Botanical Journal of the Linnean Society, 2016.

With 4 figures

The palm family (Arecaceae): a microcosm of sexual

system evolution
SOPHIE NADOT1*, ELODIE ALAPETITE1, WILLIAM J. BAKER3, JAMES W. TREGEAR4
and ANDERS S. BARFOD1,2
1
Laboratoire Ecologie, Systematique et Evolution UMR 8079, Univ. Paris-Sud, CNRS, AgroParisTech,
Universit e Paris-Saclay, Orsay Cedex 91405, France
2
Department of Bioscience, Aarhus University, Ny Munkegade 114, Aarhus C DK-8000, Denmark
3
Royal Botanic Gardens, Kew, Richmond, Surrey TW9 3AB, UK
4
Institut de Recherche pour le Developpement (IRD), UMR DIADE, 911 Avenue Agropolis BP 64501,
Montpellier Cedex 5, 34394, France

Received 5 January 2016; revised 2 March 2016; accepted for publication 19 April 2016

Among the 416 angiosperm families, palms (Arecaceae) are striking in possessing almost all possible
combinations of hermaphroditic and/or unisexual flowers, making them a particularly interesting subject for
studies of the evolution of plant sexuality. The purpose of this review is to highlight the amazing diversity of
sexual expression in palms with a view to proposing scenarios to explain the evolution of this character, drawing
on the numerous advances that have been made over the last 20 years in palm systematics, ecology,
developmental biology, phylogenetics and genomics. We provide an overview of the variability of sexual
expression in palms, with illustrations of the associated morphological diversity and its significance to
reproduction. We discuss the evolution of sexual systems using the most recent phylogenetic framework available
for palms. Finally, we review advances made towards unravelling the genetic basis of sexual expression in palms.
Our survey opens new perspectives for understanding how plant sexuality has evolved in angiosperms as a
whole. © 2016 The Linnean Society of London, Botanical Journal of the Linnean Society, 2016

ADDITIONAL KEYWORDS: dioecy – floral traits – Palmae.

INTRODUCTION ‘polygamy’ that defines various intermediate states

between the first three groups. Hermaphroditism, in
The diversity of reproductive structures and strate-
which all flowers bear functional male and female
gies in flowering plants has long fascinated botanists
organs, is the most common situation in angios-
and evolutionary biologists. Reproductive systems in
perms. Dioecy, where male and female flowers are
plants may relate to pollination, sexual expression in
borne on different plants, is found in 6% of angios-
space and time, and mating (or breeding) systems,
perm species distributed across half of all families
i.e. outcrossing vs. selfing (Barrett, 2010, 2013). The
(Renner & Ricklefs, 1995). Monoecy, with male and
term ‘sexual systems’ (or sometimes sex systems) is
female flowers borne on the same plant, has received
employed to describe the spatial arrangement of
less attention than hermaphroditism and dioecy. In a
sexes (both sexes in the same flower, in different
survey of various temperate and tropical floras,
flowers borne on the same plant or each sex on a dif-
monoecy was estimated to occur in 3–19% of angios-
ferent plant) (e.g. (Barrett, 1998; Charlesworth,
perm species (de Jong, Shmida & Thuijsman, 2008).
2006). Sexual expression in flowering plants is classi-
Darwin (1877) considered monoecy to be an adapta-
cally broken down into three main categories (her-
tion to favour outcrossing due to greater spatial sep-
maphroditism, dioecy and monoecy), with an
aration between pollen and stigmas, whereas dioecy
additional minor category, often termed as
guarantees that selfing will never occur. Renner &
Ricklefs (1995) noted that monoecy, like dioecy, is
*Corresponding author. E-mail: [email protected] often associated with abiotic pollination. In

© 2016 The Linnean Society of London, Botanical Journal of the Linnean Society, 2016 1
2 S. NADOT ET AL.

monoecious plants, outcrossing may be further 2011). Palms are conspicuous, typically large-bodied
favoured by temporal separation of the sexual organisms that often have a major ecological impact
phases, a state referred to as dichogamy. in the plant communities in which they occur
Hermaphroditism is generally considered as the (Couvreur & Baker, 2013). Although somewhat con-
ancestral state in angiosperms, although neither the strained by their relatively simple modular body
fossil record nor ancestral state reconstruction gives plan, they display an impressive variability in the
clear evidence for this (Endress & Doyle, 2009). architectural design of their leaves, stems and floral
Advantages of hermaphroditism include the sharing parts (Tomlinson, 2006). In the flowers, male and
by male and female functions of costs for the produc- female functions are often separated in various ways
tion of non-sexual organs such as perianth and through space and time (Henderson, 2002), presum-
reward-producing structures involved in pollinator ably to favour cross-pollination in association with
attraction, and the possibility of self-fertilization the pollinating vectors (Henderson, 1986, 2002; Bar-
when pollen transfer is inefficient (Obbard, Harris & fod, Hagen & Borchsenius, 2011). This variation in
Pannell, 2006). Conversely, unisexual flowers allow sexual systems, combined with the fact that there is
dimorphism and consequently specialization in the a robust comprehensive phylogenetic framework
number, shape, size and positioning of male and (Baker et al., 2009) available for the family, makes
female flowers, with implications for pollination and palms an appropriate group in which to explore
fruiting in particular. Dioecy and monoecy are con- hypotheses relating to the evolution of sexual
sidered to be derived from hermaphroditism through strategies.
pathways that may involve intermediate stages. The As a prelude to future research on sexual systems
most likely pathway to explain the transition from in palms and other angiosperms, we provide here a
hermaphroditism to monoecy is through andromo- review of sexual expression across Arecaceae. We
noecy (male and hermaphrodite flowers produced on synthesize current knowledge of the variability of
the same plant) rather than gynomonoecy (female sexual expression in palms, in terms of morphology,
and hermaphrodite flowers produced on the same space and time, and its developmental and genetic
plant), as suggested by a sex allocation model (de bases. Building on this information, we discuss
Jong et al., 2008) and corroborated by patterns of putative pathways for the evolution of sexual
sexual systems changes in monocotyledons (Weiblen, expression in the family in its phylogenetic frame-
Oyama & Donoghue, 2000). Various hypotheses have work.
been proposed to account for the origin of dioecy in
angiosperms, focusing on different drivers such as
pollination mechanisms (Ashman, 2000), ecological
OVERVIEW OF SEXUAL EXPRESSION IN
factors (Vamosi, Otto & Barrett, 2003) and evolution-
PALMS
ary pathways [reviewed by Barrett (2013) and Ren-
ner (2014)]. Two main evolutionary pathways have SEXUAL EXPRESSION IN SPACE
been suggested (Barrett, 2013). One involves In Fig. 1, we provide an overview of the proportions
monoecy and hypothesizes a gradual transition in of the different types of sexual systems in palms.
the relative proportions of male and female flowers, The genera considered in this study follow Palmweb
eventually leading to sexual separation between (www.palmweb.org), a regularly updated online
male and female plants. Evidence for this pathway source of information on palms based largely on the
has been found in a few taxa. The other hypothetical universally acknowledged reference work Genera
pathway leading to dioecy in angiosperms proceeds Palmarum (Dransfield et al., 2008). The latter is
through gynodioecy, which involves male sterility therefore the backbone source of information
mutations, invasion of hermaphrodite populations by throughout this review. The number of genera (188)
female individuals and further selection of the male. has been reduced to 181 recently (Baker & Drans-
A phylogenetic analysis of dioecy in monocots sug- field, 2016), but this difference does not significantly
gested that dioecy probably evolved more often affect the proportions given hereafter. The number of
directly from hermaphroditism than from monoecy, genera and species follows Palmweb (www.
meaning that monoecy would have evolved indepen- palmweb.org). Hermaphroditism is found in 17% of
dently (Weiblen et al., 2000). species (16% of genera), monoecy in 52% of species
Among the angiosperms, palms stand out as being (61% of genera) and dioecy in 30% of species (21% of
particularly diverse in terms of sexual expression genera). More rarely, combinations of bisexual and
patterns in both space and time. The family com- unisexual flowers are found on the same individual
prises c. 2600 species (Baker & Dransfield, 2016; this or on separate plants, a phenomenon often referred
issue), of which > 90% are found in tropical rain- to as polygamy (Dransfield et al., 2008; Kholia, 2009;
forests (Henderson, 2002; Couvreur, Forest & Baker, Beentje, 2010). The spatial separation of male and

