See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/331024960

Bat pollination in Bromeliaceae

Article in Plant Ecology & Diversity · February 2019

DOI: 10.1080/17550874.2019.1566409

CITATIONS READS

0 119

6 authors, including:

Pedro Adrián Aguilar-Rodríguez Thorsten Krömer

Universidad Veracruzana Universidad Veracruzana
16 PUBLICATIONS 27 CITATIONS 135 PUBLICATIONS 1,504 CITATIONS

SEE PROFILE SEE PROFILE

Jose G Garcia-Franco María Cristina Mac Swiney González

Institute of Ecology INECOL Universidad Veracruzana
106 PUBLICATIONS 2,066 CITATIONS 56 PUBLICATIONS 487 CITATIONS

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

BIOVERA - exploration and explanation of biodiversity patterns along gradients of climate, soil and disturbance in Central Veracruz, Mexico View project

Plant systematics and floristics View project

All content following this page was uploaded by María Cristina Mac Swiney González on 11 February 2019.

The user has requested enhancement of the downloaded file.

PLANT ECOLOGY & DIVERSITY
https://doi.org/10.1080/17550874.2019.1566409

ARTICLE

Bat pollination in Bromeliaceae

Pedro A. Aguilar-Rodrígueza, Thorsten Krömer a, Marco Tschapka b,c
, José G. García-Francod,
Jeanett Escobedo-Sartie and M.Cristina MacSwiney G. a
a
Centro de Investigaciones Tropicales, Universidad Veracruzana, Xalapa, Veracruz, Mexico; bInstitute of Evolutionary Ecology and
Conservation Genomics, University of Ulm, Ulm, Germany; cSmithsonian Tropical Research Institute, Balboa Ancón, Panamá, República
de Panamá; dRed de Ecología Funcional, Instituto de Ecología, Xalapa, Veracruz, México; eFacultad de Ciencias Biológicas y
Agropecuarias, Universidad de Colima, Crucero de Tecomán, Tecomán, Colima, México

ABSTRACT ARTICLE HISTORY

Background: Chiropterophily encompasses the floral traits by which bats are attracted as the Received 2 May 2017
main pollinators. Among the chiropterophilous flowering plants of the New World, Accepted 3 January 2019
Bromeliaceae is one of the most ecologically important families; however, information KEYWORDS
about the chiropterophilous interaction in this family is still scarce. Anoura; bromeliads;
Aims: We present a comprehensive review of bat pollination in bromeliads, covering floral chiropterophily; floral scent;
traits, rewards offered to pollinators, floral attractants and the identity of visiting bat species. nectar; pollination; Werauhia
Methods: We discuss traits shared among chiropterophilous bromeliads and present general
trends in an evolutionary context. We constructed a phylogenetic tree to elucidate the
ancestral pollination syndromes of the 42 extant bromeliad species (ca. 1% of total) known
to be bat-pollinated.
Results: Most of the species within the ten genera reported belong to the Tillandsioideae
subfamily, with three genera appearing to be exclusively bat-pollinated. Floral visitors include
19 bat species of 11 genera from the Phyllostomidae.
Conclusions: Our analysis indicated that chiropterophilous floral features originated multiple
times in bromeliad evolution, most probably from ornithophilous. The evidence for floral
traits associated with bat pollination and the chiropterophilous syndrome presented by
certain Bromeliaceae indicate the important role played by bats in the evolution of this
plant family.

Introduction 500 plant species (Dobat and Peikert-Holle 1985;

Fleming et al. 2009). The chiropterophily syndrome
The floral syndrome concept has been criticised
may encompass the following characteristics (van der
for failing to consider the complete diversity of
Pijl 1961; von Helversen 1993; von Helversen et al.
floral traits, as well as the entire spectrum of floral
2000; Tschapka and Dressler 2002; von Helversen and
visitors (Ollerton et al. 2009). Nevertheless, it
Winter 2003; Krömer et al. 2008; Rosas-Guerrero et al.
remains a useful tool for predicting the main pol-
2014): (1) nocturnal or crepuscular anthesis; (2) pale
linators of plants, especially in tropical regions
or dullish colours in the perianth (predominantly
(Rosas-Guerrero et al. 2014; Ashworth et al. 2015
white, pale green, yellow) without nectar lines; (3)
and references therein). This seems to be the case
flagelliflory (flowers suspended on long stalks, situated
for bat-pollinated plant species, which tend to
at a distance from the branches and leaves) and/or
possess flowers that are morphologically distinct
cauliflory (flowers emerging directly from the tree
from those of other pollination guilds (von
trunk or from larger branches); (4) tubular, zygo-
Helversen 1993; Fleming and Muchhala 2008).
morphic or radially symmetrical flowers that fit the
Chiropterophily, or the syndrome of pollination via
snout of the bat like a ‘mask’, and also ‘brush-type’
interaction with bats, encompasses the floral traits (e.g.
flowers/inflorescences with protruding stamens; (5)
size, shape, colour, nectar characteristics and scent)
large and sturdy flowers; (6) musty or unpleasant
shared by many angiosperms in order to attract bats
odour, described as garlic- or onion-like; (7) relatively
(Mammalia: Chiroptera) as their main pollinators
large amounts of dilute nectar with an average sugar
(Tschapka and Dressler 2002; von Helversen and
concentration of ca. 17%, and a predominantly hex-
Winter 2003; Fleming et al. 2009). Bat pollination
ose-rich composition (e.g. low sucrose content, high
occurs in the tropical and subtropical regions of both
levels of glucose and fructose); (8) large quantities of
the Old and New World and is found in more than

CONTACT M. Cristina MacSwiney G. [email protected]

Supplemental data for this article can be accessed here.
© 2019 Botanical Society of Scotland and Taylor & Francis
2 P. A. AGUILAR-RODRÍGUEZ ET AL.

pollen and, in some cases, (9) edible plant tissues as are the second most common vertebrate pollinators of
rewards (see Cox 1984; Cunningham 1995). However, the Bromeliaceae, more frequently reported than pas-
these characteristics are highly variable among plant serine birds or opossums, and surpassed only by hum-
families and some of the allegedly chiropterophilous mingbirds (Trochilidae) (Martinelli 1994, 1997;
traits (such as nocturnal anthesis) are not exclusive to Kessler and Krömer 2000; Araujo et al. 2004; Canela
plants pollinated by bats. A possible explanation for and Sazima 2005; Hornung-Leoni and Sosa 2006;
this is that bats present relatively low floral constancy, Krömer et al. 2006; Queiroz et al. 2016). Most of our
so a bat can visit flowers of many species in a single current knowledge of bat-pollinated bromeliads,
night and will thus frequently deliver mixed pollen defined as species in which bats perform most of the
loads to receptive stigmas (Fleming et al. 2005; legitimate visits to the flowers, comes from opportu-
Muchhala et al. 2008; but see evidence for differential nistic observations. Comprehensive and quantitative
pollen placement by bats in Muchhala and Thomson information remains scarce, and a detailed under-
2012; Stewart and Dudash 2016). In this sense, most standing of the reproductive biology exists for only a
bat-pollinated plants have evolved as a pollination- few species (e.g. Pitcairnia albiflos, Pseudalcantarea
guild, comprising non-related species that share a macropetala, Werauhia gladioliflora; Aguilar-
pollinator with similar behaviour and morphology Rodríguez et al. 2014; Wendt et al. 2001; Cascante-
(Muchhala and Jarrín-V 2002) rather than being spe- Marín et al. 2005; Tschapka and von Helversen 2007).
cialised to one bat species pollinating only one plant Variation in floral adaptations exhibited by
species (but see Centropogon nigricans and the bat Bromeliaceae probably accounts for their adaptability
Anoura fistulata; Muchhala et al. 2005; Muchhala in terms of changing traits associated with different
and Thomson 2009). pollinators and reproductive systems, even between
The study of bat-pollination has developed closely related species (Gardner 1986; Varadarajan
considerably since the early observations of and Brown 1988; Benzing 2000; Givnish et al. 2014).
Moseley (1870) and Burck (1892). The beginning To consolidate the rather scattered information
of formal studies dates back to the 1930s and 1960s currently available, we present a comprehensive
(e.g. Porsch 1932 or Vogel 1958, 1969; see also van review based on all available reports of bat polli-
del Pijil 1961). Classic studies by Faegri and van nation in the Bromeliaceae, including the identity
der Pijl (1979) and Dobat and Peikert-Holle of bat species involved where available. We
(1985) then provided detailed information about obtained information about chiropterophilous
the chiropterophilous syndrome and confirmed a bromeliads from an extensive review of the avail-
number of bat-pollinated plants, to which other able literature, as well as common scientific data-
examples have since been added (e.g. Fleming et bases including ISI Web of Science, SciELO and
al. 2009; Geiselman and Defex 2015). Redalyc and internet search engines such as
The Bromeliaceae is one of the most frequently Google Scholar, using operators such as ‘chirop-
reported bat-pollinated plant families (von Helversen terophily’ and ‘bromeliads’ or ‘Bromeliaceae’;
1993; Fleming et al. 2009). It includes more than 3500 ‘bat-pollination’ and ‘Anoura’, ‘Glossophaga’
herbaceous species from 70 genera (Barfuss et al. 2016; and ‘Lonchophylla’; ‘Alcantarea’, ‘Encholirium’,
Gouda et al. 2017). The specialised nectarivorous bats ‘Pitcairnia’, ‘Puya’, ‘Vriesea’ and ‘Werauhia’’,
responsible for most reports of bat-pollination in the either in the title, keywords or abstract. In addi-
Neotropics are found within the subfamilies tion, we searched the grey literature (e.g. theses)
Glossophaginae and Lonchophyllinae, among the and cross-checked references between studies,
highly diverse leaf-nosed bats (Phyllostomidae). including those provided in Fleming et al. (2009)
These bats comprise 20 genera and more than 50 and Geiselman and Defex (2015), as well as asking
species (Simmons 2005; Mantilla-Meluk and Baker certain authors to suggest publications related to
2010; Bolzan et al. 2015; Moratelli and Dias 2015; the topic (see Acknowledgements). Furthermore,
Cirranello et al. 2016). in order to improve our understanding of the
Pollination by vertebrates occurs within the evolution of chiropterophily in the Bromeliaceae,
Bromeliaceae more frequently than insect pollination we constructed a phylogenetic tree for the family,
(Benzing 2000; Kessler and Krömer 2000; Canela based on chloroplast (cp) DNA sequence varia-
and Sazima 2005; Krömer et al. 2006) and some tion, and used this tree as a means to examine
species, especially among the Tillandsioideae and the origin of chiropterophily in different clades
Bromelioideae subfamilies, present a mixed pollina- of the family and to determine which pollination
tion system (Benzing 2000; Givnish et al. 2014). Bats systems might be ancestral across these clades.
 PLANT ECOLOGY & DIVERSITY 3

