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

net/publication/309384099

Pollination of Campomanesia phaea (Myrtaceae) by night-active bees: A new

nocturnal pollination system mediated by floral scent

Article  in  Plant Biology · October 2016

DOI: 10.1111/plb.12520

CITATIONS READS

43 937

4 authors:

Guaraci Duran Cordeiro Mardiore Pinheiro

University of Salzburg Universidade Federal da Fronteira Sul
34 PUBLICATIONS   701 CITATIONS    20 PUBLICATIONS   191 CITATIONS   

SEE PROFILE SEE PROFILE

Stefan Dötterl Isabel Alves-dos-Santos

University of Salzburg University of São Paulo
175 PUBLICATIONS   4,722 CITATIONS    121 PUBLICATIONS   2,215 CITATIONS   

SEE PROFILE SEE PROFILE

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

PLANT-BEE INTERACTIONS AND POLLEN FLUX IN RESTORED AREAS OF ATLANTIC FOREST View project

Local pollinator adaptation in Arum maculatum View project

All content following this page was uploaded by Guaraci Duran Cordeiro on 04 November 2017.

The user has requested enhancement of the downloaded file.

 Plant Biology ISSN 1435-8603

RESEARCH PAPER

Pollination of Campomanesia phaea (Myrtaceae) by night-active

bees: a new nocturnal pollination system mediated by floral
scent
€ tterl3 & I. Alves-dos-Santos4
G. D. Cordeiro1, M. Pinheiro2, S. Do
1 Faculdade de Filosofia, Ci^
encias e Letras de Ribeir~
ao Preto, Universidade de S~ao Paulo, Ribeir~ao Preto, S~ao Paulo, Brazil
2 Universidade Federal da Fronteira Sul, Cerro Largo, Rio Grande do Sul, Brazil
3 Department of Ecology & Evolution, University of Salzburg, Salzburg, Austria
4 Departamento de Ecologia, Instituto de Bioci^ encias, Universidade de S~ao Paulo, S~ao Paulo, Brazil

Keywords ABSTRACT
1-Octanol; cambuci; Colletidae; floral
volatiles; Halictidae; nocturnal anthesis;
pollinator effectiveness. • Bees are the most important diurnal pollinators of angiosperms. In several groups of
bees a nocturnal/crepuscular habit developed, yet little is known about their role in
Correspondence pollination and whether some plants are adapted specifically to these bees. We used a
€tterl, Hellbrunnerstrasse 34, 5020,
S. Do multidisciplinary approach to investigate the reproductive biology and to understand
Salzburg, Austria. the role of nocturnal/crepuscular bees in pollination of Campomanesia phaea (Myr-
E-mail: Stefan.Doetterl@sbg.ac.at taceae), popularly named cambuci.

Editor
• We studied the floral biology and breeding system of C. phaea. We collected the floral
visitors and tested the pollinators’ effectiveness. We also determined the floral scents
A. Dafni released at night and during daytime, and studied behavioural responses of crepuscu-
lar/nocturnal bees towards these scents.
Received: 29 June 2016; Accepted:
19 October 2016
• The flowers of cambuci were self-incompatible and had pollen as the only resource for
flower visitors. Anthesis lasted around 14 h, beginning at 04:30 h at night. The flowers
released 14 volatile compounds, mainly aliphatic and aromatic compounds. We col-
doi:10.1111/plb.12520
lected 52 species of floral visitors, mainly bees. Nocturnal and crepuscular bees (four
species) were among the most frequent species and the only effective pollinators. In
field bioassays performed at night, nocturnal/crepuscular bees were attracted by a syn-
thetic scent blend consisting of the six most abundant compounds.
• This study describes the first scent-mediated pollination system between a plant and
its nocturnal bee pollinators. Further, C. phaea has several floral traits that do not allow
classification into other nocturnal pollination syndromes (e.g. pollinator attraction
already before sunrise, with pollen as the only reward), instead it is a plant specifically
adapted to nocturnal bees.

Frankie et al. 1983; Proencßa 1992; Wilms 1995; Souza 1996;

INTRODUCTION
Bees are the most important pollinators of angiosperms in both
natural (Ollerton et al. 2011) and crop (Klein et al. 2007) areas.
While this is true and well known for bees with a diurnal habit,
little is known about the importance of nocturnal bees as polli-
nators. Nocturnal and crepuscular activity in bees has arisen
independently in at least four of the seven families of bees, i.e.
Andrenidae, Apidae, Colletidae and Halictidae, with at least
250 species (Wcislo & Tierney 2009). Most nocturnal species
are tropical or subtropical (Warrant 2007). The visual adapta-
tions of nocturnal bees to low light, such as their larger com-
pound eyes and ocelli when compared to diurnal bees (Jander
& Jander 2002; Greiner et al. 2004a,b; Warrant et al. 2004;
Berry et al. 2011), allow them to collect resources from flowers
during the night without competing with diurnal flower visi-
tors (Wcislo et al. 2004; Wcislo & Tierney 2009).
The interaction of nocturnal bees with plants has been
recorded through direct observations of visits (Janzen 1968;
Hopkins et al. 2000; Silva & Pinheiro 2007; Faria & Stehmann
2010; Franco & Gimenes 2011; Carneiro & Martins 2012; Krug
et al. 2015) and through analyses of pollen found in their nests
(Smith et al. 2012), which both suggest that they play a role as
pollinators. However, evidence that these bees are efficient pol-
linators is lacking, and it is also unclear whether some plants
show specific adaptations to nocturnal bees or whether these
bees just make use of rewards primarily offered to other noc-
turnal or diurnal pollinators. Recently, Carvalho et al. (2012)
and Knoll & Santos (2012) attracted nocturnal bees (Megalopta
spp.) to euglossine scent traps, suggesting that these bees use
olfactory cues to find flowers. However, as the scent cues
released by plants visited by nocturnal bees are unknown, noth-
ing is yet known about the chemical communication between
nocturnal bees and their specific host plants.
The dish flower of Myrtaceae allows visits by all kinds of
flower visitors (Faegri & Van Der Pijl 1971; Endress 1994), such
as bats, birds, insects, marsupials and reptiles (Armstrong

