Geographical variations of odour and pollinators, and test for local adaptation by reciprocal transplant of two European Arum species

Geographical variations of odour and pollinators, andtest for local adaptation by reciprocal transplant of twoEuropean Arum speciesMarion Chartier1,2, Laurent P�elozuelo3, Bruno Buatois4, Jean-Marie Bessi�ere4

and Marc Gibernau*,1

1Joint Research Unit Ecology of Guiana Forests, CNRS-UMR 8172, Campus agronomique, BP 316, 97379 Kouroucedex, France; 2Laboratory of Evolution and Biological Diversity, Bât. 4R1, Université Paul Sabatier, 118 route deNarbonne, 31062 Toulouse cedex 9, France; 3Laboratory of Functional Ecology and Environment, Bât. 4R1, UniversitéPaul Sabatier, 118 route de Narbonne, 31062 Toulouse cedex 9, France; and 4Center for Functional and EvolutiveEcology, Université Montpellier 2, 1919 route de Mende, 34293 Montpellier, France

Summary

1. Interactions between entomophilous plants and their pollinators are one of the major

factors shaping the evolution of floral features. As species are distributed in more or less

connected populations, they have evolved in a geographical mosaic of co-evolution were the

outcome of the plant–pollinator interaction is likely to vary as a result of local adaptations.

2. Arum italicum and Arum maculatum are two species of Araceae which deceive their fly poll-

inators by mimicking the odour of their oviposition sites. Whereas A. italicum is known to be

pollinated by flies belonging to different families (i.e. opportunist), A. maculatum relies on only

two pollinating species of the family Psychodidae throughout its European repartition area

(i.e. specialist).

3. The interannual and geographical variations of pollinators and pollinator-attractive odours

were described in several populations of the two species over two consecutive years. Further-

more, local adaptation to pollinators was tested by transplanting inflorescence-bearing plants

between two different sites and by recording the number and composition of the insect fauna

trapped inside the inflorescences during anthesis as a measure of a fitness component.

4. Pollinators and pollinator-attractive odours of the two Arum species varied in time and

space, but there was no clear odour structure between populations. When transplanted, inflo-

rescences of both species trapped the same composition and number of insects as native inflo-

rescences at a given site; this indicates that pollinator composition is highly dependent on the

local availability of insects.

5. No pattern of local adaptation was found for these two species, but local pollination condi-

tions were shown to strongly affect the degree of geographical variations of these interactions.

The lack of a clear odour geographical structure might be due to high gene flow or to similar

selective pressures exerted by pollinators, and the high interindividual odour variation may be

linked to the deceptive strategy adopted by the two plant species.

Key-words: deception, diptera, floral scent, geographical mosaic, Psychodidae, sapromyophily,

transplant experiment

Introduction

It is believed that the interactions that evolved between

many plants species and their insect pollinators have

shaped angiosperm diversification (reviewed by Johnson

2006; Herrera, Castellanos & Medrano 2006). For

instance, the observation that plants pollinated by the

same group of pollinators often converge in their traits,

such as flower colour, odour, size and/or shape, leads to

the concept of pollination syndromes (Stebbins 1970;

Faegri & Van Der Pijl 1971; Proctor, Leo & Lack 1996;

Fenster et al. 2004). Pollinators of a given species can

vary among different populations, leading to plant and*Correspondence author. E-mail: [email protected]

© 2013 The Authors. Functional Ecology © 2013 British Ecological Society

Functional Ecology 2013 doi: 10.1111/1365-2435.12122

pollinator species interactions among populations under

different selective pressures, resulting in a geographical

mosaic of co-evolution (Thompson 2005; Gomez et al.

2009). These populations are more or less interconnected

by gene flows that control the degree of differentiation

among populations. In the case of sufficient divergence of

one of these populations, speciation is likely to occur

(Levin 2000; Thompson 2005). This process may for

instance occur through pollinator shifts: pollinators,

through visitation preference, can isolate a plant genotype

from another (Gould & Johnston 1972; Kiester, Lande &

Schemske 1984; Bradshaw & Schemske 2003). As pollina-

tors are likely to vary from a site to another, they may be

an important factor shaping geographical variations of flo-

ral traits due to local adaptation. Thus, studies on those

variations among populations might help understand the

mechanisms leading to speciation and evolution (Herrera,

Castellanos & Medrano 2006). Recently, some of these

studies revealed striking correlations between pollinator

shifts or changes in pollinator traits and variations of

floral features (e.g. Kato et al. 2000; Johnson & Steiner

1997; Elle & Carney 2003; Valiente-Banuet et al. 2004;

Anderson & Johnson 2007; Gomez et al. 2008; Cosacov,

Nattero & Cocucci 2008; Schlumpberger et al. 2009;

Brown, Downs & Johnson 2011; but see Ellis & Johnson

2009; Nattero, Cocussi & Medel 2010). It was also shown

that floral traits or pollinator composition can vary accord-

ing to abiotic conditions (Blionis & Vokou 2002; Hodgins

& Barrett 2008; Esp�ındola, Pellissier & Alvarez 2011).

Here, we investigated the geographical variations of a

plant–pollinator interaction, focusing on the main attrac-

tive signal: floral odour. The emission of floral odours is a

widespread trait in flowers, and its major function is plant

recognition by pollinators (Pichersky & Gershenzon 2002;

Knudsen et al. 2006; Schaeffer & Ruxton 2011). In some

plant species, it has been highly suspected or shown that

floral odours are the main attractive cues. Such is the case,

for example, in the very specialized interactions between

fig and fig wasps (Gibernau 1997), orchids and male bees

(Schiestl & Schl€uter 2009), yucca and yucca moth (Svens-

son, Pellmyr & Raguso 2011) or Araceae and euglossine

bees (Hentrich, Kaiser & Gottsberger 2010). This is also

the case for flowers attracting pollinators at dusk or during

the night, when visual cues are poorly informative (Maia

et al. 2012, 2013). Floral odours are labile and can vary in

compound composition, in the relative amount of com-

pounds, or their overall quantities (D€otterl, Wolfe &

J€urgens 2005; Raguso 2008). In the context of the

geographical mosaic of co-evolution (as described by

Thompson 2005), local adaptation to pollinators may thus

cause floral scent variations. Floral scent variations were

tested in a few studies, among which gene flow and

geographical distance have been more often proposed to

explain the observed odour variations or stability than

adaptation to local pollinator preferences (Ackerman,

Mel�endez-Ackerman & Salguero-Faria 1997; Knudsen

2002; Solers et al. 2011). So far, a limited number of stud-

ies have investigated geographical variations of both poll-

inators and floral scents (Svensson et al. 2005; Svensson

et al. 2006; Pettersson & Knudsen 2001; Schlumpberger &

Raguso 2008; Ibanez et al. 2010).

Arum italicum Mill. and Arum maculatum L. (Fig. 1) are

two species from the Araceae family, found in temperate

woodlands on the forest floor (Boyce 2006). They are polli-

nated by insects (mainly Diptera) which they lure with

floral odours mimicking ovipositing sites (sapromyophilous)

and entrap within isolated floral chambers for about 24 h

over 2 days (Lack & Diaz 1991; Albre, Quilichini & Giber-

nau 2003; Gibernau, Macquart & Przetak 2004). Up to

now, it has been shown that A. maculatum is mainly polli-

nated by one of only two species from the Psychodidae

family, Psychoda phalaenoides and Psycha grisescens,

according to the population (Prime 1960; Rohacek, Beck-

Haug & Dobat 1990; Lack & Diaz 1991; Diaz & Kite

2002; Chartier, Pelozuelo & Gibernau 2011; Esp�ındola,

Pellissier & Alvarez 2011). Contrastingly, the insect com-

position found in association with inflorescences of A. ital-

icum fluctuates greatly between sites (Gibernau, Macquart

& Przetak 2004; Chartier, Pelozuelo & Gibernau 2011):

different Psychodidae species were found in Spain and in

the South of France, as well as diverse Diptera species

from the families Ceratopogonidae, Sciaridae and Chiro-

nomidae (M�endez & Obeso 1992; Diaz & Kite 2002; Albre,

Quilichini & Gibernau 2003; Albre & Gibernau 2008;

Chartier, Pelozuelo & Gibernau 2011).

The pollinator-attractive odours of A. italicum and

A. maculatum have been studied in England (Kite 1995; Kite

et al. 1998; Diaz & Kite 2002). Recently, it has been pro-

posed that geographical variations of the odour profiles of

these two species were linked to their different degree of spec-

ificity in several populations in France (Chartier, Pelozuelo

& Gibernau 2011): floral odour profiles of A. italicum were

not geographically structured among populations, suggesting

(a) (b)

Fig. 1. Inflorescence of (a) Arum maculatum L., (b) A. italicum

Mill. in Bagn�eres-de-Bigorre, France. (Pictures: M. Chartier).

© 2013 The Authors. Functional Ecology © 2013 British Ecological Society, Functional Ecology

2 M. Chartier et al.

a high level of gene flow or adaptation to a fluctuant guild of

pollinators or to different pollinators having the same olfac-

tory preferences. On the other hand, odour profiles ofA. mac-

ulatum varied between studied populations, suggesting a

lower level of gene flow or adaptation to different local polli-

nator preferences. These hypotheses, nonetheless, remained

to be tested, and the chemical composition of the floral blends

was yet to be elucidated in these populations (Chartier,

Pelozuelo&Gibernau 2011).

In this paper, floral odours and pollinators of A. itali-

cum and A. maculatum were studied in native populations

within a wide geographical range. Furthermore, local

adaptation to pollinators was tested by (i) comparing

odours and pollinator geographical structures for both

species, and (ii) transplanting inflorescences between two

different sites and recording the number and the composi-

tion of insects trapped by the transplanted and native

inflorescences, as a measure of a fitness component. Hence,

if local adaptation to pollinators does occur (through vari-

ations in the attractive odour between populations as an

adaptation to local pollinator preferences), we would

expect that the fitness is lower for the transplanted plants

than for the native plants in each site.

The purpose of this long-term work was to answer three

main questions: (i) How do pollinator and odour composi-

tions vary in time and space in Arum italicum and A. mac-

ulatum? (ii) Is there a link between odour and pollinator

variations for both species? (iii) Are pollinator variations

among sites due to differences in plant attractiveness or in

pollinator availability?