© 2016 The Linnean Society of London, Botanical Journal of the Linnean Society, 2016
 SEXUAL SYSTEMS IN PALMS 3

Figure 1. Genus- and species-level survey of sexual system in palms, at the family level (bottom line) and in each of
the five subfamilies. Left-hand side: relative proportions of each of the four different sexual systems at the genus level,
in percentages of the total number of genera. Right-hand side: relative proportions of each of the four different sexual
systems at the species level, in percentages of the total number of species. Detailed numbers and proportions are given
in Appendix 1.

female functions takes place at various structural Jeanson, 2011) (Fig. 2G). In Elaeis guineensis Jacq.,
levels: within flower clusters (between flowers); these functionally unisexual inflorescences are pro-
within inflorescences (between regions of the inflo- duced in male and female cycles (Adam et al., 2005).
rescence); within individuals (between inflorescences) In Calamoideae, monoecy occurs only in Raphia
in some monoecious species; and within populations P.Beauv. and Oncocalamus (G.Mann & H.Wendl.)
(between individuals) in dioecious species (Fig. 2). H.Wendl., with both male and female flowers borne
on each rachilla. In Nypa fruticans Wurmb, the only
Monoecy species in subfamily Nypoideae, the female flowers
The vast majority of monoecious palms occur in Are- are borne on the club-shaped apex of the inflores-
coideae, the largest subfamily which contains more cence and the male flowers are densely packed on
than half of all palm species. In this almost entirely rachillae borne on lateral branching systems
monoecious clade, most species produce flowers gath- (Fig. 2F).
ered in triads or clusters derived from triads. The
triad is a highly condensed sympodial flower cluster Dioecy
consisting of a female flower flanked by two male Complete separation of sexes on different individuals
flowers (Fig. 2D, E) (Uhl, 1966, 1976). Male and is the predominant sexual system in Calamoideae
female functions may be further separated by the and Ceroxyloideae. Remarkable arrangements occur
production of fully developed triads at the base of in the ceroxyloid tribe Phytelepheae, in which four to
the rachillae and only dyads of male flowers or single nine male flowers are gathered in monopodial clus-
male flowers (interpreted as degenerate triads in ters densely inserted on spike-like inflorescences,
which the female flower is lacking) present in the and the solitary female flowers are spirally arranged
distal part. In large-fruited species, as found in tribe on the club-shaped apex of the inflorescences (Bar-
Cocoseae, this concentration of female flowers in the fod, 1991) (Fig. 2H). In the otherwise monoecious
proximal end of the rachillae may have been selected Arecoideae, tribe Chamaedoreeae is exceptional in
as a cost-effective way to lend biomechanical containing the only two truly dioecious genera pre-
strength to the infructescence. Outside Arecoideae, sent in the subfamily, Wendlandiella Dammer and
triads occur in tribe Caryoteae (subfamily Coryphoi- Chamaedorea Willd.
deae). In some monoecious genera, inflorescences
may be unisexual by suppression of the expression of Hermaphroditism
one sex in the triads, even though inflorescences of Truly hermaphroditic palms are found mostly in Cor-
both sexes occur on the same plant [Lepidorrhachis yphoideae (e.g. Fig. 2A, B) and Calamoideae. Pseu-
(H.Wendl. & Drude) O.F.Cook, Arenga Labill. ex dophoenix H.Wendl. ex Sarg., the only non-dioecious
DC., Caryota L., Wettinia Poepp. ex Endl., Marojejya genus in subfamily Ceroxyloideae, has hermaphrodi-
Humbert and Elaeis Jacq. (Baker & Hutton, 2006; tic flowers on the proximal part of each rachilla. Of
Dransfield et al., 2008)]. For example, in Arenga and the three hermaphroditic genera of Calamoideae,
Caryota the male flowers abort before anthesis in Eremospatha (G.Mann & H.Wendl.) Schaedtler and
some inflorescences that become functionally female, Laccosperma (G.Mann & H.Wendl.) Druded (tribe
whereas the female flower of the triad never devel- Lepidocaryeae) bear flowers in dyads, whereas
ops in other inflorescences that become functionally Korthalsia Blume (tribe Calameae) has solitary
male (Dransfield & Mogea, 1984; Basu & Basu, 1993; flowers.

© 2016 The Linnean Society of London, Botanical Journal of the Linnean Society, 2016
4 S. NADOT ET AL.

Figure 2. Diversity of sexual expression in palms. A, solitary hermaphroditic protandrous flowers of Sabal palmetto. B,
solitary hermaphroditic protogynous flowers of Serenoa repens. C, solitary hermaphroditic (left) and male (right) flowers
of Eugeissona utilis (andromonoecious protandrous species). D, protandrous triad of two male flowers and one female
flower (centre) of Dictyocaryum sp. at male anthesis (above), and female flowers of Dictyocaryum fuscum at anthesis (be-
low); the male flowers of the triad have fallen off. E, protogynous triad of two male flowers and one female flower (cen-
tre) of Reinhardtia simplex at female anthesis. F, heads of male (left head) and female (right head) flowers surrounded
by bracts of Nypa fruticans. G, portions of rachillae bearing male (left) and female (right) flowers of Attalea amygdalina;
in this species rachillae usually bear flowers of one sex, only occasionally triads at the proximal end. H, female (left)
and male (right) inflorescences of the dioecious species Phytelephas aequatorialis. Male and female flowers in each
species are on relative scale. Artwork: Marion Ruff Sheehan.

