Content uploaded by Rafaela Forzza
Author content
All content in this area was uploaded by Rafaela Forzza on Jan 29, 2016
Content may be subject to copyright.
1 23
Plant Systematics and Evolution
ISSN 0378-2697
Plant Syst Evol
DOI 10.1007/s00606-012-0590-z
Leaf anatomy of Quesnelia (Bromeliaceae):
implications for the systematics of core
bromelioids
André Mantovani, Anna Karla Lima
da Venda, Valquíria Rezende Almeida,
Andrea Ferreira da Costa & Rafaela
Campostrini Forzza
1 23
Your article is protected by copyright and
all rights are held exclusively by Springer-
Verlag. This e-offprint is for personal use only
and shall not be self-archived in electronic
repositories. If you wish to self-archive your
work, please use the accepted author’s
version for posting to your own website or
your institution’s repository. You may further
deposit the accepted author’s version on a
funder’s repository at a funder’s request,
provided it is not made publicly available until
12 months after publication.
ORIGINAL ARTICLE
Leaf anatomy of Quesnelia (Bromeliaceae): implications
for the systematics of core bromelioids
Andre
´Mantovani •Anna Karla Lima da Venda •
Valquı
´ria Rezende Almeida •Andrea Ferreira da Costa •
Rafaela Campostrini Forzza
Received: 25 July 2011 / Accepted: 5 January 2012
ÓSpringer-Verlag 2012
Abstract Quesnelia is an endemic genus of the Atlantic
Forest in Brazil. It features 21 species and three varieties
that are distributed from Santa Catarina to Bahia, with
diversity centers in the Rio de Janeiro coastal region and
the rainforest of southern Bahia. It is divided into the
subgenera Quesnelia subg. Quesnelia and Quesnelia subg.
Billbergiopsis. In this study, the leaf anatomy of all species
of Quesnelia is characterized and compared by multivariate
analysis to determine whether leaf anatomy confirms this
subgeneric division. The results demonstrate that leaf
anatomy supports the existence of three distinct groups of
species now classified under the genus Quesnelia. When
compared to other species, the first group, which is repre-
sented by five Billbergiopsis taxa, is characterized by dis-
tinct anatomical arrangement, where the stomata are
positioned at the same level as the epidermis, the water
storage tissue is poorly developed, and extra-fascicular
fiber strands are distributed throughout the mesophyll. The
remaining groups support the subgenera Quesnelia and
Billbergiopsis, which differ basically in terms of the con-
tour of the leaf in transverse sections, size and cell type of
the adaxial water storage tissue, and the presence of extra-
fascicular fiber strands. Comparing with anatomical data
available in the literature for Bromelioideae, these results
indicate the similarity of Quesnelia with Aechmea,Cani-
strum and Billbergia, which corroborates morphological
and molecular phylogenies, and thus support future taxo-
nomic circumscriptions of these important genera from the
core Bromelioideae.
Keywords Core bromelioids Aechmea Billbergia
Canistrum
Introduction
The Bromeliaceae family is a highly diversified group in
terms of morphology, physiology, habit, habitat and animal-
plant relationships. Bromeliads also provide an important
source of food, fiber and ornamentals [e.g., Ananas
comosus (L.) Merr, Neoglaziovia variegata (Arr.Cam.) Mez
and Aechmea fasciata (Lindl.) Baker] (Benzing 2000).
Given the ecological, economic and cultural significance of
Bromeliaceae, recent phylogenetic studies have aimed at
elucidating the evolutionary steps that led to its diversifica-
tion throughout South and Central America (Crayn et al.
2000,2004; Barfuss et al. 2005; Givnish et al. 2007,2011;
Horres et al. 2007; Schulte et al. 2005,2009; Rex et al. 2009;
Reinert et al. 2003). These studies revealed that several
genera within the family are poorly circumscribed and do not
represent monophyletic groups, especially within the core
bromelioids (Sass and Spech 2010; Schulte et al. 2005,2009).
Quesnelia is placed within the subfamily Bromelioideae
(Schulte and Zizka 2008; Schulte et al. 2009), which now
encompasses 21 species endemic to the Atlantic Forest,
mainly distributed in the Rio de Janeiro coastal region and
Bahia (Martinelli et al. 2008; Forzza et al. 2010). The
genus is traditionally characterized by simple inflores-
cence, unarmed sepals and biporate pollen grains (Baker
1889; Mez 1896; Smith and Downs 1979). Mez (1896)
divided Quesnelia into three subgenera (Quesnelia,
A. Mantovani A. K. L. da Venda R. C. Forzza (&)
Jardim Bota
ˆnico do Rio de Janeiro, Pacheco Lea
˜o 915,
Bota
ˆnico, Rio de Janeiro 22460-030, Brazil
e-mail: rafaela@jbrj.gov.br
V. R. Almeida A. F. da Costa
Museu Nacional, Universidade Federal do Rio de Janeiro,
Quinta da Boa Vista, Sa
˜o Cristo
´va
˜o,
Rio de Janeiro 20940-040, Brazil
123
Plant Syst Evol
DOI 10.1007/s00606-012-0590-z
Author's personal copy
Wawrae and Billbergiopsis). Smith and Downs (1979)
considered only two subgenera, Quesnelia subg. Quesnelia
and Quesnelia subg. Billbergiopsis, including the subgenus
Wawrae, which is differentiated by the presence of simple,
strobilate, ellipsoid or cylindrical inflorescence; subligu-
late, broadly acute to truncate floral bracts; sepals
8–10 mm long; ovary slightly, if at all, costate in the
subgenus Quesnelia and simple or compound, dense or lax
inflorescence; ovate or lanceolate, acute or acuminate floral
bracts; sepals mostly more than 10 mm long; ovary often
costate in the subgenus Billbergiopsis.
In an attempt to elucidate generic and infrageneric rela-
tionships in Quesnelia, Almeida et al. (2009) developed a
morphological phylogeny. In this paper, the authors confirm
Faria et al. (2004), where Quesnelia species emerge as
polyphyletic, indicating that species of the Quesnelia sub-
genus Quesnelia form a clade, while this is not the case for
the Quesnelia subgenus Billbergiopsis. Almeida et al. (2009)
indicated that biporated pollen grains occur in all species of
the genus Quesnelia (except for Q. kautskyi), as well as
unarmed sepals, probably appearing independently in dis-
tinct lineages. Only two molecular phylogenetic analyses
included Quesnelia among the studied species (Schulte et al.
2009; Sass and Spech 2010). Quesnelia again emerges as
polyphyletic, with its species related to Canistrum and
Aechmea, rather than intimately related to Billbergia,as
suggested by the morphological phylogenies of Faria et al.
(2004) and Almeida et al. (2009).
Anatomical features are considered potentially useful to
support taxonomic circumscriptions within Bromeliaceae
(Robinson 1969;Sajoetal.1998; Taylor and Robinson
1999; Aoyama and Sajo 2003; Arruda and Costa 2003;
Sousa et al. 2005;Monteiroetal.2011). Horres et al. (2007)
specifically used anatomical features to differentiate sub-
genera within Aechmea, showing that the arrangement of
tissues at the leaf level creates different anatomical types
that corresponded to genetic groups found in an AFLP
analysis, and were therefore regarded as potentially valuable
for future systematic revisions for this and related genera.
This study describes the anatomy of all the species of
Quesnelia, with the objective of evaluating whether anatomical
differences support the actual taxonomic division into two
subgenera. Based on the arrangement of tissues at leaf level,
results indicate that three, rather than two, groups exist, a dis-
covery which could be relevant for future taxonomic circum-
scriptions of important genera within the core Bromelioideae,
such as Quesnelia,Canistrum,Aechmea and Billbergia.