Materials and methods

INFLORESCENCE V IS ITORS

Insects trapped inside the floral chambers of A. italicum and

A. maculatum were collected at seven locations: inflorescences of

A. italicum were sampled in 2009 and 2010 in Chantonnay

(Vend�ee, France, 46°40′N 1°06′O), Smarves (Vienne, France,

46°30′N 0°22′E), Toulouse (Haute-Garonne, France, 43°33′N1°28′E), and Bagn�eres-de-Bigorre (Midi-Pyr�en�ees, France, 43°04′N0°09′E), and in 2009 additionally in Pierrelatte (Drome, France,

44°22′N 4°14′E) and Igeldo (Gipuzkoa, Spain, 43°18′N 2°04′O). A

total of three populations of A. maculatum were sampled in La

Loubati�ere (2009, Aude, France, 43°24′N 2°15′E), Bagn�eres-

de-Bigorre and Smarves (2009 and 2010). Note that Bagn�eres-de-Bigorre and Smarves are sites where A. maculatum and

A. italicum grow intermingled.

Inflorescence visitors were collected in each population in the

morning of the second day of flowering. At this phenological

stage, insects are captive inside the floral chamber. Trapped insects

were collected by pouring 70% ethanol into the floral chamber

and then opening the spathes with a scalpel. The insects were con-

served in 70% ethanol until determination at the family level

under a stereomicroscope, with precious help from a Diptera

taxonomist (Prof. Alain Thomas, University of Toulouse).

As Psychodidae were the most numerous insects trapped, and

the main pollinators of A. maculatum, we estimated their diversity

in each population. When available, four Psychodidae females per

inflorescence from five inflorescences per population were identi-

fied at the species level under a microscope based on their genitalia

and antenna shapes (Vaillant 1988; Withers 1989; Je�zek 1990). In

three populations of A. italicum, <20 Psychodidae were trapped in

all the collected inflorescences. In these cases, all the Psychodidae

were identified.

FLORAL SCENT COLLECT ION

Odours of four to 10 inflorescences of A. maculatum and A. itali-

cum from each of the studied sites were collected in 2009 and 2010.

Inflorescence odours were collected in the evening, between 8 p.m.

and 11 p.m. At this stage, the spathe is widely open, the appendix

is warm and the floral odour is strong. For odour collecting, each

sampled inflorescence (spathe and spadix) was wrapped in a plastic

inert bag (Nalophan NA colourless, calibre 90, available from ETS

Charles-Fr�eres, Saint-Etienne, France) to create an ‘open static

headspace’: the bottom of the bag was kept close under the floral

chamber with a bond, isolating the inflorescence from the leaves

and soil. The top of the bag was kept open 10 cm above the spathe,

to avoid any condensation due to the heat of the appendix. Volatile

organic compounds (VOCs) were collected by solid phase microex-

traction (SPME): VOCs are absorbed and desorbed from a fibre

stored within the needle of a modified syringe. We used Stable-

FlexTM SPME Fibre, 65 lm Polydimethylsiloxane/Divinylben-

zene, handled with a manual holder (available from Supelco,

Bellefonte, PA, USA). The fibre was introduced in the nalophan

bag through a little slit and maintained 0�5–1�0 cm distant from

the appendix for 20 min. Closed empty bags containing ambient

air from 3–4 m away the inflorescences were used as controls

to discard putative VOCs not originating from inflorescences.

Fibres were stored in a freezer (�20 °C) until analyses by gas

chromatography–mass spectometry.

FLORAL SCENT ANALYSES

Gas chromatography–mass spectometry analyses were performed

at the Platform for Chemical Analyses in Ecology of the ‘SFR 119

Montpellier Environnement Biodiversit�e’, at the ‘Centre d’�Ecolo-

gie Fonctionnelle et �Evolutive (Montpellier, France)’. SPME

Fibres were desorbed for 5 min at 250 °C into a 1177 Split/Split-

less injector of a CP-3800 gas chromatograph (Varian Inc., Palo

Alto, CA, USA) coupled to a Saturn 2000 ion trap mass spec-

trometer (Varian Inc.). The carrier gas was helium with a constant

flow rate of 1�0 mL min�1. A split ratio of 1:4 was used. The

temperature of the column (fused silica capillary column, 30 m

9 0�25 mm 9 0�25 lm, CP-Sil 8 CB lowbleed MS, Varian Inc. in

2009; Optima 5 Accent, Macherey-Nagel, D€uren, Germany in

2010) was maintained at 50 °C for 2 min after injection, increased

to 200 °C at a rate of 5 °C min�1, and then increased to 250 °Cat a rate of 10 °C min�1 and maintained at 250 °C for 1 min.

Mass spectra were recorded in scan mode from 38 to 300 m/z with

an electronic impact (EI) at 70 eV. The chemical compounds were

identified by comparison with the NIST98 MS and Adams 2007

mass spectral libraries, and retention indices were found in

libraries and published data (Adams 2007). All chromatogram

peaks were manually integrated and the relative percentage area

of each peak was calculated for each chromatogram (i.e. each

sampled inflorescence). Only peaks representing more than 1% of

the total peaks area per chromatogram were kept for the analyses.

RECIPROCAL TRANSPLANT EXPER IMENT

Plants were transplanted between the sites of Bagn�eres-de-Bigorre

and Toulouse in 2008 and 2009. Entire plants bearing an inflores-

cence were put in pots in the morning, before the opening of the

spathes and transported between sites in the early afternoon. Pots


Geographical variations of Arum pollinators 3

were randomly disposed among the native plants to avoid any

environmental bias. Inflorescences opened in the evening, and the

trapped insects were collected in the following morning. Insects

were preserved in 70% ethanol until taxonomic determination at

the family level. In 2008, plants of A. italicum were transplanted

from Toulouse to Bagn�eres-de-Bigorre and plants of A. macula-

tum were moved from Bagn�eres-de-Bigorre to Toulouse. In 2009,

plants of A. italicum were transplanted from Toulouse to

Bagn�eres-de-Bigorre and plants of the two species from Bagn�eres-

de-Bigorre to Toulouse. The composition and number of the

insects trapped inside inflorescences of these transplanted plants

were compared to those recorded in inflorescences of native plants

at the same sites and during the same days (data from native inflo-

rescences in 2008 are from Chartier, Pelozuelo & Gibernau 2011;

data from 2009 are from the present study).

The composition of Psychodidae trapped inside the trans-

planted inflorescences was estimated in 2009 by identifying female

specimens at the species level under a microscope based on their

genitalia and antenna shapes (Vaillant 1988; Withers 1989; Je�zek

1990). Four female specimens were selected per inflorescence, from

five inflorescences per treatment.

Note that all trapped insects were harvested from inside the

transplanted inflorescences; thus, no pollen from these inflores-

cences could be transported to native inflorescences.

STAT IST ICAL ANALYSES

The quantity of insects trapped inside inflorescences was com-

pared between species or populations of Arum with Wilcoxon

Mann–Whitney (WMW) and Kruskal–Wallis (KW) tests using the

functions wilcox.test() and kruskal.test() from stats package in R

version 2.15.1 (R Core Team, 2012). Insect proportions were com-

pared between groups by nonparametric multivariate analyses of

variance (npMANOVA) with the function adonis() from vegan pack-

age in R (Anderson 2001) using the factors ‘population’, ‘year’

and ‘species’ and their interactions.

For transplantation tests, the insect composition was compared

between groups (native and transplanted inflorescences) with the

same test for each transplantation event. Post hoc tests consisted

of npMANOVAs with a Bonferroni correction. In the results section,

only the statistic values for global tests are given.

To visualize individual variations of odours, VOCs were

grouped by metabolic pathways into four classes: benzenoids,

monoterpenoids, sesquiterpenoids and aliphatics. As indole, a

nitrogen-containing compound, was one of the main compounds

emitted by A. maculatum, it was added as a fifth compound class.

For each species, the relative amount of each compound class for

all individuals was compared between each group (plants grouped

per species, population and year) with a npMANOVA using the func-

tion adonis() in R.

Floral scents were compared between species and between

populations for each species and year according to the relative

percentage of each compound. The significance of the differences

between the different groups was assessed with a npMANOVA using

the function adonis() in R and the factors ‘population’, ‘year’ and

‘species’ and their interactions. NpMANOVAs were used as post hoc,

with a Bonferroni correction.

In addition, the odour variability was compared between the

two species by comparing the mean Jaccard distances among

individuals per species (Ackerman, Cuevas & Hof 2011).

To assess whether there was a correlation between scent compo-

sition and pollinator assemblage, a Mantel test was performed on

the insect and the odour matrices from inflorescences for which

both the odour and pollinators had been identified (across all

populations). As A. maculatum inflorescences trapped primarily

P. phalaenoides and P. grisescens, an initial analysis was per-

formed taking into account the non-Psychodidae families plus the

total number of Psychodidae trapped per inflorescence; in a

second analysis, the total number of Psychodidae was replaced by

an estimation of the number of P. phalaenoides and of P. grises-

cens trapped per inflorescence (extrapolated for each inflorescence

from the percentage of insects from both Psychodidae species

identified in five inflorescences per populations, see section ‘Inflo-

rescence Visitors’). The tests were performed with 10 000 repeti-

tions with the function mantel.randtest() from ade4 package in

R on three data sets: A. italicum and A. maculatum, A. italicum

alone and A. maculatum alone.

Results

OVERALL POLL INATOR VAR IAT IONS

In all studied populations in 2009 and 2010, inflorescences

of Arum italicum and A. maculatum trapped mainly

Diptera from nematoceran families: Psychodidae (10 964

insects), Chironomidae (2007), Ceratopogonidae (107),

Sciaridae (34), plus some Brachycera (209, Diptera) and

Staphylinidae (62, Coleoptera). A total of 51 other insects

were not identified. Inflorescences of A. italicum trapped in

average 18�1 � 2�87 insects (n = 240), including 8�70 � 1�27Psychodidae and 8�01 � 2�09 Chironomidae. Inflorescences

of A. maculatum trapped significantly more insects (WMW:

W = 15�866, P < 1�10�3), with a mean of 84�93 � 17�07insects per inflorescence (n = 107), including 82�94 � 17�02Psychodidae.

The number of insects trapped by inflorescences of both

A. italicum and A. maculatum varied geographically

between populations, but also interannually for a given

population (Table 1). In 2009, inflorescences of A. italicum

trapped two to 49 times more insects in Smarves,

Bagn�eres-de-Bigorre and Igeldo than in Chantonnay, Pier-

relatte and Toulouse (KW: v25, 153 = 65�86, P < 1�10�12).