Plant Biology © 2016 German Botanical Society and The Royal Botanical Society of the Netherlands 1
Pollination of a Myrtaceae by nocturnal bees
1979; Hopper 1980; Crome & Irvine 1986; Whitaker 1987;
Roitman et al. 1997). Overall, diurnal bees have been shown to
be the most common pollinators (Proencßa & Gibbs 1994;
Lughadha & Proencßa 1996; Gressler et al. 2006; Fidalgo &
Kleinert 2009; Giannini et al. 2015), although Proencßa & Gibbs
(1994) warned that nocturnal and crepuscular bees might also
be important pollinators. They have simply not been recorded
as frequently because they are active during hours when sam-
pling is not commonly carried out. Campomanesia phaea (O.
Berg) Landrum 1984 is a Myrtaceae endemic to the Atlantic
Forest in southeastern Brazil (Lorenzi et al. 2006). This species,
known popularly as cambuci, is currently included in the list of
the vulnerable species by the International Union for Conser-
vation of Nature (IUCN 2015) due to recurrent deforestation.
Nowadays it is cultivated on a small scale for the production of
homemade sweets, ice creams, liquors and juices (Kawasaki &
Landrum 1997; Vallilo et al. 2005). In C. phaea we detected, in
preliminary observations, nocturnal anthesis and also visits by
nocturnal/crepuscular bees.
To study the floral and reproductive biology of C. phaea, as
well as the chemical ecology of its pollination, we used a multi-
disciplinary approach to address the following questions: (i)
what is the floral biology and breeding system of the plant; (ii)
what are the diurnal and nocturnal visitors and pollinators of
the flowers; (iii) which scent compounds do cambuci flowers
emit during day and at night and in which absolute amounts;
and (iv) are nocturnal/crepuscular bees attracted by the main
components of the floral scent?
MATERIAL AND METHODS
Studied species
Campomanesia phaea grows in the Atlantic Forest biome. Tree
height at reproductive age ranges from 2 to 5 m. The flowers
measure about 3 cm and the fruits are 5–8 cm in diameter
(Fig. 1). The flowering period is in the warmest and most
humid months from September to December. A voucher speci-
men was deposited at HUMC (Herbario Mogiense, Universi-
dade de Mogi das Cruzes, SP), with the number 4481.
Study area
We carried out the study over three flowering periods (2011–
2014), from September to December, in a natural habitat site
A B
Fig. 1. (A) Flower with the style indicated by arrow, surrounded by c. 500 stamens. (B) Fruit of cambuci (Campomanesia phaea).
2 Plant Biology © 2016 German Botanical Society and The Royal Botanical Society of the Netherlands
€tterl & Alves-dos-Santos
Cordeiro, Pinheiro, Do
in Neblinas Park (only flower visitors, see below), and in a crop
area site with ca. 300 small-sized trees surrounded by natural
vegetation, both in the municipality of Mogi das Cruzes, state
of S~ao Paulo, Brazil (23°250 S, 46°100 W, 900 m a.s.l.). These
localities are within the Atlantic Forest biome of Ombrophilous
Dense vegetation, which climate is type Af and characterised as
humid tropical with over 2000 mm of annual rainfall (K€ oppen
1948).
Floral biology and reproductive system
We monitored anthesis time, stigma receptivity, floral
resources, pollen viability and the breeding system of cambuci.
Stigma receptivity was tested hourly, from flower opening to
senescence, in ten individuals (one flower/individual) using
hydrogen peroxide (Dafni et al. 2005), and in five individuals
(one flower/individual) with peroxtesmo KO (Dafni & Maues
1998).
The number of pollen grains produced per flower was deter-
mined through optical microscopy, by counting the pollen
grains of five flowers (from five different individuals). In order
to discriminate the morphology of C. phaea pollen from pollen
of co-flowering plants, the flowers were fixed in acetic acid for
later preparation of slides using the acetolysis method (Erdt-
man 1966). Pollen viability was tested through cytoplasm stain-
ing with acetic carmine 1.2% in around 100 anthers from five
flowers (from five different individuals) previously fixed in
FAA 70% (Dafni et al. 2005).
The breeding system was determined using manual pollina-
tion treatments: agamospermy, cross-pollination, manual self-
pollination and spontaneous self-pollination (Dafni et al.
2005). For each treatment we bagged 50 pre-anthesis flower
buds in ten trees (five flowers/individual). On the same trees,
we marked 100 flower buds to test the formation of fruits
under natural conditions (open pollination control). In all
treatments, fruit set was monitored.
Flower visitors and pollinator effectiveness
We captured flower visitors with an entomological net during
the period, where flowers offer pollen rewards (04:30 h to
12:00 h; see Results) on 60 non-consecutive days in the crop
area site (450 recorded hours) and 30 non-consecutive days in
the natural habitat (225 recorded hours). To determine the
abundance of flower visitors, we selected ten flowers from each
 €tterl & Alves-dos-Santos
Cordeiro, Pinheiro, Do
of six trees (each tree observed once on a different day) and
observed the number of visits by each insect (without catching
them) during 30-min intervals throughout the anthesis. In this
experiment, we also monitored whether visitors contacted the
stigma. In parallel, we recorded the light intensity (lux) with a
digital lux meter (MLM-1011). Pollinator effectiveness was
assessed with the most frequent visitor species through two dif-
ferent approaches using newly opened flowers that were previ-
ously bagged at the bud stage: (i) after the first visit by an
insect, we removed the pistil of the flower to count the number
of cambuci pollen grains deposited on the stigma (single-visit
deposition, SVD); (ii) after a single visit by an insect, we
bagged the flower and monitored fruit set (Dafni et al. 2005;
Ne’eman et al. 2010). Pollen deposition was measured on 63