Polygamy one on Trachycarpus takil Becc. (tribe Trachy-

A few unrelated palm genera, in Calamoideae and carpeae), which revealed an unstable sexual system
Coryphoideae (mainly in tribe Trachycarpeae), have with age-related changes from mainly male expres-
complex combinations of bisexual and unisexual flow- sion to mainly female expression (Kholia, 2009).
ers, which are referred to collectively as ‘polygamous’ Environmentally dependent plasticity in the expres-
(Dransfield et al., 2008). The two polygamous genera sion of male vs. female functions has been reported
of Calamoideae, Eugeissona Griff. (tribe Eugeis- in two polygamous species of Attalea Kunth (tribe
soneae) and Metroxylon Rottb. (tribe Calameae), pro- Cocoseae) (Voeks, 1988; Barot et al., 2005). In A.
duce dyads composed of one male flower and one funifera Mart., there is a shift from male to female
hermaphroditic flower, which corresponds to expression as the plants grow taller and are more
andromonoecy (Fig. 2C). In the otherwise dioecious exposed to direct sunlight (Voeks, 1988). Changes in
genus Rhapis L.f. ex Aiton, a similar sexual system sexual expression are also commonly observed in the
was found in R. laosensis Becc. [=R. cochinchinensis cyclical production of female and male inflorescences
(Lour.) Mart.] (Giddey, Spichiger & Stauffer, 2009). in African oil palm, Elaeis guineensis (Adam et al.,
Andromonoecy is not uncommon in flowering plants, 2005). The length of these cycles is influenced by soil
occurring in c. 4000 species (Miller & Diggle, 2002). and light conditions, and consequently the female/
male inflorescence ratio is a variable feature that
reflects local abiotic conditions and affects yields in
PLASTICITY IN SEXUAL EXPRESSION the plantation (Williams & Thomas, 1970; Freeman
Little is known about sexual plasticity in angios- et al., 1981). Male and female phases have also been
perms in general and palms are no exception, few recorded in Marojejya (Marcus, 2010), a member of
detailed studies having been conducted apart from tribe Areceae, like Elaeis.

© 2016 The Linnean Society of London, Botanical Journal of the Linnean Society, 2016
 SEXUAL SYSTEMS IN PALMS 5

SEXUAL EXPRESSION IN TIME carpels to a single carpel (Uhl & Moore, 1971;
Temporal separation of sexual expression operates at Rudall, Ryder & Baker, 2011). In unisexual flowers
various spatial levels: within the flower in the case non-functional reproductive organs (pistillode in
of hermaphroditic species (dichogamy sensu stricto) male flowers and staminodes in female flowers) vary
(Fig. 2A–C); between flowers within flower clusters in their degree of differentiation and may sometimes
and between inflorescences within plants in the case be absent, this attribute having evolved repeatedly
of monoecious taxa (Fig. 2D–G); between plants; and in the family. In addition to this variation, floral sex-
even between populations in the case of dioecious ual dimorphism (referring here to the difference in
taxa (Fig. 2H). In many palm inflorescences, sexual appearance between male and female flowers) needs
expression occurs in pulses or cycles, probably as an to be taken into account to fully appreciate the huge
adaptation to the characteristics of the pollinating diversity of sexual expression (Fig. 3). Morphological
vector (Borchsenius, 1997; Henderson, 2002; Barfod, differences between male and female flowers in
Burholt & Borchsenius, 2003). In palms bearing only palms concern size, shape and structural attributes.
hermaphroditic flowers, protandry (male organs At one extreme, male and female flowers may be
mature before female organs) and protogyny (the almost morphologically identical, as in some Aus-
opposite) can be found, protandry being more wide- tralian species of Livistona R.Br. (Dowe, 2009), Onco-
spread (Dransfield et al., 2008). The polarity of calamus and Asterogyne H.Wendl. ex Hook.f.
flower opening along inflorescence axes varies (Dransfield et al., 2008) (Fig. 3A, F). At the other
between species, as shown in Licuala Wurmb (Barfod extreme, they may be strongly dimorphic (Fig. 3D,
et al., 2003). In monoecious taxa, there is often a G). In palms female flowers tend to be slightly larger
time difference in anthesis between male and female than male flowers (Fig. 3B, C), but several groups
flowers within flower clusters. This applies, for display male flowers that are larger than the female
example, to Arecoideae and Caryoteae, in which the ones (Fig. 3D). The most noticeable size contrast is
majority of species have protandrous triads (Drans- observed in tribe Phytelepheae (Ceroxyloideae). In
field et al., 2008) (Fig. 2D). This may be due to the Aphandra natalia, the enormous and narrow female
architectural constraints of the triad, in which the flowers are up to 25 cm long, whereas male flowers
female flower is last formed in the ontogenetic series, are ten times shorter, despite the huge number of
and therefore most likely to open last (Loo et al., stamens (c. 1000) densely packed on the enlarged flo-
2006). In many cases, the male and female antheses ral receptacle (Barfod, 1991).
are separated by only a matter of days, whereas in Selective pressures due to pollinator interactions
others the development of the female flower is sub- may drive patterns of flower size in palms, as in
stantially delayed. In the wind-pollinated species of other angiosperm taxa, in which differences in male
Howea Becc., an entire year elapses between male and female flower sizes have been linked to attrac-
and female antheses, representing an extreme situa- tiveness in animal-pollinated species, in relation to
tion (Savolainen et al., 2006; Babik et al., 2009). sexual selection theory (Willson, 1991; Kriebel,
2014). In four Calamus species from peninsular
Thailand with presumed pollination by deceit
(Boegh, 1996), a size difference between male and
MORPHOLOGY OF SEXUAL EXPRESSION
female flowers may have been counter-selected due
Despite the huge variability of their sexual systems, to the behaviour of the pollinating bees that mainly
the floral groundplan of palms contains few innova- collect pollen to nourish their brood. As these bees
tions with respect to the typical monocot groundplan. are attracted to inflorescences by visual cues and are
Palm flowers generally have an actinomorphic able to detect and remember even minor morphologi-
trimerous perianth, differentiated into calyx and cor- cal differences between male and female flowers
olla, and a superior ovary with a single ovule per (Barfod et al., 2011), they would only visit male flow-
locule. However, the flowers vary extensively among ers if there was a difference in appearance. In some
species in terms of size [from a few millimetres long Chamaedorea species (Arecoideae), the pistillode of
in Synechanthus warscewiczianus H.Wendl. to sev- the male flower is larger than the gynoecium of the
eral centimetres in Aphandra natalia (Balslev & female flower and may play an important role in
A.J.Hend.) Barfod], organ numbers [from just one interactions with pollinating insects (Askgaard et al.,
stamen in Dypsis lantzeana Baill. (Dransfield & 2008).
Beentje, 1996; Rudall et al., 2003) to > 1000 in
Ammandra decasperma O.F.Cook (Barfod, 1991)]
and organ fusion. Compared with other monocotyle- A NOTE ON SELF-INCOMPATIBILITY

donous taxa, the gynoecium structure is unusually Palms are usually regarded as self-compatible, with
variable, ranging mostly from three fused or free no report to date of gametophytic or sporophytic

© 2016 The Linnean Society of London, Botanical Journal of the Linnean Society, 2016
6 S. NADOT ET AL.