In 2010, the inflorescences of Bagn�eres-de-Bigorre trapped

51 to 67 times more insects than in the other studied popula-

tions of Toulouse, Chantonnay and Smarves (KW: v23, 85 =50�85, P < 1�10�10). Moreover, interannual variations were

documented in A. italicum for the three studied popula-

tions. Hence, the number of insects trapped per inflores-

cence in 2010 compared to 2009 was three times higher in

Bagn�eres-de-Bigorre (WMW: W = 209, P = 0�003), 33

times lower in Smarves (WMW: W = 343, P < 1�10�6) and

four times lower in Chantonnay (WMW: W = 252,

P < 1�10�3). Geographical variations also occurred for

A. maculatum: inflorescences of this species trapped 11 to

99 times more insects in Bagn�eres-de-Bigorre in 2009 (KW:

v22, 49 = 13�24, P < 1�10�3) and 2010 (MW: W = 788,

P < 1�10�8) than in Smarves and la Loubati�ere. Interannu-

al variations also occurred: the number of insects trapped

per inflorescence in 2010 compared to 2009 was 1�5 times

higher in Bagn�eres-de-Bigorre (WMW: W = 183,

P = 0�031) and 13 times lower in Smarves (WMW:

W = 341, P < 1�10�4).

Pollinator composition exhibited significant differences

between species, populations and years (Figs 2 and 3).

Geographical and interannual variations of pollinator



composition were different for each species of Arum (effect

population*species and year*species were significant in the

npMANOVA), and temporal variations were different accord-

Table 1. Mean number (mean � SE) of insects trapped per inflorescence of A. italicum and A. maculatum in the studied populations from

2008 to 2010

Species Year Site N Psychodidae Chironomidae Others Total

Arum italicum 2008 Toulouse 141 0�34 � 0�06 0�33 � 0�07 1�34 � 0�15 2�02 � 0�21Bagn�eres 14 9�56 � 4�04 16�04 � 6�84 1�12 � 0�56 26�72 � 10�46Toulouse TR 15 14�73 � 4�36 4�67 � 1�44 5 � 1�97 24�4 � 6�65

2009 Bagn�eres 26 13�41 � 2�65 7�67 � 2�51 2�15 � 0�95 23�22 � 4�84Smarves 14 48�57 � 12�91 0�29 � 0�16 0�64 � 0�37 49�5 � 13�16Toulouse 39 0�31 � 0�12 0�15 � 0�08 0�21 � 0�3 1�67 � 0�36Chantonnay 19 4�47 � 0�95 0�21 � 0�12 1�47 � 0�6 6�16 � 1�39Pierrelatte 22 4 � 0�77 0 � 0 1�59 � 0�45 5�59 � 1�04Igeldo 34 4�41 � 1�54 8�06 � 2�25 0�65 � 0�26 13�12 � 3�22Toulouse TR 20 9�86 � 2�27 0�68 � 0�43 1�05 � 0�2 11�59 � 2�51Bagn�eres TR 6 0�17 � 0�09 0 � 0 0�61 � 0�29 0�78 � 0�37

2010 Bagn�eres 29 23�62 � 4�53 48�69 � 15�02 3�55 � 0�87 75�86 � 18�14Smarves 25 0�88 � 0�3 0�28 � 0�09 0�32 � 0�13 1�48 � 0�36Toulouse 15 0 � 0 0 � 0 1�13 � 0�62 1�13 � 0�62Chantonnay 16 0�31 � 0�2 0�56 � 0�22 0�5 � 0�26 1�38 � 0�46

Arum maculatum 2008 Bagn�eres 21 104�68 � 20�84 6�36 � 1�42 4�5 � 2�04 115�55 � 23�22Bagn�eres TR 9 2�3 � 0�5 0 � 0 0�4 � 0�31 2�7 � 0�47

2009 Bagn�eres 24 145�96 � 52�94 1�42 � 0�96 0�38 � 0�12 147�75 � 52�88Smarves 11 16�82 � 7�57 0�73 � 0�41 1�82 � 0�67 19�36 � 7�56Loubati�ere 14 5�36 � 1�4 0�21 � 0�11 0�93 � 0�32 6�5 � 1�44Bagn�eres TR 5 3�8 � 0�66 0 � 0 0�8 � 0�37 4�6 � 0�81

2010 Bagn�eres 24 212�25 � 39�78 1�12 � 0�31 2�75 � 1�15 216�12 � 39�89Smarves 34 0�53 � 0�15 0�35 � 0�13 0�59 � 0�19 1�47 � 0�35

N=number of sampled inflorescences in each site; Total=mean of the total number of insects trapped per inflorescence; Bagn�eres = Bag-

neres-de-Bigorre. Toulouse TR= inflorescence from Toulouse transplanted in Bagn�eres-de-Bigorre; Bagn�eres TR=inflorescences from

Bagn�eres-de-Bigorre transplanted in Toulouse. Data in italics are from the transplanted inflorescences.

Data from ‘Bagn�eres’ and ‘Toulouse’ in 2008 are taken from Chartier, Pelozuelo & Gibernau (2011).

(a)

(b)

Fig. 2. Composition of the inflorescence visitors of Arum italicum

identified at the family level (mean � standard errors of the rela-

tive insects proportions per inflorescence). The same letters indi-

cate groups not significantly different according to the npMANOVA

post hoc test. Bagn�eres = Bagneres-de-Bigorre.

(a)

(b)

Fig. 3. Composition of the inflorescence visitors of Arum macula-

tum identified at the family level (mean � standard errors of the

relative insects proportions per inflorescence). The same letters indi-

cate groups not significantly different according to the npMANOVA

post hoc test. Bagn�eres = Bagneres-de-Bigorre.



ing to populations (effect population*year was significant,

Table 2). There was no significant combined effect of year,

population and species (npMANOVA: df = 1,280, F = 0�88,r2 = 0�002, P = 0�44) and this factor was thus removed

from the model.

In 2009, different populations of A. italicum showed dif-

ferent associated pollinator compositions. Inflorescences

trapped 64–95% of Psychodidae in Smarves, Chantonnay,

Pierrelatte and Bagn�eres-de-Bigorre (Fig. 2a). Insect com-

position found in association with A. italicum inflorescences

in Igeldo and Toulouse was significantly different than that

of all other populations: inflorescences trapped mainly

Chironomidae in Igeldo (61�75%), whereas in Toulouse,

they trapped few Psychodidae (27�22%) and Chironomidae

(7�98%), but a high diversity of other insects (64�8% of all

other categories). In 2010, populations of A. italicum

attracted a lower proportion of Psychodidae than in 2009

and showed two different tendencies (Fig. 2b). In Smarves,

Chantonnay and Bagn�eres-de-Bigorre, inflorescences of

A. italicum attracted a majority of Psychodidae and Chiro-

nomidae, with a high proportion of Brachycera in Chanton-

nay (28�1%) and of unidentified insects in Smarves (20�0%).

The insect composition found in inflorescences from Tou-

louse was significantly different than that of all other popu-

lations, with 66�7% of Sciaridae and 33�3% of Brachycera

(Fig. 2b). Insect compositions varied significantly between

2009 and 2010 in Smarves and Chantonnay, and marginally

significantly in Bagn�eres-de-Bigorre and Toulouse (Fig. 2).

In 2010, inflorescences from all populations of A. macul-

atum trapped 77–90% of Psychodidae, except in Smarves.

The insect composition recovered from inflorescences in

Smarves was intermediate between Bagn�eres-de-Bigorre

and La Loubati�ere. However, in 2010, the insect

composition in Smarves was different than that of

Bagn�eres-de-Bigorre, as inflorescences trapped 42�0%of Psychodidae, 23�6% of Chironomidae and 34�5% of

insects from the other categories (Fig. 3a,b). Interannual

variations were not significant in Bagn�eres-de-Bigorre and

Smarves according to post hoc tests (Fig. 3a,b).

PSYCHODIDAE COMPOSIT ION

All identified Psychodidae belonged to six species: P. gris-

escens (98 insects), P. phalaenoides (81 insects), Psychoda

crassipenis (41 insects), Apsycha pusilla (18 insects), Logima

surcoufi (four insects) and L. albipennis (= Psychoda

parthenogenetica) (one insect) (Je�zek 1990; Je�zek & H�ajek

2007).

All Psychodidae species but Logima albipennis were

recovered inside inflorescences of A. italicum. P. grisescens

was found inside inflorescences of all populations. In

Smarves in 2009 and 2010, the main trapped Psychodidae

species was P. crassipenis (respectively 90 and 94%),

whereas in Bagn�eres-de-Bigorre in 2009 and 2010, it was

P. grisescens (respectively 70% and 85%). In Pierrelatte,

the main trapped species was A. pusilla (90%). In the other

populations, inflorescences trapped similar proportions of

P. grisescens and P. phalaenoides, with in addition

P. crassipenis in Chantonnay and Toulouse in 2009, and

L. surcoufi in Igeldo and Toulouse in 2009 (Fig. 4).

Inflorescences of A. maculatum only trapped Psychodi-

dae from the species P. phalaenoides and P. grisescens,

with one single specimen of Logima albipennis found in

one inflorescence from Smarves in 2009. Inflorescences

trapped mainly P. grisescens in Smarves in 2009 and 2010

(respectively 60% and 100%), or P. phalaenoides in

Bagn�eres-de-Bigorre in 2009 and 2010 and in La Loubat-

i�ere in 2009 (respectively 85%, 95% and 71%, Fig. 4).

(a)

(b)

Fig. 4. Psychodidae diversity found in the inflorescences of Arum

italicum and A. maculatum in 2009 and 2010 in seven different

sites. When possible, four insects per inflorescence from five

different inflorescences were identified per site and species.

Bagn�eres = Bagn�eres-de-Bigorre.

Table 2. Summary table for the npMANOVA on the insect composi-

tion trapped in inflorescences of Arum italicum and A. maculatum

in the studied sites in 2009 and 2010. d.f. = degree of freedom.