flowers visited by nocturnal visitors and 54 flowers visited by
diurnal visitors, and fruit set on 51 flowers visited by nocturnal
and diurnal visitors each.
The insect vouchers were deposited at the CEPANN collec-
tion (Colecß~ao Entomologica Paulo Nogueira Neto) at the Insti-
tute of Bioci^encias, University of S~ao Paulo under the
numbers: 42963–43374.
Volatile sampling and analyses
Flower volatiles were collected using dynamic headspace, as
described in D€otterl et al. (2005). To determine the scent bou-
quet released and how it varies over time, 14 scent samples
were collected from three different individuals. A single flower
was sampled from each of two individuals, and five flowers
were sampled from a third individual. All flowers were sampled
twice: at night (morning) immediately after opening (05:00 h)
and additionally during daytime 5 h after the beginning of
anthesis. To obtain the scent samples, flowers were enclosed in
polyester oven bags (Toppitsâ) for 2 min and then adsorbent
tubes were inserted into the bags to collect the volatiles for
2 min using a membrane pump (G12/01 EB; Rietschle Thomas,
Puchheim, Germany). Flow was adjusted to 200 mlmin1 with
a flowmeter. The adsorbent tubes (quartz vials, length: 25 mm,
inner diameter: 2 mm) were filled with 1.5 mg Tenax-TA
(mesh 60–80) and 1.5 mg Carbotrap B (mesh 20–40, both
Supelco, Bellefonte, PA, USA). The adsorbents were fixed in
the tubes using glass wool. Dynamic headspace samples of
green leaves were collected with the same method and were
used to discriminate vegetative from flower-specific scent com-
ponents.
Scent samples were analysed using GC/MS (gas chro-
matography/mass spectrometry) to: (i) identify the flower-
specific compounds in the scent samples, and (ii) determine
their absolute and relative amounts. The system consisted of
an automated thermal desorption system (model TD-20;
Shimadzu, Kyoto, Japan) coupled to a GC/MS (model
QP2010 Ultra EI; Shimadzu) equipped with a ZB-5 fused
silica column (5% phenyl polysiloxane; 60-m long; inner
diameter 0.25 mm; film thickness 0.25 lm; Phenomenex,
Torrance, CA, USA), as described in Mitchell et al. (2015).
The column flow (carrier gas: helium) was set to
1.5 mlmin1. The GC oven temperature started at 40 °C,
then increased by 6 °Cmin1 to 250 °C and held for 1 min.
The MS interface worked at 250 °C. Mass spectra were
taken at 70 eV (in EI mode) from m/z 30 to 350. The GC/
MS data were processed using the GCMSolution package,
Plant Biology © 2016 German Botanical Society and The Royal Botanical Society of the Netherlands
Pollination of a Myrtaceae by nocturnal bees
version 2.72 (Shimadzu 2012). The identification of com-
pounds was carried out using the mass spectral libraries
Wiley 9, Nist 2011, FFNSC 2 and Adams 2007, as well as
the database available in MassFinder 3. The identity of some
of the compounds was confirmed by comparison of mass
spectra and retention times with authentic standard com-
pounds.
Field bioassays
After identification of the scent compounds in the five night
samples collected from a single individual, a mixture of the six
main compounds (benzyl alcohol, 2-phenylethanol, hexanal,
1-hexanol, (Z)-3-hexen-1-ol, 1-octanol), resembling the rela-
tive abundance of these compounds in the samples (as deter-
mined by dynamic headspace and GC/MS, see also below), was
prepared to obtain a synthetic scent of cambuci flowers for
field bioassays.
Bioassays were carried out in December 2014, twice a week
(eight bioassays), from 05:00 h to 06:30 h during the period
of activity of the nocturnal/crepuscular bees. Field nine-choice
bioassays were performed at night during the flowering period
to test if the identified floral volatiles are able to attract the
nocturnal/crepuscular bees. In the field we tested the synthetic
scent mixture of cambuci at different concentrations (pure;
103 and 105 diluted in paraffin oil), pure samples of the
four most abundant compounds (1-hexanol, 1-octanol, benzyl
alcohol, 2-phenylethanol) and a negative control (paraffin
oil). Additionally, pure linalool, a compound well known
from plants pollinated by diurnal bees (Dobson 2006), was
presented in the field bioassays. For each treatment, we
offered 500 ll of the compound/mixture on filter paper
(diameter 3 cm) placed on a Petri dish, with each sample
distanced 1 m from each other in a line. Dynamic headspace
and GC/MS analyses of the scent released from the filter paper
revealed that the different concentrations of the synthetic
mixture are equivalent to 5000 flowers = 120 flowering plant
individuals (pure; flower visitors will not be confronted with
such a high concentration in nature), 50 flowers = one flower-
ing plant individuals (103), and less than one flower (105).
Bees flying within 10 cm around a Petri dish, or actually
landing on it, were collected with an insect net. Responding
bees were collected and kept in a cool box until the end of the
assay, when they were released. This prevents counting
responding bees twice.
Data analysis
For analysis of semi-quantitative (percentage contribution of
single compounds to total scent) differences in scent emission
between time periods (night and day), we performed a
PERMANOVA (fixed factor: time class; random factor: individual,
to control for the fact that some flowers were from a same
individual; 10,000 permutations) analysis based on pair-wise
Bray–Curtis similarities (Primer 6 version 6.1.15; Permanova
version 1.0.5; Clarke and Gorley 2006; Anderson et al. 2008).
The field bioassay results were analysed by exact binomial
tests of goodness-of-fit using the spreadsheet provided at
http://udel.edu/~mcdonald/statexactbin.html (accessed 20 Jan.
2015). Responses towards specific scent stimuli were compared
to responses towards the negative control when at least eight
3
Pollination of a Myrtaceae by nocturnal bees €tterl & Alves-dos-Santos
Cordeiro, Pinheiro, Do