Figure 3. Illustration of the diversity of floral sexual dimorphism in palms. A, triad of Oncocalamus macrospathus with
morphologically identical male and female (centre) flowers. B, male (left) and female (right) flowers of Hydriastele
chaunostachys; female flowers are shorter than the male flowers. C, male (left) and female (right) flowers of Hyphaene
coriacea; female flowers are shorter and rounder than the male flowers. D, male (left) and female (right) flowers of
Lodoicea maldivica; female flowers are twice as long as the male flowers and rounder. E, male (left) and female (right)
flowers of Borassus aethiopum; female flowers are four times longer than the male flowers and much rounder. F, male
(left) and female (right) flowers of Asterogyne spicata with staminodes (arrow in female flower) similar in size (although
in lower number) to stamens, and pistillode (arrow in male flower) similar in size to ovary. G, male (left) and female
(right) flowers of Pinanga coronata with strong sexual dimorphism (no pistillode in male flowers, no staminodes in
female flowers). Male and female flowers in each species are on relative scale. Artwork: Marion Ruff Sheehan.

self-incompatibility (Allen & Hiscock, 2008). How- conditions and end points of sexual system evolution
ever, it should be noted that few experiments have (Barrett, 2013). There is evidence, however, that
been conducted (Barfod et al., 2011). Experience from dioecy is not always an evolutionary dead-end and it
horticulture indicates that hermaphroditic or monoe- can give rise to hermaphroditism (Case et al., 2008).
cious palms can set fruit in cultivation, even when The predominance of monoecy and dioecy in palms
only single individuals are present in a particular (> 80% of the species) and the diversity of sexual sys-
garden. One notable exception is the Mauritian palm tems in the early-diverging subfamilies raise ques-
Hyophorbe amaricaulis Mart. (Maunder et al., 2002), tions about the ancestral sexual system of Arecaceae.
which is extinct in the wild and persists as a single Ancestral state reconstruction of this character con-
individual in the botanic garden at Curepipe (Mauri- ducted on an earlier version of the palm phylogenetic
tius), where it fails to set viable fruit. tree had resulted in an equivocal ancestral state (Wei-
blen et al., 2000). Since then, considerable advances
have been made in resolving phylogenetic relation-
ships in the family, providing the opportunity to re-
EVOLUTION OF SEXUAL EXPRESSION
examine the evolution of sexual systems in a more
Few hypotheses have been published that relate to the detailed and robust framework. Using the comprehen-
sexual system of the ancestral palm flower. It has been sive genus-level phylogenetic tree for Arecaceae
assumed that it is hermaphroditic (Moore & Uhl, (Baker et al., 2009), we reconstructed the evolution of
1982), perhaps because at the level of angiosperms as sexual systems in the family (Fig. 4A) and obtained a
a whole monoecy and dioecy are considered as derived hypothetical evolutionary scenario for this character

© 2016 The Linnean Society of London, Botanical Journal of the Linnean Society, 2016
 SEXUAL SYSTEMS IN PALMS 7

Chamaedoreeae
A
Chamaedorea seifrizii
Gaussia maya
Wendlandiella gracilis
ARECOIDEAE Synechanthus warscewiczianus
Hyophorbe verschaffeltii

Hermaphroditism Ceroxylon quindiuense

Monoecy Juania australis
Dioecy Oraniopsis appendiculata Ceroxyleae
Ravenea hildebrandtii
Polygamy
CEROXYLOIDEAE Aphandra natalia
Phytelephas aequatorialis
Phytelepheae
Ammandra decasperma
Pseudophoenix vinifera Cyclospatheae
Guihaia argyrata
Rhapis excelsa
Maxburretia rupicola
Rhapidophyllum hystrix
Trachycarpus geminisectus
Chamaerops humilis
Brahea aculeata
Johannesteijsmannia altifrons
Pholidocarpus macrocarpus
Saribus jeanneneyi Trachycarpeae
Licuala grandis
Livistona speciosa
Acoelorrhaphe wrightii
Serenoa repens
Pritchardia thurstonii
Copernicia prunifera
Washingtonia robusta
Colpothrinax wrightii
Phoenix dactylifera Phoeniceae
Hemithrinax ekmaniana
Leucothrinax morrisii
Coccothrinax argentata
Thrinax radiata
Zombia antillarum Cryosophileae
Schippia concolor
Trithrinax brasiliensis
Chelyocarpus ulei
Cryosophila warscewiczii
Itaya amicorum
Sabal palmetto Sabaleae
CORYPHOIDEAE Borassodendron machadonis
Borassus flabellifer
Lodoicea maldivica
Latania verschaffeltii
Bismarckia nobilis
Borasseae
Satranala decussilvae *
Hyphaene coriacea
Medemia argun
Corypha utan Corypheae
Arenga pinnata
Wallichia disticha * Caryoteae
Caryota mitis
Kerriodoxa elegans
Tahina spectabilis
Chuniophoeniceae
Nannorrhops ritchiana
Chuniophoenix hainanensis *
Nypa fruticans * NYPOIDEAE
Retispatha dumetosa
Pogonotium ursinum
Daemonorops grandis
Calamus aruensis
Ceratolobus concolor
Pigafetta elata
Myrialepis paradoxa
Calameae
Plectocomiopsis geminiflora
Plectocomia mulleri
Metroxylon salomonense
Salacca ramosiana
Eleiodoxa conferta
Korthalsia cheb
Mauritia flexuosa
Mauritiella armata
Lepidocaryum tenue
CALAMOIDEAE Raphia farinifera * Lepidocaryeae
Laccosperma acutiflorum
Eremospatha macrocarpa
Oncocalamus mannii *
Eugeissona tristis Eugeissoneae
Typha latifolia *
Kingia australis
Tradescantia ohiensis
Hanguana malayana OUTGROUPS
Anigozanthos flavidus
Musa acuminata

Figure 4.

© 2016 The Linnean Society of London, Botanical Journal of the Linnean Society, 2016
8 S. NADOT ET AL.

Figure 4. Phylogenetic analysis of sexual systems evolution in palms. A, supertree from the analysis of Baker et al.
(2009) showing the evolution of sexual systems in palms optimized using parsimony. In subfamily Arecoideae, all tribes
except Chamaedoreeae are monoecious and the subfamily is therefore only represented by Chamaedoreeae in this tree.
Boxes at the tips of branches are coloured according to the actual character state: white = hermaphroditism, light
grey = monoecy, dark grey = dioecy, black = andromonoecy. Branch colours correspond to the inferred ancestral sate.
Several colours on the same branch denote ambiguity in the ancestral state. The lack of both pistillode and staminodes
is indicated by a star (in Arecoideae, this relates to three genera: Iriartella, Pinanga and Sclerosperma). The sexual sys-
tem of the species used in the optimization is typical of the sexual system of the whole genus with exceptions, in which
case the species name is in bold (in Arecoideae there is no polymorphism). Phylogenetic relationships and tribe names
are according to Baker et al. (2009). B, scenario for the evolution of sexual systems in palms as suggested by the opti-
mization shown in A. The tribes (italics) and subfamilies in which dioecy, hermaphroditism and andromonoecy (the most
common type of polygamy) evolved are indicated on the right of each category. Subfamilies in which secondary transi-
tions occur are underlined. Pistils are shown in black, pistillodes in grey. Stamens are shown in white, staminodes in
grey.