Total d.f. = 280

Factor d.f. F r2 P

Population 6 15�78 0�22 <1�10�4

Year 1 24�06 0�055 <1�10�4

Species 1 16�40 0�037 <1�10�4

Population * Species 1 17�29 0�039 <1�10�4

Population * Year 3 5�30 0�036 <1�10�4

Species * Year 1 2�93 0�007 0�037



RECIPROCAL TRANSPLANT EXPER IMENT : OVERALL

INSECT COMPOSIT ION

In 2008 and 2009, Arum italicum inflorescences trans-

planted from Toulouse to Bagn�eres-de-Bigorre trapped the

same composition of insects as the native A. italicum inflo-

rescences in Bagn�eres-de-Bigorre, and a significantly differ-

ent composition than in their native site (overall test

npMANOVA: F = 10�64, r2 = 0�113, P < 1�10�4 in 2008,

F = 9�67, r2 = 0�29, P < 1�10�4 in 2009; Fig. 5). In

Bagn�eres-de-Bigorre, transplanted and native inflorescenc-

es trapped a high percentage of Psychodidae (62�6% and

48�71%), contrary to native inflorescences in Toulouse

(19�7%). In both years, the proportion of Chironomidae

trapped in the transplanted inflorescences in Bagn�eres-

de-Bigorre (23�5% in 2008, 6�4% in 2009) was slightly lower

than in the native inflorescences (41�7% in 2008, 24�5% in

2009), as they attracted a significantly lower number of

Chironomidae (KW: v22 = 30�8, P < 1�10�6 in 2008,

v22 = 58�0, P < 1�10�11 in 2009). In both years, the total

number of insects trapped was six to 16 times higher in

Bagn�eres-de-Bigorre in the transplanted and native inflo-

rescences than in the native inflorescences in Toulouse

(KW: v22 = 30�8, P < 1�10�6 in 2008, v22 = 58�0,P < 1�10�11 in 2009, Fig. 6).

In the same way, Arum italicum inflorescences trans-

planted from Bagn�eres-de-Bigorre to Toulouse in 2009

trapped the same insect composition as the native A. itali-

cum inflorescences in Toulouse, and a significantly different

composition of insects than in their native site (overall test

npMANOVA: F = 9�67, r2 = 0�29, P < 1�10�4, Fig. 5). They

trapped a low proportion of Psychodidae (16�7%) and 30

times less insects than in their native site (MW: v22 = 58�0,P < 1�10�11, Fig. 6).

The insect composition recovered from inflorescences of

A. maculatum transplanted from Bagn�eres-de-Bigorre to

Toulouse was not significantly different than that of native

A. maculatum inflorescences in Bagn�eres-de-Bigorre (Fig. 5c;

npMANOVA: F = 2�61, r2 = 0�09, P = 0�063 in 2008; F = 3�35,r2 = 0�11, P = 0�051 in 2009). In all cases, inflorescences

trapped more than 85% of Psychodidae. In both years, the

number of insects trapped by the transplanted inflorescenc-

es in Toulouse was 25–55 times lower than that recovered

inside native inflorescences in Bagn�eres-de-Bigorre

(Table 1, Fig. 6). Note that A. maculatum does not

naturally occur in Toulouse, but the transplanted inflores-

cences of this species trapped more insects than those of

A. italicum in this locality.

RECIPROCAL TRANSPLANT EXPER IMENT :

PSYCHODIDAE COMPOSIT ION

Inflorescences of A. italicum transplanted from Toulouse

to Bagn�eres-de-Bigorre trapped the same proportions of

Psychodidae species as the native A. italicum inflorescences

in Bagn�eres-de-Bigorre (65% of P. grisescens and 35% of

P. phalaenoides, Fig. 7). As inflorescences of A. italicum

trapped few Psychodidae in Toulouse, we were only able to

determine three specimens from inflorescences of Bagn�eres-

de-Bigorre transplanted in Toulouse: one L. surcoufi and

two P. crassipenis, which was consistent with the composi-

tion of Psychodidae found in the native inflorescences of

A. italicum in Toulouse. The inflorescences of A. maculatum

transplanted from Bagn�eres-de-Bigorre to Toulouse trapped

89% of P. phalaenoides, similarly to what was observed in

(a)

(b)

(c)

Fig. 5. Composition of the visitors of the native and transplanted

inflorescences of Arum italicum and A. maculatum in Toulouse

and Bagn�eres-de-Bigorre in 2008 and 2009 (mean � standard

errors of the relative insects proportions per inflorescence).

Bagn�eres = Bagn�eres-de-Bigorre. Toulouse TR = inflorescences

from Toulouse transplanted in Bagn�eres-de-Bigorre. Bagn�eres

TR = inflorescences from Bagn�eres-de-Bigorre transplanted in

Toulouse. The same letters indicate groups not significantly differ-

ent (one statistical npMANOVA test was performed per transplant

experiment).



their native site, but the minor trapped species were

A. pusilla and an unknown Psychodidae species (Unknown

sp.), whereas in their native sites, they trapped P. grisescens

as a minor species.

FLORAL BLEND COMPOSIT ION

A total of 44 different VOCs were identified in the blends

of the two Arum species; most of them were mono- and

sesquiterpenoids (Tables 3 and 4). A total of 19 VOCs

were common to both species. In A. italicum, seven

dominant compounds represented 75% of the odour: 3,

7-dimethylocta-1,6-diene (b-citronellene) (31�81% � 1�89),b-caryophyllene (12�55% � 1�85), 3,7-dimethyloct-1-ene

(12�51% � 1�39), p-cresol (11�12% � 1�53), 2,6-dimethyl-

3-octene (5�61% � 0�84) and an unidentified dihydroses-

quiterpene (RI = 1395, 9�27% � 1�14). The remaining 21

compounds each represented <5% of the average blend.

Arum maculatum differed mainly from A. italicum by the

emission of indole (21�43% � 3�67). In A. maculatum,

three dominant compounds represented 45% of the odour:

indole, limonene (16�68% � 3�76) and a-pinene (7�33% �1�86); the remaining 30 compounds each represented <5%of the total average blend.

FLORAL BLEND VAR IAT IONS

According to the variations of the five classes of com-

pounds among individuals (Fig. 8), inflorescence odour

profiles differed significantly between populations of the

two Arum species, but no population was significantly

different from all the others according to the post hoc tests.

In Arum italicum (npMANOVA: F = 4�61, r2 = 0�40, P =1�10�4), Bagn�eres-de-Bigorre in 2009 significantly differed

from Toulouse in 2009 and Bagn�eres-de-Bigorre in 2010.

In A. maculatum (npMANOVA: F = 5�28, r2 = 0�43,P = 2�10�4), inflorescence odours in Smarves 2010 signifi-

cantly differed from those of La Loubati�ere 2009 and

Smarves 2009; inflorescence odours in Bagn�eres-de-Bigorre

2010 differed from those of La Loubati�ere 2009.

When considering the relative percentage of each

compound, geographical and interannual variations were

different for each species (effect population*species and

year*species were significant in the npMANOVA), and inter-

annual variations were different according to populations

(effect population*year was significant, Table 5). There

was no significant combined effect of year, population and

species (npMANOVA: df = 1,80, F = 2�31, r2 = 0�013,

Fig. 7. Psychodidae diversity found in the native and transplanted

inflorescences of Arum italicum and A. maculatum in Toulouse

and Bagn�eres-de-Bigorre in 2009. When possible, identifications

were done on four insects per inflorescence, from five different infl-

orescences per site and species. Bagn�eres = Bagn�eres-de-Bigorre.

Toulouse TR = inflorescences from Toulouse transplanted in

Bagn�eres-de-Bigorre. Bagn�eres TR = inflorescences from

Bagn�eres-de-Bigorre transplanted in Toulouse.

(a)

(b)

(c)

Fig. 6. Number of insects trapped the native and transplanted infl-

orescences of Arum italicum and A. maculatum in Toulouse and

Bagn�eres-de-Bigorre in 2008 and 2009 (mean � standard errors).

Bagn�eres = Bagn�eres-de-Bigorre. Toulouse TR = inflorescences

from Toulouse transplanted in Bagn�eres-de-Bigorre. Bagn�eres

TR = inflorescences from Bagn�eres-de-Bigorre transplanted in

Toulouse. Letters indicate populations nonsignificantly different

according to post hoc tests (one statistical test was performed per

transplant experiment).



Table

3.Meanrelativeamounts

ofVOCsproducedbyArum

italicum

instudiedpopulationsin

2009and2010

VOCs

RI

RT

Arum

italicum

Bagn� eres-de-Bigorre

Smarves

Igeldo

Toulouse

Chantonnay

2009(N

=10)

2010(N

=5)

2009(N

=7)

2010(N

=4)

2009(N

=10)

2009(N

=5)

2009(N

=7)

Mean

SE

OMean

SE

OMean

SE

OMean

SE

OMean

SE

OMean

SE

OMean

SE

O

Benzenoids

p-cresol*

1083

11,79

7�53

3�02

713�73

1�85

519�91

3�48

74�09

4�09

17�89

2�33

922�58

7�87

56�07

1�78

6

Monoterpenoids

Linear

3,7-dim

ethyloct-1-ene*

908

6,71

17�61

2�61

10

4�28

0�61

511�31

2�51

66�82

2�92

317�57

3�02

10

2�71

0�97

415�36

4�95

7

2,6-dim

ethylocta-1,7-diene

(a-citronellene)

928

7,28

3�48

0�44

10

2�26

0�5

53�49

0�49

71�5

1�5

12�83

0�49

10

2�55

1�38

33�15

0�5

7

3,7-dim

ethylocta-1,6-diene*

(b-citronellene)

941

7,66

36�87

2�76

10

18�38

2�04

532�34

3�13

747�63

10�16

431�25

2�41

10

15�05

5�49

537�35

4�25

7

2,6-dim

ethyl-3-octene

965

8,31

9�01

1�78

10

3�07

0�95

41�47

0�28

62�43

1�74

27�56

2�01

10

1�59

1�59

18�58

2�94

7

myrcene

987

8,94

––

––

––

0�33

0�33

10�68

0�68

1–

––

––

––

––

2,6-dim

ethylocta-2,6-diene*

995

9,15

––

–1�16

0�53

3–

––

4�39

4�39

1–

––

––

––

––

3,7-dim

ethylocta-2,6-diene

997

9,22

5�93

1�6

9–

––

0�44

0�28

2–

––

3�68

0�92

70�97

0�97

14�45

26

3,7-dim

ethyloct-1-en-6-ol*

(dihydromyrcenol)

1074

11,52

1�62

0�78

4–

––

0�78

0�4

3–

––

2�11

0�75

60�54

0�33

20�84

0�42

3

linalool

1099

12,27

––

–1�38

0�39

4–

––

3�22

2�44

2–

––

––

––

––

Cyclic limonene*

1027

10,12

––

–9�29

2�24

51�46

1�46

111�54

7�98

3–

––

––

–1�22

0�36

5

dihydromonoterpene

(menthene?

