81 (RFA), EF110643,
EF110689, EF110667. Aechmea subg. Lamprococcus (Beer) Baker
2A. capixabae L.B.Sm.: Brazil, Espı
´rito Santo, Faria 90 (RFA),
EF110633, EF110680, EF110657. A. carvalhoi E. Pereira &
Leme: Brazil, Bahia, Leme 579 (RFA), EF110641, EF110687,
EF110665. A. miniata (Beer) Hort. ex Baker: Brazil, Bahia,
Sousa 436 (RFA), EF110638, EF110685, EF110662. A. pedicel-
lata Leme & H. Luther: Brazil, Espı
´rito Santo, in Selby
Garden collection, Bello s.n. (SEL, RFA), EF110629, EF110677,
EF110653. A. racinae L.B.Sm.: Brazil, Espı
´rito Santo, Faria 80
(RFA), EF110635, EF110682, EF110659. A. warasii E. Pereira:
Brazil, Espı
´rito Santo, Faria 76 (RFA), EF110634, EF110681,
EF110658. Aechmea subg. Macrochordion (deVriese) Baker 2A.
turbinocalyx Mez: Brazil, Bahia, Wendt 461 (RFA), EF110622,
EF110669, EF110645. Aechmea subg. Ortgiesia (Regel) Mez
2A. gamosepala Wittm: Brazil, Rio de Janeiro, Wendt 351
(RFA), EF110631, no sequence obtained, EF110655. A.gracilis
Lindm.: Brazil, Rio de Janeiro, Sousa 464 (RFA), EF110642,
EF110688, EF110666. Araeococcus subg. Araeococcus Brongn. 2
A. goeldianus L.B.Sm.: French Guiana, in Selby Garden
collection, Moonen s.n. (SEL), EF110623, EF110670, EF110646.
Araeococcus subg. Pseudaraeococcus Mez 2A. parviflorus
(Mart. ex Schult. & Schult.f.) Lindm.: Brazil, Bahia, Sousa 453
(RFA), EF110624, EF110671, EF110647. Cryptanthus Otto &
A.Dietr. 2C. beuckeri E. Morren: Brazil, Bahia, Wendt 489
(RFA), AF53965, EF110672, EF110648. C. bromelioides Otto &
A.Dietr : Brazil, Rio de Janeiro, Roc¸as s.n. (RFA), EF110637,
EF110684, EF110661. Lymania Read 2L. alvimii (L.B.Sm. &
Read) Read: Brazil, Bahia, Read s.n. (RFA), EF110625,
EF110673, EF110649. L. azurea Leme: Brazil, Bahia, Sousa
406 (RFA), EF110636, EF110683, EF110660. L. brachycaulis
(Baker) L.F. Sousa: Brazil, Bahia, Sousa 443 (RFA), EF110639,
no sequence obtained, EF110663. L. corallina (Brong. ex Beer)
Read: Brazil, Bahia, Sousa 458 (RFA), EF110640, EF110686,
EF110664. L. globosa Leme: Brazil, Bahia, Leme 2989 (RFA),
EF110626, EF110674, EF110650. L. smithii Read: Brazil, Bahia,
Sousa 407 (RFA), EF110627, EF110675, EF110651. L. spiculata
Leme & Forzza: Brazil, Bahia, Sousa 416 & 417 (RFA),
EF110628, EF110676, EF110652. Ronnbergia E.Morren & Andre
´
2R. brasiliensis E. Pereira & I. A. Penna: Brazil, Bahia, Leme
3017 (RFA), EF110644, EF110690, EF110668. R. neoregelioides
Leme: Brazil, Bahia, Leme 4368 (SEL), EF110632, EF110679,
EF110656.
APPENDIX 2. Morphological characters used in the cladistic
analyses of Lymania (Bromeliaceae).
1. Rosette shape: 05broadly infundibuliform; 1 5
narrowly infundibuliform; 2 5lageniform; 3 5flattened. 2.
Number of leaves per rosette: 05,10; 1 510 to 15; 2 5.