We summarise and discuss some of the traits Rica (3000 m a.s.l.; Salas 1973). Most studies, how-
that characterise chiropterophily within brome- ever, have been conducted at elevations between
liads, presenting an overview of documented 1000 and 1600 m a.s.l. (e.g. Sazima et al. 1989;
bat-pollinated species and their geographic distri- Vogel 1969; Martinelli 1997; Kaehler et al. 2005;
butions. Finally, we highlight some additional Fabián et al. 2008; Aguilar-Rodríguez et al. 2014;
bromeliad species that appear to exhibit the chir- Aguilar-Rodríguez et al. 2016; see Tables S2, S3).
opterophilous syndrome and which may also Most reports regarding bromeliads with chir-
therefore be pollinated by bats. opterophilous traits are from species of the sub-
family Tillandsioideae in humid montane forests
(i.e. Aguilar-Rodríguez et al. 2014; Salas 1973;
Studies of bat-pollinated bromeliads
Ramírez and Seres 1994; Sazima et al. 1995; Seres
Following early comments about chiropterophi- and Ramírez 1995; Martinelli 1997; Muchhala and
lous floral traits in the bromeliad family by Müller Jarrín-V 2002; Krömer 2003; Cascante-Marín et
(1897) and Porsch (1932, 1934)), initial field al. 2005; Kaehler et al. 2005; Krömer et al. 2005,
observations were reported by Vogel (1969), with 2007). Other studies have been conducted in tro-
more detailed studies conducted in the 1990s (e.g. pical rainforest habitats (e.g. von Helversen 1993;
Martinelli 1994, 1997; Sazima et al. 1999). Martinelli 1994, 1997; Sazima et al. 1995, 1999;
Bromeliads with apparently chiropterophilous Tschapka and von Helversen 2007), while some
flowers are most commonly found in wet lowland have taken place on granitic/rocky outcrops, espe-
forests (Kessler and Krömer 2000; Fleming et al. cially for species of the subfamily Pitcairnioideae
2005, 2009), where the diversity of flower-visiting (Fischer 1994; Martinelli 1994; Wendt et al. 2001;
bats is also highest. However, this contrasts with Christianini et al. 2013), in deserts (Kessler and
the diversity of confirmed bat-pollinated brome- Krömer 2000), the Brazilian ‘cerrado’ (Sazima et
liads, which peaks at high elevations such as in the al. 1989) and in tropical coastal ‘restinga’ and
humid montane forests of the Andean region ‘caatinga’ (e.g. Fischer 1994; Queiroz et al. 2016)
(Benzing 2000; Givnish et al. 2011). This apparent habitats.
contradiction might be due to sampling bias, with According to our literature review, bats have
researchers focusing on few and frequently (or been confirmed as floral visitors for 42 bromeliad
recurrently) studied regions. Only a handful of species belonging to four of the eight
studies cover large study areas or a broad altitudi- Bromeliaceae subfamilies (Table 1). Most reports
nal range (e.g. Martinelli 1994, 1997; Sazima et al. suggest that bats are the only probable pollinator
1999; Kessler and Krömer 2000; Krömer et al. of these bromeliads (Table S2). Bat pollination is
2006). Lowland records come from Brazilian also suggested for a number of additional species
Vriesea (5–500 m asl; Martinelli 1994; 1997; (based on their floral characteristics) and, thus, a
Sazima et al. 1995, 1999) and Costa Rican total of more than 100 species of bromeliads from
Werauhia (40 m; Tschapka and von Helversen 13 genera and five subfamilies (Tables S2, S3) are
2007), while records from high altitude have thought to be bat-pollinated. Most of these bro-
been obtained from the Bolivian Andes (over meliads belong to the genera Vriesea and
2000 m asl; Kessler and Krömer 2000) and Costa Werauhia of the subfamily Tillandsioideae (see

Table 1. The number of bromelia species confirmed or suggested to be bat-pollinated (i.e. with chiropterophilous floral traits) in the
literature.
Subfamily Genus Confirmed Putative Total
Bromelioideae Billbergia 1 1 2
Brocchinioideae Brocchinia 0 1 1
Pitcairnioidea Encholirium 5 0 5
Pitcairnia 6 10 16
Puyoideae Puya 2 2 4
Tillandsioideae Alcantarea 2 7 9
Guzmania 0 13 13
Lutheria 1 0 1
Pseudalcantarea 3 0 3
Stigmatodon 0 1 1
Tillandsia 2 0 2
Vriesea 13 18 31
Werauhia 7 28 35
Total 42 81 123
4 P. A. AGUILAR-RODRÍGUEZ ET AL.

Figure 1. Examples of some bat-pollinated bromeliads: (a) Alcantarea imperialis, (b) Billbergia horrida, (c) Encholirium spectabile,
(d) Pitcairnia albiflos, (e) Pitcairnia recurvata, (f) Tillandsia heterophylla, (g) Puya ferruginea, (h) Vriesea longiscapa, (i), Werauhia cf.
sanguinolenta visited by Lonchophylla robusta. Note the pale-colored corolla in all species, and the predominance of the cup-like
floral morphology in the members of the Tillandsioideae subfamily (e, f, h, i). Photos: A, C, D, H: Elton M. C. Leme. B: Ana Paula
G. Faria. E, F, I, J: Pedro A. Aguilar-Rodríguez. G: Thorsten Krömer. I: Marco Tschapka.

Figure 1), and most (Table S2) are reported to cacti and Agave in North America (Fleming 2004),
possess a ‘typical’ chiropterophilous flower and may therefore receive many secondary polli-
(short, wide and large flowers with pale petals; nators or opportunistic floral visitors.
Figure 1) and are visited mainly by bats of the Even if many bromeliad species have the potential
genus Anoura, especially at elevations of 1000 m for self-pollination (Matallana et al. 2010: but see
a.s.l. or higher in tropical montane forests. Encholirium spp.; Christianini et al. 2013; Hmeljevski
However, Alcantarea and Encholirium species et al. 2017), pollinators such as far-ranging bats are
possess less restrictive floral morphologies that valuable in terms of promoting gene flow among plant
may also allow visits by non-specialised nectari- populations (see Vriesea gigantea in Palma-Silva et al.
vorous bats. The dry and hot environment and the 2009). Such occasional cross-pollination by animal
reduced presence of bats in habitats where these pollen vectors, including bats, could be important
Alcantarea and Encholirium species occur might in terms of mitigating the potential disadvantages
act to promote these more generalist floral fea- of a reproductive system that is based mainly on
tures. In addition, these species are frequently self-pollination (Shivanna and Tandon 2014), as well
locally abundant and may thus serve as an impor- as maintaining the boundaries between species
tant resource for many different flower-visiting (see Pitcairnia spp. in Palma-Silva et al. 2011).
animals, a role analogous to that of the columnar Nevertheless, Barbará et al. (2007) suggested that
 PLANT ECOLOGY & DIVERSITY 5

pollinators of Alcantarea imperialis (which probably a tube-like purple or yellow corolla and diurnal
include Anoura caudifer and Artibeus lituratus; anthesis. Ornithophily seems to be the main pollina-
Martinelli 1994) are likely to be insufficient in number tion mode within all genera, including the allegedly
to maintain connectivity among populations of this chiropterophilous bromeliads (Givnish et al. 2014;
species in the Brazilian ‘inselbergs’, having found Figure 2), except for the genera Encholirium
that such populations were highly differentiated (Givnish et al. 2007; but see Encholirium heloisae;
genetically. However, further studies are required to Christianini et al. 2013) and Werauhia (Grant
confirm this hypothesis. In species that produce many 1995). The clades comprising the hummingbird-pol-
flowers per day, for example Alcantarea spp., foraging linated species are three to five times richer in species.
bats might promote geitonogamy over allogamy. This, Ornithophilous bromeliad lineages have higher
in addition to short pollen movement distances rates of diversification than species that possess floral
between individuals within populations (see discussion traits associated with other syndromes, as demon-
in Hmeljevski et al. 2017) could lead to a certain level of strated by Givnish et al. (2014). This might be a result
inbreeding. of the synergy between 1) the occurrence in humid
montane habitats, which 2) favours epiphytism, and
3) having a water-holding ‘tank-like’ body (phyto-
Evolution of bat-pollinated bromeliads
telma), all three of which promoted the development
The Bromeliaceae originated in South America in of avian pollination (Givnish et al. 2014).
the Guiana shield region over 70–100 million years It is noteworthy that the putative pollination
ago (mya) (Givnish et al. 2007, 2011), but extant syndromes of bromeliad species included in phy-
lineages did not appear until 15–19 mya. This logenetic analyses are not always mentioned (see
coincides with the origin of epiphytism in the Givnish et al. 2014 as an exception), and where
family and the rise of the Andean mountain they are, most species are reported as bird-polli-
chain (Barfuss et al. 2005; Givnish et al. 2007) in nated, leading to a deficit of species with other
the so-called ‘bromeliad revolution’ (ca. 15 mya), pollination syndromes.
when the family spread to other parts of tropical To examine the evolution of chiropterophylous
and subtropical America, such as the Northern bromeliad species in more detail, we generated a
Andes and Central America (Givnish et al. 2014). phylogenetic tree of bromeliads and mapped the
Our understanding of the phylogeny of the pollination syndromes of these species on to the
Bromeliaceae has changed repeatedly over the last tree. For this purpose, we modified the cpDNA
two decades, revealing that some groups formerly sequence super-matrix used in Escobedo-Sarti et
defined by floral morphology and habit are in fact al. (2013) with additional sequences downloaded
paraphyletic (e.g. Terry et al. 1997; ; Barfuss et al. from GenBank and selected a wide spectrum of
2005; Givnish et al. 2007, 2011, 2014; Escobedo- bromeliad species with different pollination syn-
Sarti et al. 2013; see the revision in Palma-Silva et dromes, including more chiropterophilous species
al. 2016). However, the monophyly of (totalling 41) than in previous phylogenetic studies
Bromeliaceae has been commonly accepted on the of the family. All eight subfamilies of Bromeliaceae
basis of morphological, anatomical and molecular (Tillandsioideae, Bromelioideae, Brocchinioideae,
characters (Gilmartin and Brown 1987; Crayn et al. Lindmanioideae, Hechtioideae, Puyoideae,
2004; Givnish et al. 2004, 2007). Navioideae and Pitcairnioidea s.str.) proposed by
The ancestral pollination mode of bromeliads Givnish et al. (2007), and which are mainly char-
seems to be entomophily, but ornithophilous flow- acterised by molecular data and the morphology of
ers evolved independently on two occasions: first in flowers, fruits and seeds (Givnish et al. 2007, 2011),
the ancestral Tillandsioideae (ca. 15.4 mya) were represented in the analysis. The final matrix
and secondly, in the common ancestor of included 136 taxa and 9049 characters (ca. 6.13%
Bromelioideae/Pitcairnioideae/Puyoideae (ca. 14.4 of gaps, with most taxa having complete sequences
mya; Givnish et al. 2007, 2014). for all cpDNA regions), based on the regions: atpB-
Within the family, there are more bromeliads with rbcL, ndhF, psbA-trnH, rpl32-trnL, rps16, trnK-
ornithophilous than with chiropterophilous charac- matK and trnL-F (see Table S1 for vouchers of all
ters (60–80% vs. 10–20%; see Martinelli 1994, 1997; sequences). In this matrix, 135 species belong to
Sazima et al. 1999; Kessler and Krömer 2000; Araujo the Bromeliaceae, while Rapatea paludosa
et al. 2004). The characters of bird pollination include (Rapateaceae) was selected as an outgroup (see
red floral bracts, showy but small flowers and anthers, Givnish et al. 2007, 2011, 2014).
6 P. A. AGUILAR-RODRÍGUEZ ET AL.