*)

1021

9,94

0�31

0�21

20�29

0�29

1–

––

––

––

––

––

–0�18

0�18

1

Sesquiterpenoids

isocaryophyllene

1404

20,77

1�43

0�54

64�76

1�23

50�17

0�17

10�41

0�41

12�1

0�59

74�87

1�53

41�38

0�66

3

b-caryophyllene*

1420

21,18

8�85

2�46

927�14

3�6

54�92

2�64

55�66

3�08

412�81

2�72

10

24�54

11�78

510�03

3�45

7

a-humulene*

1456

22,08

0�64

0�27

42�62

0�79

40�18

0�18

1–

––

0�9

0�38

43�92

1�25

41�04

0�54

3

a-selinene

1497

23,09

––

–0�22

0�22

1–

––

0�52

0�52

1–

––

0�6

0�37

2–

––

d-cadinene*

1517

23,58

0�12

0�12

10�87

0�36

30�44

0�44

1–

––

––

–1�79

0�57

40�21

0�21

1

spathulenol

1580

25,06

––

–0�38

0�38

1–

––

0�29

0�29

1–

––

––

––

––

alloaromadendrene*

1461

22,19

––

–0�43

0�43

1–

––

––

––

––

0�67

0�41

20�16

0�16

1

tetrahydrosesquiterpene

1366

19,77

––

––

––

1�36

0�7

3–

––

0�33

0�33

10�42

0�42

10�45

0�45

1

dihydrosesquiterpene

1395

20,56

4�63

1�84

97�98

1�6

518�65

4�6

75�57

2�84

38�74

1�47

10

12�06

3�77

58�29

27

2sesquiterpenes

1476

22,56

––

–0�37

0�37

1–

––

––

––

––

0�28

0�28

1–

––

Carotnoid

derivatives

6-m

ethylhept-5-en-2-one*

987

8,93

––

––

––

0�47

0�47

10�49

0�49

10�15

0�15

11�03

1�03

1–

––

Fattyacidderivatives

2-heptanone*

893

6,33

1�84

0�8

60�36

0�36

11�2

0�39

5–

––

2�08

0�58

72�18

0�78

41�07

0�54

3

2-m

ethylundecane

1154

13,92

––

––

––

––

–2�38

1�74

2–

––

––

––

––



P = 0�05) and this factor was thus removed from the

model. No population was different from all the others

when considering post hoc tests for both species.

In 2009, the inflorescence odours of A. italicum were dif-

ferent in Bagn�eres-de-Bigorre, Smarves and Chantonnay

when compared to Igeldo; and in Toulouse, they were

intermediate between the latter and Chantonnay. In 2010,

there was no significant difference between inflorescence

odours in Smarves and Bagn�eres-de-Bigorre. Floral odour

did not vary significantly between 2009 and 2010 in Smar-

ves, but did so in Bagn�eres-de-Bigorre, a result which may

be due to an increase of the emission of limonene and b-caryophyllene (Table 3).

In 2009, there was no significant difference in the inflo-

rescence odours of A. maculatum between the populations

of Bagn�eres-de-Bigorre, Smarves and La Loubati�ere, nor

in 2010 between Smarves and Bagn�eres-de-Bigorre. Odours

varied significantly between 2009 and 2010 in Smarves, but

not in Bagn�eres-de-Bigorre (Table 4).

The mean Jaccard distances among individuals were 0�44� 0�01 for A. italicum and 0�68 � 0�01 for A. maculatum.

CORRELAT ION BETWEEN SCENT AND POLL INATOR

COMPOSIT IONS

When integrating the total number of Psychodidae in the

analysis, we found a significant correlation between scent

dissimilarities and pollinator dissimilarities for A. italicum

alone (Mantel test: n = 38, obs = 0�217, P = 0�003), but

not for A. maculatum alone (Mantel test: n = 28,

obs = 0�156, P = 0�200). There was also no significant

correlation when integrating data from both Arum species

(Mantel test: n = 66, obs = 0�035, P = 0�207). When inte-

grating the number of P. phalaenoides and P. grisescens in

the analysis, we found a significant correlation between

scent dissimilarities and pollinator dissimilarities for

A. maculatum alone (Mantel test: n = 14, obs = 0�229,P = 0�045) and when integrating data from both Arum

species (Mantel test: n = 31, obs = 0�515, P < 10�3); there

was no significant correlation for A. italicum alone

(Mantel test: n = 17, obs = 0�380, P = 0�073).

Discussion

In this study, it was demonstrated that differences in floral

scent composition of two species of Arum influenced the

composition of the pollinator fauna associated with them.

However, there was no clear correlation between pollinator

variations among populations and odour composition. In

fact, pollinator attraction in each site was highly depen-

dent on local insect availability and diversity during the

transplant experiments, indicating no clear pattern of local

adaptation of the two Arum species to pollinators. The

two studied pollination systems varied between years and

sites. These variations were high for A. italicum, pollinated

by insects belonging to six different families (and by five

different Psychodidae species according to the site), andTable

3(C

ontinued)

VOCs

RI

RT

Arum

italicum

Bagn� eres-de-Bigorre

Smarves

Igeldo

Toulouse

Chantonnay

2009(N

=10)

2010(N

=5)

2009(N

=7)

2010(N

=4)

2009(N

=10)

2009(N

=5)

2009(N

=7)

Mean

SE

OMean

SE

OMean

SE

OMean

SE

OMean

SE

OMean

SE

OMean

SE

O

decanal

1203

15,36

––

––

––

––

–1�12

1�12

1–

––

0�56

0�56

1–

––

Unknowncompounds

unknown1

1645

26,52

0�12

0�12

1–

––

0�18

0�18

1–

––

––

––

––

––

–Groups

ab

acd

bde

efbcf

ab

RI=

retentionindex;RT=

retentiontime;

N=

number

ofsampledinflorescences;

SE=

standard

error;

O=

number

ofchromatogrammswheretheVOC

wasrecorded;VOC

=volatile

organic

compounds.

Compoundsrepresentingmore

than5%

oftheaveragetotalblendare

highlightedin

bold.In

thebottom

line,

thesamelettersindicate

groupsnotsignificantlydifferent.

*Compoundsalsodescribed

byKiteet

al.(1998)orDiaz&

Kite(2002).



lower for A. maculatum, mainly pollinated by two species

of Psychodidae. As reported in similar studies, we found

that the different populations of these two species grow

under different pollination evolutionary contexts according

to their geographical location (Thompson 2005; e.g. Kato

et al. 2000; Elle & Carney 2003; Valiente-Banuet et al.

2004; Anderson & Johnson 2007; Cosacov, Nattero &

Cocucci 2008; Schlumpberger et al. 2009). The geographi-

cal and interannual variations of these two pollination

systems may be shaped by local insect availability and/or

insect odour preferences and floral scent composition.

Transplantation tests with both A. italicum and A. macula-

tum showed that for the two Arum species, both the

quantity and the composition of trapped insects varied

according to the local insect availability, rather than to the

plant population. As these species are strictly allogamous,

the quantity of trapped insects is a direct component of

their fitness (Morgan 2006; Albre & Gibernau 2008;

Gomez et al. 2009). Moreover, as insect pollinating efficiency

differs (Albre, Quilichini & Gibernau 2003), differences in

both composition and quantity of insects trapped inside

transplanted and native plants should reflect the degree of

local adaptation of the plants to their local pollinator

fauna.

Local adaptation for pollinators has already been dem-

onstrated by flower manipulations (Johnson & Steiner

1997) or by tests for pollinator preferences (Elle &

Carney 2003; Gomez et al. 2008), but reciprocal transplant

tests were performed only in a few studies. Waterman

et al. (2011) showed that seed set decreased in three pairs

of recently diverged orchids when individuals were recip-

rocally transplanted out of their native sites, and hypoth-

esized that it was due to local adaptations to pollinators

(see also Campbell 2003). Similarly, Gomez et al. (2009)

showed that transplanted Erysimum plants (Brassicaceae)

coming from evolutionary hot spots exhibited a higher

attractiveness for pollinators than those coming from

cold spots. On the other way, in our study, inflorescences

from transplanted plants trapped the same number and

composition of pollinators than did those from native

plants, thus neither attractiveness nor pollinator composi-

tion seemed to depend on plants’ genotypes (e.g. popula-

tions), but rather on pollinator availability in each site.

Such an observation implies that at least A. italicum (and

maybe also A. maculatum) shows high ecological adaptive

plasticity. This plasticity may partially contribute to the

widespread distribution of these two species (Linz et al.

2010), since adaptation to local pollinators is a factor

that reduces species’ fitness, especially when it is coloniz-

ing or is introduced in a new habitat that presents new

pollinator guilds (Geber & Eckhart 2005; Angert, Brad-

shaw & Schemske 2008). In the related species Arum

palaestinum, variations in the occurrence of two main

pollinators in different populations were associated with

differences in pollinator availability (according to the

rurality of the sampled sites) (St€okl et al. 2010). In popu-

lations of A. maculatum, it has been suggested that polli-

nator distribution varies according to climatic conditions:

in a large-scale study covering its distribution in Europe,

Esp�ındola, Pellissier & Alvarez (2011) found that A. mac-

ulatum was mainly pollinated by P. phalaenoides in

Northern and Central Europe, and that the proportions

of P. grisescens trapped inside inflorescences increased in

the Mediterranean area and in north-west of France. Our

results are coherent with these results, as inflorescences

of A. maculatum caught a majority of P. phalaenoides in

the two populations sampled in non-Mediterranean south

of France, and a majority of P. grisescens in populations

in the north-west of France. An insufficient number of

populations of A. italicum were sampled to test for a

geographical gradient of pollinator composition, but, as

suggested by Chartier, Pelozuelo & Gibernau (2011), the

populations in the north of France tended to attract more

Psychodidae than the southern populations, similarly to

what was observed by Diaz & Kite 2002 in England.

