INVITED REVIEWS AND SYNTHESES
On the rarity of dioecy in flowering plants
JOS K €AFER,* GABRIEL A. B. MARAIS* and JOHN R. PANNELL†
*Laboratoire de Biometrie et Biologie Evolutive, CNRS, UMR 5558, Universite Lyon 1, B^at. Gregor Mendel 43, bd du 11
novembre 1918, 69622 Villeurbanne cedex, France, †Department of Ecology and Evolution, University of Lausanne, Biophore
Building, 1015 Lausanne, Switzerland
Abstract
Dioecy, the coexistence of separate male and female individuals in a population, is a
rare but phylogenetically widespread sexual system in flowering plants. While research
has concentrated on why and how dioecy evolves from hermaphroditism, the question
of why dioecy is rare, despite repeated transitions to it, has received much less attention.
Previous phylogenetic and theoretical studies have suggested that dioecy might
be an evolutionary dead end. However, recent research indicates that the phylogenetic
support for this hypothesis is attributable to a methodological bias and that there is no
evidence for reduced diversification in dioecious angiosperms. The relative rarity of
dioecy thus remains a puzzle. Here, we review evidence for the hypothesis that dioecy
might be rare not because it is an evolutionary dead end, but rather because it easily
reverts to hermaphroditism. We review what is known about transitions between
hermaphroditism and dioecy, and conclude that there is an important need to consider
more widely the possibility of transitions away from dioecy, both from an empirical
and a theoretical point of view, and by combining tools from molecular evolution and
insights from ecology.
Keywords: angiosperms, Dollo’s law, evolutionary dead end, leaky dioecy, reproductive assurance,
sexual systems
Received 23 May 2016; revision received 27 December 2016; accepted 28 December 2016
Introduction
Most flowering plant species (angiosperms) are hermaphroditic
(Yampolsky & Yampolsky 1922; Barrett
2002; Charlesworth 2006) but deviations from this general
pattern are frequent (see Fig. 1). Dioecy, that is the
occurrence of separate male and female individuals, is
found in about 6% of angiosperms (Renner & Ricklefs
1995; Weiblen et al. 2000; Renner 2014), and is thus
strikingly rare in comparison with the equivalent sexual
system (gonochorism) in animals, of which approximately
95% of all species have separate females and
males (Jarne & Auld 2006; Eppley & Jesson 2008). However,
while dioecy is rare in plants, it has nevertheless
frequently evolved from hermaphroditism; in a recent
assessment, Renner (2014) estimated that the 15 600
dioecious species may have evolved in between 900
and 5000 independent transitions from hermaphroditic
ancestors. Dioecious species thus typically form isolated
lineages and rarely large species-rich clades (Darwin
1884; Yampolsky & Yampolsky 1922; Renner & Ricklefs
1995; Heilbuth 2000; Renner 2014), as becomes clear
from its phylogenetic distribution: almost half of all
families and the majority of orders contain dioecious
species (Fig. 2), yet dioecy is almost always rare within
these higher taxa.
The multiple origins of dioecy have attracted a great
deal of attention from both empiricists and theoreticians,
because they expose questions concerning the
evolution and determination of separate sexes per se
(Charnov et al. 1976; Bull 1983; Kohn 1988; Beukeboom
& Perrin 2014), of contrasting allocation strategies and
sex ratios (Hardy 2002; West 2009), as well as questions
about sexual dimorphism and sexual selection in plants
(Charlesworth 1999; Webb 1999; Barrett 2002; Moore &
Pannell 2011; Barrett & Hough 2013). Some dioecious
plants have also recently evolved sex chromosomes,
either XY or ZW systems, which are found in many animals,
too (Westergaard 1958; Ming et al. 2011; Bachtrog
Correspondence: Jos K€afer, Fax: (+33)(0)4 72 43 13 88;
E-mail: jos.kafer@univ-lyon1.fr
© 2017 John Wiley & Sons Ltd
Molecular Ecology (2017) 26, 1225–1241 doi: 10.1111/mec.14020
et al. 2014), and with which they share some evolutionary
processes (Charlesworth et al. 2005; Charlesworth
2016; Muyle et al. 2017). Ever since Darwin (1884) discussed
dioecy from an evolutionary perspective, a number
of hypotheses have been proposed and examined
for why dioecy has evolved (cf., among many others,
reviews by Bawa 1980; Thomson & Brunet 1990; Sakai
& Weller 1999; Charlesworth 1999; Renner 2014). In
contrast, questions arising from the relative rarity of
dioecy and its scattered phylogenetic distribution have
received much less attention.
Given the frequent evolution of dioecy and its overall
rarity in angiosperms, the view that dioecious species
are less evolutionary successful than hermaphrodites
has dominated the literature. In his now classic review
of plant sex chromosomes, Westergaard (1958) dismissed
dioecy as a “failure” in plants. This view, which
has only seldom been challenged (e.g. Bawa 1980; Renner
2014), was bolstered by a quantitative comparative
analysis conducted by Heilbuth (2000), who also concluded
that dioecy was an “evolutionary dead end”,
perhaps easily acquired but doomed to failure as a
result of higher lineage extinction rates, resulting in
shorter persistence times of those lineages. Similar reasoning
and analysis has explained the distribution of
self-compatible plant species (Stebbins 1950; Igic et al.
2008; Goldberg et al. 2010), or of asexual species (Maynard
Smith 1978; Barraclough et al. 2003; Schwander &
Crespi 2009): while self-incompatibility and sexuality
are complex traits that are more easily lost than
regained, they nevertheless are represented in the
majority of extant species due to higher extinction rates
of the species that lost them.
Recently, however, two studies have called into question
the view that dioecy is an evolutionary dead end.
First, building upon an improved test for sister clade
comparisons (K€afer & Mousset 2014), K€afer et al. (2014)
documented a small positive effect of dioecy on species
richness, suggesting that diversification might be more
rapid in at least some of the dioecious lineages than in
the hermaphrodite lineages from which they are
derived. Second, using recently developed phylogenetic
tools (Maddison et al. 2007; FitzJohn et al. 2009), Sabath
et al. (2015) found no general effect of dioecy on diversification
rates in a number of genera having both dioecious
and nondioecious species, concluding that
different lineages reveal different patterns. If dioecy
does evolve frequently, and has no consistently negative
effect on diversification, why is it not more common
than it is?
We here promote the view that the rarity of dioecy in
flowering plants might be due not to its frequent loss
by extinction, but because dioecious species sometimes,
perhaps often, revert back to an hermaphroditic state.
Although the evolution of dioecy has frequently been
considered irreversible, partially based on comparisons
with gonochorism in animals (Bull & Charnov 1985; but
see Bachtrog et al. 2014), recent studies suggest that
reversals might be common in plants (Barrett 2013;
K€afer et al. 2014; Renner 2014). We review the sparse literature
bearing on the question of reversion of dioecy
to hermaphroditism in flowering plants and argue that
this topic needs more attention. We begin with a short
overview of ideas about the evolution of dioecy, the
paths that the transition has taken, and the (selective)
drivers that might be responsible. In this context, we
argue that evolution has taken different routes towards
dioecy, and that quite different evolutionary forces have
probably been involved in different lineages. We should
therefore be cautious about trying to generalize across
all flowering plants. Next, we explain the basis of the
dead-end hypothesis for dioecy, and argue that, at least
for a majority of lineages, it provides a poor
Fig. 1 Schematic depiction of the various conformations of
combined vs. separate sexes in plant populations. Estimates for
the percentage of angiosperm species corresponding to each
conformation are given below each diagram, taken from Renner
(2014). Note that the term “hermaphroditism” is ambiguous,
because it can either refer to the flowers or the
individuals. In the latter sense, all monomorphic species can be
considered hermaphroditic. Conformations having more than
two types of flowers or individuals exist, but are probably rare,
and are not indicated here. The arrows correspond to the main
pathways that have been identified (see Box 1): plain arrows
indicate the gynodioecy and the monoecy–paradioecy pathways,
that have received considerable attention, and dashed
arrows the pathways for which more limited evidence exists.
© 2017 John Wiley & Sons Ltd
1226 J. K €AFER, G. A. B. MARAIS and J. R. PANNELL
Fig. 2 Distribution of dioecious species among the angiosperm orders. The angiosperm tree is reproduced from Stevens (2001
onwards). Differences in clade size, varying from 1 (Amborellales) to 36 135 species (Asparagales), are represented by the size of the
triangles. The percentage of dioecious species, if any, is indicated in parentheses.
© 2017 John Wiley & Sons Ltd
RARITY OF DIOECY IN PLANTS 1227
explanation for the phylogenetic distribution of dioecy
among flowering plants. We then examine the alternative
hypothesis that dioecy is rare because of frequent
reversions to hermaphroditism, first exploring the phylogenetic
evidence for this idea and then considering
what mechanisms might have been involved.