(Fig. 4B). Both parsimony and maximum likelihood patterns seen for female and male flowers in the
optimizations (Fig. 4A and Supporting Information, monoecious state, the question of whether this state is
respectively; details of the methods used are given in homologous between different clades of palms may be
Appendix 2) resulted in monoecy as the ancestral con- raised. Despite this apparent diversity, it should be
dition for the family, an unusual situation compared noted that the basic unit of palm inflorescences is gen-
with the rest of angiosperms. From this state, dioecy erally sympodial, with dyads, triads and acervuli rep-
and hermaphroditism were found to have evolved sev- resenting different forms of cincinni (Uhl, 1988). It
eral times in the family, whereas polygamy evolved may actually require few evolutionary steps to switch
from dioecy, hermaphroditism or monoecy, suggesting from one form to another, supporting a homologous
a high degree of sexual flexibility during palm evolu- nature for the various forms of monoecy. The variable
tion. In the same scheme, dioecy was found to have degree of sexual differentiation observed between
evolved directly from monoecy in Arecoideae and male and female flowers of different clades (Fig. 3)
Calamoideae and from hermaphroditism in Coryphoi- would have been further selected in response to selec-
deae. Similarly, monoecy evolved secondarily from tion pressures related to pollination strategies (see
hermaphroditism in Coryphoideae. In Arecoideae, ‘Morphology of sexual expression’ above).
dioecy is found exclusively in tribe Chamaedoreeae
and the strong similarities in sexual differentiation
patterns between monoecious and dioecious species
ECOLOGICAL CORRELATES
suggest that shared molecular mechanisms govern
flower unisexuality (Casta~ no et al., 2014). Considering General conclusions regarding the ecological corre-
the extraordinary diversity of spatial distribution lates of sexual expression in palms, particularly with

© 2016 The Linnean Society of London, Botanical Journal of the Linnean Society, 2016
 SEXUAL SYSTEMS IN PALMS 9

respect to plant–pollinator interactions, should be et al., 2013; Zehdi-Azouzi et al., 2015), allowing the
drawn with caution due to the limited number of characterization of three genetically linked loci that
studies and the inherent complexity of plant–animal appear to be located on a Y-type sex chromosome.
interactions (Barfod et al., 2011). Only a few studies Further studies of the date palm genes present in
have demonstrated a link between temporal and/or this genomic region should allow identification of the
spatial separation of sexual expression in palms and molecular determinants of sex in this species.
pollination strategies (Henderson, 2002). One exam- Clearly, it will be of great interest to compare data
ple concerns the dioecious species Phytelephas see- between different palm species with contrasting sex-
mannii O.F.Cook, in which sexual expression is ual systems, including species such as oil palm which
displaced in the flowering population, typically with display environmentally regulated sex determination
diurnal onset of male anthesis and nocturnal onset (Adam et al., 2011). In this context, the recently pub-
of female anthesis. Bernal & Ervik (1996) considered lished oil palm genome (Singh et al., 2013a, b) is an
this to be an adaptation to minimize the competition invaluable resource and has already been used to
for pollinators between male and female plants. Dis- identify a genomic region of importance to sex regu-
placement of sexual expression has also been lation (Somyong et al., 2015). The region in question
recorded on more extended time scales in the wind- contains a putative aldo-keto reductase gene that
pollinated monoecious palm Howea forsteriana might be involved in abiotic stress response. At a
(F.Muell.) Becc. In populations of this species, ende- physiological level, a possible role for hormones has
mic to Lord Howe Island, the male flowering peak been suggested for oil palm and date palm (e.g. Cor-
was found to occur 2 weeks before the peak of female ley 1976; Masmoudi-Allouche et al., 2009). These
receptivity (Savolainen et al., 2006). This phenologi- studies provide a mere hint of what may well be an
cal difference was suggested to be linked to geology important role for such molecules in palm sex deter-
as it is not observed on all soil types (Savolainen mination.
et al., 2006; Babik et al., 2009).

SEXUAL EXPRESSION IN PALMS:

GENETIC AND MOLECULAR BASIS OF IMMEDIATE AND LONG-TERM PROSPECTS
SEXUAL SEPARATION
Exploring, cataloguing and understanding the great
With regard to the temporal separation of male and diversity of sexual expression in palms is a major
female expression, no genes involved in dichogamy challenge to researchers, but the rewards are likely
have as yet been identified in palms, or indeed in to be equally great, both from a purely scientific
any angiosperms, although progress has been made point of view and also on account of the numerous
in Carya illinoinensis (Wangenh.) K.Koch (Beedana- economically important species belonging to the
gari et al., 2005) through the identification of a group. An essential prerequisite for any investiga-
genetic locus regulating this character. In the wider tion of sexual system evolution is a deep and broad
field of angiosperm flower and inflorescence develop- knowledge of biodiversity studied from morphologi-
ment, much progress has been made in elucidating cal, phylogenetic and ecological perspectives. In the
genetic determinants over the last two decades case of palms, this vital element is provided by the
(Theissen & Melzer, 2007; Teo et al., 2014). To date, milestone work Genera Palmarum (Dransfield et al.,
the only member of Arecaceae for which functional 2008). Molecular approaches, which have already
studies of floral organ identity genes have been pub- revolutionized phylogenetic analyses of the family,
lished is Elaeis guineensis, which appears to display are now being applied to whole genomes, and an
a general conservation of the ABC model (Coen & exciting range of new tools centred around next-gen-
Meyerowitz, 1991; Adam et al., 2007). No palm sex eration sequencing (NGS) technologies that have
determination genes have been reported to date, become available in recent years (Varshney et al.,
however, and even for angiosperms as a whole, this 2009) are now being applied to palms (Heyduk
has only been achieved for monoecious Zea mays L. et al., 2015). The power of such approaches has
and Cucumis melo L. and for dioecious Diospyros already been demonstrated using oil palm, for which
lotus L. (Chuck et al., 2007; Martin et al., 2009; genes determining fruit structure and colour have
Akagi et al., 2014). A major breakthrough was made been identified (Singh et al., 2013a, b, 2014). To
for Arecaceae with the sequencing of the date palm capitalize on these new technologies, it will be vital
(Phoenix dactylifera L.) genome (Al-Dous et al., 2011; to pursue studies of flower and inflorescence diver-
Al-Mssallem et al., 2013). Microsatellite markers sity in the family, so as to provide a bedrock of
identified in the former study were tested for sex- knowledge for future evolutionary and molecular
linkage on a range of male and female palms (Cherif investigations.