Materials and methods
In this study, all 21 species currently recognized in the
genus Quesnelia were studied. Altogether, 44 accessions
were investigated: 11 taxa were represented by 3
accessions each, 1 taxon by 2 accessions, and 9 taxa
were represented by a single accession. The studied
species are listed in Table 1, along with their respective
distribution by Brazilian state and habitat. The samples
were collected in the field and in living collections, and
others were fixed in ethanol 70% or obtained from her-
barium material. In this case, the material was hydrated
in heated water-glycerin solution (9:1, v/v) for 1 h
before fixation.
Cross-sections of the middle third of the leaf blade were
obtained through Campden 752 M vibratome or embed-
ment in hydroxyethyl methacrylate resin (Mantovani et al.
2010). For vibratome sections, 5 ml agarose solution (6%
weight/volume in distilled water) was heated until it
became translucent and then placed in plastic vials (1 ml
volume capacity) with 0.15 cm
2
leaf samples (Ruzin 1992).
After the solution had cooled down and hardened, blocks
were unmolded, and cross-sections, around 5 lmin
thickness, were obtained, cleared in sodium hypochlorite,
stained with an aqueous solution of 0.5% Safranin-0.5%
Toluidine blue (4:6, v/v) and mounted on slides with 50%
glycerin (Mantovani et al. 2010). For embedding in resin,
2.0-cm
2
squares were fixed in a solution of 4% glutaral-
dehyde and 1% formaldehyde in a sodium phosphate buffer
(0.1 M, pH 7.2), dehydrated in ethylic series and included
in hydroxyethyl methacrylate (Gerrits and Smid 1983).
Sections 2–4 lm in thickness were cut with a Spencer
microtome and stained with Toluidine blue O (O’Brien and
Mccully 1981).
Digital photomicrographs were taken using a Photo-
metrics Coolsnap-Pro digital camera coupled with an
Olympus BX-50 optical microscope. Cell and tissue clas-
sification followed Tomlinson (1969). For the purposes of
comparative analysis among species, 21 anatomical char-
acters were established (Table 2).
To evaluate the similarity among species of Quesnelia
and their respective inclusions in the subgenera Quesnelia
and Billbergiopsis, a multivariate analysis was applied.
Initially, a similarity matrix was calculated for the quali-
tative data from Table 2through the PCord 5 program
(MCune and Mefford 1999), using Euclidean distance as
the coefficient. Subsequently, a multidimensional scaling
analysis (MDS) was applied, restricting distribution to a
two-dimensional space axis. Analysis of similarity (ANO-
SIM) was applied to determine the existence of potential
differences among groups of species based on the global R,
using p\0.01. When significant differences were detected,
a similarity percentage analysis (SIMPER) was employed to
evaluate the main anatomical features responsible for dif-
ferentiation. ANOSIM and SIMPER analysis were per-
formed using the Primer 5.0 program (Clarke and Gorley
2011).
A. Mantovani et al.
123
Author's personal copy
Results
Quesnelia species can be found in different vegetational
physiognomies of the Atlantic Forest (sandy coastal plains,
mangroves, tropical rain forests and seasonal forests), and
they may also present distinct life forms, such as terrestrial,
epiphytic or rupicolous plants (Table 1). Through ana-
tomical analysis of multiple accessions for 12 of the 21
species investigated, intraspecific variations were detected
in 7 out of 12 taxa (Table 2). Quesnelia arvensis
showed intraspecific variation in 6 of 21 characters, while
Q. lateralis showed intraspecific variation in 3 characters,
and Q. quesneliana and Q. marmorata showed variation
in 2 characters. Quesnelia edmundoi var. edmundoi,
Q. indecora and Q. testudo showed variation in one char-
acter each. It is worth noting that the variable characters
were not the same for the species cited above.
The contour of the adaxial and abaxial surfaces was
smooth to slightly undulate in all species studied (Fig. 1),
except for Q. alvimii,Q. arvensis,Q. quesneliana,Q. testudo
(adaxial) and Q. marmorata (abaxial), which presented rib-
bed surfaces (Fig. 1d). Intraspecific variation occurred for
the adaxial surface of Q. quesneliana and both surfaces of
Q. arvensis, slightly undulate in two of three specimens of
both species and ribbed surfaces in the remainder.
For the thickening of the epidermal cells of the adaxial
surface, only Q. edmundoi var. edmundoi,Q. edmundoi
var. rubrobracteata,Q. humilis and Q. imbricata showed
slightly thickened cell walls, delimiting a wide lumen. For
the abaxial surface, only Q. edmundoi var. edmundoi and
Q. edmundoi var. rubrobracteata exhibited this character-
istic (Fig. 1c). In most species studied, both surfaces had
epidermal cells with anticlinal and strongly thickened inner
periclinal walls and reduced lumen (Fig. 1a, b). Variability
in this trait was observed in Q. marmorata, where two of
the three specimens analyzed presented an abaxial surface
composed of cells with slightly thickened cell walls and a
wide lumen instead.
The stomata in cross-sectional view in Bromeliaceae can
be classified according to their position relative to the level
Table 1 Systematic leaf anatomy of Quesnelia
AG Distribution Habitat Habit
SC PR SP MG RJ ES BA R RF SF M T E R
Q. arvensis (Vell.) Mez Q X X X X X X X X X X
Q. conquistensis Leme Q X X X X
Q. quesneliana (Brongn.) L.B.Sm. Q X X X X X X X X X X
Q. testudo Lindm. Q X X X X X X
Q. clavata Amorim & Leme A X X X
Q. dubia Leme A X X X
Q. edmundoi L.B.Sm. var. edmundoi AXXX
Q. edmundoi var. rubrobracteata E.Pereira A X X X
Q. koltesii Amorim & Leme A X X X
Q. alvimii Leme B X X X
Q. augusto-coburgii Wawra B X X X X X
Q. humilis Mez B X X X X X X
Q. imbricata L.B.Sm. B X X X X X X
Q. indecora Mez B X X X X X
Q. kautskyi C.M.Vieira B X X X X
Q. lateralis Wawra B X X X X X
Q. liboniana (De Jongle) Mez B X X X X X X
Q. marmorata (Lem.) R.W.Read B X X X X X X
Q. seideliana L.B.Sm. B X X X X
Q. strobilispica Wawra B X X X X X X
Q. tubifolia Leme & L.Kollmann B X X X
Q. violacea Wand. & S.L.Proenc¸a B X X X X
Leaf anatomy, distribution, habitat and habit of Quesnelia species. Anatomical groups (Q, B, A see text); Distribution, habitat and habit of
Quesnelia species. Distribution by State of occurrence (RJ Rio de Janeiro, SP Sa
˜o Paulo, MG Minas Gerais, BA Bahia, ES Espı
´rito Santo, SC
Santa Catarina, PR Parana
´); Habitat [Rsandy coastal plain (restinga), RF rainforest, SF seasonal forest, Mmangrove]; habit (Tterrestrial,
Eepiphyte, Rrupicolous). Data gathered from Vieira (1999,2006), Amorim and Leme (2009), Leme (2008), Wanderley and Proenc¸a (2006),
Forzza et al. (2010) and Leme and Kollmann (2011)