Figure 2. Phylogenetic tree of bromeliads and their pollination syndromes. Different colours on the nodes indicate different
pollination syndromes (green: entomophily, red: ornithophily, blue: chiropterophily, pink autogamous, yellow: generalist
pollination). Notes: Alcantarea duarteana is not mentioned in the literature as bat-pollinated, but we consider that it possible
is; however, since Siqueira Filho (2003) suggested also moth-pollination within the genus, we provisionally refer to it as bimodal.
Tillandsia malzinei is referred as a species with chiropterophilous flowers; it was considered a member of the former Vriesea
subg. Xiphion (currently, the accepted name would be sect. Synandra; see Barfuss et al. 2016). We describe Tillandsia
heterophylla with a bimodal pollination syndrome since diurnal floral visitors could serve as pollinators, even if most of the
visits are by bats (Aguilar-Rodríguez et al. 2016). The form of Pitcairnia flammea represented in this figure with chiropter-
ophilous floral traits refers only to P. flammea var. pallida.

We aligned sequences of the regions using Muscle among regions, and additionally used the phyloge-
(Edgar 2004) and analysed the data with JModelTest netic tree of Escobedo-Sarti et al. (2013) to define
2.7.1 in order to estimate the evolutionary model topological constraints, assuming an accurate topol-
(Darriba et al. 2012). The best-fitting model for the ogy that would reflect the phylogeny of bromeliads.
regions was GTR+I + G. We concatenated the data in Pollination syndromes of bromeliad species
a supermatrix that was analysed in GARLI (Zwickl according to the literature, and our field observations,
2006) on the CIPRES Science Gateway platform were coded as follows: 0, entomophily; 1, ornithoph-
(Miller et al. 2010). We used independent models ily; 2, chiropterophily; 3, autogamous; 4, generalist
 PLANT ECOLOGY & DIVERSITY 7

Figure 3. Examples of some bromeliads with diurnal anthesis: A) Alcantarea geniculata, bird-pollinated, B) Aechmea tilland-
sioides, bimodal pollination by birds and insects, probably bees, C) Billbergia issingiana, bird-pollinated, D) Catopsis sessiliflora,
probably entomophilic, E) Pitcairnia ringens, bird-pollinated, F) Puya riparia, probably bird-pollinated, G) Tillandsia multicaulis,
bird-pollinated, H) Tillandsia violacea, bird-pollinated, I) Vriesea platynema, probably bird-pollinated. Note the bright-coloration,
principally red, of the bird-pollinated species, and the tubular shape of the corolla. Photos: A, I: Elton M. C. Leme. B, C, D, F, G, H:
Thorsten Krömer. E: Pedro A. Aguilar-Rodríguez. This image was elaborated by J. Escobedo Sarti, one of the coauthors.

pollination (species with two or more complementary Our phylogenetic tree (Figure 2) strongly suggests
pollinators from different functional groups; see that chiropterophilous floral features have appeared
Schmid et al. 2011; Aguilar-Rodríguez et al. 2016). multiple times from ornithophilous ancestors in the
We used this information to assemble a character Bromeliaceae, especially within Tillandsioideae.
matrix and selected the most appropriate evolution- Thus, all clades of the bromeliads with chiroptero-
ary model for pollination syndrome, based on this philous traits are embedded within groups of
matrix and the ML tree obtained with GARLI, as well ornithophilous species, and in general, chiroptero-
as the Geiger package and APE (Paradis et al. 2004; philous floral traits appear derived from ornithophi-
Harmon et al. 2008), implemented in R (R Core Team lous or sphingophilous floral traits (Tschapka and
2016). The test consisted of calculating the plausibility Dressler 2002; von Helversen and Winter 2003;
of all of the evolutionary models available in Geiger. Fleming et al. 2009).The transition within plant
Based on the Akaike Information Criterion (AIC), families from ornithophilous to chiropterophilous
we then identified the Automatic Relevance traits seems to be more common than in the opposite
Determination regression (ARD) model to present direction, and also more common than transitions
the best fit, before reconstructing ancestral characters from other syndromes (e.g. melittophily or sphin-
with more plausibility and the ARD model. gophily) to chiropterophily (van der Niet and
8 P. A. AGUILAR-RODRÍGUEZ ET AL.

Johnson 2012; Rosas-Guerrero et al. 2014), suggest- develop similar, highly specific adaptations to bat-pol-
ing that it may constitute an evolutionary ‘dead-end’ lination as in other, much older plant families such as
(Tripp and Manos 2008). Within the Bromeliaceae, the Asparagaceae (Agave) and Cactaceae (Fleming et
this transition seems to have occurred independently al. 2009; Rosas-Guerrero et al. 2014). In this section,
in different genera, particularly within the we examine the chiropterophilous floral traits present
Tillandsioideae subfamily (Versieux et al. 2012; in bromeliads.
Givnish et al. 2014). As a whole, chiropterophilous
flowers have appeared independently many times in
clades with different evolutionary histories; in differ- Phenology related to bat visitation
ent genera among the ‘core’ members of the
Anthesis in the bat-pollinated bromeliads mainly starts
Tillandsioideae subfamily, once in the Encholirium
at dusk, although some species open their flowers
clade, independently in Pitcairnia within the
during mid-afternoon (i.e. Encholirium spectabile, E.
Pitcairnioideae, and separately in Puya and
vogelii, Tillandsia heterophylla, Vriesea bituminosa
Billbergia, each within its own subfamily. However,
var. bituminosa, V. hydrophora, V. longicaulis, V. long-
it is important to acknowledge that the taxonomy of
iscapa and Werauhia ororiensis; Salas 1973; Martinelli
this family has proved difficult to resolve (Givnish et
1994, 1997; Seres and Ramírez 1995; Christianini et al.
al. 2007; Barfuss et al. 2016).
2013; Aguilar-Rodríguez et al. 2016). Duration of
Some bat-pollinated bromeliads may use
anthesis varies considerably between species; from
diurnal visitors as secondary pollinators (see
the shortest in Pitcairnia (5 to 10 h in P. albiflos and
examples in Billbergia, Encholirium, Guzmania,
P. flammea; Martinelli 1994; Wendt et al. 2001) to the
Pitcairnia and Tillandsia; Krömer 2003;
longest in Billbergia horrida Regel (24 h; Marques et al.
Christianini et al. 2013; Marques et al. 2015;
2015) and Encholirium subsecundum (3 d; Sazima et al.
Aguilar-Rodríguez et al. 2016; Queiroz et al.
1989).
2016; Silva Jorge et al. 2018), with humming-
The crepuscular anthesis of these species might be
birds being the most frequently reported floral
a remnant of ornithophilous ancestors with diurnal
visitors/secondary pollinators (Table S1). In
anthesis (except in Pitcairnia recurvata, which opens
many angiosperms, secondary pollinators fre-
flowers late at night, P. A. Aguilar-Rodríguez et al.,
quently coincide with the probable ancestral
unpublished data). In addition, an extended anthesis
pollinator group (Rosas-Guerrero et al. 2014),
could increase the probability of pollination by other
and this is reflected in our inference that chir-
animals, such as hummingbirds, although these
opterophilous features in bat-pollinated brome-
might not be as effective pollinators as bats
liads are derived from ornithophilous traits
(Christianini et al. 2013; Marques et al. 2015;
present in hummingbird-pollinated species.
Aguilar-Rodríguez et al. 2016; Queiroz et al. 2016;
Changes in the corolla, as well as the time of
Silva Jorge et al. 2018; but see Tschapka and von
anthesis/flower duration, might be important
Helversen 2007). Hummingbirds, moths and, to a
modifications to restrict hummingbird visits.
lesser extent, various species of bees, are often fre-
However, there is evidence of ‘intermediate’
quent alternative visitors of bat-pollinated brome-
floral traits in some species, suggesting either
liads (Table S1); however, the timing of visits in
that this transition is ongoing or that this is a
relation to stigma receptivity, and floral morphology
steady state in which different functional groups
may preclude pollination by bees (Aguilar-Rodríguez
of pollinators complement each other in contri-
et al. 2014; Christianini et al. 2013; Marques et al.
buting to the reproductive success of the bro-
2015; Aguilar-Rodríguez et al. 2016).
meliad (e.g. Billbergia horrida, Tillandsia
heterophylla; Marques et al. 2015; Aguilar- Nearly all bat-pollinated bromeliad species studied
Rodríguez et al. 2016). present a ‘steady-state’ flowering pattern (i.e. each
individual in the population produces one or two
flowers per night, over many days or weeks; sensu
The chiropterophilous syndrome within Gentry 1974), with an annual or biannual flowering
Bromeliaceae
period (Martinelli 1997). Two exceptions are
The comparably recent origin of both Neotropical Alcantarea imperialis and Vriesea gigantea with large
nectarivorous bats and bromeliads (Givnish et al. inflorescences that open up to 18 and 10 flowers per
2007, 2011; Baker et al. 2012) indicates that there has day, respectively (Martinelli 1994; Araujo et al. 2004).
been only a relatively short time span in which to Another exception, Pitcairnia flammea, exhibits a
 PLANT ECOLOGY & DIVERSITY 9