Evolution of dioecy in flowering plants: multiple
paths and drivers
Not only is dioecy widely distributed among higher
taxa of flowering plants (Fig. 2), it is also associated
with a number of different traits or features (cf. Bawa
1980; Givnish 1980; Muenchow 1987; Thomson & Brunet
1990; Charlesworth 1993; Renner & Ricklefs 1995;
Sakai & Weller 1999; Vamosi et al. 2003). For instance,
in comparison to non-dioecious species, dioecious species
are more likely to be trees, shrubs or lianas than
annual herbs; they have higher frequencies in the tropics
and on islands than in temperate regions or on continents;
they are more often pollinated by wind or
water than by insects and particularly rarely by bats or
birds; they rely more often on animal-mediated seed
dispersal than on dispersal by other means; they tend
to have small, simple whitish or greenish flowers in
large inflorescences rather than large flowers with complex
morphology; and they tend to occur in clades
where species with combined sexes are monoecious,
that is, individuals have separate male and female flowers.
The number of different traits, the fact that none of
them is found in a large majority of dioecious species
and the dependence of several traits upon one another
(e.g. trees more often have animal-dispersed fruits and
occur in the tropics) make the search for a general reason
for the evolution of dioecy challenging.
Dioecy could have two main advantages over hermaphroditism
for the organisms that possess it, and
these advantages could explain at least some of the
associations. First, dioecy could evolve as a mechanism
to ensure outcrossing, which is thought to be beneficial
for most species (Darwin 1876; Glemin 2007; Wright &
Barrett 2010). That outcrossing is a major driving force
for plant reproductive system evolution is also suggested
by the fact that at least half of the flowering
plant species have some mechanism to avoid selfing
(Igic & Kohn 2006). Thus, dioecy, as a mechanism that
prevents selfing, could replace genetic self-incompatibility
(SI) under certain conditions (Lloyd 1974; Charlesworth
& Charlesworth 1978). In the case of wind-pollinated
plants, or in the case of large or clonal plants that present
many open flowers at the same time, pollen may
be less effectively transferred from one plant to the other,
leading to higher selfing rates (Bawa 1980; Harder &
Barrett 1995; Barrett 2015). Genetic self-incompatibility
in such plants would prevent self-fertilization, but would
not prevent self-pollination and pollen discounting, that
is, it may still compromise outcrossing opportunities
through the loss of pollen to self-pollination (Holsinger
1988; Harder & Wilson 1998), as well as the clogging of
stigmas. Also, as SI is probably much harder to evolve
than to lose (Goldberg et al. 2010; Barrett 2013), the loss
of SI might set the stage for the evolution of dioecy as
an alternative outcrossing mechanism, as appears to
have happened after the loss of SI due to polyploidization
(Miller & Venable 2000; Ashman et al. 2013). Moreover,
SI might be lost for the sake of reproductive
assurance when species colonize isolated areas such as
Oceanic islands, and subsequent selection against selfing
might explain the high incidence of dioecy on
many such islands (Baker 1955; Carlquist 1966; Pannell
2015).
Second, dioecy could be a way to optimize resource
allocation for the female and the male functions, which
is an important hypothesis for the evolution of gonochorism
in animals (Maynard Smith 1978; Charnov
1982). Charnov et al. (1976) showed that hermaphroditism
is favoured when the male or female fitness of individuals
flattens with accrued investment; indeed, the
immobility of plants is likely to limit male fitness more
than in mobile animals (Eppley & Jesson 2008). Furthermore,
the investment into attractive flowers may benefit
both sexual functions, while the main costs of male and
female functions are separated in time, favouring hermaphroditism
(Charnov et al. 1976). Sex allocation theory
could thus explain why wind-pollinated species,
which invest less in showy flowers, are more often
dioecious than insect-pollinated species (de Jong &
Klinkhamer 2005), or why species with fleshy fruits, in
which the costs of reproduction are unevenly distributed
between the sexes, have a greater tendency to
become dioecious (Bawa 1980; Charnov 1982). Whether
dioecy should mainly evolve to avoid selfing or to optimize
sex allocation has been the subject of heated
debates in the past (Charlesworth & Charlesworth 1978;
Willson 1979; Bawa 1980; Thomson & Barrett 1981;
Charnov 1982; Givnish 1982; Charlesworth 1985), but
recent surveys and modelling studies suggest that the
evolution of dioecy probably results from an interplay
of several forces that include inbreeding avoidance and
advantages associated to sexual specialization (Dufay &
Billard 2012; Spigler & Ashman 2012; Dornier & Dufay
2013; Golenberg & West 2013; Litrico & Maurice 2013;
Sinclair et al. 2013).
The question of why dioecy evolves is intimately
linked with the question of how it evolves. A simple
and insightful view is that the evolution of separate
sexes from ancestors with perfect flowers requires at
least two steps (see Fig. 1, Box 1): either dioecy evolves
© 2017 John Wiley & Sons Ltd
1228 J. K €AFER, G. A. B. MARAIS and J. R. PANNELL
with the spread of individuals of one sex first, and then
the other (gynodioecy and androdioecy pathways), or it
evolves starting with a transition of perfect-flowered
hermaphrodites to monoecy, where individuals have
unisexual flowers of both sexes, and then by the evolution
of unisexual individuals (monoecy–paradioecy
pathway). As discussed in Box 1, modelling suggests
that the gynodioecy pathway is particularly likely to
occur when the selfing rate and the level of inbreeding
depression are both high; androdioecy, by contrast, cannot
easily evolve in response to selection for inbreeding
avoidance, because selfing severely compromises male
siring opportunities and thus selects against males. The
monoecy–paradioecy pathway, for which fewer theoretical
studies exist, might proceed in response to selection
for sexual specialization (Charnov et al. 1976; Charlesworth
& Charlesworth 1981; Charnov 1982).
Which of the pathways is more important to explain
the current distribution of dioecious species remains an
open question. The fact that males of dioecious species
more often develop female flowers or organs than the
inverse (i.e. males are typically more likely to be ‘inconstant’
in their sex expression) indicates that males often
have evolved after females via a gynodioecious intermediate
state (Lloyd & Bawa 1984; Barrett 1992a; Ehlers &
Bataillon 2007). Furthermore, detailed studies of gynodioecious
species have revealed that there is substantial
variation in the maleness of hermaphrodites, suggesting
that different species may be at different intermediate
stages along this pathway (Delph 1990; Barrett 1992a;
Delph & Wolf 2005; McCauley & Bailey 2009). However,
while the statistical association between dioecy and gynodioecy
at higher taxonomical levels has been shown
recently (Dufay et al. 2014), the support for the association
between dioecy and monoecy is stronger (Renner &
Ricklefs 1995). Furthermore, many gynodioecious species
are probably not en route to becoming fully dioecious
(Darwin 1884; Renner 2016; Rivkin et al. 2016), and
gynodioecy is associated with quite different traits than
is dioecy, as gynodioecy is for example more frequent in
herbaceous plants and temperate climates (Dufay et al.
2014; Caruso et al. 2016; Rivkin et al. 2016), whereas
dioecy is more common in woody species in the tropics
(Renner & Ricklefs 1995). Monoecy, on the other hand,
is associated with some of the same traits as dioecy,
such as wind pollination (Charlesworth 1993; Renner &
Ricklefs 1995; Linder 1998; Friedman & Barrett 2008),
suggesting that dioecy has more often evolved under
conditions that favour monoecy than those that favour
gynodioecy. It should finally be noted that the evolution
of dioecy from hermaphroditism does not necessarily
follow two distinct steps; evolution can be gradual as
well or require more steps (Golenberg & West 2013;
Renner 2016). For example, dioecy in the aquatic
perennial Sagittaria latifolia seems to have evolved from
monoecy via gynodioecy (Dorken & Barrett 2004), where
hermaphroditic individuals in such gynodioecious populations
do not have perfect flowers, but separate male
and female flowers (cf. Renner & Won 2001). In contrast
to this hypothesis, recent evidence suggests that gynodioecy
in this species may also be the result of hybridization
between monoecious and dioecious populations
(Yakimowski & Barrett 2016) and might thus not have
been a direct step towards dioecy from monoecy.
End of the dead-end hypothesis?
The dead-end hypothesis for the evolution of dioecy
posits that while dioecy may evolve frequently, under a
wide range of circumstances, dioecious lineages do not
persist and become evolutionary sinks or dead ends,
much as is thought to be the case for species that lose
self-incompatibility (Stebbins 1950; Escobar et al. 2010;
Goldberg et al. 2010) or those that evolve asexual reproduction
(Maynard Smith 1978; Paland & Lynch 2006;
Schwander & Crespi 2009). Support for the dead-end
hypothesis was bolstered by Heilbuth (2000), who provided
quantitative evidence showing that dioecious
clades contained fewer species than their nondioecious
sister clades, and thus apparently diversify less well
(see Box 2).