Leaf anatomy of Quesnelia
123
Author's personal copy
Table 2 Systematic leaf anatomy of Quesnelia
VOUCHER A B C D E F G H I J K L M N O P Q R S T U
Q. alvimi Seidel 1132 1 1 0 0 1 1 0 0 0 1 1 0 1 0 1 0 NA 1 0 1 1
Q. arvensis Martinelli 15031 1100010110111 1 10 NA0 110
Q. arvensis Silva 7 0000010000011 1 0NA NA NA 1 1 0
Q. arvensis Vieira 1234 0000010000011 1 0NA NA NA 1 1 0
Q. augusto-coburgii Leme 3346 0 0 0 0 0 1 1 0 0 1 1 1 1 0 1 0 NA 0 1 1 0
Q. augusto-coburgii Martinelli 15515 0 0 0 0 0 1 1 0 0 1 1 1 1 0 1 0 NA 0 1 1 0
Q. augusto-coburgii Sucre 3084 0 0 0 0 0 1 1 0 0 1 1 1 1 0 1 0 NA 0 1 1 0
Q. clavata Jardim 5440 0 0 0 0 1 0 0 0 0 0 0 0 1 1 1 1 1 0 1 1 0
Q. dubia Leme 5152 0 0 0 0 1 0 0 0 0 0 0 0 1 0 1 1 0 0 1 1 0
Q. conquistensis Leme 5667 0 1 0 0 0 1 0 0 0 0 0 1 1 1 0 NA NA NA 1 1 1
Q. edmundoi var.
edmundoi
Silva 126 0 0 1 1 1 0 0 0 0 0 0 0 1 0 1 1 1 1111
Q. edmundoi var.
edmundoi
Vieira 873A 0 0 1 1 1 0 0 0 0 0 0 0 1 0 1 1 1 0111
Q. edmundoi var.
edmundoi
Vieira 873 0 0 1 1 1 0 0 0 0 0 0 0 1 0 1 1 1 0111
Q. edmundoi var.
rubrobracteata
Martinelli 15747 0 0 1 1 1 0 0 0 0 0 0 0 1 0 1 1 1 1 1 1 1
Q. edmundoi var.
rubrobracteata
Vieira 966 0 0 1 1 1 0 0 0 0 0 0 0 1 0 1 1 1 1 1 1 1
Q. edmundoi var.
rubrobracteata
Martinelli 8500 0 0 1 1 1 0 0 0 0 0 0 0 1 0 1 1 1 1 1 1 1
Q. humilis Leme 3473 0 0 1 0 1 1 0 0 0 1 1 1 1 0 1 0 NA 0 0 1 0
Q. humilis Ferreira 506 0 0 1 0 1 1 0 0 0 1 1 1 1 0 1 0 NA 0 0 1 0
Q. humilis Custodio Filho 2268 0 0 1 0 1 1 0 0 0 1 1 1 1 0 1 0 NA 0 0 1 0
Q. imbricata Leme 1661 0 0 1 0 1 1 0 0 0 1 1 1 0 NA 0 NA NA NA 0 1 0
Q. indecora Leme 1543 0 0 0 0 0 1 1 0 01111 0 10 NA0 110
Q. indecora Abreu s/n (CESJ
44034)
0000011011111 0 10 NA0 110
Q. indecora Alves 4268 0 0 0 0 0 1 1 0 01111 0 10 NA0 110
Q. kautskyi Martinelli 15725 0 0 0 0 0 1 1 1 1 1 1 1 1 0 1 0 0 0 0 1 0
Q. kautskyi Martinelli 15184 0 0 0 0 0 1 1 1 1 1 1 1 1 0 1 0 0 0 0 1 0
Q. koltesii Amorim 5443 0 0 0 0 0 0 0 0 0 1 0 0 1 1 1 1 0 0 1 1 0
Q. lateralis Bocayuva 129 0 0 0 0 0 1 1 111111 1 10 NA0 110
Q. lateralis Vieira 1065 0 0 0 0 0 1 1 001111 1 10 NA0 110
Q. lateralis Pessoa 177 0 0 0 0 0 1 1 001111 1 10 NA0 110
Q. liboniana Almeida 43 0 0 0 0 0 1 0 0 0 0 1 1 1 0 1 0 NA 1 1 1 0
Q. liboniana Marquete 655 0 0 0 0 0 1 0 0 0 0 1 1 1 0 1 0 NA 1 1 1 0
Q. liboniana Vieira 1210 0 0 0 0 0 1 0 0 0 0 1 1 1 0 1 0 NA 1 1 1 0
Q. marmorata Martinelli 15600 0 1 0 0010001111 0 10 NA1 010
Q. marmorata Nadruz 478 0 1 0 1010001111 0 10 NA1 010
Q. marmorata Giordano 2105 0 1 0 1010001111 0 10 NA1 011
Q. quesneliana Almeida 8 11000100000 11 1 10NA0111
Q. quesneliana Silva 460 0100010000011 1 0NA NA NA 1 1 1
Q. quesneliana Gurken 1545 01000100 00011 1 0NA NA NA 1 1 0
Q. seideliana Leme 3284 0 0 0 0 0 1 0 0 0 1 1 1 1 0 1 0 NA 1 1 1 0
Q. strobilispica Leme 946 0 0 0 0 0 1 0 0 0 1 1 1 1 0 1 0 NA 0 1 1 0
Q. strobilispica Costa 409 0 0 0 0 0 1 0 0 0 1 1 1 1 0 1 0 NA 0 1 1 0
Q. strobilispica Leme 3560 0 0 0 0 0 1 0 0 0 1 1 1 1 0 1 0 NA 0 1 1 0
A. Mantovani et al.
123
Author's personal copy
of the epidermis and the presence or absence of occluded
substomatal chambers. Stomata at the same level as the
epidermal cells are found in Q. kautskyi,Q. edmundoi var.
rubrobracteata,Q. edmundoi var. edmundoi,Q. dubia,
Q. clavata and Q. koltesii (Fig. 2a). In the remaining
species, the stomata are sunken, positioned below the epi-
dermis (Fig. 2b). No stomata were found above the epider-
mis in Quesnelia. Only Q. augusto-coburguii,Q. indecora
and Q. lateralis (Fig. 2b) had thickened mesophyll cells
surrounding and occluding substomatal chambers, while for
the remaining species, substomatal chambers were occluded
by the enlargement of abaxial hypodermal cells, forming
lobes that reach into the stoma. For these characters, no
intraspecific variability was detected.
In Bromeliaceae, the peltate scales are attached to the
epidermis by one or more cells arranged in a uniseriate
structure, termed a stalk. The number of stalk cells varied
among Quesnelia species. The stalk is composed of more
than two cells in Q. alvimii,Q. dubia,Q. edmundoi var.
edmundoi,Q. edmundoi var. rubrobracteata,Q. humilis,
Q. imbricata,Q. koltesii and Q. clavata (Fig. 2c), while it
is composed of only two cells in the other species (Fig. 2d).
No intraspecific variability was observed for this character.
The leaf hypodermis in Bromeliaceae can be differen-
tiated in mechanical and water storage tissues. In the
majority of the Quesnelia species, the mechanical
hypodermis is composed of only one layer of sclerified
cells, with the lumen ranging from small to large (Fig. 1a,
c). However, in a few species, the mechanical hypodermis
was composed of several layers of highly sclerified cells
with greatly thickened walls and reduced lumen. Such
features occurred on the adaxial and abaxial surfaces of
Q. lateralis and Q. arvensis, and on the abaxial surface of
Q. kautskyi (Fig. 1b). In these three species, there is often a
transitional cell layer internal to the sclerified hypodermis.
This layer is composed of cells with less thickened walls
and a larger lumen, when compared to the mechanical
hypodermis above, but clearly more sclerified than the
underlying mesophyll cells below (Fig. 1a, c). Intraspecific
variation is noted for the adaxial and abaxial mechanical
hypodermis of two specimens of Q. arvensis and Q. late-
ralis, and for the abaxial mechanical hypodermis of
Q. indecora, which consisted of only one layer of strongly
sclerified cells, followed by a transitional cell layer.