‘cornucopia’ pattern (i.e. various individuals in the paniculata; Beaman and Judd 1996; Barfuss et al.
population present many flowers simultaneously, 2016). Ornithophilous bromeliads share several
over a short period of time; Martinelli 1994), while floral features with chiropterophilous species, and
individuals of E. spectabile may flower over 30 days the latter pollination syndrome may have evolved
with an entire population of E. subsecundum some- quite easily from the former, as indicated by several
times flowering for three to four months (Sazima et al. evolutionary shifts observed among other species
1989). (van der Niet and Johnson 2012) (see also exam-
Sympatric bromeliad species that share polli- ples among the genus Alcantarea; Versieux et al.
nators may present staggered flowering seasons 2012).
to avoid competition for pollinators (Fischer Floral morphology plays an important role in pol-
1994; Araujo et al. 2004), while simultaneously len placement on the body of the bat, and in reducing
preventing hybridisation. However, this stag- the amount of pollen wasted on heterospecific stig-
gered flowering pattern is not present in all bro- mas (Stewart and Dudash 2017). In most species of
meliad communities (see Martinelli 1997; Vriesea and Werauhia, the anthers are positioned on
García-Franco et al. 2001; Versieux et al. 2012). the dorsal or ventral side of the corolla, but in some
species, such as Vriesea gigantea and V. limae, or in
Stigmatodon spp., the anthers are radially or laterally
Floral morphology
oriented (von Helversen 1993; Sazima et al. 1995;
While the flower buds of many Tillandsioideae are Siqueira Filho 2003). One important difference
initiated in two opposite rows along the inflorescence, between diurnal and chiropterophilous bromeliad
they undergo reorientation during ontogeny to face species is that the former present tubular corollas
the same direction at anthesis (e.g. Werauhia spp.), with stamens arranged in a bundle around the centre
thus allowing a bat to visit the inflorescence over of the flower such that pollen is deposited on the beak
several days always from the same direction. Most of perching and hovering birds; in contrast, chirop-
bat-pollinated bromeliads have pale yellow, green, terophilous species have larger, broader, cup-like and
creamy or white petals, occasionally with a reddish generally quite open flowers with stamens spreading
tint on the calyx, for example some species of from the centre. Stamen position during anthesis is
Alcantarea and Vriesea (Martinelli 1994, 1997; an important floral trait that separates chiropterophi-
Sazima et al. 1999; Moura and Costa 2014; Versieux lous from ornithophilous Alcantarea species (spread-
and Wanderley 2015). At least four distinct flower ing vs. bundle forming; Versieux et al. 2012). The
shapes can be distinguished among such bromeliads: model of Muchhala (2007) suggests that wider cor-
i) the zygomorphic tube-type of some Billbergia, ollas would be selected in flowering species where
Vriesea, Pitcairnia and Puya species (Sazima et al. bats conduct more than 44% of all visits, in order to
1999; Kessler and Krömer 2000; Schmid H 2000; match the bat morphology and reduce pollen waste
Kowalski and Tardivo 2015; Macías-Rodríguez et al. by hummingbirds. This might apply to flowers of
2007; Scultori Da Silva 2009; Figure 1(b,c,g)); ii) the Werauhia (Tschapka and von Helversen 2007) and
zygomorphic bell-shape type present in Vriesea and even to the spirally twisted petals of P. macropetala.
Werauhia and even Tillandsia (Utley 1983; Grant In the latter species, the petals form an open corolla,
1995; Leme 1995; Tschapka and von Helversen 2007; in which only a bat can contact both the exerted
Figure 1(f,h,i)); iii) the heliciform actinomorphic anthers and the stigma simultaneously during a visit
flower with strap-like petals and protruding stamens to the flower, while hummingbirds fail to do this due
in Alcantarea and Pseudalcantarea (Martinelli 1994; to their visitation behaviour and because of changes
Krömer et al. 2012; Versieux et al. 2012; Aguilar- to floral parts following anthesis, for example the
Rodríguez et al. 2014; Figure 1(a)); and iv) the brush- stigma pointing downwards due to turgor loss in
type flower of Encholirium (Sazima et al. 1989; the style, so the hummingbird does not come into
Christianini et al. 2013; Queiroz et al. 2016; Gomes direct contact with it (Aguilar-Rodríguez et al. 2014).
et al. 2018; Figure 1(c); but see Encholirium horridum Pollen of bell-shaped chiropterophilous bromeliad
in Hmeljevski et al. 2017). flowers, for example in Werauhia and Vriesea, is
These distinct flower shapes could be related to placed on the head of the bat (Figure 1(k)), while in
intrinsic differences in the floral morphology actinomorphic helicoiform flowers, such as in
between subfamilies and genera (but see the simi- Pseudalcantarea macropetala, it is deposited on the
larities between flowers in Alcantarea, ventral side of the wings (Aguilar-Rodríguez et al.
Pseudalcantarea, Tillandsia baliophylla and T. 2014).
10 P. A. AGUILAR-RODRÍGUEZ ET AL.

The position of the stigma below the anthers and environmental factors may also affect nectar mea-
facing downwards in some Vriesea and Werauhia surements (Jakobsen and Kritjánsson 1994;
species might allow the stigma to make contact with Willmer 2011). Another explanation could be that
the bat fur before it reaches the anthers, thus favour- the rather short evolutionary time span of the
ing outcrossing when the flower is fully opened, in mutualism between bats and bromeliads may not
spite of self-compatibility (Salas 1973). yet have allowed the development of a clear nectar
production pattern (Rodríguez-Peña et al. 2016).
In general, bat-pollinated bromeliads present lower
Nectar characteristics
nectar volumes and sugar concentrations than other
Nectar is one of the main rewards offered to bat-pollinated plants (Tschapka and Dressler 2002;
flower-visiting bats in the Neotropics, with the Fleming et al. 2009). The study of Krömer et al.
nectar of bat-pollinated species possessing charac- (2008) showed that chiropterophilous species had the
teristics that distinguish it from that provided by lowest sugar concentration of all bromeliads studied
plant species to other floral visitors ((von (11.5 ± 4.0%; but see W. gladioliflora; Tschapka and
Helversen and Winter 2003), see also Mosti et al. von Helversen 2007). The low nectar sugar concentra-
2013 for information about nectar secretion in tion in bat-pollinated bromeliads (Table S1) might
bromeliads). In fact, some of the most robust evi- require visitors to consume rather high quantities of
dence for bat-pollination in bromeliads is based on nectar to meet their energetic needs (von Helversen
the particular characteristics of nectar. The and Reyer 1984), thus necessitating an increased num-
sucrose/hexose ratio of various bromeliad species ber of flower visits. More dilute nectar may evolve
studied by Krömer et al. (2008) indicated that the when competition for food among bats is higher
nectar of chiropterophilous species is hexose-rich, (Nachev et al. 2017), but we still lack sufficient infor-
leading them to suggest that some Bolivian mation to confirm this hypothesis.
Guzmania species which produce hexose-rich nec-
tar are bat-pollinated, even though they present
Floral scent
brightly-coloured floral bracts more often asso-
ciated with bird-pollinated species. Many brome- Floral scent probably serves as a long-distance
liads with chiropterophilous floral traits within attractant for bats, but will also aid close-range
Tillandsioideae show tank-forming rosettes, in location of flowers (von Helversen et al. 2000;
which water and detritus are collected. Givnish et Gonzalez-Terrazas et al. 2016). Many authors
al. (2014) suggested that the tank-habit might ori- report that chiropterophilous bromeliads present
ginally have favoured ornithophily, since the a characteristic floral scent described as ‘musky’
water-filled tanks may facilitate the production of or garlic-like (Table S1, Table S2). However, to
relatively large amounts of nectar. However, the date, the scent volatiles of only two species of
same reasoning holds true for chiropterophilous bat-pollinated bromeliads have been identified: in
species since nectarivorous bats require an even Werauhia gladioliflora (Bestmann et al. 1997) and
higher amount of nectar than hummingbirds Pseudalcantarea macropetala (Aguilar-Rodríguez
(Tschapka and Dressler 2002). Nectar volume var- et al. 2014). In the former, dimethyl disulphide
ies widely among chiropterophilous bromeliads was present, a volatile that is innately attractive to
species (Table S1, Table S2), ranging from 4 µl in New World nectarivorous bats (von Helversen et
Encholirium vogelli (Christianini et al. 2013) to al. 2000). Sulphur-containing compounds are com-
1129 µl in W. gladioliflora (Tschapka and von mon in the scent of chiropterophilous plants, but
Helversen 2007). Moreover, sugar concentration not omnipresent (Knudsen and Tollsten 1995).
ranges from 4% in Encholirium subsecundum Thus, in the case of P. macropetala, no sulphur-
(Sazima et al. 1989) to 21% in Vriesea atra var. containing volatiles were found. It is possible that
atra (Fischer 1994). This high variability in nectar this specific attractant has evolved in Werauhia, a
volume and constituency could, at least partly, be genus highly specialised in bat-pollination, but not
attributed to the different methodologies used in yet in P. macropetala. The volatiles identified from
different studies. For example, measurement of 13 species of different Bromeliaceae genera of all
accumulated nectar over the entire life of the pollination syndromes are highly variable, even
flower vs. standing crop measurements (Corbet within the same genus (Hilo de Souza et al.
2003). Furthermore, there is interplant variation 2016), and some of the volatile compounds
of nectar traits (e.g. Hodges 1993) and detected in these diurnally flowering species are
 PLANT ECOLOGY & DIVERSITY 11

shared with those identified for W. gladioliflora by 2011) and comprises the ‘core’ species of the nectar-
Bestmann et al. (1997) and P. macropetala in ivorous bat guild at higher elevations (Fleming et al.
Aguilar-Rodríguez et al. (2014). 2005; Moras et al. 2013), as well as being the bat genus
most frequently reported to visit bromeliads (Table
S1). In addition to other characteristics (i.e. long
Acoustic signals
dense fur, reduced uropatagium, hairy feet and toes,
A study by von Helversen et al. (2003) has sug- small ears and a higher body mass than the lowland
gested that W. gladioliflora might have flowers with nectarivorous bats; Soriano et al. 2002), the basal
special echo-reflecting properties. Unlike other metabolic rate of most Anoura species (Figure 4(a,b))
bromeliads, W. gladioliflora presents flowers that reflects their tolerance to lower environmental tem-
are largely embedded within the stalk of the inflor- peratures (Ortega-García et al. 2017; but see Anoura
escence, covered by bracts, and only exposed when caudifer in the lowland of the Amazonian region) and
fully opened. The only section protruding from the allows them to remain active during cold nights at
stalk is the distal portion of the corolla, as well as high elevations. This might be the reason for the
the stamens and stigma. The cup-like corolla of frequent reports of members of this bat genus as floral
Werauhia could produce a distinctive echo that visitors of bromeliads at high elevations. In contrast, at
might help a bat investigate the inflorescence and low elevations in Mexico and Central America, the
locate the flower entrance (von Helversen et al. ‘core’ nectarivorous bat-fauna is formed by the genus
2003). Glossophaga (Fleming et al. 2005, 2009), as shown by
studies carried out in Costa Rica on W. gladioliflora
(Tschapka 2004; Tschapka and von Helversen 2007).
Characteristics of bats visiting bromeliads
To date, a total of 19 species from ten phyllostomid
The bat family Phyllostomidae of New World Leaf- genera have been recorded visiting bromeliad flowers
nosed bats originated about 30.3 mya (Rojas et al. (Table S1), 15 of which belong to the specialised
2016). Nectarivorous feeding habits evolved twice nectarivorous subfamilies Glossophaginae and
within the family (Baker et al. 2012; Tschapka et al. Lonchophyllinae: Anoura (4 spp.), Glossophaga (2
2015); firstly, around 21.55 mya (23.4–19.7; subfamily spp.), Hylonycteris (1 sp.), Lichonycteris (1 sp.),
Glossophaginae) and secondly around 11.13 mya Lonchophylla (5 spp.), Platalina (1 sp.) and
(11.4–10.9; subfamily Lonchophyllinae) in South Xeronycteris (1 sp.) (Figure 4). Furthermore, three
America, with the exception of some genera frugivorous species from the subfamily
(Brachyphylla, Erophylla, Phyllonycteris, Monophyllus, Stenodermatinae are reported: Artibeus lituratus and
Leptonycteris and Glossophaga; Rojas et al. 2016) that Pygoderma bilabiatum, visiting bromeliads with open
originated in the Antilles. The oldest fossil of a or brush-like corollas (Alcantarea imperialis, A. regina,
Neotropical nectarivorous bat is from Palynephyllum Vriesea bituminosa var. bituminosa and V. hoehneana;
antimaster (about 12–13 mya; Morgan and Czaplewki Martinelli 1994; Kaehler et al. 2005); and Carollia
2012) and the majority of extant nectarivorous bat perspicillata visiting the ‘tube-like’ flower of
genera originated around 10–5 mya (Rojas et al. Pitcairnia paniculata (Maguiña et al. 2012). In addi-
2016). There is evidence of the presence of Anoura tion, Phyllostomus discolor from the Phyllostominae
in the Peruvian Andes dating from the Pleistocene subfamily visits and pollinates flowers of Encholirium
(at least 2.5 mya; Shockey et al. 2009). Thus, specialised spectabile (Queiroz et al. 2016).
nectarivorous bats occurred at the same time and It has been proposed that bats re-visit flowers over
location as the ‘bromeliad revolution’, with most of the course of the night in a behaviour known as ‘trap-
the extant nectarivorous genera already present by the lining’ (sensu Janzen 1971; von Helversen 1993;
time of the origin of some bromeliad genera such as Fleming et al. 2009). Almost all authors who observed
Encholirium, Pseudalcantarea and Werauhia (Givnish bat visits to bromeliads have suggested the occurrence
et al. 2014). of such a foraging mode; however, no study to date has
The Andean region is particularly important for experimentally confirmed this behaviour. The dura-
species diversity of genera comprising bat-pollinated tion of the actual flower visit is extremely short (less
bromeliads, as well as for the bat genus Anoura than one second; Aguilar-Rodríguez et al. 2014;
(Patterson et al. 1996; Muchhala et al. 2008; Sazima et al. 1995; Wendt et al. 2001; Tschapka and
Mantilla-Meluk et al. 2009; Shockey et al. 2009; von Helversen 2007) and the bats usually insert their
Mantilla-Meluk and Baker 2010). This bat genus head into the corolla or lap the nectar directly from the
diversified over the last 10 mya or less (Rojas et al. inflorescence (as in Encholirium; Sazima et al. 1989).
12 P. A. AGUILAR-RODRÍGUEZ ET AL.