Subsequent work suggested that the reduced diversification
rate of dioecious lineages was mainly due to
higher extinction rates of dioecious species. Heilbuth
et al. (2001) proposed a demographic model that might
explain the poor success of dioecious lineages. If seeds
are dispersed over short distances into small seed shadows
where establishing individuals experience strong
competition, dioecious species would suffer a ‘seed-shadow
handicap’, because only females disperse seeds
and the potential seed shadow of males remains unexploited
(Heilbuth et al. 2001). Dioecious species may
consequently be more prone to extinction via demographic
stochasticity. This demographic disadvantage
could be overcome if females produce sufficiently more
and/or better-dispersed seeds than hermaphrodite competitors,
a prediction that might account for the frequent
occurrence of fleshy fruits in dioecious species
(Vamosi et al. 2007).
In a second model, Vamosi & Otto (2002) considered
the demographic consequences of the possible effect of
dioecy on interactions with pollinators. Competition
among males for pollinators could render males more
attractive than females (Bateman’s principle; Bateman
1948; Moore & Pannell 2011), to the extent that visits to
females may become insufficient to guarantee seed set,
which increases the population extinction risk (Vamosi
& Otto 2002). Indeed, experimental studies suggest that
© 2017 John Wiley & Sons Ltd
RARITY OF DIOECY IN PLANTS 1229
Box 1. Pathways to dioecy
It has been long acknowledged that many dioecious species have recent hermaphroditic ancestors (Darwin 1884),
because they are closely related to hermaphroditic taxa (see Introduction), and because many unisexual flowers still
show residues of the aborted organs of the other sex (Mitchell & Diggle 2005). Early on, it was suggested that other
sexual systems (see Fig. 1) might sometimes be intermediate steps in the evolution of dioecy from hermaphroditism
(Yampolsky & Yampolsky 1922). Influential studies in the second half of the 20th century identified gynodioecy
and monoecy as the most probable candidates for these intermediate stages (Charlesworth & Charlesworth
1978; Renner & Ricklefs 1995).
Gynodioecy pathway
According to the model by Charlesworth & Charlesworth (1978), a nuclear male-sterility mutation can spread if the
product of the inbreeding depression and the selfing rate exceeds 0.5. Thus, the main driving force for the first step
of this pathway is the existence of inbreeding depression in a population of partially selfing hermaphrodites (Lewis
1941; Lloyd 1974; Charlesworth & Charlesworth 1978). However, the complete loss of the male reproductive function
severely decreases the individual’s contribution to the next generation, and the mere gain of seed quality
through outcrossing may often be not enough to compensate this reduction. Females may also be able to invest
more resources to produce more and/or better seeds, which could probably easily be achieved because they do
not produce pollen.
Females need a much smaller advantage to invade hermaphroditic populations in the case of cytoplasmic male
sterility (CMS), because feminizing cytoplasmic selfish elements are transmitted to all offspring of a female (Lewis
1941; Lloyd 1974). The frequency of females in such gynodioecious populations can be high, which could facilitate
the next step towards dioecy, that is either the invasion of males or selection to increase the maleness of the hermaphrodites
(Maurice et al. 1994; Schultz 1994). There is indeed evidence that many cases of gynodioecy are caused
by CMS (Bailey & Delph 2007; Delph et al. 2007), but there is little direct evidence that CMS-mediated gynodioecy
has preceded the evolution of dioecy (Spigler & Ashman 2012).
Androdioecy
The spread of a female-sterility mutation in a hermaphroditic population, and thus the establishment of an androdioecious
population, occurs under much more stringent conditions than the spread of male sterility, as pure males
would have to compete with selfing hermaphrodites to sire ovules (Lloyd 1975c; Charlesworth & Charlesworth
1978). Also, male sterility immediately prevents self-fertilization of the ovules of an individual, while female sterility
does not: gynodioecy thus more effectively promotes outcrossing than androdioecy (Lloyd 1976). The relative
frequencies of gynodioecy and androdioecy are consistent with the theory: the former is much more common than
the latter, which is extremely rare and found only in a handful of species, mostly resulting from the breakdown of
dioecy (Charlesworth 1984; Pannell 2002). Nevertheless, there is now convincing evidence that androdioecy can
evolve from hermaphroditism through the spread of female-sterility elements due to previously unanticipated mating
patterns (Pannell & Korbecka 2010; Saumitou-Laprade et al. 2010; Billiard et al. 2015), yet this seems to be rather
an exception than a rule.
Monoecy–paradioecy pathway
In this pathway, individuals first evolve separate male (staminate) and female (pistillate) flowers, and selection
subsequently favours a gradual tendency of some individuals to accentuate their male and others their female
function (Lloyd 1972a, 1980; Renner & Won 2001). There has been much less interest in this pathway by theoreticians
than in the gynodioecy pathway. It is nevertheless thought that the evolution of dioecy via monoecy might
proceed in response to selection for increased sexual specialization (Charnov et al. 1976; Charlesworth & Charlesworth
1981; Charnov 1982), and not in response to selection for inbreeding avoidance, which should favour a gynodioecious
step, as noted above. The separation of the sexes in this pathway is more gradual than in the
gynodioecy pathway: female flowers are converted into male flowers on some individuals, and male flowers into
female flowers on others, such that there is no sudden reduction in reproductive output due to sterility mutations
(Lloyd 1972a, 1980). This putative path has been put forward as an explanation for the statistical association at the
© 2017 John Wiley & Sons Ltd
1230 J. K €AFER, G. A. B. MARAIS and J. R. PANNELL
male flowers are often preferred by pollinators (e.g. Bell
1985; Waelti et al. 2009), and comparative evidence indicates
that insect-pollinated plants are under-represented
among dioecious species (see above). Together, these
models thus not only suggest that dioecious species
might be more likely to go extinct than their hermaphroditic
counterparts, but they also hint at how they
could escape that fate through the evolution of life history
or dispersal traits that reduce their handicaps.
Indeed, Vamosi & Vamosi (2004) found a tendency for
dioecious clades to have more species if they were tropical,
or if their species had fleshy fruits, a woody
growth form and less ornate flowers.
Although the dead-end hypothesis has frequently been
cited to explain patterns of species richness (e.g. Magallon
& Castillo 2009; Schlessman et al. 2014), as examples
of species selection (e.g. Savolainen et al. 2002; Rabosky
2009), or in reviews on plant breeding systems (Barrett
2002; Charlesworth 2006), recent studies suggest that the
hypothesis may have more limited support than is
widely thought. First, K€afer & Mousset (2014) showed
that sister clade comparisons can lead to incorrect conclusions
if the trait under study is evolutionarily derived,
as is likely to be the case for dioecy, because in most
cases, it has probably evolved fairly recently from hermaphroditism.
Using a new test for sister clade pairs
with a known ancestral state (see Box 2), and trimming
the data of Renner & Ricklefs (1995) so that it included
only clade pairs for which dioecy was probably not the
ancestral state, K€afer et al. (2014) found no evidence for
the idea that transitions to dioecy reduced the diversification
rate. Rather, their analysis pointed to the possibility
that transitions to dioecy tend to lead to an increase in
diversification, instead of a decrease.
The absence of a consistently negative effect of separate
sexes on diversification was confirmed by Sabath et al.
(2015), who fitted an explicit model of speciation, extinction
and character transitions (Maddison et al. 2007; FitzJohn
et al. 2009; see Box 2) to each of 43 dioecious genera.
They too found a slight excess in genera for which dioecy
seemed to be associated with a higher diversification rate,
although the excess was not significant in their analysis.
Both the studies by K€afer et al. (2014) and Sabath et al.
(2015) found that there is large variation in the diversification
rates of dioecious clades, which could be due to one
or several other factors that were not taken into account
in the analyses. However, the most common traits associated
with dioecy (biotic dispersal, abiotic pollination,
tropical distribution and woody growth form) were
included in these studies, yet they could not explain the
observed differences in diversification rates either. Thus,
while the dead-end hypothesis has given rise to some
interesting secondary hypotheses explaining why dioecious
species might be more prone to extinction under
some conditions (Heilbuth et al. 2001; Vamosi & Otto
2002; Vamosi et al. 2007), it does not seem to provide a
general explanation for the rarity of dioecy and its scattered
phylogenetic distribution.
Frequent reversion of dioecy to
hermaphroditism?