The water storage tissue is formed of rounded or
anticlinally elongated cells. When less developed, this
tissue occupies up to one-third of the leaf blade in trans-
verse section, reaching up to a half in the more developed
state. Moreover, the water storage tissue is formed of
rounded parenchymatic cells in Q. dubia,Q. clavata,
Q. edmundoi var. edmundoi,Q. edmundoi var. rubrob-
racteata,Q. koltesii,Q. quesneliana and Q. testudo
Table 2 continued
VOUCHER A B C D E F G H I J K L M N O P Q R S T U
Q. testudo Martinelli 15897 0 1 0 0 0 1 0 0 0 0 11 1 1 0 NA NA NA 1 1 0
Q. testudo Vieira 903 0 1 0 0 0 1 0 0 0 0 01 1 1 0 NA NA NA 1 1 0
Q. testudo Mello-Silva s/n (RB
270279)
0100010000011 1 0NANANA110
Q. tubifolia Leme 8210 0 1 0 0 1 1 0 0 0 1 1 0 1 0 0 NA NA NA 0 1 1
Q. violacea Martinelli 15777 0 0 0 0 0 1 0 0 0 1 1 1 0 0 0 NA NA NA 0 1 0
The distribution of anatomical characters is shown for 1–3 specimens per species, based on presence (1) and absence (0), when not specified. The
impossibility of application is shown by ‘‘NA’’ (not applicable). Data in bold indicate intraspecific variation. Characters are (A) adaxial surface
contour: (0) smooth to slightly undulate and (1) ribbed; (B) abaxial surface contour: (0) smooth to slightly undulate and (1) ribbed; (C) wall
thickening of the adaxial epidermal cell: (0) inner wall markedly thickened, cell lumen almost occluded, more or less U-shaped in transverse
section and (1) slightly thickened walls delimiting a wide lumen; (D) wall thickening of the abaxial epidermal cell: (0) inner wall markedly
thickened, cell lumen almost occluded, more or less U-shaped in transverse section and (1) slightly thickened walls delimiting a wide lumen;
(E) number of stalk cells of peltate trichomes: (0) two and (1) more than two; (F) stomata position in transverse section, in relation to the
epidermis: (0) superficial and (1) sunken; (G) type of modified substomatal cells occluding stomatal chamber: (0) thickened mesophyll cells
surrounding substomatal chamber and (1) abaxial hypodermal cells forming lobes that reach into the stoma; (H) adaxial sclerenchymatous
hypodermis: (0) composed of one layer of slightly thickened cells and (1) composed of one to several layers of sclerenchymatic thick-walled
cells; (I) abaxial sclerenchymatous hypodermis: (0) composed of one layer of slightly thickened cells and (1) composed of one to several layers
of sclerenchymatic thick-walled cells; (J) adaxial water storage tissue, size: (0) up to 1/3 of the leaf blade in transverse section and (1) larger than
1/3 of the leaf blade in transverse section; (K) adaxial water storage tissue, cell shape: (0) rounded and (1) anticlinally elongated; (L) transition
between the water storage tissue and the chlorenchyma: (0) absent and (1) present; (M) air-lacunae chlorenchyma: (0) absent and (1) present;
(N) air-lacunae chlorenchyma, cell form: (0) short arms, enclosing less developed intercellular lacunae, and (1) large arms, enclosing large
intercellular lacunae, (NA); (O) extra-fascicular fibrous strands: (0) absent and (1) present; (P) position of the extra-fascicular fibrous strands: (0)
only close to abaxial surface and (1) close to adaxial and abaxial surfaces; (NA); (Q) number of extra-fascicular fibrous strand layers close to
adaxial surface: (0) one layer and (1) more than one layer; (NA); (R) number of extra-fascicular fibrous strand layers close to abaxial surface: (0)
one layer and (1) more than one layer; (NA); (S) smaller vascular bundles (2nd and 3rd order), fiber caps form: (0) larger than they are wide and
(1) wider than they are large; (T) fiber caps at larger vascular bundles: (0) absent and (1) present; (U) parenchymatic sheath cells of the vascular
bundles, radial arrangement: (0) absent and (1) present
Leaf anatomy of Quesnelia
123
Author's personal copy
(Fig. 1f). In the remaining species, the water storage cells are
anticlinally elongated (Fig. 1e). Intraspecific variation for
this trait was found only in Q. arvensis and Q. testudo. The
adaxial water storage tissue covers up to one-third of the
mesophyll in Q. arvensis,Q. clavata,Q. dubia,Q. con-
quistensis,Q. edmundoi var. edmundoi,Q. edmundoi var.
rubrobractea,Q. liboniana,Q. quesneliana and Q. testudo.
In other species, the water storage tissue occupies more than
one-third of the mesophyll (Fig. 1e). No intraspecific vari-
ability was detected for this trait among different specimens.
The transition from the adaxial water storage tissue to
the chlorenchyma immediately below does not always
Fig. 1 Systematic leaf anatomy of Quesnelia.aQ. kautskyi: adaxial
epidermis cells with inner periclinal wall markedly thickened, cell
lumen almost occluded, followed by one layer of slightly thickened
cells and followed, in turn, by one layer of hypodermal cells less
sclerified. bQ. lateralis: adaxial epidermis cells highly sclerified,
followed by a hypodermis composed by several layers of scleren-
chymatic thick-walled cells. cQ. edmundoi var. edmundoi: abaxial
epidermis and hypodermis consist of cells with slightly thickened
walls delimiting a wide lumen. dQ. marmorata: abaxial surface,
ribbed. eQ. augusto-coburgii: water storage tissue constituted
by anticlinally elongated cells, abruptly followed by chlorenchyma.
fQ. koltesii: water storage tissue constituted by rounded cells, without
abrupt transition to the chlorenchyma. Scale bar 30 lm
A. Mantovani et al.
123
Author's personal copy
happen abruptly, but may be continuous when the cells of
both tissues have similar shape and size. Abrupt transition
was observed in most taxa (Fig. 1e), except in Q. alvimii,
Q. clavata,Q. dubia,Q. edmu ndoi var. edmundoi,Q. edmundoi
var. rubrobracteata,Q. koltesii and Q. tubifolia (Fig. 1f).
The intercostal chlorenchyma is formed of armed cells
that delimit air lacunae. The armed cells may be long,
delimiting large air spaces, or short, delimiting small ones.
They occur intercalating with vascular bundles, and air
channels are often connected with the substomatal cham-
bers. Cells with long arms were found in Q. arvensis,
Q. clavata,Q. koltesii,Q. lateralis,Q. quesneliana and
Q. testudo (Fig. 2e). The remaining species showed cells
with short arms. This air-lacunae chlorenchyma is present
Fig. 2 Systematic leaf anatomy of Quesnelia.aQ. edmundoi var.
rubrobracteata: stomata positioned at the same level of the epidermis
surface; substomatal chamber without differentiated cells. bQ.
augusto-coburgii: sunken stomata with occluded substomatal cham-
ber. cQ. edmundoi var. edmundoi: peltate scale with stalk composed
of more than two cells. dQ. testudo: peltate scale with stalk
composed of two cells. eQ. quesneliana: air-lacunae chlorenchyma
constituted by long arm cells. fQ. imbricata: compact mesophyll, air-
lacunae absent. Scale bar 30 lm
Leaf anatomy of Quesnelia
123
Author's personal copy
in all species, except Q. imbricata and Q. violacea, where it
is replaced by rounded parenchymatic cells, arranged in a
highly compact manner (Fig. 2f). No intraspecific vari-
ability was observed for this character.
The vascular bundles are collateral, surrounded by a
double sheath of cells, the internal one constituted of
sclerified cells and the external one of parenchymatic cells.