bees (pooled over replicate assays) responded towards a specific

Flower visitors and pollinator effectiveness
stimulus.
We collected a total of 411 insect visitors on the flowers, repre-
senting 52 species: including 36 bee species, five dipterans, two
RESULTS wasps and one butterfly (Tables 2 and S1). Thirteen of the spe-
cies occurred at both sites, eight only in the natural habitat site
Floral biology and reproductive system
and 31 only in the crop area site. The most abundant species,
Campomanesia phaea flowers opened their petals and extended with the exception of a few stingless bee species, were present
their filaments and styles before sunrise, from 04:30 to 05:00 h. at both sites. As quantified in the crop area site, the occurrence
At the moment of full flower opening, the stigma was receptive and abundance of different species of flower visitor varied with
and pollen available. The flower and stigmatic receptivity lasted time of day. At the beginning of anthesis, before sunrise, the
until 19:00 h. Pollen was the only resource offered to the visi- nocturnal bees Megalopta sodalis, Megommation insigne, Ptilo-
tors. The anthers had available pollen until 12:00 noon. The glossa latecalcarata, and Zikanapis seabrai were the only visi-
average pollen viability was 90.3  4.7% (N = 5 flowers). tors, and their abundance was high until 06:00 h (Figs 2 and
Flowers did not produce fruits asexually (agasmospermy 3). Apis mellifera Linnaeus, 1758 (Fig. 2D) appeared shortly
treatment) nor following manual and spontaneous self-pollina- before sunrise, and reached similar high abundances as the
tion, but had a similar fruit set in the cross-pollination (38%) nocturnal/crepuscular bees at and shortly after sunrise, and
and natural conditions (28%) treatments (Table 1). then decreased in abundance until noon. Other diurnal bees
visited the flowers only after sunrise and were, with lower or
similar abundances to the honeybee, also present until noon
(Fig. 3).
Table 1. Percentage fruit set in manual and open (control) treatments in We measured the contact of bees, ventral part of thorax or
C. phaea. In each treatment, flowers from ten different trees were used. abdomen, with the stigma. The large-size bees, such as the noc-
turnal bees P. latecalcarata and Z. seabrai, as well as the day-
treatments N fruit set (%)
active bees of Bombus (two species), Centris (one species) and
cross-pollination 50 38 Epicharis (one species) contacted the stigma in 100% of the vis-
manual self-pollination 50 0 its (Fig. 4). Nocturnal M. sodalis and M. insigne contacted the
spontaneous self-pollination 50 0 stigma in about 60% of visits. All other recorded bees, such as
agamospermy 50 0 A. mellifera, touched the stigma in <30% of visits. Insects other
open pollination 100 28 than bees did not contact the stigma (Fig. 4).
During the pollination efficiency tests, 21 bee species visited
N = total number of flowers.
the flowers. Among these, four nocturnal and seven diurnal
species deposited pollen during a single visit to a flower. How-
Table 2. Number and identity of insects collected on flowers of cambuci
ever, flowers set fruits only following a visit by a nocturnal bee
(C. phaea) at the study sites (crop area, natural habitat). Nocturnal bee spe-
species (Table 3).
cies are in bold. Only species with at least five individuals in any site are
shown. A complete list of insects collected on flowers of cambuci is given in
Table S1.

number of individuals A B

species crop natural habitat

Apidae
Apini
Apis mellifera Linnaeus, 1758 142 37
Bombini
Bombus brasiliensis Lepeletier, 1836 5 5
Meliponini
Melipona quadrifasciata Lepeletier, 1836 12
Paratrigona subnuda Moure, 1947 16 3 C D
Partamona helleri Friese, 1900 23
Trigona spinipes Fabricius, 1793 24
Tapinotaspidini
Trigonopedia sp.1 7
Colletidae
Diphaglossini
Ptiloglossa latecalcarata Moure, 1945 25 8
Zikanapis seabrai Moure, 1953 5 5
Halictidae
Augochlorini
Fig. 2. Flower visitors of C. phaea. A–C. Nocturnal and crepuscular bees:
Megalopta sodalis (Vachal, 1904) 15 5
Megalopta sodalis (A), Megommation insigne (B), Ptiloglossa latecalcarata
Megommation insigne (Smith, 1953) 6 5
(C). Individuals of day-active Apis mellifera (D).

4 Plant Biology © 2016 German Botanical Society and The Royal Botanical Society of the Netherlands
 €tterl & Alves-dos-Santos
Cordeiro, Pinheiro, Do Pollination of a Myrtaceae by nocturnal bees

25 3000

2500
20

N. individuals/30min.

Light intensity (lux)

2000
15
1500
10
1000

5
500

0 0

5 - 5:30 h

6 - 6:30 h

7 - 7:30 h

8 - 8:30 h

9 - 9:30 h

10:30 - 11 h

11:30 - 12 h
5:30 - 6 h

6:30 - 7 h

8:30 - 9 h

9:30 - 10 h

10 - 10:30 h

11 - 11:30 h
Fig. 3. Number (mean SD) of individuals of the flower
visitors of C. phaea (n = 6 trees), and ambient light
intensity (black line, mean of six replicates) over time,
from the beginning of anthesis until noon. : Indication
of sunrise. Nocturnal/Crepuscular bees Apis mellifera Other diurnal species

diurnal bees. These bees, however, were not identified as polli-

Floral volatiles
nators of cambuci. Nevertheless, we cannot exclude that they
The flowers emitted strong scents before and after sunrise, as play a minor role in pollination, not detected by our approach.
perceived by the human nose, and our chemical analyses The four pollinating visitor species were from four different
revealed roughly 1–10 lg of scent per flower and hour bee genera, all of which are well known as being nocturnal
(Table 4). (Warrant 2007).
In total 14 compounds of four chemical classes were found Species from the four pollinating genera were found visiting
in the night scent samples, and with the exception of a trace flowers in other species at night. Ptiloglossa was found on
compound, all of them also occurred in the day-time samples. another Myrtaceae (Siphoneugena densiflora Berg., Proencßa
The volatiles consisted of aliphatics (ten compounds, nine in 1992) and a Melastomataceae (Myrcia pubipetala Miq, Wilms
day samples), aromatics (two compounds), monoterpenes 1995). Megalopta and Megommation visited species of several
(one compound) and N-containing compounds (one com- families other than Myrtaceae (Janzen 1968; Wilms 1995; Hop-
pound; Table 4). The most abundant compounds during both kins et al. 2000; Franco & Gimenes 2011; Krug et al. 2015).
day and night were the aliphatics 1-hexanol and 1-octanol, and Zikanapis visited two species of Campomanesia (including
the aromatics benzyl alcohol and 2-phenylethanol. There was C. phaea) and several species of other families (Wilms 1995).
no overall (all compounds considered) differences in scent Further information on the plant spectrum visited by two
composition between the two time periods (PERMANOVA: Megalopta species comes from analyses of pollen samples from
Pseudo-F1,13 = 0.34, P = 0.76). nests of these bees. In such analyses, pollen of 40 different spe-
cies was detected, among them Bombacaceae, Anacardiaceae
and Guttiferae (Smith et al. 2012). Overall, these data show
Field bioassays
that nocturnal/crepuscular bees visit a broad spectrum of host
Two nocturnal bee species responded in field bioassays. M. plants, they are not specialised to a specific plant taxon but are
insigne and P. latecalcarata flew towards two different concen- generalist foragers (see also Frankie et al. 1983).
trations of the synthetic scent mixture, and to pure 1-octanol, Diurnal bees were not identified as pollinators even though
whereas a significant attractive effect (compared to the negative some deposited pollen onto the stigma. It is possible that diur-
control) could only be detected for the two highest concentra- nal visitors did not promote cross-pollination, which is a con-
tions of the synthetic mixture (P < 0.01 each). The other com- dition for fruit set of C. phaea. For example, the non-native
pounds and the negative control did not attract insects honeybee Apis mellifera was the most abundant diurnal visitor
(Table 5). on cambuci flowers and deposited large amounts of pollen
grains onto the stigma, but was not an effective pollinator. As
in other systems, it may have transferred pollen especially
DISCUSSION
within plant individuals (Goulson 2003; Dohzono & Yokoyama
Our study demonstrates that cambuci flowers are visited by 2010; Aizen et al. 2014), resulting in no fruits due to the self-
nocturnal and diurnal bees, as well as by other diurnal insects, incompatibility of cambuci. Further, by depleting the pollen,
but that only night-active bees are effective pollinators. The the honeybee might have affected potential native diurnal polli-
flowers emit a strong scent before and after sunrise, and nators (e.g. Bombus spp., Melipona spp.) through exploitative
nocturnal bees were attracted by the main scent components. competition (Butz-Huryn 1997; Roubik & Villanueva-
The flowers of cambuci are self-incompatible, which seems Gutierrez 2009).
to be most common among Myrtaceae (Beardsell et al. 1993; The compounds identified in the scent samples of cambuci
Lughadha & Proencßa 1996). They depend on pollinators for are widespread among floral scents (Knudsen et al. 2006; but
setting fruit and, as in other Myrtaceae species, were visited by see octyl formate). Interestingly, floral scent of cambuci