The dead-end hypothesis predicts that dioecious lineages
are more likely to go extinct than their hermaphroditic
counterparts, but it also supposes that
evolutionary paths to dioecy are one-way streets, that is
that reversions are unlikely (although this assumption
is not always made explicitly). This idea conforms to
level of the angiosperms between monoecy and dioecy (Renner & Ricklefs 1995), which is indeed stronger than the
one between gynodioecy and dioecy (Dufay et al. 2014).
Other pathways
Although gynodioecy and monoecy are seen as the two most likely alternative intermediate stages in the evolution
of dioecy from hermaphroditism, dioecy may also evolve from distyly or heterodichogamy, in which populations
are initially polymorphic for the spatial or temporal separation of the sexes on plants, respectively (Darwin 1884;
Lloyd 1979; Ross 1982; Webb 1999; Pannell & Verdu 2006). In distylous species, hermaphroditic flowers either have
long styles and short stamens, or long stamens and short styles, with mating usually restricted to crosses between
rather than within each of the morphs by a linked self-incompatibility system (Baker 1958; Ornduff 1966; Barrett
1992b). Here, dioecy may evolve if one of the morphs (usually that with long styles) evolves towards increasing
female function and the other (usually the short-styled morph) becomes increasingly male. In heterodichogamous
species, populations comprise protandrous and protogynous individuals (usually at equal frequencies), where half
the individuals flower first as a male and then as a female, and vice versa (Renner 2001). Disruptive selection can
give rise to the evolution of dioecy via a gradual shift towards increased maleness (e.g. in protandrous individuals)
or femaleness (in protogynous individuals; Pannell & Verdu 2006).
Box 1. Continued
© 2017 John Wiley & Sons Ltd
RARITY OF DIOECY IN PLANTS 1231
Dollo’s law, which states that evolution never reverts to
exactly the same state as the ancestor (Gould 1970; Bull &
Charnov 1985; Collin & Miglietta 2008). To some extent,
the idea makes intuitive sense: the evolution of dioecy typically
involves the spread of male- and female-sterility
mutations (see Box 1), and the loss of a function should be
more likely than regaining it. However, although sterility
mutations need to increase in frequency in a population
evolving towards dioecy, they cannot of course go to fixation,
as both sexual functions must continue to be
expressed for sexual reproduction to take place. The
recovery of the lost sexual function by members of one sex
thus need not be so difficult if it simply requires the
expression of genes that have continued to be expressed in
Box 2. The study of diversification of dioecious clades
The analysis of Heilbuth (2000) that led to the dead-end hypothesis for dioecy was based on sister clade comparisons
(cf. Mitter et al. 1988; Vamosi & Vamosi 2005). These consist of pairwise comparisons of the number of species
between clades having the trait of interest and their most closely related clades that lack this trait. The two
clades in each pair have the same age, which allows one to use species numbers alone to estimate differences in
diversification, without having to worry about variation in diversification rates through time (Ricklefs 2007). Second,
the correction for phylogenetic inertia is automatic, as one can reasonably suppose that both clades share a
majority of traits with their common ancestor. The use of a large number of clade pairs and of statistical testing
removes the effect of incidental associations between traits.
Despite substantial methodological improvements (cf. Vamosi & Vamosi 2005), sister clade comparisons did not
explicitly consider the timing of the evolution of the trait under study; they thus implicitly assumed that the
appearance of the trait coincided with the split of both clades from their common ancestor. Heilbuth’s (2000) analysis
of dioecy through sister clade comparisons also made this implicit assumption. However, we know that dioecy
has mostly evolved recently from a hermaphrodite ancestor; it is thus likely to have evolved somewhere along the
branch leading to the dioecious clade. This puts an extra constraint on the dioecious clade, namely that it should
not have diversified before dioecy evolved. The assumption that both clades diversified for equal times, as mentioned
above, is thus likely often violated.
K€afer & Mousset (2014) formalized the point that dioecy might evolve subsequent to a lineage split. They found
that, in the case of derived traits like dioecy, the null expectation under equal diversification rates is that the clades
with the derived trait should be smaller than their sisters. The new test they proposed corrects for the sampling
bias that is introduced by “cherry-picking” clade pairs, but requires that ancestral states for all clade pairs are identified.
K€afer et al. (2014) thus only used a set of 115 clade pairs for which the ancestral state was inferred to be nondioecious.
With this data set, including more than 50% of all known dioecious species, a small positive effect of
dioecy on diversification was detected.
Sister clade comparisons use only a minimal amount of information that is present in phylogenies. More sophisticated
methods that are explicitly based on molecular phylogenies have been developed in recent years. One particularly
popular method, “BiSSE” (Binary State Speciation and Extinction model; Maddison et al. 2007; FitzJohn et al.
2009), allows the simultaneous estimation of speciation, extinction and transition rates. This method has, for example,
been successfully applied to show that the loss of self-incompatibility in Solanaceae decreases the lineage
diversification rate (Goldberg et al. 2010), and thus that self-compatibility is an evolutionary dead end. The rarity
of dioecy in angiosperms, however, makes dioecy less suitable for the application of the BiSSE method. First, the
power of BiSSE drops substantially when a trait is represented by less than 10% of the species in the data set
(Davis et al. 2013). Furthermore, this rarity forces one to use large phylogenetic trees in order to have enough taxa
for testing. However, the larger the phylogenetic trees, the higher the probability that other factors influence the
diversification rates in the tree. These heterogeneities could lead to false inferences if they are not explicitly taken
into account (Rabosky & Goldberg 2015).
Sabath et al. (2015) addressed these difficulties by focussing on a set of genera in which both dioecy and other sexual
systems occur. They fitted a BiSSE model to each genus separately, minimizing the problems due to rate
heterogeneity, and they corrected for the effect of residual heterogeneities through simulations. Furthermore, the
average proportion of dioecious species in these genera is substantially higher than the angiosperm average, which
provides a solution to the loss of power at low frequencies of the state. Sabath et al. (2015) did not find a significant
signal for the hypothesis that dioecy would increase diversification rates, in contrast with the results of K€afer et al.
(2014), but both studies refute the dead-end hypothesis as a general explanation for the rarity of dioecy.
© 2017 John Wiley & Sons Ltd
1232 J. K €AFER, G. A. B. MARAIS and J. R. PANNELL
the genome of individuals of the other sex. Reversions
from dioecy back towards hermaphroditism should therefore
be developmentally and evolutionarily possible.
Although the topic has received too little attention,
evidence for the breakdown of dioecy through the reacquisition
of the lost sexual function by each respective
sex is accumulating. Recent research has established
that dioecy most probably was the ancestral state of
several large families, or even orders, in which hermaphroditism
has re-evolved. For instance, Datwyler &
Weiblen (2004) found dioecy to be ancestral in the Moraceae
and the Urticaceae (Rosales); Zhang et al. (2006)
assigned dioecy to the ancestor of the Begoniaceae,
Cucurbitaceae, Datiscaceae and Tetramelaceae (Cucurbitales);
Schlessman (2010) inferred that dioecy was the
ancestral state in the Apiales; and K€afer et al. (2014)
concluded that the ancestor of the Euphorbiaceae, Rafflesiaceae
and Peraceae (Malpighiales) was likely dioecious.
The fact that these clades are species-rich, and
that dioecy only occurs in a minority of species in them,
argues against the idea that the evolution of dioecy is
irreversible, or an evolutionary dead end. In the wellstudied
family Cucurbitaceae, there is strong phylogenetic
evidence that dioecy has been lost several times
(e.g. Volz & Renner 2008; Schaefer & Renner 2010). It
would thus seem that clades that start out as dioecious
can, and sometimes do, end up being predominantly
hermaphroditic following one or more reversions.
Of course, the loss of dioecy does not necessarily imply
a return to the ancestral hermaphroditic state in which
individuals have ‘perfect’ flowers, but may lead to the
evolution of monoecy, in which individuals have separate
male and female flowers. Of the clades just cited, only in
the Apiales does dioecy seem to have reverted to hermaphroditism
with perfect flowers (Schlessman 2010); the
other clades are characterized by a high percentage of
monoecious species. That monoecy often results from the
loss of dioecy rather than being a precursor to dioecy was
also inferred by Weiblen et al. (2000) in their analysis of
sexual-system evolution in the monocots (comprising 11
orders, and about 60 000 species, cf. Fig. 2). They found
that most transitions to dioecy were from hermaphroditic
ancestors (26–35 transitions), while only a few (3–4) were
from monoecious ancestors. On the other hand, monoecy
was about equally likely to evolve from hermaphroditism
(9–17 transitions) as from dioecy (9–11 transitions); transitions
from dioecy to hermaphroditism were slightly rarer
(6). Thus, the strong statistical association between dioecy
and monoecy found by Renner & Ricklefs (1995) might
often be due to evolutionary transitions from dioecy to
monoecy, rather than the reverse.