The external parenchymatic cells are rounded in most
species (Fig. 3a, b). However, in Q. alvimii,Q. conquist-
ensis,Q. edmundoi var. edmundoi,Q. edmundoi var. rub-
robracteata and Q. quesneliana, these outer parenchymatic
cells are clearly radially elongated (Fig. 3e, f).
The small-diameter vascular bundles (2nd and 3rd order)
have caps near the xylem poles that can be longer than they
Fig. 3 Systematic leaf anatomy of Quesnelia.aQ. seideliana: small
vascular bundles with fiber caps at xylem poles taller than they are
wide. bQ. imbricata: small vascular bundles with fiber caps at xylem
poles wider than they are tall. cQ. edmundoi var. rubrobracteata:
double layers of extra-fascicular fiber strands near the adaxial surface.
dQ. seideliana: one layer of extra-fascicular fiber strands near the
abaxial surface. eQ. kautskyi: vascular bundle sheath composed of
non-radiating parenchymatic cells. fQ. edmundoi var. rubrobracte-
ata: vascular bundle sheath composed of radiating parenchymatic
cells. Scale bar 30 lm
A. Mantovani et al.
123
Author's personal copy
are wide, as found in most species (Fig. 3a), or wider than
they are long (Fig. 3b), as observed in Q. marmorata,
Q. kautskyi,Q. imbricata,Q. humilis,Q. alvimii and Q. viola-
cea. No intraspecific variability was detected for this character.
Extra-fascicular fibrous strands were found in all spe-
cies, except Q. arvensis,Q. conquistensis,Q. imbricata,
Q. quesneliana,Q. testudo and Q. violacea. These fibrous
strands showed a distinct location, being close to the adaxial
and abaxial surface in Q. clavata,Q. dubia,Q. edmundoi var.
edmundoi,Q. edmundoi var. rubrobracteata and Q. koltesii,
or restricted to the abaxial side in the other species. Intra-
specific variation was found only for Q. arvensis and
Q. quesneliana, which presented a few sparse extra-fascicular
strands in each of the three specimens.
When near the adaxial surface, the extra-fascicular
fibrous strands are immersed in the water storage tissue,
arranged in a single layer in Q. dubia and Q. koltesii or
more than one layer in Q. edmundoi var. edmundoi,
Q. edmundoi var. rubrobracteata and Q. clavata (Fig. 3c).
When near the abaxial surface, these strands are found
immersed in the chlorenchyma, arranged in more than
one layer in Q. alvimii,Q. edmundoi var. edmundoi,
Q. edmundoi var. rubrobracteata,Q. liboniana,Q. mar-
morata and Q. seideliana, but in only one layer in the other
species that present this character (Fig. 3d). Intraspecific
variation for this trait was only observed in one specimen
of Q. edmundoi var. edmundoi, which showed extra-
fascicular fibrous strands distributed in more than one layer
on the abaxial surface.
The MDS analysis was generated using the characters
listed in Table 2, except for characters N–R, which are not
applicable to all species, and character T, which is present
in all species. There was spatial separation of groups,
generated with a low stress of only 0.07. This spatial
separation was significant using the ANOSIM test (global
R=0.704, p=0.001), with the distribution of variables
on two axes explaining 84.6% of the data variation. With
this result, we could separate the species into three distinct
groups, here denoted as groups Q, B and A (Fig. 4). Such
groups are detailed below, along with the contribution from
the major anatomical features, as calculated by SIMPER
analysis, to group respective species (Fig. 5):
Group Q includes Q. arvensis,Q. conquistensis,
Q. quesneliana and Q. testudo (Fig. 4a). The main ana-
tomical features of this group are the ribbed abaxial surface
(12%), sunken stomata (21%), abrupt transition between
the water storage and chlorenchyma tissues (21%), the
presence of air-lacunae chlorenchyma (21%) and the extent
of the vascular bundle sheath, which is longer than it is
wide (21%). These features together explain 96% of this
group, which includes all species belonging to the subge-
nus Quesnelia.
Group B includes Q. alvimii,Q. augusto-coburgii,
Q. humilis,Q. imbricata,Q. indecora,Q. kautskyi,
Q. lateralis,Q. liboniana,Q. marmorata,Q. seideliana,
Q. strobilispica,Q. tubifolia and Q. violacea (Fig. 4b). The
main anatomical features belonging to the species in this
group are the sunken stomata (20%), the water storage
tissue occupying more than one-third of the mesophyll
(15%) with anticlinally elongated cell (20%), the abrupt
transition between the water storage and chlorenchyma
tissues (18%), the presence of air-lacunae chlorenchyma
(16%) and the extent of the vascular bundle sheath, which
is longer than it is wide (7%). These features explain 96%
of the data variation in this group.
Group A includes Q. edmundoi var. edmundoi,
Q. edmundoi var. rubrobracteata,Q. dubia,Q. clavata and
Fig. 4 Systematic leaf anatomy of Quesnelia. Groups revealed by the
MDS analysis, based on leaf anatomical features. aGroup Q,
Quesnelia arvensis.bGroup B, Quesnelia seideliana.cGroup A,
Quesnelia edmundoi var. edmundoi.Scale bar 100 lm
Leaf anatomy of Quesnelia
123
Author's personal copy
Q. koltesii (Fig. 4c). The main anatomical features that
approximate the taxa of this group are the contour and low
cellular sclerification of the adaxial (8%) and abaxial (8%)
surfaces, enclosing large cell lumen, the number of stalk
cells from the scale (20%), the presence of air-lacunae
chlorenchyma (26%), the vascular bundle sheath extension,
which is longer than it is wide (26%), and the presence of
radially elongated parenchymatic cells from the vascular
bundle sheath (8%). These features together explain 96%
of the grouping for these species.
Note that many of the features present in one group may
contribute, but in different ways, to the composition of the
two remaining groups, e.g., the presence of air-lacunae
chlorenchyma (character 6) ?bundle sheath extension
(character 19). This is especially demonstrated by the small
dissimilarity between groups B and Q (ANOSIM,
R=0.48, p=0.001), which are closer to each other than
they are to group A (Table 3). Groups Q and B are sepa-
rated mainly by the surface contour, the shape and thick-
ness of the water storage cells, and the bundle sheath
extension. While species from group Q have a ribbed
surface and poorly developed water storage tissue, consti-
tuted of rounded cells, besides vascular bundle sheath
extension that is longer than it is wide, species from group
B have a smooth to slightly undulate surface, large water
storage tissue constituted of elongated cells and sheath
extensions that are wider than they are long. Such features
explain nearly 60% of dissimilarity between these two
(2)
(3)
(1) (1)
(2)
(3)
(2)
(1)
(1)
(2)
(1)
(2)
(3)
(1)
(3)
(2)
(3)
(1)
(1)
(2)
(3)
(1)
(1)
(2)
(3)
(1)
(2)
(3)
(1)
(2)
(3)
(1)
(2)
(3)
(2)
(3)
(1)
var. (1)
var. (2)
var. (3)
var. (1)
var. (2)
var. (3)
Fig. 5 Systematic leaf anatomy of Quesnelia. Multidimensional scaling (MDS) with species distributed along two axes. Symbols indicate
groups Q (open square), B (filled triangle) and A (filled circle)
Table 3 Systematic leaf anatomy of Quesnelia
Anatomical types Q B
B 0.484
(10, 11, 2, 19, 21,
7, 5, 1, 9, 8)
\92.3%[
A 0.925
(6, 12, 5, 2, 3, 4,
21, 11, 1)
\95.2%[
0.893
(6, 11, 12, 10, 5,
21, 4, 3, 19, 7)
\93.2%[
Pairwise comparison among anatomical groups Q, B and A. Data are
global R statistics (ANOSIM test, p=0.001) for each comparison.