15. 3. Rosette forming a water impounding tank: 05yes; 1
5no. 4. Habit: 05primarily epiphytic; 1 5primarily
terrestrial or saxicolous; 2 5facultative. 5. Stolon formation:
05primarily from between leaves; 1 5primarily from the
base of the rosette. 6. Form of clonal growth: 05elliptical
cluster; 1 5linear cluster; 2 5pendent stolon; 3 5climbing
stolon. 7. Mature leaf length (include blade and sheath, cm):
05#45; 1 5.45 to 65; 2 5$65. 8. Width of leaf blade
(cm): 05,2.5; 1 5$2.5 to 5; 2 5.5. 9. Leaf blade base
and sheath: 05not or slightly differentiated; 1 5distinctly
differentiated but not petiolate; 2 5petiolate between blade
and sheath. 10. Leaf blade channeling: 05absent; 1 5
present. 11. Shape of the leaf apex: 05acute or attenuate; 1
5rounded or obtuse. 12. Leaf margin armament: 05none
and entire, to obscurely serrate; 1 5serrate. 13. Scape: 05
erect and hidden in the center of the rosette; 1 5erect or
slightly curved and exceeding the rosette; 2 5pendulous and
exceeding the rosette. 14. Scape bracts present: 05yes,
persistent; 1 5yes, soon disintegrating; 2 5no. 15.
Distribution of scape bracts: 05erect and imbricate,
covering the scape; 1 5erect, not imbricate, the scape
exposed; 2 5spreading, the scape exposed. 16. Scape bract
margin: 05entire; 1 5armed. 17. Scape bract color: 05
greenish or pale (tan); 1 5rose or red. 18. Scape bract apex:
05acute to attenuate; 1 5rounded to obtuse. 19.
Inflorescence length (cm): 05#10; 1 5.10 to 20; 2 5.
20. 20. Inflorescence and scape position in relation to the
leaves: 05nidular, not exceeding the leaf-sheaths; 1 5
exceeding the leaf-sheaths, but not exceeding the leaf-blades;
25completely exceeding the leaf-blades. 21. Inflorescence
branching: 05none, simple, never branched; 1 5simple or
rarely branched at the base; 2 5bipinnate; 3 5tripinate or
more; 4 5corymbiform. 22. Flowers congested on rachis or
inflorescence branches: 05no, rachis visible; 1 5yes, rachis
not visible. 23. Flower arrangement along the rachis or
inflorescence branches: 05polystichous; 1 5distichous; 2
5in fascicles; 3 5polystichous on main axis, and distichous
on the branches. 24. Primary bracts: 05absent; 1 5present.
25. Primary bract margins: 05armed; 1 5entire. 26.
Primary bract color: 05greenish or pale; 1 5rose or red. 27.
Primary bract texture: 05coriaceous; 1 5membranaceous
2007] DE SOUSA ET AL.: PHYLOGENETIC OF LYMANIA 269
or chartaceous. 28. Primary bract apex: 05acute to long
attenuate; 1 5rounded to obtuse. 29. Floral bracts: 05
present; 1 5absent or inconspicuous. 30. Floral bract
margins: 05entire; 1 5armed. 31. Floral bract color: 05
greenish, pale or tan; 1 5rose or red; 2 5blue, lilac or purple.
32. Floral bract apex, excluding the terminal spine: 05
rounded to obtuse; 1 5attenuate to acute. 33. Terminal spine
of floral bract apex: 05very short and inconspicuous; 1 5
conspicuous, but shorter than the bract; 2 5equaling or
longer than the bract. 34. Floral bract shape (length/width
ratio): 05broadly ovate (1:1); 1 5narrow ovate (3:1); 2 5
ovate (2:1). 35. Floral bracts carinate: 05no; 1 5yes, one
keel. 36. Flower length (excluding pedicel, cm): 05,0.5; 1
5$0.5 to 2; 2 5.2 to 3.5; 3 5.3.5 cm to 5; 4 5.5. 37.
Pedicel: 05absent; 1 5present. 38. Sepal color: 05white,
greenish or tan; 1 5blue, lilac, or purple; 2 5red or rose; 3 5
yellow or orange; 4 5distinctly bicolored. 39. Sepal apex,
excluding terminal spine: 05rounded to obtuse; 1 5acute
to attenuate. 40. Sepal terminal spine: 05absent or
inconspicuous; 1 5present, but shorter than the sepal; 2 5
equaling or exceeding the sepal. 41. Sepal symmetry: 05
symmetric or slight asymmetric; 1 5strongly asymmetric. 42.
Sepals carinate: 05yes; 1 5no. 43. Sepal connation: 05
free or mostly free; 1 5connate by 1/3 or more of its length.