It is also interesting to note that in the syntopic popula-

tions of Smarves and Bagn�eres-de-Bigorre, the two Arum

species growing intermingled were associated with a

different preferential species of Psychoda. The majority of

P. grisescens were entrapped by inflorescences of A. itali-

Fig. 8. Odour profile of sampled inflorescence of Arum italicum and A. maculatum, ordered per populations and years. Stacked bars repre-

sent the relative amount of the main compound classes. Bagn�eres = Bagneres-de-Bigorre.



cum in Bagn�eres-de-Bigorre and by inflorescences of

A. maculatum in Smarves. This suggests that there might

be a mechanism of avoidance of hybridization or/and

competition, which could reinforce interspecific odour

divergence, favouring pollinator shift when the two species

grow in syntopy (Levin & Anderson 1970; Waser 1978;

Mitchell et al. 2009). Odour variations between sympatric

related species and their hybrids can be involved in specia-

tion and competition avoidance and are thus particularly

interesting in the context of the geographical mosaic of

Table 4. Mean relative amounts of VOCs produced by Arum maculatum in 2009 and 2010

VOCs RI RT

Arum maculatum

Bagn�eres-de-Bigorre Smarves La Loubati�ere

2009 (N = 9) 2010 (N = 5) 2009 (N = 8) 2010 (N = 4) 2009 (N = 7)

Mean SE O Mean SE O Mean SE O Mean SE O Mean SE O

Benzenoids

acetophenone 959 8,15 – – – 1�03 0�46 3 – – – 0�55 0�55 1 – – –p-cresol* 1083 11,79 1�12 0�78 3 6�88 4�21 3 5�16 3�4 2 – – – 1�49 1�05 2

N-containing compounds

indole* 1298 17,97 26�88 7�61 7 20�85 11�15 5 26�77 8�33 7 1�99 1�19 2 15�99 4�55 6

Monoterpenoids

Linear

3,7-dimethyloct-1-ene 908 6,71 3�1 2�41 4 0�36 0�36 1 – – – – – – 0�82 0�53 2

3,7-dimethylocta-

1,6-diene

941 7,66 8�03 4�45 7 4�8 0�82 5 3�86 1�28 7 3�05 1�78 2 0�97 0�49 3

myrcene* 987 8,94 0�5 0�38 2 1�51 0�11 5 1�48 0�64 5 0�41 0�41 1 1�7 0�63 5

3,7-dimethylocta-

2,6-diene

997 9,22 1�37 1�17 2 – – – 0�39 0�27 2 – – – 0�16 0�16 1

linalool 1099 12,27 – – – 4�62 1�18 4 – – – 13�89 2�81 4 – – –Cyclic

limonene* 1027 10,12 6�86 3�28 5 41�99 11�16 4 4�29 2�16 5 50�53 4�8 4 5�05 2�58 5

a-pinene* 933 7,42 11�43 5�05 5 2�4 0�73 4 5�28 1�82 6 1�36 0�79 2 9�71 4�06 6

b-pinene 974 8,57 – – – 2�3 0�61 4 – – – 1�21 0�71 2 – –Sesquiterpenoids

isocaryophyllene 1404 20,77 0�55 0�39 2 1�06 0�8 2 2�27 0�71 7 – – – 1�26 0�94 2

b-caryophyllene* 1420 21,18 2�44 0�96 5 1�54 1�54 1 7�7 1�09 8 – – – 5�84 4�33 3

a-humulene* 1456 22,08 2�36 0�95 5 – – – 6�21 0�57 8 – – – 1�54 0�87 3

a-copaene* 1375 20,01 0�29 0�29 1 – – – 1�74 0�57 5 – – – 0�23 0�23 1

a-selinene 1497 23,09 1�94 0�72 5 0�3 0�3 1 4�8 1�27 7 0�42 0�42 1 2�9 2�41 2

d-cadinene* 1517 23,58 5�95 3�25 6 – – – 5�34 0�64 8 – – – 2�82 1�11 5

alloaromadendrene* 1461 22,19 2�01 0�82 5 0�24 0�24 1 4�18 1�21 7 – – – 0�67 0�67 1

2 sesquiterpenes 1476 22,56 1�8 0�72 5 – – – 4�34 0�79 8 – – – 0�3 0�3 1

dihydrosesquiterpene 1395 20,56 5�3 1�27 8 3�25 1�02 4 1�8 0�83 5 – – – 1�98 0�84 4

Carotenoid derivatives

geranylacetone* 1445 21,79 0�71 0�56 2 0�25 0�25 1 0�31 0�2 2 0�89 0�52 2 3�93 1�18 6

6-methylhept-5-

en-2-one*987 8,93 – – – – – – 0�96 0�58 3 – – – 11�63 3�83 7

Fatty acid derivatives

2-heptanone* 893 6,33 5�26 3�76 7 – – – 4�26 1�46 7 – – – 8�48 3�55 5

alcane 996 9,2 – – – – – – – – – 2�13 0�92 3 – – –2-methylundecane 1154 13,92 – – – – – – 5�69 2�13 6 7�94 4�59 2 5�84 2�08 5

nonanal 1006 9,48 – – – – – – – – – – – – 2�04 0�92 4

decanal* 1203 15,36 9�17 3�57 6 1�94 0�5 4 1�98 0�64 6 4�93 0�67 4 9�56 3�15 7

alcane 1481 22,7 – – – – – – – – – – – – 3�59 3�59 1

Unknowns compounds

unknown 1330 18,82 – – – – – – – – – 0�79 0�48 2 – – –unknown 1342 19,14 1�63 1�63 1 – – – – – – 3�83 2�41 2 1�07 1�07 1

unknown 1392 20,48 – – – – – – – – – 0�95 0�58 2 – – –unknown 1401 20,71 – – – 1�02 0�43 3 – – – 1�04 0�6 2 – – –

Groups abc bd ce d ae

RI= retention index; RT= retention time; N= number of sampled inflorescences; SE= standard error; O= number of chromatogramms

where the VOC was recorded; VOC= volatile organic compounds. Compounds representing more than 5% of the average total blend are

highlighted in bold.

In the bottom line, the same letters indicate groups not significantly different.

*Compounds also described by Kite et al. (1998) or Diaz & Kite (2002).



co-evolution. Such is the case, for example, of some decep-

tive orchids (e.g. Schiestl & Ayasse 2002; St€okl et al. 2008)

and Silene species (Waelti et al. 2008).

Whereas the lack of clear geographical structure of

odour is consistent with the results from Chartier, Pelozu-

elo & Gibernau (2011) for A. italicum, we found no

geographical structure in the odour of A. maculatum either

in 2009 and 2010, surprisingly in contradiction with our

results from 2008 (Chartier, Pelozuelo & Gibernau 2011).

The geographical structure found in 2008 would have been

consistent with the hypothesis of Esp�ındola, Pellissier &

Alvarez (2011), suggesting that the floral odour of A. mac-

ulatum could be adapted to attract P. grisescens in some

populations and P. phalaenoides in others. This geographi-

cal structure was anyway poorly supported as only two

populations were sampled for odours in 2008, one of them

represented by only three individuals (Chartier, Pelozuelo

& Gibernau 2011). Instead, we found no clear odour varia-

tion/structure between populations of the two Arum

species, but their interindividual variations were high. The

lack of interpopulation odour variations can be either

explained as (i) a consequence of high gene flow among

populations of the two species, preventing a genetic differ-

entiation and thus the divergence of odour in some popu-

lations (Knudsen 2002; e.g. Svensson et al. 2005); or (ii) a

consequence of similar selective pressures for pollinator

attraction in distant populations (Svensson et al. 2006).

This could be the case for the attraction of P. phalaenoides

by inflorescences of A. italicum: the five major compounds

dominating the floral odour of A. italicum in our study

only slightly differed from the ones identified in British

populations primarily pollinated by Psychodidae (mainly

P. phalaenoides) and Chironomidae (Smittia pratorum)

(Diaz & Kite 2002).

The mean Jaccard distance indices between individual

floral odours of A. italicum (0�44) and A. maculatum (0�68)were both closer to the mean interindividual Jaccard index

of deceptive flowers (0�55) than to rewarding flowers

(0�28). These results were issued in a comparative study

over 12 species of rewarding flowers vs. deceptive flowers

(Ackerman, Cuevas & Hof 2011), indicating that deception

might be a factor selecting for high interindividual varia-

tions in the attractive odour. The odours of both Arum

species were mainly composed of sesquiterpenoids and

monoterpenoids, with some aliphatics (in higher propor-

tions in the floral odour of A. maculatum), and an aromatic

heterocyclic organic compound containing nitrogen, indole,

exclusively found in the floral odour of A. maculatum.

Among these compounds, some are very common

components of flower scents, such as a-pinene, limonene

and b-caryophyllene (Knudsen et al. 2006), and 2-hepta-

none, p-cresol, indole, a-pinene, limonene and b-caryo-phyllene are typical of the odours of sapromyophilous

plants (J€urgens, D€otterl & Meve 2006; Urru, Stensmyr &

Hansson 2011), and were also found in the odour of dung

from sheep, cow, horse or boar (Kite 1995; Dormont et al.

2010; Johnson & J€urgens 2010). In addition, 2-heptanone,

p-cresol and indole were shown to be attractive to psycho-

did flies, even more when they were mixed together (Kite

et al. 1998). St€okl et al. (2010), in their detailed study of

the pollination mechanism of the deceiving A. palaestinum,

proposed that the deceptive fermentation odour of the

plant was constituted of various attractive compounds

forming a ‘super-attractive mixture’ acting on Drosophila

innate preferences for fermentation-associated volatiles.

The mechanism is likely to be similar in inflorescences of

A. italicum and A. maculatum. The sapromyophilous flies

they deceive are known to look for a resource to oviposit,

but apparently inflorescences of neither studied species are

breeding sites for attracted flies, as no eggs, larvae or

pupae were ever observed in them (M. Chartier, pers.

obs.). This might suggest that a strong selective pressure is

likely to influence the behaviour of insects associated with

these aroids, favouring the ones able to ‘recognize’ and

avoid the inflorescences (Renner 2006). Arum italicum and

A. maculatum might thus be subjected to a balancing selec-

tion in favour of interindividual variations in the composi-

tion of the ‘super-attractive mixture’, to compensate any

evolutionary response from pollinators (Thompson 2005;

Renner 2006). In other plant species, it has been suggested

that high odour variations between individuals resulted

from few selective pressures from pollinators, for instance

if other cues play an important role in pollinator attraction

(Ibanez et al. 2010). This is not likely the case of Arum

italicum and A. maculatum were the major pollinator

attractive feature is known to be the floral odour (Lack &

Diaz 1991; reviewed by Gibernau, Macquart & Przetak

2004 and Urru, Stensmyr & Hansson 2011). High individ-

ual odour variations can also conceal the stability of physi-

ologically active compounds, which may represent a more

or less important proportion of the total variable odour

(Ayasse et al. 2000; Mant, Peakall & Schiestl 2005; Ibanez

et al. 2010), or be an adaptive response to high variations

of pollinators from year to year (Geber & Moeller 2006).