Between brackets, in decrescent order of contribution, the most rel-
evant anatomical characters contributing to the indicated dissimilarity
(for characters see Table 1). The final accumulated contribution of all
respective anatomical characters of each pairwise comparison is
indicated between ‘‘\[.’’ Note that Q and B present lower dissimi-
larity when compared to A
A. Mantovani et al.
123
Author's personal copy
groups. Group A is separated from groups B and Q
(ANOSIM, R[0.90, p=0.001) mainly by the presence
of stomata at the same level as a smooth epidermis, by
poorly developed water storage tissue provided by rounded
cells, in addition to the lack of abrupt transition between
this tissue and chlorenchyma. These characteristics explain
almost 50% of this differentiation. Finally, the dissimilarity
of group A compared to groups B and Q is also reinforced
by the distribution, when present, of extra-fascicular
fibrous strands, which are distributed throughout all the
mesophyll, while restricted to the abaxial surfaces in
groups B and Q.
Discussion
The circumscription of the genera within Bromeliaceae,
especially within the Bromelioideae, has led researchers
from different fields to combine their efforts to find natural
groups using different sources of characters. Since the
beginning of systematic studies in Bromeliaceae, authors
such as Baker (1889), Wawra (1880) and Mez (1896) have
expressed the difficulty in morphologically differentiating
species from the Quesnelia,Aechmea and Billbergia gen-
era. This can be explained in part by the artificiality of
Quesnelia, as pointed out by several authors, such as Faria
et al. (2004), Leme (2005), Almeida et al. (2009) and
Amorim and Leme (2009).
Some of the aforementioned authors agree that the spe-
cies of Quesnelia subg. Billbergiopsis would be more related
to Billbergia,whilethetaxaplacedinQuesnelia subg.
Quesnelia would have more affinity with Aechmea species.
Moreover, Q. edmundoi var. edmundoi,Q. edmundoi var.
rubrobracteata,Q. dubia,Q. clavata and Q. koltesii were
described within Quesnelia based on the difficulty of placing
them in any other Bromelioideae genera (Amorim and Leme
2009). The only taxon of this group (Q. edmundoi)included
in molecular phylogenies (Schulte et al. 2009) shows affinity
with the genus Canistrum.
The two varieties, Q. edmundoi var. edmundoi and
Q. edmundoi var. rubrobracteata, appear here next to the
species Q. dubia,Q. clavata and Q. koltesii, forming group
A, confirming what was suggested by Leme (2005) and
Amorim and Leme (2009), who stated that these taxa do
not possess characteristics typical of Q. quesneliana, the
type species of the genera. Indeed, our group A is ana-
tomically very different from the groups formed of species
of the subgenus Quesnelia (group Q) and the subgenus
Billbergiopsis (group B). This separation is achieved by the
simultaneous occurrence of stomata at the same level as a
smooth epidermis, by the adaxially less prominent water
storage tissue and by the distribution of extra-fascicular
fibers throughout the mesophyll. This arrangement, found
only in group A, is identical to that described for ‘‘type III’’
in the classification of Horres et al. (2007) found in a
species of Aechmea subgenus Aechmea. Sajo et al. (1998)
described the same leaf anatomical arrangement for three
Canistrum species, supporting the affinity with the part of
genera Quesnelia with Canistrum, indicated in molecular
phylogenies (Schulte et al. 2009).
The species of group A have a restricted distribution
(Rio de Janeiro and Bahia states), occurring as epiphytes in
the rainforest, never reaching habitats more prone to
drought, as do the other Quesnelia species, which occur as
terrestrial plants in sandy coastal plains or as epiphytes in
mangroves and seasonal forests. Thus, in comparison with
other Quesnelia species, the arrangement of type A seems
to indicate a retraction of xeromorphic characters (under-
developed water storage tissue rather than developed with
anticlinally elongated cells; epidermis smooth rather than
ribbed, superficial rather than sunken stomata), which was
probably linked to the survival in the wettest epiphytic
environments.
We could expect that other co-occuring core bromeli-
oids with phylogenetic affinities (e.g., Aechmea,Cani-
strum,Billbergia) also present the same leaf anatomical
arrangement as Group A species. This phenomenon, which
is called concerted convergence, is characterized by the
convergence of different characters in organisms native to
similar habitats that share the same set of ecological con-
ditions (Patterson and Givnish 2002). Givnish et al. (2007)
report that this phenomenon for xeromorphical characters
(mechanical and aquiferous adaxial hypodermises, absence
of palisade parenchyma) evolved independently many
times in Hechtia,Puya and in the xeric clade (Abromeiti-
ella-Deuterocohnia–Dyckia–Encholirium), along with
several invasions of the dry terrestrial environments in the
neotropical region. However, the coexistence of core bro-
melioids with distinct adaptive anatomical characters could
also be expected (Sass and Spech 2010), induced by the
multiple distinct niches found on tree canopies (Benzing
1990). Further studies are needed to evaluate this possible
relationship between the onset and absence of different
anatomical types and potential different niches on a smaller
spatial scale and within phylogenetic clades that share
geographical proximity (Horres et al. 2007; Sass and Spech
2010).
In turn, group B (subgenus Billbergiopsis) was identified
by thick, well-developed water storage tissue, with abrupt
transition between this and chlorenchyma, besides sunken
stomata and extra-fascicular fibers near the abaxial surface.
Such an arrangement is very similar to that found in species
of Billbergia (Vargens 2008; Proenc¸a and Sajo 2007), and
it confirms the ideas of Wawra (1880) and Mez (1896), as
well as the phylogenies of Faria et al. (2004) and Almeida
et al. (2009). However, it is noteworthy that the hypothesis
Leaf anatomy of Quesnelia
123
Author's personal copy
obtained from molecular characters by Sass and Spech
(2010), which addressed only two species (Q. lateralis and
Q. liboniana) of this subgenus, does not sustain this inti-
mate relationship.
Compared with group A, Quesnelia species from
group B are the most widely distributed and may occur
either as terrestrial (including sandy coastal plains) or,
sometimes, as rupicolous, as well as epiphytes in man-
grove areas. These characteristics seem to indicate that
the presence of an internal reservoir of water associated
with sunken stomata is related to species survival in
potentially more stressful environments in terms of water
availability. This could also be extended to Q. arvensis,
Q. conquistensis,Q. quesneliana and Q. testudo,which,
together, comprise the group with anatomic arrangement
of type Q. Such species are widespread and may occur as
terrestrial, epiphytic or rupicolous plants in environments
subject to water stress, such as sandy coastal plains and
mangroves.
There are clear similarities between the anatomical
arrangement of ‘‘type IV’’ (sensu Horres et al. 2007),
characteristic of species of the Aechmea subgenus Pothu-
ava, and the arrangement of the group we classified as
group Q with ribbed leaf surfaces, adaxially developed
water storage tissue (but less prominent than observed in
type B) constituted of rounded cells and lack of extra-
fascicular fibrous strands. In fact, Almeida et al. (2009)
obtained a clade that unifies the species of subgenus
Quesnelia (herein group Q) along with the species Aech-
mea vanhoutteana (Van Houtte) Mez from the subgenus
Pothuava. Similarly, Givnish et al. (2011) also obtained a
clade formed of Q. quesneliana (type species of Quesnelia)
and A. nudicaulis (type species of Aechmea subg. Pothu-
ava), which confirms the morphological results presented
here. However, the affinity between Q. arvensis and
A. distichantha Lem., as indicated by molecular phyloge-
nies (Schulte et al. 2009; Sass and Spech 2010), is partially
supported by leaf anatomy (Derwiduee
´and Gonzalez
2010).