It is not clear whether the preferential breakdown of
dioecy to monoecy rather than to hermaphroditism may
be attributable to developmental constraints. For
example, in the Cucurbitaceae, perfect flowers are very
rare but not absent altogether (Schaefer & Renner 2011),
and double loss-of-function mutants yield perfect flowers
in the mostly monoecious melon (Boualem et al.
2015). The rarity of perfect flowers in this clade thus
seems to reflect a lack of selection rather than a developmental
constraint. In the Euphorbiaceae, inflorescences
of unisexual flowers have formed so-called
pseudanthia, clusters of highly reduced female and
male flowers that, together, function as a floral unit
similar to an hermaphrodite flower (cf. Prenner &
Rudall 2007). This suggests that potential developmental
constraints in this family may have prevented the
reappearance of perfect flowers, but that these constraints
also can be bypassed if selection favours hermaphroditic
structures. Yet even in the Euphorbiaceae,
perfect-flowered variants have sometimes been
observed (Durand 1963), suggesting that the constraint
cannot be absolute. It may thus be that monoecy is
favoured over hermaphroditism after a history of
dioecy in a given lineages because of evolved associations
between unisexuality and other traits.
We currently lack a global picture and quantitative
data on reversions from dioecy across the phylogeny of
flowering plants. Phylogenetic studies are needed to
address this lack of knowledge, but these will only
yield indirect evidence of reversals, and macroevolutionary
models of sexual-system evolution often yield
ambiguous results that are difficult to interpret (e.g.
Renner & Won 2001; Volz & Renner 2008; Torices et al.
2011). The fact that a large body of work exists on certain
pathways (i.e. the monoecy and gynodioecy pathways
towards dioecy) might influence what model is
finally retained as more plausible and might obscure
less well-known pathways. We thus expect that
improved phylogenies and explicit acceptance of the
idea that the evolution of dioecy is reversible may
reveal many new reversals that could be investigated in
greater detail. Of course, in some clades (e.g., the families
Ebenaceae, Myristicaceae, Pandanaceae and Menispermaceae;
Renner & Ricklefs 1995; Heilbuth 2000;
K€afer et al. 2014; Renner 2014), dioecy does not appear
to revert to hermaphroditism or monoecy, or does so
only rarely; such clades are characterized by a large
number of dioecious species, but few of them have been
well studied (Renner 2014). It is thus likely that, in
some lineages, a transition to dioecy does indeed reflect
a one-way street, but it is not known why lineages
should differ in this sense, too.
By what mechanisms could dioecy be lost?
Although we have a poor understanding from a phylogenetic
point of view as to why some lineages do seem
© 2017 John Wiley & Sons Ltd
RARITY OF DIOECY IN PLANTS 1233
to lock into a dioecious state whereas others allow
reversion, there are three mechanistic hypotheses that
might explain reversions when they do occur. The first
two hypotheses are directly inspired by the selective
forces that have been identified to play a role in the
evolution of dioecy from hermaphroditism (discussed
above), and a third may be specific only to the loss of
separate sexes.
First, just as the shape of fitness gain curves in sex
allocation theory can explain why dioecy evolves in
some cases (Charnov 1982; West 2009), so fitness gain
curves could also account for the evolution of
hermaphroditism from dioecy. Models indicate that
hermaphrodite phenotypes could invade a dioecious
population under conditions where the male and/or
female gain curves become saturating, or flatten off
with investment (Charnov et al. 1976). There are reasons
to suspect that male and female fitness gain
curves must often be saturating in plants, notably if
pollen grains from the same parent compete with one
another to fertilize a restricted pool of ovules (Hamilton
1967; Taylor & Bulmer 1980; Charnov 1982; Lloyd
1983, 1984), or if seeds are dispersed into a restricted
seed shadow, leading to competition among related
seedlings (Clark 1978; Lloyd 1984). The loss of, or a
change in, effective pollinators or seed dispersers, or a
change in the density of individuals in plant populations
could therefore favour hermaphroditic over male
or female strategies. Unfortunately, data bearing on
these ideas lag far behind the theory, and we still have
a poor idea of the actual shape of male or female fitness
gain curves. This is an area where the tools of
molecular ecology could be profitably employed to
infer parental fitness gains from the detailed genetic
structure and genealogy of populations.
Second, in lineages in which dioecy is maintained as
an outcrossing mechanism, it is conceivable that selffertile
hermaphrodites could invade a population if circumstances
arose that caused a reduction in inbreeding
depression. Genetic bottlenecks can purge a population
of its genetic load (Glemin 2003), for example through
recurrent colonization in a range expansion (Pujol et al.
2009; Peischl et al. 2013). The purging of genetic load
and its (re-) accumulation can take place rapidly
(Glemin 2003; Pujol et al. 2009), so one would need to
compare patterns of inbreeding and inbreeding depression
among dioecious and nondioecious populations of
species displaying within-species variation in their sexual
systems. Alternatively, when other means of
outcrossing are acquired, the maintenance of the
inbreeding depression need not to be sufficient to prevent
the breakdown of dioecy. Such a possibility is suggested
by studies on the Hawaiian genus Schiedea
(Caryophyllaceae), in which sexual dimorphism
(gynodioecy and dioecy) is associated with a transition
to wind pollination, possibly as a mechanism to avoid
inbreeding (Weller et al. 1998). The occurrence of a hermaphroditic
insect-pollinated species, Schiedea lydgatei,
within the dimorphic, wind-pollinated clade, indicates
that the transition might be easily reversible here.
Although S. lydgatei suffers strong inbreeding depression,
as its gynodioecious sister species, the fact that it
relies on biotic pollination seems to increase its
outcrossing rate, such that the product of selfing and
inbreeding depression is not enough to maintain
females in the population (Norman et al. 1997).
Finally, the evolution of self-fertile hermaphroditism
from dioecy might be favoured under selection for
reproductive assurance under conditions of mate and/
or pollinator limitation (Maurice & Fleming 1995; Wolf
& Takebayashi 2004; Pannell 2008). Dioecy represents
an obligate outcrossing mechanism in which the
absence of potential mates of the opposite sex should
render an individual effectively sterile. If such situations
should arise frequently, for example in species
with a colonizing habit or where population densities
diminish (Baker 1955; Pannell 2015; Pannell et al. 2015),
males or females that produce a few hermaphroditic
flowers, or flowers of the opposite gender, would have
a clear advantage over those that are fully male or
female (Pannell 1997a, 2000). The selection of hermaphroditism
on the basis of the selfing ability of such
‘leaky’ or ‘inconstant’ males or females under conditions
of mate limitation seems to represent the most
likely scenario for the breakdown of dioecy and is starting
to be supported by models and observations (Ehlers
& Bataillon 2007; Crossman & Charlesworth 2014).
The classical examples for the breakdown of dioecy
via inconstancy in sex expression are provided by the
dioecious and monoecious species and populations in
the genus Leptinella (Asteraceae; formerly named Cotula)
studied by Lloyd (1972a,b, 1975a,b). Sex inconstancy is
typically low in the dioecious populations, and plants
with a few florets of the other sex are still functionally
unisexual (Lloyd 1975a). However, in small populations
where one of the sexes is lost, these inconstant individuals
can have high relative fitness through the inconstant
sex. If females are the heterogametic sex, as they
are in these Leptinella populations (Lloyd 1975a), the
selfing by inconstant females after a loss of males
would result in the segregation of males in their progeny
and thus the immediate restoration of dioecy.
However, if the females are lost from such populations,
selfing by males will yield only male progeny, and
monoecy can evolve gradually as a result of selection
on males to increase the production of female flowers.
Due to secondary differences between male and female
florets, Lloyd (1975b) was able to infer that
© 2017 John Wiley & Sons Ltd
1234 J. K €AFER, G. A. B. MARAIS and J. R. PANNELL
monomorphic populations with high levels of inconsistency
were indeed mostly derived from inconstant males
after the loss of females. Lloyd’s experiments and observations
suggest that loss of dioecy has occurred several
times independently in different species of Leptinella.
Observations on the Mercurialis annua species complex
(Euphorbiaceae) in Europe and the Mediterranean Basin
also suggest that selection for monomorphism or dimorphism
may depend on population densities and dynamics.
Most Mercurialis species are dioecious perennials,
and monoecy in the annual clade, including polyploid
M. annua, is derived from ancestral dioecy (Durand &
Durand 1992; Kr€ahenb€uhl et al. 2002; Pannell et al. 2008).