Plant Biology © 2016 German Botanical Society and The Royal Botanical Society of the Netherlands 5
Pollination of a Myrtaceae by nocturnal bees €tterl & Alves-dos-Santos
Cordeiro, Pinheiro, Do

100
Frequency of stigma contact (%)

90
80
70
60
50
40
30
20
10
0

Fig. 4. Percentage of visits with stigma contact in

C. phaea. The insects are listed in order of decreasing
frequency of stigma contact. Nocturnal/crepuscular bees
are represented with blue bars and diurnal insects with
black bars.

Table 3. Number of pollen grains deposited ( SD) onto the stigma during Table 4. Relative amount of each compounds detected in the night and
a single visit by individuals of the most frequent visitor species (all bees) and day floral scent samples of C. phaea (N = 3 individuals; mean of pseudo-
the likelihood of forming a fruit after a single visit. Nocturnal and diurnal replicate samples from several flowers of one individual was used for calcu-
species are in alphabetical order. lations). Compounds are listed according to chemical class. KRI: Kovats
Retention Index. The four most abundant compounds are in bold. Identity
pollen deposition formation of fruits of compounds marked with an asterisk was confirmed with synthetic
standards.
no. pollen
no. grains no. fruit night day
bee species flowers deposited flowers set (%) KRI median (min/max) median (min/max)
nocturnal species total absolute 1648 (886/2788) 829 (540/1096)
Megalopta seabrai 10 85  80 12 50 amount of scent
Megommation insigne 21 39  66 19 53 (ngh1flower1)
Ptiloglossa latecalcarata 28 70  69 17 41 compounds
Zikanapis seabrai 4 61  54 3 33 Aliphatics
diurnal species hexanal* 802 1.38 (1.05/2.73) 5.26 (2.85/8.44)
Apis mellifera 14 81  72 26 0 (Z)-3-hexen-1-ol* 856 4.86 (3.86/8.44) 7.48 (4.80/11.19)
Augochlorini spp. 7 95 5 0 1-hexanol* 868 8.76 (2.50/43.26) 32.81 (18.47/33.00)
Bombus spp. 9 20  40 12 0 2-heptanol* 899 0.84 (0.25/3.41) 0.00 (0.00/0.47)
Eucerini spp. 3 15  6 3 0 hexyl acetate* 1013 1.73 (0.64/2.11) 1.63 (0.63/2.54)
Megachile spp. 2 10  3 1-octanol* 1072 9.00 (2.53/38.46) 14.24 (2.86/15.80)
Melipona spp. 4 14  20 5 0 octyl formate 1130 0.86 (0.14/3.52) 0.38 (0.00/1.36)
Tapinotaspidini spp. 2 25 hexyl butyrate* 1192 0.00 (0.00/0.27) 0.14 (0.00/0.86)
Paratrigona subnuda 3 0 octyl acetate* 1211 0.65 (0.08/7.41) 0.10 (0.00/2.96)
Partamona helleri 5 0 hexyl hexanoate* 1388 0.00 (0.00/0.01) –
Trigona spinipes 5 0 Aromatics
benzyl alcohol* 1039 7.53 (4.14/40.22) 10.25 (7.80/21.68)
2-phenylethanol* 1121 24.62 (6.59/64.40) 24.96 (15.31/34.46)
Monoterpenes
differs from scent released by other plants pollinated at night. linalool* 1100 0.00 (0.00/0.28) 0.97 (0.00/14.06)
This is especially true for plants pollinated by beetles (Gotts- N-containing compounds
berger et al. 2012; Maia et al. 2012; Pereira et al. 2014), cock- indole* 1144 1.24 (0.06/2.14) 0.26 (0.00/2.02)
roaches (Vlasakova et al. 2008), bats (Dobson 2006;
Piechowski et al. 2010) and rodents/shrews (Johnson et al.
2011). Some overlap, however, is evident with plants polli- species (Levin et al. 2001), only released in trace/minor
nated by moths. Many such flowers emit benzyl alcohol and amounts (Dobson 2006; Knudsen et al. 2006). Our beha-
2-phenylethanol, sometimes also as major components (Dob- vioural experiments demonstrate that a mixture of the main
son 2006). The aliphatics 1-hexanol and 1-octanol, however, compounds is capable of attracting nocturnal bee pollinators.
are often not emitted by plants pollinated by nocturnal visi- A few bees also responded to 1-octanol alone. These data are
tors or are, with the exception of 1-hexanol in a single a first proof of the importance of olfactory cues involved in

6 Plant Biology © 2016 German Botanical Society and The Royal Botanical Society of the Netherlands
 €tterl & Alves-dos-Santos
Cordeiro, Pinheiro, Do Pollination of a Myrtaceae by nocturnal bees

Table 5. Number of bees attracted to scent treatments in field bioassays (N = 8). For each assay and each treatment, we offered 500 ll.