These two hypotheses could also explain the pattern found

in A. italicum and A. maculatum.

Future research should focus on coupling biotests on

pollinators and accumulating genetic information on plants

across a broad geographical population range. This data

will be decisive in assessing which VOCs are actually

attractive for pollinators, and what are the relative effects

of gene flow/genetic drift and pollinator selective pressures

Table 5. Summary table for the npMANOVA on the odour variabil-

ity for A. italicum and A. maculatum inflorescences in the studied

sites in 2009 and 2010. d.f. = degree of freedom, Total d.f. = 80

Factor d.f. F r2 P-value

Species 1 37�374 0�374 <1�10�4

Population 5 2�726 0�076 <1�10�4

Year 1 12�7 0�07 <1�10�4

Species * Population 1 3�398 0�019 0�009Population * Year 1 5�178 0�029 0�001Species * Year 1 8�085 0�045 <1�10�4



on odour variability among populations and individuals

(Volis 2011).

Acknowledgements

The authors thank Gael Grenouillet for his help doing the statistical analy-

ses, Professor Alain Thomas for Diptera identification, Josselin Cornuault

and Suzanne Liagre for their help in the field. This work received funding

from the CNRS-GDREC (French National Research Group in Chemical

Ecology). We also thank Stefan D€otterl, Artur Maia and an anonymous

reviewer for their constructive comments on the manuscript.

References

Ackerman, J.D., Cuevas, A.A. & Hof, D. (2011) Are deception-pollinated

species more variable than those offering a reward? Plant Systematics

and Evolution, 293, 91–99.Ackerman, J.D., Mel�endez-Ackerman, E.J. & Salguero-Faria, J. (1997)

Variation in pollinator abundance and selection on fragrance phenotypes

in an epiphytic orchid. American Journal of Botany, 84, 1383–1390.Adams, R.P. (2007) Identification of Essential Oil Components by Gas Chro-

matography/Mass Spectroscopy, 4th edn. Allured Publishing, Carol

Stream.

Albre, J. & Gibernau, M. (2008) Reproductive biology of Arum italicum

(Araceae) in the South of France. Botanical Journal of the Linnean

Society, 156, 43–49.Albre, J., Quilichini, A. & Gibernau, M. (2003) Pollination ecology of

Arum italicum (Araceae). Botanical Journal of the Linnean Society, 141,

205–214.Anderson, M.J. (2001) A new method for non-parametric multivariate

analysis of variance. Austral Ecology, 26, 32–46.Anderson, B. & Johnson, S.F. (2007) The geographic mosaic of coevolution

in a plant-pollinator mutualism. Evolution, 62, 220–225.Angert, A.L., Bradshaw, H.D. & Schemske, D.W. (2008) Using experimen-

tal evolution to investigate geographic range limits in monkey flowers.

Evolution, 62, 2660–2675.Ayasse, M., Schiestl, F.P., Paulus, H.F., L€ofstedt, C., Hannson, B., Ibarra, F.

& Francke, W. (2000) Evolution of reproductive strategies in the sexually

deceptive orchid Ophrys sphegodes: how does flower-specific variation of

odor signals influence reproductive success? Evolution, 54, 1995–2006.Blionis, G.J. & Vokou, D. (2002) Structural and functional divergence of

Campanula spatulata subspecies on Mt Olympos (Greece). Plant System-

atics and Evolution, 232, 89–105.Boyce, P.C. (2006) Arum – a decade of change. Aroideana, 29, 132–137.Bradshaw, H.D. & Schemske, D.W. (2003) Allele substitution at a flower

colour locus produces a pollinator shift in monkeyflowers. Nature, 426,

176–178.Brown, M., Downs, C.T. & Johnson, S.D. (2011) Covariation of flower traits

and bird pollinator assemblages among pollinations of Kniphofia linearifo-

lia (Asphodelaceae). Plant Systematics and Evolution, 294, 199–206.Campbell, D.R. (2003) Natural selection in Ipomopsis hybrid zones: impli-

cations for ecological speciation. New Phytologist, 161, 83–90.Chartier, M., Pelozuelo, L. & Gibernau, M. (2011) Do floral odor profiles

geographically vary with the degree of specificity for pollinators? Investi-

gation in two sapromyophilous Arum species (Araceae). Annales de la

Soci�et�e Entomologique de France, 47, 71–77.Cosacov, A., Nattero, J. & Cocucci, A.A. (2008) Variation of pollinator

assemblages and pollen limitation in a locally specialized system: the oil-

producing Nierembergia liariifolia (Solanaceae). Annals of Botany, 102,

723–734.Diaz, A. & Kite, G.C. (2002) A comparison of the pollination ecology of

Arum maculatum and A. italicum in England. Watsonia, 24, 171–181.Dormont, L., Jay-Robert, P., Bessi�ere, J.-M., Rapior, S. & Lumaret, J.-P.

(2010) Innate olfactory preferences in dung beetles. The Journal of

Experimental Biology, 213, 3177–3185.D€otterl, S., Wolfe, L.M. & J€urgens, A. (2005) Qualitative and quantitative

analyses of flower scent in Silene latifolia. Phytochemistry, 66, 203–213.Elle, E. & Carney, R. (2003) Reproductive assurance varies with flower size

in Collinsia parviflora (Scrophulariaceae). American Journal of Botany,

90, 888–896.Ellis, A.G. & Johnson, S.D. (2009) The evolution of floral variation with-

out pollinator shifts in Gorteria diffusa (Asteraceae). American Journal of

Botany, 96, 793–801.

Esp�ındola, A., Pellissier, L. & Alvarez, N. (2011) Variation in the propor-

tion of flore visitors of Arum maculatum along its distributional range in

relation with community-based climatic niche analyses. Oikos, 120,

728–734.Faegri, K. & Van Der Pijl, L. (1971) The Principles of Pollination Ecology.

Pergamon Press, Oxford.

Fenster, C.B., Armbruster, W.S., Wilson, P., Dudash, M.R. & Thomson,

J.D. (2004) Pollination syndromes and floral specialization. Annual

Reviews of Ecology, Evolution and Systematics, 35, 375–403.Geber, M.A. & Eckhart, V.M. (2005) Experimental studies of adaptation

in Clarkia xantiana. II. Fitness variation across a subspecies border.

Evolution, 59, 521–531.Geber, M.A. & Moeller, D.A. (2006) Pollinator responses to plant commu-

nities and implications for reproductive character evolution. Ecology and

Evolution of Flowers (eds L.D. Harder & S.C.H. Barrett), pp. 102–119.Oxford University Press, Oxford.

Gibernau, M. (1997) Odeurs et sp�ecificit�e dans les mutualismes figuier polli-

nisateur: Le cas de Ficus carica L. et de Blastophaga psenes L. PhD

thesis, University of Montpellier, Montpellier.

Gibernau, M., Macquart, D. & Przetak, G. (2004) Pollination in the genus

Arum – a review. Aroideana, 27, 148–166.Gomez, J.M., Bosh, J., Perfectti, F., Fern�andez, J.D., Abdelaziz, M. &

Camacho, J.P.M. (2008) Spatial variation in selection on corolla shape

in a generalist plant is promoted by the preference patterns of its local

pollinators. Proceedings of the Royal Society B, 275, 2241–2249.Gomez, J.M., Abdelaziz, M., Camacho, J.P.M., Munoz-Pajares, A.J. &

Perfectti, F. (2009) Local adaptation and maladaptation to pollinators in

a generalist geographic mosaic. Ecology Letters, 12, 672–682.Gould, S.J. & Johnston, R.F. (1972) Geographic variation. Annual Review

of Ecology, Evolution, and Systematics, 3, 457–498.Hentrich, H., Kaiser, R. & Gottsberger, G. (2010) Floral biology and

reproductive isolation by floral scent in three sympatric Aroid species in

French Guyana. Plant Biology, 12, 587–596.Herrera, C.M., Castellanos, M.C. & Medrano, M. (2006) Geographical

context of floral evolution: towards an improved research programme in

floral diversification. Ecology and Evolution of Flowers (eds L.D. Harder

& S.C.H. Barrett), pp. 278–294. Oxford University Press, Oxford.

Hodgins, K.A. & Barrett, C.H. (2008) Geographic variation in floral mor-

phology and style-morph ratios in a sexually polymorphic daffodil.

American Journal of Botany, 95, 185–195.Ibanez, S., D€otterl, S., Anstett, M.-C., Baudino, S., Caissard, J.C., Gallet,

C. & Despr�es, L. (2010) The role of volatile organic compounds, mor-

phology and pigments of globeflowers in the attraction of their specific

pollinating flies. New Phytologist, 188, 451–463.Je�zek, J. (1990) Redescriptions of nine common palaeartic and holartic

species of Psychodini End. (Diptera: Psychodidae). Acta Entomologica

Musei Nationalis Pragae, 43, 33–83.Je�zek, J. & H�ajek, J. (2007) Psychodidae (Diptera) of the Orlick�e hory

protected landscape area and neighbouring areas with descriptions of

two new species from the Czech Republic. Acta Entomologica Musei

Nationalis Pragae, 47, 237–285.Johnson, S.D. (2006) Pollinator-driven speciation in plants. Ecology and


Johnson, S.D. & J€urgens, A. (2010) Convergent evolution of carrion and

faecal scent mimicry in fly-pollinated angiosperm flowers and a stink-

horn fungus. South African Journal of Botany, 76, 796–807.Johnson, S.D. & Steiner, K.E. (1997) Long-tongued fly pollination and evo-

lution of floral spur length in the Disa draconis complex (Orchidaceae).