Where populations are small and population turnover
(extinctions and colonization) is particularly frequent,
monoecious individuals have a strong advantage over
males and/or females through their ability to self-fertilize
(Baker 1955; Pannell & Barrett 1998; Dorken & Pannell
2007; Pannell 2015). The mating system in
monoecious M. annua is strongly density-dependent
(Eppley & Pannell 2007), with isolated plants setting full
seed set upon selfing but individuals in dense populations
largely outcrossing (Hesse & Pannell 2011;
Labouche et al. 2017). This allows males to co-occur with
monoecious plants in androdioecious populations when
populations grow in size, illustrating not only the selection
of combined sexes when densities are low, but also
the evolution of separate sexes when densities increase
(Pannell 1997a; Pannell et al. 2014). In both dioecious and
androdioecious M. annua, males are XY and females or
monoecious individuals are XX (Russell & Pannell 2015),
so that monoecious individuals can be interpreted as
being modified females in which their ‘leaky’ sex expression
(also observed in dioecious populations Yampolsky
1930; Kuhn 1939) has presumably been enhanced and
canalized by selection. A scenario has been proposed
where dioecious populations were located in southern
refugia during the last ice age; monoecious individuals
then colonized northwards when the climate got cooler,
and finally, males are now spreading north among the
dense monoecious populations (Pannell et al. 2014).
These examples show that whether and how selection
can act to change the sexual system depends on the
availability of inconstant sex expression in males or
females, as well as on the mechanisms of sex determination.
The breakdown of dioecy in Leptinella probably
occurred through selection for enhanced leaky sex
expression in homogametic males (Lloyd 1975b) and for
selection of male flowers on homogametic females in
M. annua (Pannell 1997b). While inconstant sex expression
might be more frequent in species that have
recently evolved dioecy, it also occurs naturally in species
in which dioecy is old enough to have led to relatively
strong sexual dimorphism and heteromorphic sex
chromosomes, such as in Silene latifolia (Frick & Cavers
1989) and Humulus (Ainsworth 2000). Mutations of the
Y chromosome give rise to fertile hermaphrodites in
Carica papaya (Ming et al. 2011), a species in a predominantly
dioecious family (Carvalho & Renner 2012), as
well as in Silene latifolia (Westergaard 1958; Lardon et al.
1999; Fujita et al. 2012). In these species, however, the Y
chromosome is too degenerated to render the YY genotype
viable, and models have shown that returns to
monomorphism are not likely in this case (Ehlers &
Bataillon 2007; Crossman & Charlesworth 2014).
Conclusion and future research directions
The widespread but scattered distribution of dioecy
among flowering plants has puzzled evolutionary biologists:
while the evolution of separate sexes has clearly
been favoured in a wide range of circumstances, rarely
has it led to large dioecious clades in plants, in sharp
contrast to animals. While there does not seem to be a
clear evolutionary advantage to dioecy for flowering
plants, and while dioecy might even be a disadvantage
for short-lived, insect-pollinated species with poor seed
dispersal (Vamosi et al. 2003; Vamosi & Vamosi 2004),
the idea that dioecy is typically an evolutionary dead
end has recently been found wanting (K€afer et al. 2014;
Sabath et al. 2015). We suggest that the distribution of
dioecy in flowering plants, that is both its rarity and its
occurrence in many families and orders, could be due
not only to frequent gains, but also to frequent losses
due to reversion rather than to extinction. While much
effort has been invested in studying why and how
dioecy should evolve, the question of whether, how
and why it might be lost requires more attention.
More phylogenetic studies are needed to estimate
how often reversions from dioecy to hermaphroditism
have occurred, in which groups they are most frequent,
and what ecological or life history traits might influence
these reversals. The few case studies and models that
are currently available point to the importance of leakiness,
its heritability and the mechanisms of sex determination
in dioecious species. Both field observations and
experimental evolution studies could help to determine
the frequency of leakiness in sex expression in the individuals
of dioecious species, and the extent to which
selection can act on this leakiness to bring about a
change in the sexual system. It will be particularly
important to know how inconstant sex expression
affects the female and, especially, the male components
of reproductive success and indeed the extent to which
inconstancy is heritable rather than just an expression
of developmental instability. However, as the developmental
and genetic basis of sex determination in plants
remain largely unknown, the basis of inconstancy is
© 2017 John Wiley & Sons Ltd
RARITY OF DIOECY IN PLANTS 1235
also obscure. For only two dioecious species (Persimmon,
Diospyros lotus, Ebenaceae; Akagi et al. 2016; and Asparagus
officinalis, Asparagaceae; Murase et al. 2016) have the
sex-determining genes been identified. Work on monoecious
melon and cucumber has identified pathways of
sex determination of individual flowers and shown that
loss-of-function mutations may result in dioecy as well as
in the restoration of hermaphroditism (Boualem et al.
2015), but whether the same genes are responsible for sex
determination in related dioecious lineages is questionable
(Ma & Pannell 2016). It is however likely that different
and multiple genes are involved in sex determination
in the different groups of dioecious species (Tanurdzic &
Banks 2004; Diggle et al. 2011; Pannell 2017), a hypothesis
that has implications not only for the evolution of dioecy
(Golenberg & West 2013; Renner 2016), but also for its
breakdown.
If sex is determined by the segregation of sex chromosomes
with a degenerate Y or W, the breakdown of
dioecy will depend on whether selection for increased
leakiness occurs on the homogametic or the heterogametic
sex, as well as on the age of the sex chromosomes
and the extent of degeneration. Surprisingly little is
known about sex determination of the vast majority of
dioecious species; only about 40 species are known to
have sex chromosomes (Westergaard 1958; Ming et al.
2011). Much could be learned from a description of the
sex-determining systems of any of the many yet ill-studied
species. Tools are being developed for the discovery
of sex chromosomes through sequencing of nonmodel
species (reviewed in Muyle et al. 2017), and these could
help identify the sex-determining genes that are expected
to reside on the sex chromosomes (Ming et al. 2011; Charlesworth
2016; Muyle et al. 2017). The knowledge of sexdetermining
systems in families with several groups of
dioecious species might also contribute to the better
reconstruction of ancestral states: if rather distant species
share the same sex chromosomes and sex-determining
genes, dioecy is likely to be ancestral. However, this is
only possible if the turnover of sex chromosomes is not
too frequent, but little is known about this turnover rate
at present. The emerging idea that sex chromosomes can
be lost and regained (cf. Bachtrog et al. 2014; Muyle et al.
2017) is yet another point that supports the view that
dioecy is probably neither frequently a stable end-point
in sexual-system evolution nor a dead end, but a state
that may often be lost to the re-evolution of monoecy or
hermaphroditism.
Note added in proof
Goldberg et al. (2017), using a phylogenetic approach to study
sexual system transitions in 40 angiosperm genera, find that
transitions away from dioecy towards monoecy or hermaphroditism
are approximately as common as transitions towards
dioecy. The authors report large variation between genera in
the dominant polarity of transitions between combined and
separate sexes, suggesting that different selective forces and
mechanisms likely operate in different clades. Their paper
stresses a need for more research into the mechanisms of the
loss of dioecy, and provides a list of species that probably have
lost dioecy in their recent history. Their results and conclusions
support the majority of the hypotheses and research directions
put forward here.
Acknowledgements
We thank Susanne Renner and one anonymous referee for
comments and suggestions that helped to clarify our arguments.
JK and GABM acknowledge funding by the Agence
Nationale de Recherche (grant numbers: ANR-11-BSV7-013,
ANR-11-BSV7-024; ANR-14-CE19-0021); JRP acknowledges
support from the Swiss National Science Foundation and the
University of Lausanne.
References
Ainsworth C (2000) Boys and girls come out to play: the
molecular biology of dioecious plants. Annals of Botany, 86,
211–221.
Akagi T, Henry IM, Tao R, Comai L (2014) Plant genetics. A Ychromosome
encoded small RNA acts as a sex determinant
in persimmons. Science, 346, 646–650.
Ashman TL, Kwok A, Husband BC (2013) Revisiting the dioecypolyploidy
association: alternate pathways and research
opportunities. Cytogenetic and Genome Research, 140, 241–255.
Bachtrog D, Mank JE, Peichel CL et al. (2014) Sex determination:
why so many ways of doing it? PLoS Biology, 12,
e1001899.
Bailey MF, Delph LF (2007) A field guide to models of sex-ratio
evolution in gynodioecious species. Oikos, 116, 1609–1617.
Baker HG (1955) Self compatibility and establishment after
long distance dispersal. Evolution, 9, 347–349.
Baker H (1958) Studies in the reproductive biology of West
African Rubiaceae. Journal of the West African Science Association,
4, 9–24.
Barraclough TG, Birky CW Jr, Burt A (2003) Diversification in
sexual and asexual organisms. Evolution, 57, 2166–2172.