Figure 4. Bats identified as pollinators of Bromeliaceae: (a) Anoura caudifer (Glossophaginae subfamily), (b) Anoura geoffroyi
(Glossophaginae), (c) Glossophaga commissarisi (Glossophaginae), (d) Glossophaga soricina (Glossophaginae), (e) Hylonycteris
underwoodi (Glossophaginae), (f) Lichonycteris obscura (Glossophaginae), (g) Lonchophylla bokermanni (Lonchophyllinae), (h)
Artibeus lituratus (Stenodermatinae), (i) Pygoderma bilabiatum (Stenodermatinae), (j) Phyllostomus discolor (Phyllostominae).
Photographic credits: A, B, C, D, E, F, H: Marco Tschapka. G, I: Lena Geise. J: Pedro A. Aguilar-Rodríguez.

Some non-glossophagine bats, such as P. discolor density offered by this plant in Costa Rica reaches
(Queiroz et al. 2016; and see also reports of frugivor- up to 349.3 kJ/ha/day during the peak flowering
ous bats in; Martinelli 1994, 1997), cannot forage for season, thus constituting one of the most efficient
nectar during hovering-flight, and instead perch foraging options for bats (Tschapka 2004). Up to
upside down from the inflorescence to gain access. 28% of the chiropterophilous plants found within
High plant abundance and high nectar produc- the home ranges of Glossophaga commissarisi in
tion make bromeliads an important resource for Costa Rica were W. gladioliflora, making it the
bats, birds and insects (Howell and Burch 1974; most important plant resource for this bat
Siqueira Filho 2003; Maguiña Conde 2016; (Rothenwöhrer et al. 2011).
Maguiña and Amanzo 2016; Morales 2016;
Cordero-Schmidt et al. 2017). In the Maquiné
river valley in Brazil, Anoura caudifer were found Conclusions and future directions
to feed on an unidentified Vriesea species for six Chiropterophily has evolved in different lineages of the
months, accounting for over 50% of their total diet Bromeliaceae family, with evidence for convergent
during the austral summer and almost 18% of their evolution occurring across different phylogenetic
diet over the course of the year (Barros et al. 2013). clades. Key traits of chiropterophily, such as nocturnal
Due to the local high abundance of Werauhia anthesis and the length of floral structures, allow bats
gladioliflora and its high nectar volume, the energy to act as the principal or occasionally the secondary
 PLANT ECOLOGY & DIVERSITY 13

pollinator of chiropterophilous species (Rosas- Lexer et al. 2016) or even their role in hybridisation
Guerrero et al. 2014); however, most studies of such between species sharing habitats or pollinators
species have not specifically evaluated the contribu- (Wendt et al. 2001; Palma-Silva et al. 2011;
tions of other floral visitors in terms of pollination Versieux et al. 2012; Queiroz et al., 2015). The
success. The short evolutionary time span of the bat- effect of pollinators on species cohesion could be
bromeliad interaction has probably led to the main- particularly important considering the lack of other
tenance of some ‘generalist’ floral characteristics that prezygotic reproductive barriers among many bro-
currently correspond to a bimodal pollination system meliad species (Wendt et al. 2008, but see also
between primarily nocturnal and diurnal pollinators. Matallana et al. 2016).
In a few clades, more specialised flowers seem to have The presence of bat pollination among species
evolved relatively rapidly; for example, in Werauhia, of the Bromeliaceae family is associated with envir-
possibly the most specialised bat-pollinated genus in onmental conditions and elevation (Kessler and
the family, which is only ca. 5 mya (Givnish et al. 2014). Krömer 2000 ; Kessler 2002) and most of the
Bat visitation has been reported in more than 40 reported species are epiphytic or lithophytic.
species from four of the eight recognised bromeliad Borges et al. (2016), suggest that plant families
subfamilies. Some bat-pollinated species may have living in habitats with low water availability are
been overlooked in other subfamilies or presumed more likely to develop crepuscular or nocturnal
to be moth-pollinated (Gardner 1986; Rauh 1986, anthesis as a strategy to reduce nectar loss through
1987, 1990; Aguilar-Rodríguez et al. 2016). Only 24 evaporation. In addition, species living in arid
(35%) of the references presented in Table S1 actually regions (including Encholirium and Puya) and epi-
represent in-depth studies on the reproductive biol- phytic species grow under environmental water
ogy of Bromeliaceae. Of these, only seven (10%) were stress (Ruzana Adibah and Ainuddin 2011). In
conducted during the last 10 years, but even this this context, future work should focus on whether
relatively small number of studies yielded five chir- high altitude and precipitation are truly associated
opterophilic species previously unreported in the with pollination by bats among epiphytic species,
literature. This emphasises the importance of field- and whether drier environmental conditions
work for determining the extent of the distribution of favour bat pollination in large bromeliad species
this pollination syndrome within the family. such as those of the genera Alcantarea and
Pollination by vertebrates is considered to be an Encholirium.
important driver of speciation within the In order to understand the role of bat pollination in
Bromeliaceae (Givnish et al. 2014) and nectar-feed- a bromeliad species visited by several pollinators over
ing bats may therefore have played a significant role its geographical distribution, it is important to quan-
in the origin of some species, although this remains tify the relative contribution of each pollinator species
to be investigated in detail. The genus Werauhia, to reproduction within local populations. If a brome-
with 92 species (Gouda et al. 2017), seems to be liad flower has extended anthesis, lasting from after-
particularly well adapted to bat pollination, (see also noon until early morning, it may be visited by animals
Utley, 1983; von Helversen and von Helversen, 1975) other than or in addition to bats. With regard to
and offers interesting options for comparisons nectar-feeding bats, different species may carry a spe-
between species from differing regions, elevations cies-specific pollen load (King et al. 2013) and dissim-
and morphology; currently, detailed information on ilarities in size and behaviour might result in
the pollination system is only available for Werauhia differences in the relative contributions of each species
gladioliflora. In addition, the speciose genus to seed set, and thus may ultimately modify the breed-
Stigmatodon, formerly included in Vriesea, shows ing system of the species. There are bromeliad species
floral characteristics, including nocturnal anthesis, that seem to be self-incompatible and therefore
pale and greenish petals and musty scent (Barfuss et depend on pollinators for reproduction (e.g.
al. 2016), which suggest that all of its 18 species might Encholirium spp.; Christianini et al. 2013; Hmeljevski
be bat-pollinated, although supporting field evidence et al. 2017), while others are self-compatible and even
to date is limited. capable of autonomous self-pollination (T. hetero-
As is the case with almost all plant-pollinator phylla; Aguilar-Rodríguez et al. 2016, W. gladioliflora,
interactions, only scarce information exists regard- Tschapka and von Helversen 2007).
ing the contribution of bats to genetic population Finally, it is to be emphasised that the interaction
structure within bromeliad species (i.e. Wendt et between bromeliads and nectar-feeding bats presents
al. 2002; Barbará et al. 2007; Paggi et al., 2007; a multitude of interesting ecological and
14 P. A. AGUILAR-RODRÍGUEZ ET AL.

evolutionary aspects, ranging from pollinator beha- M.Cristina MacSwiney G. studies the ecology of bats and
viour to local adaptations and population genetic rodents. She is interested in the composition of bat assem-
consequences. Bromeliad species occur in different blage in different tropical habitats.
habitats all over the Neotropics, in high numbers and
often large population sizes. These characteristics
allow interesting studies that will help further our ORCID
understanding of wider patterns and ecological roles Thorsten Krömer http://orcid.org/0000-0002-1398-8172
within different habitats as well as the evolutionary Marco Tschapka http://orcid.org/0000-0002-1398-8172
relationships that exist within the bromeliad family. M.Cristina MacSwiney G. http://orcid.org/0000-0002-
9007-4622

Acknowledgements
References
We are grateful to Marlies Sazima, Elton M. C. Leme, Stefan
Vogel, Michael Barfuss, Leonardo M. Versieux, Gustavo Aguilar-Rodríguez PA, Krömer T, García-Franco JG,
Martinelli, Andrea Araujo, Marília Barros, Ana Paula Gelli MacSwiney GMC. 2016. From dusk till dawn: nocturnal
de Faria, Karina Hmeljevski, Leandro Freitas, Eric Gouda and and diurnal pollination in the epiphyte Tillandsia hetero-
Lena Geise for providing information and references men- phylla (Bromeliaceae). Plant Biol. 18:37–45.
tioned in this work, comments in the main text and about Aguilar-Rodríguez PA, MacSwiney GMC, Krömer T,
bromeliad systematics, as well as some photos of Brazilian García-Franco JG, Knauer A, Kessler M. 2014. First
bromeliads and bats. We thank Lilia Ruiz for preparing the record on bat-pollination in the species-rich genus
photo plates. We thank Keith MacMillan and six anonymous Tillandsia (Bromeliaceae). Ann Bot. 113:1047–1055.
reviewers to help improve this manuscript. This study was Araujo AC, Fischer E, Sazima M. 2004. As Bromélias na
funded by a CONACyT grant awarded to PAAR (No. 59406). região do Rio Verde. In: Marques O, Duleba W, editors.
Estação Ecológica Juréia-Itatins: ambiente físico, flora e
fauna. Ribeirão Preto: Holos; p. 162–171.
Disclosure statement Ashworth L, Aguilar R, Martén-Rodríguez S,
Lopezaraiza-Mikel M, Avila-Sakar G, Rosas-Guerrero
No potential conflict of interest was reported by the V, Quesada M. 2015. Pollination syndromes: A global
authors. pattern of convergent evolution driven by the most
effective pollinator. In: Pontarotti P, editor.
Evolutionary biology: biodiversification from genotype
Funding to phenotype. Switzerland: Springer International
Publishing; p. 203–224.
This work was supported by the Consejo Nacional de Baker RJ, Bininda-Emonds ORP, Mantilla-Meluk H, Porter
Ciencia y Tecnología (CONACyT) by the grant number CA, Van Den Bussche RA. 2012. Molecular timescale of
59406 awarded to P.A.A.-R. diversification of feeding strategy and morphology in New
World leaf-nosed bats (Phyllostomidae): A phylogenetic
perspective. In: Gunnell GF, Simmons NB, editors.
Notes on contributors Evolutionary history of bats: fossils, molecules and morphol-
ogy. Cambridge: Cambridge University Press; p. 385–409.
Pedro A. Aguilar-Rodríguez recently completed his Ph.D. Barbará T, Martinelli G, Fay MF, Mayo SJ, Lexer C. 2007.
on chiropterophilous bromeliads in Mexico. Population differentiation and species cohesion in two
Thorsten Krömer studies the diversity, distribution and closely related plants adapted to Neotropical high-alti-
ecology of epiphytic plants and their conservation in tro- tude ‘inselbergs’, Alcantarea imperialis and Alcantarea
pical montane forests. geniculata (Bromeliaceae). Mol Ecol. 16:1981–1992.