© 2016 The Linnean Society of London, Botanical Journal of the Linnean Society, 2016
10 S. NADOT ET AL.

ACKNOWLEDGEMENTS Askgaard A, Stauffer F, Hodel DFR, Barfod AS. 2008.

Floral structure in the neotropical palm genus Chamae-
We are grateful to the Aarhus University Research dorea (Arecoideae, Arecaceae). Anales del Jardin Botanico
Foundation for funding S.N. as guest researcher at de Madrid 65: 197–210.
Aarhus University, Denmark. We thank Herv e Sau- Babik W, Butlin RK, Baker WJ, Papadopulos AST,
quet for giving us access to the PROTEUS database. Boulesteix M, Anstett MC, Lewer C, Hutton I, Savo-
We acknowledge the late Marion Ruff Sheehan for her lainen V. 2009. How sympatric is speciation in the
elegant and accurate drawings of palm flowers, and Howea palms of Lord Howe Island? Molecular Ecology 18:
the L.H. Bailey Hortorium for permitting use of the 3629–3638.
drawings. We would like to acknowledge the wonder- Baker WJ, Dransfield J 2016. Beyond Genera Palmarum:
ful work of two palm specialists in particular, Natalie progress and prospects in palm systematics. Botanical
Uhl and John Dransfield, who through their careful Journal of the Linnean Society doi: 10.1111/boj.12401.
descriptions of all palm genera have paved the way for Baker WJ, Hutton I. 2006. Lepidorrhachis. Palms 50: 33–38.
countless future studies of the family. Baker WJ, Savolainen V, Asmussen-Lange CB, Chase
MW, Dransfield J, Forest F, Harley MM, Uhl NW,
Wilkinson M. 2009. Complete generic-level phylogenetic
analyses of palms (Arecaceae) with comparisons of supertree
REFERENCES
and supermatrix approaches. Systematic Biology 58: 240–256.
Adam H, Collin M, Richaud F, Beul e T, Cros D, Omore Barfod AS. 1991. A monographic study of the subfamily
A, Nodichao L, Nouy B, Tregear JW. 2011. Environmen- Phytelephantoideae (Arecaceae). Opera Botanica 105: 5–73.
tal regulation of sex determination in oil palm: current Barfod AS, Burholt T, Borchsenius F. 2003. Contrasting
knowledge and insights from other species. Annals of Bot- pollination modes in three species of Licuala (Arecaceae:
any 108: 1529–1537. Coryphoideae). Telopea 10: 207–223.
Adam H, Jouannic S, Escoute J, Duval Y, Verdeil JL, Barfod AS, Hagen M, Borchsenius F. 2011. Twenty-five
Tregear JW. 2005. Reproductive developmental complexity years of progress in understanding pollination mechanisms
in the African oil palm (Elaeis guineensis, Arecaceae). Ameri- in palms (Arecaceae). Annals of Botany 108: 1503–1516.
can Journal of Botany 92: 1836–1852. Barot S, Mitja D, Miranda I, Meija GD, Grimaldi M.
Adam H, Jouannic S, Morcillo F, Orieux Y, Duval Y, 2005. Reproductive plasticity in an Amazonian palm. Evo-
Tregear JW. 2007. Functional characterization of MADS lutionary Ecology Research 7: 1051–1065.
box genes involved in the determination of oil palm flower Barrett SCH. 1998. The evolution of mating strategies in
structure. Journal of Experimental Botany 58: 1245–1259. flowering plants. Trends in Plant Science 3: 335–341.
Akagi T, Henry IM, Tao R, Comai L. 2014. A Y-chromo- Barrett SCH. 2010. Understanding plant reproductive diver-
some-encoded small RNA acts as a sex determinant in per- sity. Philosophical Transactions of the Royal Society of
simmons. Science 346: 646–650. London Series B, Biological sciences 365: 99–109.
Al-Dous EK, George B, Al-Mahmoud ME, Al-Jaber MY, Barrett SCH 2013. The evolution of plant reproductive sys-
Wang H, Salameh YM, Al-Azwani EK, Chaluvadi S, tems: how often are transitions irreversible? Proceedings of
Pontaroli AC, Debarry J, Arondel V, Ohlrogge J, Saie the Royal Society B Biological Sciences 280: 20130913.
IJ, Suliman-Elmeer KM, Bennetzen JL, Kruegger RR, Basu SK, Basu P. 1993. Sex expressions in some caryotoid
Malek JA. 2011. De novo genome sequencing and compara- palms. Journal of Economic and Taxonomic Botany 17: 49–54.
tive genomics of date palm (Phoenix dactylifera). Nature Beedanagari SR, Dove SK, Wood BW, Conner PJ. 2005.
Biotechnology 29: 521–U584. A first linkage map of pecan cultivars based on RAPD and
Allen AM, Hiscock SJ. 2008. Evolution and phylogeny of AFLP markers. Theoretical and Applied Genetics 110:
self-incompatibility systems in angiosperms. In: Franklin- 1127–1137.
Tong VE, ed. Self-incompatibility in flowering plants – evo- Beentje H 2010. The Kew plant glossary. Kew: Royal Bota-
lution, diversity, and mechanisms. Berlin: Springer-Verlag. nic Gardens.
Al-Mssallem IS, Hu S, Zhang X, Lin Q, Liu W, Tan J, Yu Bernal R, Ervik F. 1996. Floral biology and pollination of the
X, Liu J, Pan L, Zhang T, Yin Y, Xin C, Wu H, Zhang dioecious palm Phytelephas seemannii in Colombia: an adap-
G, Ba Abdullah MM, Huang D, Fang Y, Alnakhli YO, tation to staphylinid beetles. Biotropica 28: 682–696.
Jia S, Yin A, Alhuzimi EM, Alsaihati BA, Al-Owayyed Boegh A. 1996. The reproductive phenology and pollination
SA, Zhao D, Zhang S, Al-Otaibi NA, Sun G, Majrashi ecology of four Calamus (Arecaceae) species in Thailand.
MA, Li F, Tala, Wang J, Yun Q, Alnassar NA, Wang L, Principes 40: 5–15.
Yang M, Al-Jelaify RF, Liu K, Gao S, Chen K, Alkhaldi Borchsenius F. 1997. Flowering biology of Geonoma irena
SR, Liu G, Zhang M, Guo H, Yu J. 2013 Genome and G. cuneata var. sodiroi (Arecaceae). Plant Systematics
sequence of the date palm Phoenix dactylifera L. Nature and Evolution 208: 187–196.
Communications 4: 2274. Case A, Graham S, Macfarlane T, Barrett S. 2008. A phylo-
Ashman T-L. 2000. Pollinator selectivity and its implications genetic study of evolutionary transitions in sexual systems in
for the evolution of dioecy and sexual dimorphism. Ecology Australasian Wurmbea (Colchicaceae). International Journal
81: 2577–2591. of Plant Sciences 169: 141–156.