SSCF pure* SSCF 103* SSCF 105 1-hexanol 1-octanol 2-phenylethanol benzyl alcohol linalool NC

Megommation insigne 8 4 2 2
Ptiloglossa latecalcarata 4 6 2

SSCF, synthetic scent mixture (see Methods); NC, negative control (paraffin oil).
*Treatment attracted significantly more bees than the negative control (exact binomial test, P < 0.01).

host finding by (likely flower-experienced) nocturnal bees.

They confirm earlier studies performed outside a pollination
context, which found that such bees respond to olfactory
cues in the absence of visual ones (Carvalho et al. 2012; Knoll
& Santos 2012). Olfactory cues are likely also involved in
attraction of diurnal bees. This is because the main com-
pounds of cambuci, i.e. benzyl alcohol and 2-phenylethanol,
are known attractants for solitary and social bees, among
them the honeybee (D€ otterl & Vereecken 2010). Indeed, we
observed honeybee individuals responding to all synthetic
scents offered at the end of our behavioural experiments (at
or shortly after sunrise), however, we did not record them
quantitatively (G.D. Cordeiro, unpublished).
To conclude, our study describes the first scent-mediated
pollination system between a plant and its nocturnal bee polli-
nators. Since nocturnal bees are known to visit a wide range of
plants, these bees may be more important and widespread pol-
linators than known so far. Future studies are needed to test
this hypothesis. Such studies will also show whether plants pol-
linated by such bees are characterised by a common suite of
floral traits. Results from the present study do not suggest that
there is a private communication channel – as recently
described from oil-secreting flowers and oil-collecting bees
(Sch€affler et al. 2015) – between nocturnal bees and their host
plants, given the absence of highly uncommon compounds in
the scent of C. phaea. Nevertheless, we predict that nocturnal
bees rely on an innate (widely tuned) olfactory search image,
potentially modified through learning behaviour when search-
ing for their various host plants, which, however, requires test-
ing in future research.
ACKNOWLEDGEMENTS
We are grateful to Clemens Schlindwein, Claudia In^es da Silva,
Jim Cane, Marcia Maues, Patricia Morellato and Paulo Milet
Pinheiro for significant comments during the study; Mario
Cordeiro Filho, Rafael Takashi and Priscila Bittar for helping in
collecting field data; Claudia In^es da Silva, Elisa Queiros and
Gladys Flavia for helping in pollen analysis; Irmgard Sch€affler
for supporting the scent sample analysis; several taxonomists
(Antonio Aguiar, Dan uncia Urban, Eduardo Almeida, Felipe
Vivallo, Kelly Ramos, Leandro Santos, Leo Correia da Rocha,
Mirian Morales, Rodolfo Arantes, Silvia Pedro and Terry Gris-
wold) for identification of specimens. We are also grateful to
Mr. Alexandre and Mr. Nagai for allowing us to perform the
study on their farms, and Instituto Ecofuturo for allowing the
study at Parque das Neblinas. This study was supported by
Fundo de Amparo a Pesquisa do Estado de S~ao Paulo
(FAPESP; 2011/06811-5, 2013/26599-6) and CNPq (Project
Science without Borders: 401466/2014-0).
SUPPORTING INFORMATION
Additional Supporting Information may be found online in the
supporting information tab for this article:
Table S1. Identity and number of bees collected on flowers
of cambuci (Campomanesia phaea) at two study sites (crop
area, natural habitat); bees with a frequency equal to or higher
than five are in bold.

REFERENCES
Aizen M.A., Morales C.L., Vazquez D.P., Garibaldi
L.A., Saez A., Harder L.D. (2014) When mutualism
goes bad: density-dependent impacts of introduced
bees on plant reproduction. New Phytologist, 204,
322–328.
Anderson M.J., Gorley R.N., Clarke K.R. (2008) PER-
MANOVA+ for PRIMER: guide to software and sta-
tistical methods. PRIMER-E, Plymouth, UK.
Armstrong J.A. (1979) Biotic pollination mechanisms
in the Australian flora – A review. New Zealand Jour-
nal of Botany, 17, 467–508.
Beardsell D.V., O’Brien S.P., Williams E.G., Knox R.B.,
Calder D.M. (1993) Reproductive biology of Aus-
tralian Myrtaceae. Australian Journal of Botany, 41,
511–526.
Berry R.P., Wcislo W.T., Warrant E.J. (2011) Ocellar
adaptations for dim light vision in a nocturnal
bee. Journal of Experimental Biology, 214, 1283–
1293.
Butz-Huryn V.N. (1997) Ecological impacts of intro-
duced honeybees. Quarterly Review of Biology, 72,
275–297.
Carneiro L.T., Martins C.F. (2012) Africanized honey
bees pollinate and preempt the pollen of Spondias
mombin (Anacardiaceae) flowers. Apidologie, 43,
474–486.
Carvalho A.T., Maia A.C.D., Ojima P.Y., Santos A.A.,
Schlindwein C. (2012) Nocturnal bees are attracted
by widespread floral scents. Journal of Chemical Ecol-
ogy, 3, 315–318.
Clarke K.R., Gorley R.N. (2006) PRIMER v6: user man-
ual/tutorial. PRIMER-E Ltd, Plymouth, UK.
Crome F.H., Irvine A.K. (1986) Two bob each way: the
pollination breeding system of the Australian rain-
forest tree Syzygium cormiflorum (Myrtaceae).
Biotropica, 18, 115–125.
Dafni A., Maues M.M. (1998) A rapid and simple pro-
cedure to determine stigma receptivity. Sexual Plant
Reproduction, 11, 117–180.
Dafni A., Kevan P.G., Husband B.C. (2005) Practical
pollination biology. Elsevier, Ontario, Canada, p 590.
Dobson H.E.M. (2006) Relationship between floral fra-
grance composition and type of pollinator. In:
Dudareva N., Pichersky E. (Eds), Biology of floral
scent. CRC Press/Taylor and Francis, Boca Raton,
FL, USA, pp 147–198.
Dohzono I., Yokoyama J. (2010) Impacts of alien bees
on native plant–pollinator relationships: a review
with special emphasis on plant reproduction.
Applied Entomology and Zoology, 45, 37–47.
D€
otterl S., Vereecken N. (2010) The chemical ecology
and evolution of bee–flower interactions: a review
and perspectives. Canadian Journal of Zoology, 88,
668–697.
D€
otterl S., Wolfe L.M., J€ urgens A. (2005) Qualitative
and quantitative analyses of flower scent in Silene
latifolia. Phytochemistry, 66, 203–213.
Endress P.K. (1994) Diversity and evolutionary biology
of tropical flowers. Cambridge University Press, Cam-
bridge, UK, p 420.
Erdtman G. (1966) Pollen morphology and plant taxo-
nomic. Almqvist & Wilsell, Stockholm, Sweden, p
539.