In order to evaluate the potential diagnostic value of the
anatomical types of Quesnelia to the taxonomy of other
core bromelioids, it is first necessary to determine the
intraspecific stability of these characters. Although a few
works have already evaluated the intraspecific variation of
anatomical data in Bromelioideae, all of them indicate that,
while quantitative variations occur, the qualitative pattern
(types, organization and proportion of leaf tissues in
transversal section) does not show intensive intraspecific
variation along the same leaf (Braga 1977), along different
leaves from the same individual (Mantovani et al. 2005), or
along shade and sun leaves (Duarte 1998; Scarano et al.
2002; Voltolini and Santos 2011). In Quesnelia, we found
intraspecific variation in 7 of the 21 characters analyzed.
These results indicate that anatomical characters or inte-
grative anatomical typologies could be used to characterize
groups of species in Bromelioideae, but always using the
same leaf region and leaf age.
Anatomical characters occur as distinct arrangements
along distinct genera in the subfamily Bromelioideae,
separating groups of species (Derwiduee
´and Gonzalez
2010). Although high homoplasy levels are commonly
registered for morphological phylogenies in the family
(Forzza 2001; Faria et al. 2004; Almeida et al. 2009;
Monteiro 2009), homoplastic characters could be used to
compare genera and subgenera, when geographical distri-
bution is taken into account, based on the geographic
conservatism cited by Sass and Spech (2010) for Brome-
lioideae. In this sense, the integrative anatomical typolo-
gies proposed by Horres et al. (2007) could yield useful
data for future systematic analysis in this group. For
example, Ronnbergia carvalhoi Martinelli and Leme could
be integrated into group B, if extravascular fibers are
present. On the other hand, Lymania globosa Leme does
not belong to group A, only by the presence of stomata
positioned below the epidermis level (Aoyama and Sajo
2003).
Considering the low genetic variability in Bromeliaceae,
morphological studies are relevant to the understanding of
this family’s systematics (Givnish et al. 2011). Further
efforts should combine extensive intraspecific sampling,
morphological characters and anatomical typologies, in
addition to molecular analysis, in order to improve the
systematic analysis and better understand evolution in
Bromelioideae.
Acknowledgments The authors wish to thank Andre Amorim and
Elton Leme for making part of the samples used here available, as
well as Jorge Nessimian for assistance with statistical analysis; and
Conselho Nacional de Desenvolvimento Cientı
´fico e Tecnolo
´gico
(CNPq), Coordenadoria de Aperfeic¸oamento do Pessoal de Nı
´vel
Superior (CAPES) and Fundac¸a
˜o de Amparo a
`Pesquisa do Estado do
Rio de Janeiro (FAPERJ) for fellowships.
References
Almeida VR, Costa AF, Mantovani A, Gonc¸alves-Esteves V, Arruda
RCO, Forzza RC (2009) Morphological phylogenetics of
Quesnelia (Bromeliaceae—Bromelioideae). Syst Bot 34:660–
672
Amorim AMA, Leme EMC (2009) Two new species of Quesnelia
(Bromeliaceae—Bromelioideae) from the Atlantic Rainforest of
Bahia, Brazil. Brittonia 61:14–21
Aoyama EM, Sajo MG (2003) Estrutura foliar de Aechmea Ruiz &
Pav. subg. Lamproccocus (Beer) Baker e espe
´cies relacionadas.
Rev Bras Bot 26:461–473
Arruda RCO, Costa AF (2003) Foliar anatomy of five Vriesea sect.
Xyphion (Bromeliaceae) species. Selbyana 24:180–189
Baker JG (1889) Handbook of the Bromeliaceae. George Bell & Sons,
London
A. Mantovani et al.
123
Author's personal copy
Barfuss MHJ, Samuel R, Till W, Stuessy TF (2005) Phylogenetic
relationships in subfamily Tillandsioideae (Bromeliaceae) based
on DNA sequence data from seven plastid regions. Am J Bot
92:337–351
Benzing DH (1990) Vascular epiphytes. Cambridge University Press,
New York
Benzing DH (2000) Bromeliaceae: profile of an adaptive radiation.
Cambridge University Press, Cambridge
Braga MMN (1977) Anatomia foliar de Bromeliaceae da Campina.
Acta Amazonica 7(suplemento):1–74
Clarke KR, Gorley RN (2011) Software PRIMER v5. PRIMER-E,
Plymouth
Crayn DM, Terry RG, Smith AC, Winter K (2000) Molecular
systematic investigations in Pitcairnioideae (Bromeliaceae) as a
basis for understanding the evolution of crassulacean acid
metabolism (CAM). In: Wilson KL, Morrison DA (eds) Mono-
cots: systematics and evolution. CSIRO Publishing, Collingwood,
pp 569–579
Crayn DM, Winter K, Smith JAC (2004) Multiple origins of
crassulacean acid metabolism and the epiphytic habit in the
neotropical family Bromeliaceae. Proc Natl Acad Sci USA
101:3703–3708
Derwiduee
´FS, Gonzalez AM (2010) Anatomı
´a foliar en Bromelia-
ceae del nordeste argentino y Paraguay. Bonplandia 19(2):153–
173
Duarte HM (1998) Variabilidade fenotı
´pica intra e interpopulacional
em caracteres morfolo
´gicos e anato
ˆmicos em populac¸o
˜es alop-
a
´tricas de Quesnelia quesneliana Lindman (Bromeliaceae) sob
distintos regimes de luz. MSc Dissertation, Universidade Federal
do Rio de Janeiro
Faria APG, Wendt T, Brown GK (2004) Cladistic relationships of
Aechmea (Bromeliaceae-Bromelioideae) and allied genera. Ann
Mo Bot Gard 29:303–319
Forzza RC (2001) Filogenia da tribo Puyeae Wittm. e revisa
˜o
taxono
ˆmica do ge
ˆnero Encholirium Mart. ex Schult. & Schult. f.
(Pitcairnioideae—Bromeliaceae). Ph.D. Dissertation, Universid-
ade de Sa
˜o Paulo
Forzza RC, Costa A, Siqueira Filho JA, Martinelli G (2010)
Bromeliaceae. In: Lista de Espe
´cies da Flora do Brasil. Jardim
Bota
ˆnico do Rio de Janeiro. (http://floradobrasil.jbrj.gov.br/
2010/FB006335)
Gerrits PO, Smid L (1983) A new, less toxic polymerization system
for the embedding of soft tissues in glycol methacrylate and
subsequent preparing of serial sections. J Microsc 132:81–85
Givnish TJ, Millam KC, Berry PE, Systma KJ (2007) Phylogeny,
adaptive radiation and historical biogeography of Bromeliaceae
inferred from ndhF sequence data. In: Columbus JT, Friar EA,
Porter JM, Prince LM, Simpson MG (eds) Monocots: compar-
ative biology and evolution—Poales. Rancho Santa Ana Bot
Garden, Claremont, pp 3–26
Givnish TJ, Barfuss MHJ, Van EB, Riina R, Schulte K, Horre R,
Gonsiska PA, Jabaily RS, Crayn DM, Smith JAC, Winter K,
Brown GK, Evans TM, Holst BK, Luther H, Till W, Zizka G,
Berry PE, Systma KJ (2011) Phylogeny, adaptive radiation and
historical biogeography in Bromeliaceae Insights from an eight-
locus plastid phylogeny. Am J Bot 98:872–895
Horres R, Schulte K, Weising K, Zizka G (2007) Systematics of
Bromelioideae (Bromeliaceae)—evidence from molecular and
anatomical studies. In: Columbus JT, Friar EA, Porter JM, Prince
LM, Simpson MG (eds) Monocots: comparative biology and
evolution—Poales. Rancho Santa Ana Bot Garden, Claremont,
pp 27–43
Leme EMC (2005) Three new miscellaneous species of Bromeliaceae
from Bahia, Brazil. J Brom Soc 55:13–22
Leme EMC (2008) A new Quesnelia species from Bahia. J Brom Soc
58:269–271
Leme EMC, Kollmann LJC (2011) Two new species from Alto Cariri
State Park, Minas Gerais, Brazil. J Brom Soc 61:54–64
Mantovani A, Magalha
˜es N, Teixeira ML, Leita
˜o G, Staines CL,
Resende B (2005) First report on host plants and feeding habits
of the leaf beetle Acentroptera pulchella Gue
´rin-Me
´nevill
(Chrysomelidae, Hispinae). In: Konstantinov A, Tishechkin A,
Penev L (eds) Contributions to systematics and biology of
beetles. Papers celebrating the 80th birthday of Igor Konstan-
tinovich Lopatin. Pensoft Publishers, Sofia, pp 153–157
Mantovani A, Filartiga AL, Coelho MN (2010) Anatomia da folha e
espata de espe
´cies de Anthurium (Araceae) da Mata Atla
ˆntica.