© 2016 The Linnean Society of London, Botanical Journal of the Linnean Society, 2016
 SEXUAL SYSTEMS IN PALMS 11

Casta~no F, Stauffer F, Marquinez X, Cr evecoeur M, Col- genus Sabal (Arecaceae) using targeted sequence capture.
lin M, Pintaud JC, Tregear JW. 2014. Floral structure Biological Journal of the Linnean Society 117: 106–120.
and development in the monoecious palm Gaussia attenuata Jeanson M 2011. Systematics of tribe Caryoteae (Arecaceae).
(Arecaceae; Arecoideae). Annals of Botany 114: 1483–1495. Thesis: Museum national d’Histoire naturelle, Paris.
Charlesworth D. 2006. Evolution of plant breeding systems. de Jong TJ, Shmida A, Thuijsman F. 2008. Sex allocation
Current Biology 16: R726–R735. in plants and the evolution of monoecy. Evolutionary Ecol-
Cherif E, Zehdi S, Castillo K, Chabrillange N, ogy Research 10: 1087–1109.
Abdoulkader S, Pintaud JC, Santoni S, Salhi-Hanna- Kholia BS. 2009. Gender variation in a threatened and
chi A, Glemin S, Aberlenc-Bertossi F. 2013. Male-speci- endemic palm Trachycarpus takil Becc. Current Science 96:
fic DNA markers provide genetic evidence of an XY 144–148.
chromosome system, a recombination arrest and allow the Kriebel R. 2014. Sexual dimorphism in the flowers of Aegi-
tracing of paternal lineages in date palm. New Phytologist phila odontophylla (Lamiaceae). Systematic Botany 39:
197: 409–415. 1204–1207.
Chuck G, Meeley R, Irish E, Sakai H, Hake S. 2007. The Loo AHB, Dransfield J, Chase MW, Baker WJ. 2006. Low-
maize tasselseed4 microRNA controls sex determination copy nuclear DNA, phylogeny and the evolution of dichogamy
and meristem cell fate by targeting Tasselseed6/indetermi- in the betel nut palms and their relatives (Arecinae; Are-
nate spikelet1. Nature Genetics 39: 1517–1521. caceae). Molecular Phylogenetics and Evolution 39: 598–618.
Coen ES, Meyerowitz EM. 1991. The war of the whorls – Maddison W, Maddison D 2011. Mesquite: a modular sys-
genetic interactions controlling flower development. Nature tem for evolutionary analysis. Version 3.01. Available at:
353: 31–37. http://mesquiteproject.org
Corley RHV. 1976. Sex differentiation in oil palm - effects of Marcus J. 2010. Observations on the flowering of Marojejya
growth-regulations. Journal of Experimental Botany 27: darianii. Palms 54: 189–192.
553–558. Martin A, Troadec C, Boualem A, Rajab M, Fernandez
Couvreur T, Baker WJ. 2013. Tropical rain forest evolu- R, Morin H, Pitrat M, Dogimont C, Bendahmane A.
tion: palms as a model group. BMC Biology 11: 48. 2009. A transposon-induced epigenetic change leads to sex
Couvreur TLP, Forest F, Baker WJ. 2011. Origin and glo- determination in melon. Nature 461: 1135–8113.
bal diversification patterns of tropical rain forests: infer- Masmoudi-Allouche F, Ch^ aari-Rkhis A, Kria^ a W, Gar-
ences from a complete genus-level phylogeny of palms. gouri-Bouzid R, Jain SM, Drira N. 2009. In vitro her-
BMC Biology 9: 44. maphrodism induction in date palm female flower. Plant
Darwin CR. 1877. The different forms of flowers on plants of Cell Reports 28: 1–10.
the same species. London: John Murray. Maunder M, Page W, Mauremootoo J, Payendee R,
Dowe JL. 2009. A taxonomic account of Livistona R. Br. Mungroo Y, Maljkovic A, Vericel C, Lyte B. 2002. The
(Arecaceae). The Garden’s Bulletin Singapore 60: 185–344. decline and conservation management of the threatened
Dransfield J, Beentje H 1996. Lexicon palmarum. Marly- endemic palms of the Mascarene Islands. Oryx 36: 56–65.
Le-Roi: Editions Champflour in association with the Royal Miller JS, Diggle PK. 2002. Diversification of andromo-
Botanic Gardens. noecy in Solanum section Lasiocarpa (Solanaceae): the
Dransfield J, Mogea JP. 1984. The flowering behaviour of roles of phenotypic plasticity and architecture. American
Arenga (Palmae, Caryotoideae). Botanical Journal of the Journal of Botany 90: 707–715.
Linnean Society 88: 1–10. Moore HE, Uhl NW. 1982. Major trends of evolution in
Dransfield J, Uhl NW, Asmussen CB, Baker WJ, Harley palms. Botanical Review 48: 1–69.
MM, Lewis CE. 2008. Genera Palmarum - the evolution Obbard DJ, Harris SA, Pannell JR. 2006. Sexual systems
and classification of palms. Kew: Kew Publishing. and population genetic structure in an annual plant: testing
Endress PK, Doyle JA. 2009. Reconstructing the ancestral the metapopulation model. The American Naturalist 167:
angiosperm flower and its initial specializations. American 354–366.
Journal of Botany 96: 22–66. Renner SS. 2014. The relative and absolute frequencies of
Freeman DC, McArthur ED, Harper KT, Blauer AC. angiosperm sexual systems: dioecy, monoecy, gynodioecy,
1981. Influence of environment on the floral sex-ratio of and an updated online database. American Journal of Bot-
monoecious plants. Evolution 35: 194–197. any 101: 1588–1596.
Giddey A, Spichiger RE, Stauffer FW. 2009. Comparative Renner SS, Ricklefs RE. 1995. Dioecy and its correlations
floral structure and systematics in the Asian palm genus in the flowering plants. American Journal of Botany 82:
Rhapis (Arecaceae, Coryphoideae). Flora 204: 347–357. 596–606.
Henderson A. 1986. A review of pollination studies in the Rudall PJ, Abranson K, Dransfield J, Baker W. 2003.
Palmae. The Botanical Review 52: 221–259. Floral anatomy in Dypsis (Arecaceae-Areceae): a case of
Henderson A. 2002. Evolution and ecology of palms. New complex synorganization and stamen reduction. Botanical
York: NYBG Press. Journal of the Linnean Society 143: 115–133.
Heyduk K, Trapnell DW, Barrett CF, Leebens-Mack J. Rudall PJ, Ryder RA, Baker WJ. 2011. Comparative
2015. Phylogenomic analyses of species relationships in the gynoecium structure and multiple origins of apocarpy in

© 2016 The Linnean Society of London, Botanical Journal of the Linnean Society, 2016
12 S. NADOT ET AL.