Plant Biology © 2016 German Botanical Society and The Royal Botanical Society of the Netherlands 7
Pollination of a Myrtaceae by nocturnal bees
Faegri K., Van Der Pijl L. (1971) The principles of polli-
nation ecology. Pergamon Press, Oxford, UK, p 256.
Faria F.S., Stehmann J.R. (2010) Biologia reprodutiva
de Passiflora capsularis L. e P. pohlii Mast. (Decaloba,
Passifloraceae). Acta Bot^anica Brasileira, 24, 262–
269.
Fidalgo A.O., Kleinert A.M.P. (2009) Reproductive
biology of six Brazilian Myrtaceae: is there a syn-
drome associated with buzz-pollination? New Zeal-
and Journal of Botany, 47, 355–365.
Franco E.L., Gimenes M. (2011) Pollination of Cam-
bessedesia wurdackii in Brazilian campo rupestre veg-
etation, with special reference to crepuscular bees.
Journal of Insect Science, 11, 1–13.
Frankie G.W., Haber W.W., Opler P.A., Bawa K.S.
(1983) Characteristics and organization of the large
bee pollination system in the Costa Rican dry forest.
In: Jones C.E., Little R.J. (Eds), Handbook of experi-
mental pollination biology. Scientific and Academic
Editions, New York, NY, USA, pp 411–447.
Giannini T.C., Boff S., Cordeiro G.D., Cartolano E.A.
Jr, Veiga A.K., Imperatriz-Fonseca V.L., Saraiva
A.M. (2015) Crop pollinators in Brazil: a review of
reported interactions. Apidologie, 46, 209–223.
Gottsberger G., Silberbauer-Gottsberger I., Seymour
R.S., D€ otterl S. (2012) Pollination ecology of Magno-
lia ovata may explain the overall large flower size of
the genus. Flora, 207, 107–118.
Goulson D. (2003) Effects of introduced bees on native
ecosystems. Annual Review of Ecology and Systemat-
ics, 34, 1–26.
Greiner B., Ribi W.A., Warrant E.J. (2004a) Retinal
and optical adaptations for nocturnal vision in the
halictid bee Megalopta genalis. Cell and Tissue
Research, 316, 377–390.
Greiner B., Ribi W.A., Wcislo W.T., Warrant E.J.
(2004b) Neural organization in the first optic gan-
glion of the nocturnal bee Megalopta genalis. Cell
and Tissue Research, 318, 429–437.
Gressler E., Pizo M.A., Morellato L.P.C. (2006) Polin-
izacß~ao e dispers~ao de sementes em Myrtaceae do
Brasil. Revista Brasileira de Bot^anica, 29, 509–530.
Hopkins M.J.G., Fortune-Hopkins H.C., Sothers C.A.
(2000) Nocturnal pollination of Parkia velutina by
Megalopta bees in Amazonia and its possible signifi-
cance in the evolution of chiropterophily. Journal of
Tropical Ecology, 16, 733–746.
Hopper S.D. (1980) Bird and mammal pollen vectors
in Banksia communities at Cheyne Beach, Western
Australia. Australian Journal of Botany, 28, 61–75.
IUCN (2015) World Conservation Monitoring Centre.
Campomanesia phaea. The IUCN Red list of threat-
ened species. Version 2014.3. www.iucnredlist.org.
Downloaded 14 May 2015.
Jander U., Jander R. (2002) Allometry and resolution
of bee eyes (Apoidea). Arthropod Structure and
Development, 30, 179–193.
Janzen D.H. (1968) Notes on nesting and foraging
behavior of Megalopta (Hymenoptera: Halictidae) in
Costa Rica. Journal of the Kansas Entomological
Society, 41, 342–350
Johnson S.D., Burgoyne P.M., Harder L.D., D€ otterl S.
(2011) Mammal pollinators lured by the scent of a
parasitic plant. Proceedings of the Royal Society of
London, Series B, 278, 2303–2310.
8 Plant Biology © 2016 German Botanical Society and The Royal Botanical Society of the Netherlands
View publication stats
Kawasaki M.L., Landrum L.R. (1997) A rare and
potentially economic fruit of Brazil. Cambuci, Cam-
pomanesia phaea (Myrtaceae). Economic Botany, 51,
403–407.
Klein A.M., Vaissiere B.E., Cane J.H., Steffan-Dewenter
I., Cunningham S.A., Kremen C., Tscharntke T.
(2007) Importance of pollinators in changing land-
scapes for world crops. Proceedings of the Royal Soci-
ety of London, Series B, 274, 303–313.
Knoll F.R., Santos L.M. (2012) Orchid bee baits attract-
ing bees of the genus Megalopta (Hymenoptera, Hal-
ictidae) in Bauru region, S~ao Paulo, Brazil:
abundance, seasonality, and the importance of odors
for dim-light bees. Revista Brasileira de Entomologia,
56, 481–488.
Knudsen J.T., Eriksson R., Gershenzon J., St ahl B.
(2006) Diversity and distribution of floral scent.
Botanical Reviews, 72, 1–120.
K€
oppen W. (1948) Climatologia: con un estudio de los
climas de la tierra. Fondo de Cultura Econ^ omica,
Mexico, p 478.
Krug C., Garcia M.V.B., Gomes F.B. (2015) A scientific
note on new insights in the pollination of guarana
(Paullinia cupana var. sorbilis). Apidologie, 46, 164–
186.
Levin R.A., Raguso R.A., McDade L.A. (2001) Fra-
grance chemistry and pollinator affinities in Nyctagi-
naceae. Phytochemistry, 58, 429–440.
Lorenzi H., Bacher L., Lacerda M., Sartori S. (2006)