Evolution, 51, 45–53.J€urgens, A., D€otterl, S. & Meve, U. (2006) The chemical nature of fetid flo-

ral odours in stapeliads (Apocynaceae–Asclepiadoideae–Ceropegieae).New Phytologist, 172, 452–468.

Kato, M., Itioka, T., Sakai, S., Momose, K., Yamane, S., Hamid, A.A. &

Inoue, T. (2000) Various population fluctuation patterns of light-

attracted beetles in a tropical lowland dipterocarp forest in Sarawak.

Population Ecology, 42, 97–104.Kiester, A.R., Lande, R. & Schemske, D.W. (1984) Models of coevolution

and speciation in plants and their pollinators. The American Naturalist,

1242, 220–243.Kite, G.C. (1995) The floral odour of Arum maculatum. Biochemical Sys-

tematics and Ecology, 23, 343–354.Kite, G.C., Hetterscheid, W.L.A., Lewis, M.J., Boyce, P.C., Ollerton, J.,

Cocklin, E., Diaz, A. & Simmonds, M.S.J. (1998) Inflorescence odours

and pollinators of Arum and Amorphophallus (Araceae). Reproductive



Biology (eds S.J. Owens & P.J. Rudall), pp. 295–315. Royal Botanical

Gardens, Kew.

Knudsen, J.T. (2002) Variation in floral scent composition within and

between populations of Geonoma macrostachys (Arecaceae) in the wes-

tern Amazon. American Journal of Botany, 89, 1772–1778.Knudsen, J.T., Eriksson, R., Gershenzon, J. & St�ahl, B. (2006) Diversity

and distribution of floral scent. The Botanical Review, 72, 1–120.Lack, A.J. & Diaz, A. (1991) The pollination of Arum maculatum L. – a

historical review and new observations. Watsonia, 18, 333–342.Levin, D.A. (2000) The Origin, Expansion and Demise of Plant Species.

Oxford University Press, Oxford.

Levin, D.A. & Anderson, W.W. (1970) Competition for pollinators between

simultaneously flowering species. The American Naturalist, 104, 455–467.Linz, J., St€okl, J., Urru, I., Kr€ugel, T., Stensmyr, M.C. & Hansson, B.S.

(2010) Molecular phylogeny of the genus Arum (Araceae) inferred from

multi-locus sequence data and AFLPs. Taxon, 59, 405–415.Maia, A.C.D., D€otterl, S., Kaiser, R., Silberbauer-Gottsberger, I., Teichert,

H., Gibernau, M., Navarro, D.M.D.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 An-

nonaceae and Araceae. Journal of Chemical Ecology, 38, 1072–1080.Maia, A.C.D., Gibernau, M., D€otterl, S., Navarro, D.M.do.A.F., Seifert,

K., M€uller, T. & Schlindwein, C. (2013) The floral scent of Taccarum

ulei (araceae): attraction of scarab beetle pollinators to an unusual ali-

phatic acyloin. Phytochemistry, http://dx.doi.org/10.1016/j.phytochem.

2013.03.005. (in press).

Mant, J., Peakall, R. & Schiestl, F.P. (2005) Does selection on floral odor

promote differentiation among populations and species of the sexually

deceptive orchid genus Ophrys? Evolution, 59, 1449–1463.M�endez, M. & Obeso, J.R. (1992) Influencia del osmoforo en la produccion

de infrutescencias en Arum italicum Miller (Araceae). Anales del Jard�ın

Bot�anico de Madrid, 50, 229–237.Mitchell, R.J., Flanagan, R.J., Brown, B.J., Waser, N.M. & Karron, J.D.

(2009) New frontiers in competition for pollination. Annals of Botany,

103, 1403–1413.Morgan, M.T. (2006) Selection on reproductive characters: conceptual

foundations and their extension to pollinator interactions. Ecology and


Nattero, J., Cocussi, A.A. & Medel, R. (2010) Pollinator-mediated selection

in a specialized pollination system: matches and mismatches across

populations. Journal of Evolutionary Biology, 23, 1957–1968.Pettersson, S. & Knudsen, J.T. (2001) Floral scent and nectar production in

Parkia biglobosa Jacq. (Leguminosae: Mimosoideae). Botanical Journal

of the Linnean Society, 135, 97–106.Pichersky, E. & Gershenzon, J. (2002) The formation and function of plant

volatiles: perfumes for pollinator attraction and defense. Plant Biology,

5, 237–243.Prime, C.T. (1960) Lords and Ladies. Harper Collins Publishers, London.

Proctor, M., Leo, P. & Lack, A. (1996) Natural History of Pollination.

Harper Collins Publishers, London.

Raguso, R.A. (2008) Wake up and smell the roses: the ecology and evolu-

tion of floral scent. Annual Review of Ecology, Evolution, and Systemat-

ics, 39, 549–569.R Core Team (2012) R: A language and environment for statistical

computing. R Foundation for Statistical Computing, Vienna, Austria

ISBN 3-900051-07-0, URL http://www.R-project.org/University of

Chicago Press, Chicago, IL.

Renner, S.S. (2006) Rewardless flowers in the Angiosperms and the role of

insect cognition in their evolution. Plant–Pollinator Interactions: From

Specialization to Generalization (eds N.M. Waser & J. Ollerton),

pp. 123–144. University of Chicago Press, Chicago, IL.

Rohacek, J., Beck-Haug, I. & Dobat, K. (1990) Sphaeroceridae associated

with flowering Arum maculatum (Araceae) in the vicinity of T€ubingen,

SW-Germany. Senckenbergiana biologie, 71, 259–268.Schaeffer, H.M. & Ruxton, G.D. (2011) Plant-Animal Communication.

Oxford University Press, Oxford.

Schiestl, F.P. & Ayasse, M. (2002) Do changes in floral odor cause speciation in

sexually deceptive orchids? Plant Systematics and Evolution, 234, 111–119.Schiestl, F.P. & Schl€uter, P.M. (2009) Floral isolation, specialized pollina-

tion, and pollinator behavior in Orchids. Annual Review of Entomology,

54, 425–446.

Schlumpberger, B.O. & Raguso, R.A. (2008) Geographic variation in floral

scent of Echinopsis ancistrophora (Cactaceae); evidence for constraints

on hawkmoth attraction. Oikos, 117, 801–814.Schlumpberger, B.O., Cocussi, A.A., Mor�e, M., S�ersic, A.N. & Raguso,

R.A. (2009) Extreme variation in floral characters and its consequences

for pollinator attraction among populations of an Andean cactus. Annals

of Botany, 103, 1489–1500.Solers, C., Hossaert-McKey, M., Buatois, B., Bessi�ere, J.-M., Schatz, B. &

Proffit, M. (2011) Geographic variation of floral scent in a highly special-

ized pollination mutualism. Phytochemistry, 72, 74–81.Stebbins, G.L. (1970) Adaptive radiation of reproductive characteristics in

Angiosperms, I: pollination mechanisms. Annual Review of Ecology and

Systematics, 1, 307–326.St€okl, J., Strutz, A., Dafni, A., Svatos, A., Doubsky, J., Knaden, M.,

Sachse, S., Hansson, B.S. & Stensmyr, M.C. (2010) A deceptive pollina-

tion system targeting drosophilids through olfactory mimicry of yeast.

Current Biology, 20, 1846–1852.St€okl, J., Schl€uter, P.M., Stuessy, T.F., Paulus, H.F., Assum, G. & Ayasse,

M. (2008) Scent variation and hybridization cause the displacement of a

sexually deceptive orchid species. American Journal of Botany, 95,

472–481.Svensson, G.P., Pellmyr, O. & Raguso, R.A. (2006) Strong conservation of

floral scent composition in two allopatric yuccas. Journal of Chemical

Ecology, 32, 2657–2665.Svensson, G.P., Pellmyr, O. & Raguso, R.A. (2011) Pollinator attraction to

volatiles from virgin and pollinated host flowers in a yucca/moth obli-

gate mutualism. Oikos, 120, 1577–1583.Svensson, G.P., Hickman, M.O., Bartram, S., Boland, W., Pellmyr, O. &

Raguso, R.A. (2005) Chemistry and geographic variation of floral scent

in Yucca filamentosa (Agavaceae). American Journal of Botany, 92,

1624–1631.Thompson, J.N. (2005) The Geographic Mosaic of Coevolution. University

Chicago Press, Chicago, IL.

Urru, I., Stensmyr, M.C. & Hansson, B.S. (2011) Pollination by brood-site

deception. Phytochemistry, 72, 1655–1666.Vaillant, F. (1988) Dipt�eres psychodidae coprophiles et coprobiontes d’Eur-

ope. Bulletin Romand d’Entomologie, 6, 1–43.Valiente-Banuet, A., Molina-Freaner, F., Torres, A., del Coro Arizmendi,

M. & Casas, A. (2004) Geographic differentiation in the pollination sys-

tem of the columnar cactus Pachycereus pecten-aboriginum. American

Journal of Botany, 91, 850–855.Volis, S. (2011) Adaptive genetic differentiation in a predominantly self-pol-

linating species analyzed by transplanting into natural environment,

crossbreeding and Qst-Fst test. New Phytologist, 192, 237–248.Waelti, M.O., Muhlemann, J.K., Widmer, A. & Schiestl, F.P. (2008) Floral

odour and reproductive isolation in two species of Silene. Journal of

Evolutionary Biology, 21, 111–121.Waser, N.M. (1978) Interspecific pollen transfer and competition between

co-occurring plant species. Oecologia, 36, 223–236.Waterman, R.J., Bidartondo, M.I., Stofberg, J., Combs, J.K., Gebauer, G.,

Savolainen, V., Barraclough, T.G. & Pauw, A. (2011) The effects of

above- and belowground mutualisms on orchid speciation and coexis-

tence. The American Naturalist, 177, E54–E68.Withers, P. (1989) Roth flies. Diptera: Psychodidae.Dipterist Digest, 4, 1–83.

Received 21 June 2012; accepted 22 April 2013

Handling Editor: Manfred Ayasse

Supporting Information

Additional Supporting information may be found in the online

version of this article:

Table S1. Percentages of insects (mean � SE) caught in the

A. italicum and A. maculatum inflorescences from all studied sites

in 2009 and 2010, and from 2008 (data from Chartier, Pelozuelo

& Gibernau 2011).



Geographical variations of odour and pollinators, and test for local adaptation by reciprocal transplant of two European Arum species

Documents