Marco Tschapka studies the interactions between Barfuss M, Samuel MR, Till W, Stuessy TF. 2005.
Neotropical bats and their food plants, especially the role Phylogenetic relationships in subfamily Tillandsioideae
of bats in pollination and resource partitioning among (Bromeliaceae) based on DNA sequence data from seven
nectarivorous bats. plastid regions. Am J Bot. 92:337–351.
Barfuss MHJ, Till W, Leme EMC, Pinzón JP, Manzanares
José G. García-Franco is interested in the ecology of epi-
JM, Halbritter H, Samuel R, Brown GK. 2016.
phytic and parasitic vascular plants, their reproduction,
Taxonomic revision of Bromeliaceae subfam.
host specificity and the biotic and abiotic factors that the
Tillandsioideae based on a multi-locus DNA sequence
distribution of these plants is related to.
phylogeny and morphology. Phytotaxa. 279:001–097.
Jeanett Escobedo-Sarti studies plant-animal interactions in Barros MAS, Rui AM, Fabián ME. 2013. Seasonal variation
relation to reproductive biology of selected taxa. She is also in the diet of Anoura caudifer (Phyllostomidae:
interested in the evolution of key phylogenetic traits of Glossophagine) at the Southern limit of its geographic
epiphytic plants, including bromeliads. range. Acta Chiropterologica. 15:77–84.
 PLANT ECOLOGY & DIVERSITY 15

Beaman RS, Judd WS. 1996. Systematics of Tillandsia sub- Escobedo-Sarti J, Ramírez I, Leopardi C, Carnevali G,
genus Pseudoalcantarea (Bromeliaceae). Brittonia. 48:1–19. Magallón S, Duno R, Mondragón D. 2013. A phylogeny
Benzing DH. 2000. Bromeliaceae. Profile of an adaptive of Bromeliaceae (Poales, Monocotyledoneae) derived
radiation. Cambridge: Cambridge University Press. from an evaluation of nine supertree methods. J Syst
Bestmann HJ, Winkler L, von Helversen O. 1997. Evol. 51:743–757.
Headspace analysis of volatile flower scent constituents Fabián ME, Rui AM, Waechter JL. 2008. Plantas utilizadas
of bat-pollinated plants. Phytochemistry. 46:1169–1172. como alimento por morcegos (Chiroptera, Phyllostomidae)
Bolzan DP, Pessôa LM, Peracchi AL, Strauss RE. 2015. no Brasil. In: Reis NR, Peracchi AL, Santos GASD, editors.
Allometric patterns and evolution in Neotropical nec- Ecologia de morcegos. Londrina: Technical Books Editora;
tar-feeding bats (Chiroptera, Phyllostomidae). Acta p. 51–70.
Chiropterologica. 17:59–73. Faegri K, van der Pijl L. 1979. The principles of pollination
Borges RM, Somanathan H, Kelber A. 2016. Patterns and ecology. Toronto: Pergamon Press.
processes in nocturnal and crepuscular pollination ser- Fischer EA. 1994. Polinização, fenologia e distribuição
vices. Q Rev Biol. 91:389–418. espacial de Bromeliaceae numa comunidade de Mata
Burck W. 1892. Wandelingen door den Botanischen Tuin. Atlântica, Litoral Sul de São Paulo [Master´s Thesis].
T Treub’s Lands Plantentuin te Buitenzorg. Leipzig: São Paulo (Brazil): Universidade Estadual de
Batavia. Campinas.
Canela MBF, Sazima M. 2005. The pollination of Bromelia Fleming TH. 2004. Nectar corridors: migration and the
antiacantha (Bromeliaceae) in Southeastern Brazil: ornitho- annual cycle of lesser long-nosed bats. In: Nabhan GP,
philous versus melittophilous features. Plant Biol. 7:411–416. editor. Conserving migratory pollinators and nectar cor-
Cascante-Marín A, Oostermeijer JGB, Wolf JHD, Dennijs ridors in western North America. Tucson: University of
JCM. 2005. Reproductive biology of the epiphytic bro- Arizona Press; p. 23–42.
meliad Werauhia gladioliflora in a premontane tropical Fleming TH, Geiselman C, Kress WJ. 2009. The evolution
forest. Plant Biol. 7:203–209. of bat pollination: a phylogenetic perspective. Ann Bot.
Christianini AV, Forzza RC, Buzato S. 2013. Divergence on 104:1017–1043.
floral traits and vertebrate pollinators of two endemic Fleming TH, Muchhala N. 2008. Nectar-feeding bird and
Encholirium bromeliads. Plant Biol. 15:360–368. bat niches in two worlds: pantropical comparisons of
Cirranello A, Simmons NB, Solari S, Baker RJ. 2016. vertebrate pollination systems. J Biogeogr. 35:764–780.
Morphological diagnoses of higher-level phyllostomid Fleming TH, Muchhala N, Ornelas JF. 2005. New World
taxa (Chiroptera: Phyllostomidae). Acta Chiropterologica. nectar-feeding vertebrates: community patterns and pro-
18:39–71. cesses. In: Sánchez-Cordero V, Medellín RA, editors.
Corbet S. 2003. Nectar sugar content: estimating standing Contribuciones mastozoológicas en homenaje a
crop and secretion rate in the field. Apidologie. 34:1–10. Bernardo Villa. México City: Instituto de Biología e
Cordero-Schmidt E, Barbier E, Vargas-Mena JC, Oliveira Instituto de Ecología-UNAM; p. 161–184.
PP, Santos FDAR, Medellín RA, Rodríguez-Herrera B, García-Franco JG, Martínez-Burgoa D, Pérez TM. 2001.
Venticinque EM. 2017. Natural history of the Caatinga Hummingbird flower mites and Tillandsia spp.
endemic Vieira’s flower bat, Xeronycteris vieirai. Acta (Bromeliaceae): polyphagy in a cloud forest of
Chiropterologica. 19:399–408. Veracruz, Mexico. Biotropica. 33:538–542.
Cox AC. 1984. Chiropterophily and ornithophily in Gardner CS. 1986. Inferences about pollination in
Freycinetia (Pandanaceae) in Samoa. Plant Syst Evol. Tillandsia (Bromeliaceae). Selbyana. 9:76–87.
144:277–290. Geiselman CK, Defex TI. 2015. Bat eco-interactions data-
Crayn DM, Winter K, Smith JAC. 2004. Multiple origins of base. [accessed 2016 Jan 15]. www.batplant.org.
crassulacean acid metabolism and the epiphytic habit in Gentry AH. 1974. Flowering phenology and diversity in
the Neotropical family Bromeliaceae. Proc Nat Acad Sci. tropical Bignoniaceae. Biotropica. 6:64–68.
101:3703–3708. Gilmartin AJ, Brown GK. 1987. Bromeliales, related mono-
Cunningham SA. 1995. Ecological constraints on fruit cots, and resolution of relationships among
initiation by Calyptrogyne ghiesbreghtiana (Arecaceae): Bromeliaceae subfamilies. Syst Botany. 12:493–500.
floral herbivory, pollen availability and visitation by pol- Givnish TJ, Barfuss M, Van Ee B, Riina R, Schulte K,
linating bats. Am J Bot. 82:1527–1536. Horres R, Gonsiska P, Jabaily R, Crayn D, Smith J, et
Darriba D, Taboada G, Doallo R, Posada D. 2012. al. 2011. Phylogeny, adaptive radiation, and historical
jModelTest 2: more models, new heuristics and parallel biogeography in Bromeliaceae: insights from an eight-
computing. Nat Methods. 9:772. locus plastid phylogeny. Am J Bot. 98:1–24.
Dobat K, Peikert-Holle T. 1985. Blüten und Fledermäuse Givnish TJ, Barfuss M, Van Ee B, Riina R, Schulte K,
(Chiropterophilie). Frankfurt: Waldemar Kramer Verlag. Horres R, Gonsiska P, Jabaily R, Crayn D, Smith J, et
Edgar R. 2004. MUSCLE: multiple sequence alignment al. 2014. Adaptive radiation, correlated and contingent
with high accuracy and high throughput. Nucleic Acids evolution, and net species diversification in
Res. 32:1792–1797. Bromeliaceae. Mol Phylogenet Evol. 71:55–78.
16 P. A. AGUILAR-RODRÍGUEZ ET AL.