Frutas brasileiras e exoticas cultivadas: de consumo in
natura. Instituto Plantarum de Estudos da Flora, S~ao
Paulo, Brazil, p 627.
Lughadha E.N., Proencßa C. (1996) A survey of the
reproductive biology of the Myrtoideae (Myrtaceae).
Annals of the Missouri Botanical Garden, 83, 480–
503.
Maia A.C.D., D€ otterl S., Kaiser R., Silberbauer-Gotts-
berger I., Teichert H., Gibernau M., Navarro
D.M.A.F., Schlindwein C., Gottsberger G. (2012)
The key role of 4-methyl-5-vinylthiazole in the
attraction of scarab beetle pollinators: a unique
olfactory floral signal shared by Annonaceae and
Araceae. Journal of Chemical Ecology, 38, 1072–1080.
Mitchell T.C., D€ otterl S., Schaefer H. (2015) Hawk-
moth pollination and elaborate petals in Cucur-
bitaceae: the case of the Caribbean endemic Lin-
naeosicyos amara. Flora, 216, 50–56.
Ne’eman G., Jurgens A., Newstrom-Lloyd L., Potts
S.G., Dafni A. (2010) A framework for comparing
pollinator performance: effectiveness and efficiency.
Biological Reviews, 85, 435–451.
Ollerton J., Winfree R., Tarrant S. (2011) How many
flowering plants are pollinated by animals? Oikos,
120, 321–326.
Pereira J., Schlindwein C., Antonini Y., Maia A.C.D.,
D€ otterl S., Martins C., Navarro D.M.A.F., Oliveira
R. (2014) Philodendron adamantinum (Araceae)
lures its single cyclocephaline scarab pollinator with
specific dominant floral scent volatiles. Biological
Journal of the Linnean Society, 111, 679–691.
Piechowski D., D€ otterl S., Gottsberger G. (2010) Polli-
nation biology and floral scent chemistry of the
Neotropical chiropterophilous Parkia pendula. Plant
Biology, 12, 172–182.
€tterl & Alves-dos-Santos
Cordeiro, Pinheiro, Do
Proencßa C.E.B. (1992) Buzz pollination – older and
more widespread than we think? Journal of Tropical
Ecology, 8, 115–120.
Proencßa C.E.B., Gibbs P.E. (1994) Reproductive biol-
ogy of eight sympatric Myrtaceae from Central Bra-
zil. New Phytologist, 126, 343–354.
Roitman G.G., Montaldo N.H., Medan D. (1997) Polli-
nation biology of Myrrhinium atropurpureum (Myr-
taceae): sweet, fleshy petals attract frugivorous birds.
Biotropica, 29, 162–168.
Roubik D.W., Villanueva-Gutierrez R. (2009) Invasive
Africanized honey bee impact on native solitary bees:
a pollen resource and trap nest analysis. Biological
Journal of Linnean Society, 98, 152–160.
Sch€affler I., Steiner K.E., Haid M., van Berkel S.S., Ger-
lach G., Johnson S.D., Wessjohann L., D€ otterl S.
(2015) Diacetin, a reliable cue and private commu-
nication channel in a specialized pollination system.
Scientific Reports, 5, 12779.
Silva A.L.G., Pinheiro M.C.B. (2007) Biologia floral e
da polinizacß~ao de quatro especies de Eugenia L.
(Myrtaceae). Acta Bot^anica Brasileira, 21, 235–247.
Smith A.D., Quintero I.J.L., Pati~ no J.E.M., Roubik
D.W., Wcislo W.T. (2012) Pollen use by Megalopta
sweat bees in relation to resource availability in a
tropical forest. Ecological Entomology, 37, 309–317.
Souza M.A.D. (1996) Biologia reprodutiva de onze
especies de Myrtaceae em floresta de terra firme na
Amaz^ onia Central. Masters Dissertation, Instituto
Nacional de Pesquisas da Amaz^ onia/Universidade
Federal do Amazonas, Manaus, Brazil.
Vallilo M.I., Garbelotti M.L., Oliveira E., Lamardo
L.C.A. (2005) Caracterısticas fısicas e quımicas
dos frutos do Cambucizeiro (Campomanesia
phaea). Revista Brasileira de Fruticultura, 27,
241–244.
Vlasakova B., Kalinova B., Gustafsson M.H.G., Teichert
H. (2008) Cockroaches as pollinators of Clusia aff.
sellowiana (Clusiaceae) on Inselbergs in French Gui-
ana. Annals of Botany, 102, 295–304.
Warrant E.J. (2007) Nocturnal bees. Current Biology,
17, R991–R992.
Warrant E.J., Kelber A., Gisle’N A., Greiner B., Ribi
W., Wcislo W.T. (2004) Nocturnal vision and land-
mark orientation in a tropical halictid bee. Current
Biology, 14, 1309–1318.
Wcislo W.T., Tierney S.M. (2009) Behavioural envi-
ronments and niche construction: the evolution of
dim-light foraging in bees. Biological Reviews, 84,
19–37.
Wcislo W.T., Arneson L., Roesch K., Gonzalez V.,
Smith A., Fernandez H. (2004) The evolution of
nocturnal behavior in sweat bees, Megalopta genalis
and M. ecuadoria (Hymenoptera: Halictidae): an
escape from competitors and enemies? Biological
Journal the Linnean Society, 83, 377–387.
Whitaker A.H. (1987) The roles of lizards in New Zeal-
and plant reproductive strategies. New Zealand Jour-
nal of Botany, 25, 315–328.
Wilms W. (1995) Die Bienenfauna im K€ ustenregen-
wald Brasiliens und ihre Beziehungen zu Bl€ uten-
planzen: Fallstudie Boraceia, S~ao Paulo. PhD thesis,
University of T€ ubingen, Germany.