44. Petal color: 05white, cream or greenish; 1 5blue, lilac or
purple; 2 5yellow or orange. 45. Petal apex: 05rounded to
obtuse; 1 5acute to attenuate. 46. Petal shape: 05linear,
margins parallel or nearly so; 1 5spathulate, with distinct
blade and claw. 47. Petal appendages: 05absent; 1 5pouch
or sac-like; 2 5ligulate with conspicuous fringed margins; 3
5ligulate with entire margins. 48. Petal with lateral folds on
the adaxial surface: 05absent; 1 5present. 49. Petal
orientation at anthesis: 05erect or cucullate; 1 5spreading
or recurved. 50. Petal connation: 05free; 1 5connate. 51.
Filament length (cm): 05,0.5; 1 5$0.5 to ,1; 2 5$1to
,2; 3 5$2. 52. Filament shape: 05complanate; 1 5
filiform. 53. Filament fusion: 05none, filaments free; 1 5
yes, antepetalous filaments adnate to the petals; 2 5yes,
filaments connate, forming a tube. 54. Anther length (cm):
05,0.2; 1 5$0.2 to #0.5; 2 5.0.5. 55. Ovary shape in
cross-section: 05circular; 1 5trigonous. 56. Ovary wall
ornamentation: 05none, smooth and regular; 1 5furrowed;
25winged; 3 5verrucose; 4 5pilose. 57. Ovary color: 05
white or green; 1 5yellow or orange; 2 5rose or red. 58.
Stigma position in relation to anthers at anthesis: 05
equaling or slightly shorter than the anthers; 1 5exceeding
the anthers. 59. Ovules appendaged: 05yes, caudate; 1 5
no. 60. Ovule number per locule: 05.5; 1 5#5.
APPENDIX 3. MATRIX of morphological character states (from Appendix 2) used in a phylogenetic analysis of Lymania and
related species of Bromelioideae with Cryptanthus as the outgroup. Inapplicable characters or missing data are indicated with
a dash.
Character number 10 20 30 40 50 60
Aechmea capixabae 1102111111 0110000012 2030 ----1- -----10300 1111012000 1011031000
Aechmea carvalhoi 1000120111 1110201012 203111101- -----10000 1101011000 0011010001
Aechmea gamosepala 12021 -0100 1010100012 0100----00 0111010202 1111002100 10110020 -0
Aechmea gracilis 1102100110 1110000011 2001111000 1112010202 1111012100 1011002010
Aechmea lingulata 0202102200 1110010022 2001101000 0121010002 1110110110 0000000000
Aechmea miniata 1002111111 1110100002 201111101- -----10400 1101012000 1001032000
Aechmea pedicellata 110 -100011 0110100001 0000----1- -----11300 1100103000 1001041100
Aechmea racinae 1002110111 0120000002 0000----1- -----21200 1102002000 1011032000
Aechmea turbinocalyx 0102101010 1010100001 2001101000 0102010001 1100110110 0011000000
Aechmea warasii 01001-1111 1120000002 1000 ----1- -----21100 1101002000 1012032000
Araeococcus
goeldianus 1002100110 0110000012 3001101001 0012000001 0100000000 -000000 -10
Araeococcus
parviflorus 2000130110 0010100012 3001111000 0111001001 0100010010 0000000110
Cryptanthus beuckeri 3011000220 0102- -- -00 2121000000 0112130010 0010100111 3121100110
Cryptanthus
bromelioides 3111000010 0102-- - -00 4121001000 0111020011 0010000010 2101100111
Lymania alvimii 2000130111 1111100012 2001101000 0000010000 1110010111 1011010000
Lymania azurea 2000130111 1111100012 200110101- -----10000 1110010110 1011010100
Lymania brachycaulis 2000130111 0100001100 2101111100 1101020210 1010010100 1011022100
Lymania corallina 2000130111 0010001101 200111111- -----20210 1010010100 1011022100
Lymania globosa 2000130111 0000100011 200110101- -----20010 1010010111 1011020100
Lymania smithii 2000130111 0011100012 200110101- -----10000 1110010111 0011010100
Lymania spiculata 2000130111 1111100012 2001101000 0000010000 1110010111 1011010000
Ronnbergia
brasiliensis 1202101011 0010001002 2001111000 2110010101 1101100010 1011001010
Ronnbergia
neoregelioides 1000130010 0000000000 0100----00 0111020001 1110110111 2011000110
270 SYSTEMATIC BOTANY [Volume 32