Givnish TJ, Millam KC, Berry PE, Sytsma KJ. 2007. King C, Ballantyne G, Willmer PG. 2013. Why flower visita-
Phylogeny, adaptive radiation, and historical biogeogra- tion is a poor proxy for pollination: measuring single-visit
phy of Bromeliaceae inferred from ndhf sequence data. pollen deposition, with implications for pollination net-
Aliso. 23:3–26. works and conservation. Methods Ecol Evol. 4:811–818.
Givnish TJ, Millam KC, Evans TM, Hall JC, Pires JC, Berry Knudsen JT, Tollsten L. 1995. Floral scent in bat-pollinated
PE, Sytsma KJ. 2004. Ancient vicariance or recent long- plants: a case of convergent evolution. Bot J Linnean Soc.
distance dispersal? Inferences about phylogeny and 119:45–57.
South American-African disjunction in Rapateaceae Kowalski VS, Tardivo RC. 2015. O grupo Vriesea platy-
and Bromeliaceae based on ndhF sequence data. Int J nema Gaudich. (Bromeliaceae: Tillandsioideae) no
Plant Sci. 165:S35–S54. estado do Paraná, Brasil. Rodriguésia. 66:465–475.
Gomes LAC, Maas ACS, Godoy MSM, Martins MA, Pedrozo Krömer T. 2003. Fledermausbestäubung (Chiropterophilie)
AR, Peracchi AL. 2018. Ecological considerations on bei Bromelien der Gattung Guzmania in Bolivien. Die
Xeronycteris vieirai: an endemic bat species from the Bromelie. 3:60–63.
Brazilian semiarid macroregion. Mastozoología Neotrop. Krömer T, Espejo A, López-Ferrari AR, Acebey A. 2005.
25:81–88. The presence of Werauhia nutans in Mexico. J
Gonzalez-Terrazas TP, Martel C, Milet-Pinheiro P, Ayasse Bromeliad Soc. 55:280–284.
M, Kalko EKV, Tschapka M. 2016. Finding flowers in Krömer T, Espejo A, López-Ferrari AR, Acebey A. 2007.
the dark: nectar-feeding bats integrate olfaction and Werauhia noctiflorens (Bromeliaceae), una nueva especie
echolocation while foraging for nectar. R Soc Open Sci. del sureste de México y Belice. Novon. 17:336–340.
3:160199. Krömer T, Espejo-Serna A, López-Ferrari AR, Ehlers R,
Gouda EJ, Butcher D, Gouda CS. 2017. Encyclopedia of Lautner J. 2012. Taxonomic and nomenclatural status
bromeliads, version 3. University botanic gardens, of the Mexican species in the Tillandsia viridiflora com-
Utrecht. [accessed 2017 Jan 27]. http://encyclopedia.flor plex (Bromeliaceae). Acta Botánica Mexicana. 99:1–20.
apix.nl/. Krömer T, Kessler M, Herzog SK. 2006. Distribution and
Grant JR. 1995. The resurrection of Alcantarea and flowering ecology of bromeliads along two climatically
Werauhia, a new genus. Tropische und Subtropische contrasting elevational transects in the Bolivian Andes.
Pflanzenwelt. 91:7–57. Biotropica. 38:183–195.
Harmon LJ, Weir JT, Brock CD, Glor RE, Challenger W. Krömer T, Kessler M, Lohaus G, Schmidt-Lebuhn AN.
2008. GEIGER: investigating evolutionary radiations. 2008. Nectar sugar composition and concentration in
Bioinformatics. 24:129–131. relation to pollination syndromes in Bromeliaceae.
Hilo de Souza E, Massarioli AP, Moreno IAM, Souza FVD, Plant Biol. 10:502–511.
Ledo CAS, Alencar SM, Martinelli AP. 2016. Volatile Leme EM. 1995. Contribução ao estudo da seção Xiphion
compounds profile of Bromeliaceae flowers. Revista de (Vriesea) –I. Studies of the section Xiphion (Vriesea) −1.
Biología Tropical. 64:1101–1116. Bromélia. 2:24–28.
Hmeljevski K, Wolowski M, Forzza R, Freitas L. 2017. High Lexer C, Marthaler F, Humbert S, Barbará T, de la Harpe
outcrossing rates and short-distance pollination in a M, Bossolini E, Paris M, Martinelli G, Versieux LM.
species restricted to granitic inselbergs. Aus J Bot. 2016. Gene flow and diversification in a species complex
65:315–326. of Alcantarea inselberg bromeliads. Bot J Linnean Soc.
Hodges SA. 1993. Consistent interplant variation in nectar 3:505–520.
characteristics of Mirabilis multiflora. Ecology. 74:542– Macías-Rodríguez MA, Sahagún-Godínez E, Lomelí-
548. Sención JA. 2007. Pitcairnia abundans L.B.SM.
Hornung-Leoni C, Sosa V. 2006. Morphological variation (Bromeliaceae): supplementary description, range exten-
in Puya (Bromeliaceae): an allometric study. Plant Syst sion and conservation status of an endemic mexican
Evol. 256:35–53. plant. Polibotánica. 23:93–100.
Howell DJ, Burch D. 1974. Food habits of some Costa Maguiña Conde NR. 2016. Nectar bats and their flowers
Rican bats. Rev Biol Trop. 21:281–294. across cloud forests of Ecuador and the effect of artificial
Jakobsen HB, Kritjánsson K. 1994. Influence of tempera- nectar feeders [Master´s Thesis]. Missouri: University of
ture and floret age on nectar secretion in Trifolium Missouri-St. Louis.
repens L. Ann Bot. 74:327–334. Maguiña R, Amanzo J. 2016. Diet and pollinator role of the
Janzen DH. 1971. Euglossine bees as long distance pollina- long-snouted bat Platalina genovensium in Lomas eco-
tors of tropical plants. Science. 171:203–205. system of Peru. Tropl Conserv Sci. 9:1-8. October-
Kaehler M, Varassin IG, Goldenberg R. 2005. Polinização December.
em uma comunidade de bromélias em floresta atlântica Maguiña R, Amanzo J, Huamán L. 2012. Dieta de
alto-montana no estado do Paraná, Brasil. Braz J Bot. murciélagos filostómidos del valle de Kosñipata, San
28:219–228. Pedro, Cusco-Perú. Revista Peruana de Biología.
Kessler M. 2002. Environmental patterns and ecological 19:159–166.
correlates of range size among bromeliad communities Mantilla-Meluk KH, Baker RJ. 2010. New species of
of Andean forest in Bolivia. Botl Rev. 68:100–127. Anoura (Chiroptera: Phyllostomidae) from Colombia,
Kessler M, Krömer T. 2000. Patterns and ecological corre- with systematic remarks and notes on the distribution
lates of pollination modes among bromeliad commu- of the A. geoffroyi complex. Occ Papers Museum Texas
nities of Andean forests in Bolivia. Plant Biol. 2:659–669. Tech. 292:1–19.
 PLANT ECOLOGY & DIVERSITY 17

Mantilla-Meluk KH, Jiménez-Ortega AM, Baker RJ. 2009. Muchhala N, Jarrín-V P. 2002. Flower visitation by bats
Range extension of Anoura aequatorialis and notes on in cloud forest of western Ecuador. Biotropica.
distributional limits of small Anoura in Colombia. 34:387–395.
Investigación, Biodiversidad y Desarrollo. 28:107–112. Muchhala N, Mena VP, Albuja VL. 2005. A new species of
Marques JS, Tagliati MC, Faria APG. 2015. Diurnal versus Anoura (Chiroptera: Phyllostomidae) from the
nocturnal pollination success in Billbergia horrida Regel Ecuadorian Andes. J Mammal. 86:457–461.
(Bromeliaceae) and the first record of chiropterophily for Muchhala N, Thomson JD. 2009. Going to great lengths:
the genus. Anais da Academia Brasileira de Ciências. selection for long corolla tubes in an extremely specialized
87:835–842. bat–flower mutualism. Proc Rl Soc B. 276:2147–2152.
Martinelli G. 1994. Reproductive biology of Bromeliaceae in Muchhala N, Thomson JD. 2012. Interspecific competition
the Atlantic rainforest of southeastern Brazil [PhD Thesis]. in pollination systems: costs to male fitness via pollen
Saint Andrews (UK): University of Saint Andrews. misplacement. Funct Ecol. 26:476–482.
Martinelli G. 1997. Biologia reprodutiva de Bromeliaceae Müller F. 1897. Einige Bemerkungen über Bromeliaceen.
na Reserva Ecológica de Macaé de Cima. In: Lima HC, IX-XIII. Flora. 83:455–473.
Guedes-Bruni RR, editors. Serra de Macaé de Cima: Nachev V, Stich KP, Winter C, Bond A, Kamil A, Winter
diversidade Florística e Conservação em Mata Y. 2017. Cognition-mediated evolution of low-quality
Atlântica. Brazil: Instituto de Pesquisa Jardim Botânico floral nectars. Science. 355:75–78.
do Rio de Janeiro; p. 213–250. Ollerton J, Alarcón R, Waser NM, Price MV, Watts S,
Matallana G, Godinho MAS, Guilherme FAG, Belisario M, Cranmer L, Hingston A, Peter CI, Rotenberry J. 2009.
Coser TS, Wendt T. 2010. Breeding systems of A global test of the pollination syndrome hypothesis.
Bromeliaceae species: evolution of selfing in the context of Ann Bot. 103:1471–1480.
sympatric occurrence. Plant Syst Evol. 289:57–65. Ortega-García S, Guevara L, Arroyo-Cabrales J, Lindig-
Matallana G, Oliveira PE, Da Silva PRR, Wendt T. 2016. Post- Cisneros R, Martínez-Meyer E, Vega E, Schondube JE.
pollination barriers in an assemblage of Bromeliaceae in 2017. The thermal niche of Neotropical nectar-feeding
south-eastern Brazil. Bot J Linnean Soc. 181:521–531. bats: its evolution and application to predict responses to
Miller M, Pfeiffer W, Schwartz T. 2010. Creating the global warming. Ecol Evol. 7:6691–6701.
CIPRES Science Gateway for inference of large phyloge- Paggi GM, Palma-Silva C, Silveira LCT, Kaltchuk-Santos E,
netic trees. New Orleans (LA): Gateway Computing Bodanese-Zanettini MH, Bered F. 2007. Fertility of
Environments Workshop (GCE). Vriesea gigantea Gaud. (Bromeliaceae) in southern
Morales RAE. 2016. Dieta, actividad y reproducción de los Brazil. Am J Bot. 94:683–689.
murciélagos Anoura geoffroyi y Sturnira hondurensis en Palma-Silva C, Leal BSS, Chaves CJN, Fay MF. 2016.
el bosque Nublado del Parque Nacional Montecristo, El Advances in and perspectives on evolution in
Salvador [Bachelor´s Thesis]. San Salvador (El Salvador): Bromeliaceae. Bot J Linnean Soc. 181:305–322.
Universidad de El Salvador. Palma-Silva C, Lexer C, Paggi GM, Barbará T, Bered F,
Moras LM, Bernard E, Gregorin R. 2013. Bat assemblages Bodanese-Zanettini MH. 2009. Range-wide patterns of
at high-altitude area in the Atlantic forest of southeast- nuclear and chloroplast DNA diversity in Vriesea gigan-
ern Brazil. Mastozoología Neotrop. 20:269–278. tea (Bromeliaceae) a Neotropical forest species.
Moratelli R, Dias D. 2015. A new species of nectar-feeding Heredity. 103:503–512.
bat, genus Lonchophylla, from the Caatinga of Brazil Palma-Silva C, Wendt T, Pinheiro F, Barbará T, Fay MF,
(Chiroptera, Phyllostomidae). ZooKeys. 514:73–91. Cozzolino S, Lexer C. 2011. Sympatric bromeliad species
Morgan GS, Czaplewski NJ. 2012. Evolutionary history of (Pitcairnia spp.) facilitate test of mechanisms involved in
the neotropical chiroptera: the fossil record. In: Gunnell species cohesion and reproductive isolation in
GF, Simmons NB, editors. Evolutionary history of bats. Neotropical iselbergs. Mol Ecol. 20:3185–32101.
Cambridge: Cambridge University Press; p. 105–161. Paradis E, Claude J, Strimmer K. 2004. APE: analyses of
Moseley HN. 1870. Notes by a naturalist on the challenger. phylogenetic and evolution in R language.
London: Macmillan. Bioinformatics. 20:289–290.
Mosti S, Friedman CR, Pacini E, Brighigna L, Papini A. Patterson BD, Pacheco V, Solari S. 1996. Distributions of
2013. Nectary ultrastructure and secretory modes in bats along an elevational gradient in the Andes of south-
three species of Tillandsia (Bromeliaceae) that have dif- eastern Peru. J Zool (London). 240:637–658.
ferent pollinators. Botany. 91:786–798. Porsch O. 1932. Das Problem Fledermausblume. Anzeiger
Moura RL, Costa AF. 2014. Taxonomic notes on Vriesea der Akademie der Wissenschaft Wien, Mathematisch-
sect. Xiphion (Bromeliaceae) with descriptions of three Naturwissenschaftliche Klasse. 69: 27–28.
new species. Syst Botany. 39:791–803. Porsch O. 1934. Säugetiere als Blumenausbeuter und die Frage
Muchhala N. 2007. Adaptive trade-off in floral morphology der Säugetierblume. III. Mit 1 Abbildung im Text. 4:1–21.
mediates specialization for flowers pollinated by bats and Queiroz JA, Quirino ZGM, Lopes AV, Machado IC. 2016.
hummingbirds. Am Nat. 169:494–504. Vertebrate mixed pollination system in Encholirium
Muchhala N, Caiza A, Vizcete JC, Thomson JD. 2008. A spectabile: A bromeliad pollinated by bats, opossum,
generalized pollination system in the tropics: bats, birds and hummingbirds in a tropical dry forest. J Arid
and Aphelandra acanthus. Ann Bot. 103:1481–1487. Environ. 125:21–30.
18 P. A. AGUILAR-RODRÍGUEZ ET AL.