Rev Bras Bot 33:185–200
Martinelli G, Vieira CM, Gonzalez M, Leitman P, Piratininga A,
Costa AF, Forzza RC (2008) Bromeliaceae da Mata Atla
ˆntica
Brasileira: lista de espe
´cies, distribuic¸a
˜o e conservac¸a
˜o. Rodr-
igue
´sia 59:209–258
MCune B, Mefford MJ (1999) PC-ORD version 4.0: multivariate
analysis of ecological data, users guide. MjM Software Design,
Glaneden Beach. Oregon, USA
Mez C (1896) Bromeliaceae. In: Candolle ALPP, Candolle ACP (eds)
Monographiae Phanerogamarum, vol 9. G. Massom, Paris,
pp 1–990
Monteiro RF (2009) Estudos anato
ˆmicos e filogene
´ticos em Bromelia
L. (Bromeliaceae—Bromelioideae). MSc Dissertation, Jardim
Bota
ˆnico do Rio de Janeiro, Rio de Janeiro
Monteiro RF, Forzza RC, Mantovani A (2011) Leaf structure of
Bromelia and its significance for the evolution of Bromelioideae
(Bromeliaceae). Plant Syst Evol 293:53–64
O’Brien TP, Mccully ME (1981) The study of plants structure,
principles and selected methods. Termarcarphi, Melbourne
Patterson TB, Givnish TJ (2002) Phylogeny, concerted convergence,
and phylogenetic niche conservatism in the core Liliales:
insights from rbcL and ndhF sequence data. Evolution
56:233–252
Proenc¸a SL, Sajo MG (2007) Anatomia foliar de brome
´lias em a
´reas
de cerrado do estado de Sa
˜o Paulo, Brasil. Acta Bot Bras
21:657–673
Reinert F, Russo CAM, Salles LO (2003) The evolution of CAM in
the subfamily Pitcairnioideae (Bromeliaceae). Bio J Linn Soc
80:261–268
Rex M, Schulte K, Zizka G, Peters J, Vasquez R, Ibisch PL, Weising
K (2009) Phylogenetic analysis of Fosterella L.B. Sm. (Pitcair-
nioideae, Bromeliaceae) based on four chloroplast DNA regions.
Mol Phylogenet Evol 51:472–485
Robinson H (1969) A monograph on foliar anatomy of the genera
Connellia,Cottendorfia and Navia (Bromeliaceae). Sm C Bot
2:1–41
Ruzin SE (1992) Plant microtechnique and microscopy. Oxford
University Press, New York
Sajo MG, Machado SR, Camello-Guerreiro SM (1998) Aspectos
estruturais de folha de brome
´lia e suas implicac¸o
˜es no agrupa-
mento de espe
´cies. In: Pereira MV (ed) Brome
´lias da Mata
Atla
ˆntica: Canistropsis. Salamandra, Rio de Janeiro, pp 102–111
Sass C, Spech CD (2010) Phylogenetic estimation of the core
bromelioids with emphasis on the genus Aechmea (Bromelia-
ceae). Mol Phylogenet Evol 55:559–571
Scarano FR, Duarte HM, Ro
ˆc¸as G, Barreto SMB, Amado F, Reinert
F, Wendt T, Mantovani A, Lima HRP, Barros CF (2002)
Acclimation or stress sympton? An integrated study of intraspe-
cific variation in the clonal plant Aechmea bromeliifolia,a
widespread CAM tank-bromeliad. Bot J Linn Soc 140:391–402
Schulte K, Zizka G (2008) Multi locus plastid phylogeny of
Bromelioideae (Bromeliaceae) and the taxonomic utility of petal
appendages and pollen characters. Candollea 63:209–255
Schulte K, Horres R, Zizka G (2005) Molecular phylogeny of
Bromelioideae and its implications on biogeography and the
Leaf anatomy of Quesnelia
123
Author's personal copy
evolution of CAM in the family. Senckenbergiana Biologica
85:113–125
Schulte K, Barfuss MHJ, Zizka G (2009) Phylogeny of Bromelioideae
(Bromeliaceae) inferred from nuclear and plastid DNA loci
reveals the evolution of the tank habit within the subfamily. Mol
Phylogenet Evol 51:327–339
Smith LB, Downs RJ (1979) Bromelioideae (Bromeliaceae). Flora
Neotropica Monograph 14:1493–2141
Sousa GM, Estelita MEM, Wanderley MGL (2005) Anatomia foliar
de espe
´cies brasileiras de Aechmea subg. Chevaliera (Gaudich.
ex Beer) Baker, Bromelioideae—Bromeliaceae. Rev Bras Bot
28:603–613
Taylor DC, Robinson H (1999) A rejection of Pepinia (Bromeliaceae:
Pitcairnioideae) and taxonomic revisions. Harvard Pap Bot
4:203–217
Tomlinson PB (1969) Commelinales-Zingiberales. In: Metcalf CR
(ed) Anatomy of the monocotyledons. Claredon Press, Oxford,
pp 193–294
Vargens FAC (2008) Anatomia foliar de espe
´cies de Billbergia
Thunb. (Bromeliaceae, Bromelioideae). MSc Dissertation, Uni-
versidade Federal do Rio de Janeiro, Rio de Janeiro
Vieira CM (1999) Uma nova espe
´cie de Quesnelia (Bromelioideae:
Bromeliaceae) para o Brasil. Bradea 8:131–134
Vieira CM (2006) Quesnelia Gaudich. (Bromelioideae: Bromelia-
ceae) do estado do Rio de Janeiro, Brasil. Pesquisas Bota
ˆnica
57:7–102
Voltolini CH, Santos M (2011) Variac¸o
˜es na morfo-anatomia foliar
de Aechmea lindenii (E.Morren) Baker var. lindenii (Bromeli-
aceae) sob distintas condic¸o
˜es ambientais. Acta Bot Bras
25:2–10
Wanderley MGL, Proenc¸a SL (2006) Nova espe
´cie de Quesnelia
Gaudich. (Bromeliaceae) do estado de Sa
˜o Paulo, Brasil.
Hoehnea 33:111–113
Wawra HRF (1880) Die Bromeliaceen-Ausbeute von der Reise der
Prinzen August und Ferdinand von Sachsen-Cobourg nach
Brasilien 1879. Oesterr Bot Zeitschrift 30:148–151
A. Mantovani et al.
123
Author's personal copy