coryphoid palms (Arecaceae). International Journal of Plant Tomlinson PB. 2006. The uniqueness of palms. Botanical
Sciences 172: 674–690. Journal of the Linnean Society 151: 5–14.
Savolainen V, Anstett MC, Lexer C, Hutton I, Clarkson Uhl NW. 1966. Morphology and anatomy of the inflorescence
JJ, Norup MV, Powell MP, Springate D, Salamin N, axis and flowers of a new palm. Journal of the Arnold
Baker WJ. 2006. Sympatric speciation in palms on an Arboretum 47: 9–22.
oceanic island. Nature 441: 210–213. Uhl NW. 1976. Developmental studies in Ptychosperma (Pal-
Singh R, Low E, Ooi L, Ong-Abdullah M, Ting N, Nagap- mae). I. The inflorescence and the flower cluster. American
pan J, Nookiah R, Amiruddin M, Rosli R, Manaf M, Journal of Botany 63: 97–109.
Chan K, Halim M, Azizi N, Lakey N, Smith S, Budi- Uhl NW 1988. Floral organogenesis in palms. In: Leins P,
man M, Hogan M, Bacher B, Brunt AV, Wang C, Ord- Tucker SC, Endress PK, eds. Aspects of floral development.
way J, Sambanthamurthi R, Martienssen R. 2013a. Berlin: J. Cramer.
The oil palm SHELL gene controls oil yield and encodes a Uhl NW, Moore HE. 1971. The palm gynoecium. American
homologue of SEEDSTICK. Nature 500: 340–344. Journal of Botany 58: 945–992.
Singh R, Low ET, Ooi LC, Ong-Abdullah M, Nookiah R, Vamosi J, Otto S, Barrett S. 2003. Phylogenetic analysis of
Ting NC, Marjuni M, Chan PL, Ithnin M, Manaf MA, the ecological correlates of dioecy in angiosperms. Journal
Nagappan J, Chan KL, Rosli R, Halim MA, Azizi N, of Evolutionary Biology 16: 1006–1018.
Budiman MA, Lakey N, Bacher B, Brunt AV, Wang C, Varshney RK, Nayak SN, May GD, Jackson SA. 2009.
Hogan M, He D, MacDonald JD, Smith SW, Ordway Next-generation sequencing technologies and their implica-
JM, Martienssen RA, Sambanthamurthi R. 2014. The tions for crop genetics and breeding. Trends in Biotechnol-
oil palm virescens gene controls fruit colour and encodes a ogy 27: 522–530.
r2r3-myb. Nature Communications 5: 4106. Voeks RA. 1988. Changing sexual expression of a Brazilian rain
Singh R, Ong-Abdullah M, Low E, Manaf M, Rosli R, forest palm (Attalea funifera Mart.). Biotropica 20: 107–113.
Nookiah R, Ooi L, Ooi S, Chan K, Halim M, Azizi N, Weiblen GD, Oyama RK, Donoghue MJ. 2000. Phyloge-
Nagappan J, Bacher B, Lakey N, Smith S, He D, netic analysis of dioecy in monocotyledons. The American
Hogan M, Budiman M, Lee E, DeSalle R, Kudrna D, Naturalist 155: 46–58.
Goicoechea J, Wing R, Wilson R, Fulton R, Ordway J, Williams CN, Thomas RL. 1970. Observations on sex dif-
Martienssen R, Sambanthamurth R. 2013b. Oil palm ferentiation in the oil palm, Elaeis guineensis L. Annals of
genome sequence reveals divergence of interfertile species Botany 34: 937–963.
in Old and New Worlds. Nature 500: 335–339. Willson MF. 1991. Sexual selection, sexual dimorphism and
Somyong S, Poopear S, Jomchai N, Uthaipaisanwong plant phylogeny. Evolutionary Ecology 5: 69–87.
P, Ruang-areerate P, Sangsrakru D, Sonthirod C, Zehdi-Azouzi S, Cherif E, Moussouni S, Gros-Baltha-
Ukoskit K, Tragoonrung S, Tangphatsornruang S. zard M, Abbas Naqvi S, Lude~ na B, Castillo K,
2015. The AKR gene family and modifying sex ratios in Chabrillange N, Bouguedoura N, Bennaceur M, Si-
palms through abiotic stress responsiveness. Functional & Dehbi F, Abdoulkader S, Daher A, Terral J-F, Santoni
Integrative Genomics 15: 349–362. S, Ballardini M, Mercuri A, Ben Salah M, Kadri K,
Teo ZWN, Song SY, Wang YQ, Liu J, Yu H. 2014. New Othmani A, Littardi C, Salhi-Hannachi A, Pintaud
insights into the regulation of inflorescence architecture. J-C, Aberlenc-Bertossi F. 2015. Genetic structure of the
Trends in Plant Science 19: 158–165. date palm (Phoenix dactylifera) in the Old World reveals a
Theissen G, Melzer R. 2007. Molecular mechanisms under- strong differentiation between eastern and western popula-
lying origin and diversification of the angiosperm flower. tions. Annals of Botany 116: 101–112.
Annals of Botany 100: 603–619.

Appendix 1 Genus- and species-level survey of sexual system in palms

The first column indicates the total number of genera (bold) and species (italics). The other columns show the
relative proportion of genera (bold) and species (italics) for each of the four main types of sexual system. Sub-
families are assigned according to Dransfield et al. (2008). Numbers of species and genera follow www.-
palmweb.org.

Number of
genera/species Monoecy Dioecy Hermaphroditism Polygamy

Family level
Arecaceae 188/2585 61%/52% 21%/30% 16%/17% 2%/1%
Subfamily level
Calamoideae 21/645 10%/4% 66%/87% 14%/7% 10%/2%

© 2016 The Linnean Society of London, Botanical Journal of the Linnean Society, 2016
 SEXUAL SYSTEMS IN PALMS 13

Table . Continued

Number of
genera/species Monoecy Dioecy Hermaphroditism Polygamy

Nypoideae 1/1 100%/100%

Coryphoideae 47/503 6%/10% 32%/13% 60%/76% 2%/1%
Ceroxyloideae 8/47 87%/91% 13%/9%
Arecoideae 111/1389 98%/92% 2%/8%

Appendix 2 Methods used for character information. The species vs. genus approach, called
optimization on the palm phylogenetic tree the exemplar approach, avoids the problem of coding
polymorphisms whenever they occur within a genus.
We conducted a survey of 183 species, one for each
We used Mesquite (Maddison & Maddison, 2011) to
genus included in this phylogenetic tree, which rep-
perform parsimony (Fig. 4) and maximum likelihood
resent almost all of the recognized genera (Drans-
optimizations (Supplementary Information) of the
field et al., 2008; Couvreur et al., 2011). We retrieved
character ‘sexual systems’. This character was coded
descriptions of these species from the literature and
as multistate, with four states representing her-
recorded the data in the database PROTEUS (see
maphroditism, monoecy, dioecy and polygamy,
http://eflower.myspecies.info/proteus) which allows
respectively.
precise and complete tracking of the source

SUPPORTING INFORMATION
Additional Supporting Information may be found in the online version of this article:
Data S1. Maximum-likelihood optimization of sexual systems evolution.

© 2016 The Linnean Society of London, Botanical Journal of the Linnean Society, 2016