Queiroz JA, Quirino ZGM, Machado IC. 2015. Floral traits Schmid S, Schmid VS, Zillikens A, Harter-Marques B,
driving reproductive isolation of two co-flowering taxa Steiner J. 2011. Bimodal pollination system of the bro-
that share vertebrate pollinators. AoB Plants. 7:plv127. meliad Aechmea nudicaulis involving hummingbirds
doi:10.1093/aobpla/plv127 and bees. Plant Biol Suppl. 1:41–50.
R Core Team. 2016. R: A Language and Environment for Scultori da Silva HCS. 2009. Comunidade de morcegos,
Statistical Computing. Vienna (Austria): R Foundation interações com flores e estratifição vertical em Mata
for Statistical Computing. Atlântica do Sui do Brasil [Master´s Thesis]. São
Ramírez N, Seres A. 1994. Plant reproductive biology of Paulo (Brazil): Universidade Estadual de Campinas.
herbaceous monocots in a Venezuelan tropical cloud Seres A, Ramírez N. 1995. Biología floral y polinización de
forest. Plant Syst Evol. 190:129–142. algunas monocotiledóneas de un bosque nublado
Rauh W. 1986. Bromelienstudien. I. Neue und wenig Venezolano. Ann Mo Bot Gard. 82:61–81.
bekannte Arten aus Peru und anderen Ländern (18 Shivanna KR, Tandon R. 2014. Breeding systems. In:
Mitteilung). Tropische und Subtropische Pflanzenwelt. Shivanna KR, Tandon R, editors. Reproductive ecology
58:715–773. of flowering plants: a manual. New Delhi: Springer; p.
Rauh W. 1987. Bromelienstudien. I. Neue und wenig 107–123.
bekannte Arten aus Peru und anderen Ländern (19 Shockey BJ, Salas-Gismondi R, Baby P, Guyot JL, Baltazar
Mitteilung). Tropische und Subtropische Pflanzenwelt. MC, Huamán L, Clack A, Stucchi M, Pujos F, Emerson
60:907–1004. JM, et al. 2009. New Pleistocene cave faunas of the
Rauh W. 1990. Bromelien, Tillandsien und andere Andes of Central Peru: radiocarbon ages and the survi-
kulturwürdige Bromelien. 2nd ed. Ulmer: Stuttgart. val of low latitude, Pleistocene DNA. Palaeontologia
Rodríguez-Peña N, Stoner KE, Flores-Ortiz CM, Ayala- Electronica. 12:1–15.
Berdón J, Munguía-Rosas MA, Sánchez-Cordero V, Silva Jorge J, Silva Rocha LH, Silva Jorge JP, Pierote Sousa
Schondube JE. 2016. Factors affecting nectar sugar com- PH, Xavier Freire EM. 2018. Floral visitors and potential
position in chiropterophilic plants. Revista Mexicana De pollinators of a rupicolous bromeliad (Pitcairnioideae)
Biodiversidad. 87:465–473. in the Brazilian semiarid. Neotropl Biol Conserv.
Rojas D, Vale A, Ferrero V, Navarro L. 2011. When did plants 13:101–110.
become important to leaf-nosed bats? Diversification of Simmons NB. 2005. Order Chiroptera. In: Wilson DE,
feeding habits in the family Phyllostomidae. Mol Ecol. Reeder DM, editors. Mammal species of the world. A
20:2217–2228. taxonomic and geographic reference. Baltimore: Johns
Rojas D, Warsi OM, Dávalos LM. 2016. Bats (Chiroptera: Hopkins University Press; p. 312–529.
Noctilionoidea) challenge a recent origin of extant Siqueira Filho JA. 2003. Fenologia da floração, ecología da
Neotropical diversity. Syst Biol. 65:432–448. polinização e conservação de Bromeliaceae na Florest
Rosas-Guerrero V, Aguilar R, Martén-Rodríguez S, Atlâantica Nordestina [PhD Thesis]. Recife (Brazil):
Ashworth L, Lopezaraiza-Mikel M, Bastida JM, Universidade Federal de Pernambuco.
Quesada M. 2014. A quantitative review of pollination Soriano PJ, Ruiz A, Arends A. 2002. Physiological
syndromes: do floral traits predict effective pollinators? responses to ambient temperature manipulation by
Ecol Lett. 17:388–400. three species of bats from Andean cloud forests. J
Rothenwöhrer C, Becker NI, Tschapka M. 2011. Resource Mammal. 83:445–457.
landscape and spatio-temporal activity patterns of a Stewart A, Dudash MR. 2016. Differential pollen placement
plant-visiting bat in a Costa Rica lowland rainforest. J on an Old World nectar bat increases pollination effi-
Zool. 283:108–116. ciency. Ann Bot. 117:145–152.
Ruzana Adibah MS, Ainuddin AN. 2011. Epiphytic plants Stewart AB, Dudash MR. 2017. Field evidence of strong
responses to light and water stress. Asian J Plant Sci. differential pollen placement by Old World bat-polli-
10:97–107. nated plants. Ann Bot. 119:73–79.
Salas D. 1973. Una bromeliácea costarricense polinizada Terry RG, Brown GK, Olmstead RG. 1997. Examination of
por murciélagos. Brenesia. 2:5–10. subfamilial phylogeny in Bromeliaceae using compara-
Sazima I, Vogel S, Sazima M. 1989. Bat pollination of tive sequencing of the plastid locus ndhF. Am J Bot.
Encholirium glaziovii, a terrestrial bromeliad. Plant Syst 84:664–670.
Evol. 168:167–179. Tripp EA, Manos PS. 2008. Is floral specialization an
Sazima M, Buzato S, Sazima I. 1995. Polinização de Vriesea evolutionary dead-end? Pollination system transitions
por morcegos no Sudeste brasileiro. Revista de in Ruellia (Acanthaceae). Evolution. 62:1712–1737.
Sociedade Brasileira de Bromélias. 2:29–37. Tschapka M. 2004. Energy density patterns of nectar
Sazima M, Buzato S, Sazima I. 1999. Bat-pollinated flower resources permit coexistence within a guild of neo-
assemblage and bat visitors at two Atlantic forest sites in tropical flower-visiting bats. J Zool (London). 263:7–
Brazil. Ann Bot. 83:705–712. 21.
Schmid H. 2000. Puya nitida (Bromeliaceae): Generalist für Tschapka M, Dressler S. 2002. Chiropterophily: on bat-
Chiropterophilie und Ornithophilie [Thesis]. Bayreuth flowers and flower bats. Curtis Bot Mag (Series 6).
(Germany): Universität Bayreuth. 19:114–125.
 PLANT ECOLOGY & DIVERSITY 19

Tschapka M, Gonzalez-Terrazas TP, Knörnschild M. 2015. von Helversen O 1993. Adaptations of flowers to the
Nectar uptake in bats using a pumping-tongue mechan- pollination by glossophagine bats. In: Barthlott W,
ism. Sci Adv. 1:e1500525. Naumann CM, Schmidt-Loske K, Schuchmann KL,
Tschapka M, von Helversen O. 2007. Phenology, nectar editors. Animal-Plant Interactions in tropical envir-
production and visitation behaviour of bats on the onments. Results of the Symposium of the German
flowers of the bromeliad Werauhia gladioliflora in Society for Tropical Ecology held at Bonn; 1992 Feb
Costa Rican lowland rain forest. J Tropical Ecol. 13–16; Bonn: Museum Koenig; p. 41–59.
23:385–395. von Helversen O, Reyer HU. 1984. Nectar intake and
Utley J. 1983. A revision of the Middle American thecophylloid energy expenditure in a flower visiting bat. Oecologia.
Vrieseas (Bromeliaceae). Tulane Stud Zool Bot. 24:1–81. 63:178–184.
van del Pijil L. 1961. Ecological aspects of flower evolution. von Helversen O, Winkler L, Bestmann HJ. 2000. Sulphur-
II. Zoophilous flower classes. Evolution. 15:44–59. containing “perfumes” attract flower visiting bats. J
van der Niet T, Johnson SD. 2012. Phylogenetic evidence Comp Physiol A. 186:143–153.
for pollinator driven diversification of angiosperms. von Helversen O, Winter Y. 2003. Glossophagine bats and
Trend Ecol Evol. 27:353–361. their flowers: costs and benefits for plants and pollina-
Varadarajan GS, Brown GK. 1988. Morphological variation tors. In: Kunz TH, Fenton MB, editors. Bat ecology.
of some floral features of the subfamily Pitcairnioideae Chicago: The University of Chicago Press, Chicago; p.
(Bromeliaceae) and their significance in pollination biol- 346–397.
ogy. Bot Gaz. 149:82–91.
Wendt T, Canela MBF, Klein DE, Rios RI. 2002. Selfing
Versieux LM, Barbará T, Wanderley MGL, Calvente A, Fay
facilities reproductive isolation among three sympatric
MF, Lexer C. 2012. Molecular phylogenetics of the Brazilian
species of Pitcairnia (Bromeliaceae). Plant Syst Evol.
giant bromeliads (Alcantarea, Bromeliaceae): implications
232:201–212.
for morphological evolution and biogeography. Mol
Phylogenet Evol. 64:177–189. Wendt T, Coser TS, Matallana G, Guilherme FAG.
Versieux LM, Wanderley MGL. 2015. Bromélias-gigantes 2008. An apparent lack of pre-zygotic reproductive
do Brasil. Natal: Capim Macio & Offset. isolation among 42 sympatric species of Bromeliaceae
Vogel S. 1958. Fledermausblumen in Südamerika. in southeastern Brazil. Plant Syst Evol. 275:31–41.
Oesterreichische Botanische Zeitschrift. 104:491–530. Wendt T, Ferreira CMB, Gelli de Faria AP, Iglesias-Rios R.
Vogel S. 1969. Chiropterophilie in der neotropischen Flora- 2001. Reproductive biology and natural hybridization
Neue Mitteilungen III. Flora. 158:289–323. between two endemic species of Pitcairnia
von Helversen D, Holderied MW, von Helversen O. (Bromeliaceae). Am J Bot. 88:1760–1767.
2003. Echoes of bat-pollinated bell-shaped flowers: Willmer P. 2011. Pollination and floral ecology. Princeton
conspicuous for nectar-feeding bats? J Exp Biol. (NJ): Princeton University Press.
206:1025–1034. Zwickl D. 2006. Genetic algorithm approaches for the
von Helversen D, von Helversen O. 1975. Glossophaga phylogenetic analysis of large biological sequence
soricina (Phyllostomidae) Nahrungsaufnahme (Lecken). datasets under the maximum likelihood criterion
Encyclopaedia Cinematographica E. 1837/1975. [PhD Thesis]. Austin (TX): The University of Texas.

View publication stats