Barrett SCH (1992a) Gender variation and the evolution of
dioecy in Wurmbea dioica (Liliaceae). Journal of Evolutionary
Biology, 5, 423–444.
Barrett SCH (1992b) Heterostylous genetic polymorphisms:
model systems for evolutionary analysis In: Evolution and
function of heterostyly (ed. S. C. H. Barrett) pp. 1–39. New
York: Springer-Verlag.
Barrett SCH (2002) The evolution of plant sexual diversity. Nature
Reviews Genetics, 3, 274–284.
Barrett SCH (2013) The evolution of plant reproductive systems:
how often are transitions irreversible? Proceedings of the
Royal Society of London B: Biological Sciences, 280, 20130913.
Barrett SCH (2015) Influences of clonality on plant sexual
reproduction. Proceedings of the National Academy of Sciences of
the United States of America, 112, 8859–8866.
© 2017 John Wiley & Sons Ltd
1236 J. K €AFER, G. A. B. MARAIS and J. R. PANNELL
Barrett SCH, Hough J (2013) Sexual dimorphism in flowering
plants. Journal of Experimental Botany, 64, 67–82.
Bateman AJ (1948) Intra-sexual selection in drosophila. Heredity,
2, 349–368.
Bawa KS (1980) Evolution of dioecy in flowering plants. Annual
Review of Ecology and Systematics, 11, 15–39.
Bell G (1985) On the function of flowers. Proceedings of the Royal
Society of London B: Biological Sciences, 224, 223–265.
Beukeboom LW, Perrin N (2014) The Evolution of Sex Determination.
Oxford University Press, Oxford.
Billiard S, Husse L, Lepercq P et al. (2015) Selfish male-determining
element favors the transition from hermaphroditism
to androdioecy. Evolution, 69, 683–693.
Boualem A, Troadec C, Camps C et al. (2015) A cucurbit
androecy gene reveals how unisexual flowers develop and
dioecy emerges. Science, 350, 688–691.
Bull JJ (1983) Evolution of Sex Determining Mechanisms. Benjamin/Cummings,
Menlo Park, California.
Bull JJ, Charnov EL (1985) On irreversible evolution. Evolution,
39, 1149–1155.
Carlquist S (1966) Biota of long-distance dispersal. I. Principles
of dispersal and evolution. The Quarterly Review of Biology,
41, 247.
Caruso CM, Eisen K, Case AL (2016) An angiosperm-wide
analysis of the correlates of gynodioecy. International Journal
of Plant Sciences, 177, 115–121.
Carvalho FA, Renner SS (2012) A dated phylogeny of the
papaya family (Caricaceae) reveals the crop’s closest relatives
and the family’s biogeographic history. Molecular Phylogenetics
and Evolution, 65, 46–53.
Charlesworth D (1984) Androdioecy and the evolution of
dioecy. Biological Journal of the Linnean Society, 22, 333–348.
Charlesworth D (1985) Distribution of dioecy and self-incompatibility
in angiosperms In: Evolution. Essais in honour of John
Maynard Smith (eds. P. J. Greenwood, P. H. Harvey and S. C.
H. Slatkin) Cambridge University Press, Cambridge.
Charlesworth D (1993) Why are unisexual flowers associated
with wind pollination and unspecialized pollinators. The
American Naturalist, 141, 481–490.
Charlesworth D (1999) Theories of the evolution of dioecy In:
Gender and Sexual Dimorphism in Flowering Plants (eds. M. A.
Geber, T. E. Dawson and L. F. Delph) pp. 33–60. SpringerVerlag,
Berlin, Heidelberg.
Charlesworth D (2006) Evolution of plant breeding systems.
Current Biology, 16, R726–R735.
Charlesworth D (2016) Plant sex chromosomes. Annual Review
of Plant Biology, 67, 397–420.
Charlesworth B, Charlesworth D (1978) Model for evolution of
dioecy and gynodioecy. The American Naturalist, 112, 975–
997.
Charlesworth D, Charlesworth B (1981) Allocation of resources
to male and female functions in hermaphrodites. Biological
Journal of the Linnean Society, 15, 57–74.
Charlesworth D, Charlesworth B, Marais G (2005) Steps in the
evolution of heteromorphic sex chromosomes. Heredity, 95,
118–128.
Charnov EL (1982) The theory of sex allocation. Monographs in Population
Biology 18. Princeton University Press, Princeton, New
Jersey.
Charnov EL, Maynard Smith J, Bull JJ (1976) Why be an hermaphrodite?
Nature, 263, 125–126.
Clark AB (1978) Sex-ratio and local resource competition in a
prosimian primate. Science, 201, 163–165.
Collin R, Miglietta MP (2008) Reversing opinions on Dollo’s
Law. Trends in Ecology & Evolution, 23, 602–609.
Crossman A, Charlesworth D (2014) Breakdown of dioecy:
models where males acquire cosexual functions. Evolution,
68, 426–440.
Darwin C (1876) The Effects of Cross and Self Fertilisation in the
Vegetable Kingdom. J. Murray, London.
Darwin C (1884) The Different Forms of Flowers on Plants of the
Same Species, 2nd edn. John Murray, London.
Datwyler SL, Weiblen GD (2004) On the origin of the fig: phylogenetic
relationships of Moraceae from ndhF sequences.
American Journal of Botany, 91, 767–777.
Davis MP, Midford PE, Maddison W (2013) Exploring power
and parameter estimation of the BiSSE method for analyzing
species diversification. BMC Evolutionary Biology, 13, 38.
Delph LF (1990) Sex-ratio variation in the gynodioecious shrub
Hebe strictissima (Scrophulariaceae). Evolution, 44, 134–142.
Delph LF, Wolf DE (2005) Evolutionary consequences of gender
plasticity in genetically dimorphic breeding systems.
New Phytologist, 166, 119–128.
Delph LF, Touzet P, Bailey MF (2007) Merging theory and
mechanism in studies of gynodioecy. Trends in Ecology &
Evolution, 22, 17–24.
Diggle PK, Di Stilio VS, Gschwend AR et al. (2011) Multiple
developmental processes underlie sex differentiation in
angiosperms. Trends in Genetics, 27, 368–376.
Dorken ME, Barrett SCH (2004) Sex determination and the evolution
of dioecy from monoecy in Sagittaria latifolia (Alismataceae).
Proceedings of the Royal Society B, 271, 213–219.
Dorken ME, Pannell JR (2007) The maintenance of hybrid
zones across a disturbance gradient. Heredity, 99, 89–101.
Dornier A, Dufay M (2013) How selfing, inbreeding depression,
and pollen limitation impact nuclear-cytoplasmic gynodioecy:
a model. Evolution, 67, 2674–2687.
Dufay M, Billard E (2012) How much better are females? The
occurrence of female advantage, its proximal causes and its
variation within and among gynodioecious species. Annals of
Botany, 109, 505–519.
Dufay M, Champelovier P, K€afer J, Henry JP, Mousset S, Marais
GA (2014) An angiosperm-wide analysis of the gynodioecy-dioecy
pathway. Annals of Botany, 114, 539–548.
Durand B (1963) Le complexe Mercurialis annua l. s.l.: une
etude biosystematique. Annales des Sciences Naturelles, Botanique,
Paris, 12, 579–736.
Durand R, Durand B (1992) Dioecie, monoecie, polyplo€ıdie et
speciation chez les Mercuriales annuelles. Bulletin de la
Societe Botanique de France, 139, 377–390.
Ehlers BK, Bataillon T (2007) ‘Inconstant males’ and the maintenance
of labile sex expression in subdioecious plants. New
Phytologist, 174, 194–211.
Eppley SM, Jesson LK (2008) Moving to mate: the evolution of
separate and combined sexes in multicellular organisms.
Journal of Evolutionary Biology, 21, 727–736.
Eppley SM, Pannell JR (2007) Density-dependent self-fertilization
and male versus hermaphrodite siring success in an
androdioecious plant. Evolution, 61, 2349–2359.
Escobar JS, Cenci A, Bolognini J et al. (2010) An integrative test
of the dead-end hypothesis of selfing evolution in Triticeae
(Poaceae). Evolution, 64, 2855–2872.
© 2017 John Wiley & Sons Ltd
RARITY OF DIOECY IN PLANTS 1237
FitzJohn RG, Maddison WP, Otto SP (2009) Estimating traitdependent
speciation and extinction rates from incompletely
resolved phylogenies. Systematic Biology, 58, 595–611.
Frick B, Cavers PB (1989) Aberrations in sex expression in
Silene latifolia. The American Naturalist, 134, 562–573.
Friedman J, Barrett SCH (2008) A phylogenetic analysis of the
evolution of wind pollination in the angiosperms. International
Journal of Plant Sciences, 169, 49–58.
Fujita N, Torii C, Ishii K et al. (2012) Narrowing down the
mapping of plant sex-determination regions using new Ychromosome-specific
markers and heavy-ion beam irradiation-induced
Y-deletion mutants in Silene latifolia. G3
(Bethesda), 2, 271–278.
Givnish TJ (1980) Ecological constraints on the evolution of
breeding systems in seed plants – dioecy and dispersal in
gymnosperms. Evolution, 34, 959–972.
Givnish TJ (1982) Outcrossing versus ecological constraints in
the evolution of dioecy. The American Naturalist, 119, 849–
865.
Glemin S (2003) How are deleterious mutations purged? Drift
versus nonrandom mating. Evolution, 57, 2678–2687.
Glemin S (2007) Mating systems and the efficacy of selection at
the molecular level. Genetics, 177, 905–916.
Goldberg EE, Kohn JR, Lande R, Robertson KA, Smith SA, Igic
B (2010) Species selection maintains self-incompatibility.
Science, 330, 493–495.
Goldberg EE, Otto SP, Vamosi JC et al. (2017) Macroevolutionary
synthesis of flowering plant sexual systems. Evolution, in
press.
Golenberg EM, West NW (2013) Hormonal interactions and
gene regulation can link monoecy and environmental plasticity
to the evolution of dioecy in plants. American Journal of
Botany, 100, 1022–1037.
Gould SJ (1970) Dollo on Dollo’s law: irreversibility and the
status of evolutionary laws. Journal of the History of Biology, 3,
189–212.
Hamilton WD (1967) Extraordinary sex ratios. A sex-ratio theory
for sex link age and inbreeding has new implications in
cytogenetics and entomology. Science, 156, 477–488.
Harder LD, Barrett SCH (1995) Mating cost of large floral displays
in hermaphrodite plants. Nature, 373, 512–515.
Harder LD, Wilson WG (1998) A clarification of pollen discounting
and its joint effects with inbreeding depression on mating
system evolution. The American Naturalist, 152, 684–695.
Hardy ICW (ed.) (2002) Sex Ratios: Concepts and Research Methods.
Cambridge University Press, Cambridge, UK.
Heilbuth JC (2000) Lower species richness in dioecious clades.
The American Naturalist, 156, 221–241.
Heilbuth JC, Ilves KL, Otto SP (2001) The consequences of
dioecy for seed dispersal: modeling the seed-shadow handicap.
Evolution, 55, 880–888.
Hesse E, Pannell JR (2011) Density-dependent pollen limitation
and reproductive assurance in a wind-pollinated herb with
contrasting sexual systems. Journal of Ecology, 99, 1531–1539.
Holsinger KE (1988) Inbreeding depression doesn’t matter –
the genetic basis of mating-system evolution. Evolution, 42,
1235–1244.
Igic B, Kohn JR (2006) The distribution of plant mating systems:
study bias against obligately outcrossing species. Evolution,
60, 1098–1103.
Igic B, Lande R, Kohn JR (2008) Loss of self-incompatibility
and its evolutionary consequences. International Journal of
Plant Sciences, 169, 93–104.
Jarne P, Auld JR (2006) Animals mix it up too: the distribution
of self fertilization among hermaphroditic animals. Evolution,
60, 1816–1824.
de Jong TJ, Klinkhamer GL (2005) Evolutionary Ecology of Plant
Reproductive Strategies. Cambridge University Press, Cambridge,
UK.
K€afer J, Mousset S (2014) Standard sister clade comparison fails
when testing derived character states. Systematic Biology, 63,
601–609.
K€afer J, de Boer HJ, Mousset S, Kool A, Dufay M, Marais GA
(2014) Dioecy is associated with higher diversification rates in
flowering plants. Journal of Evolutionary Biology, 27, 1478–1490.
Kohn JR (1988) Why be female. Nature, 335, 431–433.
Kr€ahenb€uhl M, Yuan YM, K€upfer P (2002) Chromosome and
breeding system evolution of the genus Mercurialis (Euphorbiaceae):
implications of ITS molecular phylogeny. Plant Systematics
and Evolution, 234, 155–169.
Kuhn E (1939) Selbstbest€aubungen subdi€ocischer Bl€utenpflanzen,
ein neuer Beweis f€ur die genetische Theorie der
Geschlechtsbestimmung. Planta, 30, 457–470.
Labouche A-M, Richards SA, Pannell JR (2017) Effects of pollination
intensity on offspring number and quality in a windpollinated
herb. Journal of Ecology, 105, 197–208.
Lardon A, Georgiev S, Aghmir A, Le Merrer G, Negrutiu I
(1999) Sexual dimorphism in white campion: complex control
of carpel number is revealed by y chromosome deletions.
Genetics, 151, 1173–1185.
Lewis D (1941) Male sterility in natural populations of hermaphrodite
plants. New Phytologist, 40, 56–63.
Linder HP (1998) Morphology and the evolution of wind pollination.
In: Reproductive Biology (eds Owens SJ, Rudall PJ), pp.
123–135. Royal Botanic Gardens, Kew.
Litrico I, Maurice S (2013) Resources, competition and selfing:
their influence on reproductive system evolution. Evolutionary
Ecology, 27, 923–936.
Lloyd DG (1972a) Breeding systems in Cotula L. (Compositae,
Anthemideae). I. Array of monoclinous and diclinous systems.
New Phytologist, 71, 1181.
Lloyd DG (1972b) Breeding systems in Cotula L. (Compositae,
Anthemideae). II. Monoecious populations. New Phytologist,
71, 1195–1202.
Lloyd DG (1974) Theoretical sex-ratios of dioecious and gynodioecious
angiosperms. Heredity, 32, 11–34.
Lloyd DG (1975a) Breeding systems in Cotula. III. dioecious
populations. New Phytologist, 74, 109–123.
Lloyd DG (1975b) Breeding systems in Cotula. IV. Reversion
from dioecy to monoecy. New Phytologist, 74, 125–145.
Lloyd DG (1975c) Maintenance of gynodioecy and androdioecy
in angiosperms. Genetica, 45, 325–339.
Lloyd DG (1976) Transmission of genes via pollen and ovules
in gynodioecious angiosperms. Theoretical Population Biology,
9, 299–316.
Lloyd DG (1979) Evolution towards dioecy in heterostylous
populations. Plant Systematics and Evolution, 131, 71–80.
Lloyd DG (1980) The distributions of gender in 4 angiosperm
species illustrating 2 evolutionary pathways to dioecy. Evolution,
34, 123–134.
© 2017 John Wiley & Sons Ltd
1238 J. K €AFER, G. A. B. MARAIS and J. R. PANNELL
Lloyd DG (1983) Evolutionarily stable sex-ratios and sex allocations.
Journal of Theoretical Biology, 105, 525–539.
Lloyd DG (1984) Gender allocation in outcrossing cosexual
plants. In: Perspectives on Plant Population Ecology (eds Dirzo
R, Sarukhan J), pp. 277–300. Sinauer, Sunderland, Mas-
sachusetts.
Lloyd DG, Bawa KS (1984) Modification of the gender of seed
plants in varying conditions. Evolutionary Biology, 17, 255–338.
Ma WJ, Pannell JR (2016) Sex determination: separate sexes are
a double turnoff in melons. Current Biology, 26, R171–R174.
Maddison WP, Midford PE, Otto SP (2007) Estimating a binary
character’s effect on speciation and extinction. Systematic
Biology, 56, 701–710.
Magallon S, Castillo A (2009) Angiosperm diversification
through time. American Journal of Botany, 96, 349–365.
Maurice S, Fleming TH (1995) The effect of pollen limitation
on plant reproductive systems and the maintenance of sexual
polymorphisms. Oikos, 74, 55–60.
Maurice S, Belhassen E, Couvet D, Gouyon PH (1994) Evolution
of dioecy: can nuclear-cytoplasmic interactions select for
maleness? Heredity, 73, 346–354.
Maynard Smith J (1978) The Evolution of Sex. Cambridge
University Press, Cambridge, UK.
McCauley DE, Bailey MF (2009) Recent advances in the study
of gynodioecy: the interface of theory and empiricism.
Annals of Botany, 104, 611–620.
Miller JS, Venable DL (2000) Polyploidy and the evolution of
gender dimorphism in plants. Science, 289, 2335–2338.
Ming R, Bendahmane A, Renner SS (2011) Sex chromosomes in
land plants. Annual Review of Plant Biology, 62, 485–514.
Mitchell CH, Diggle PK (2005) The evolution of unisexual flowers:
morphological and functional convergence results from
diverse developmental transitions. American Journal of Botany,
92, 1068–1076.
Mitter C, Farrell B, Wiegmann B (1988) The phylogenetic study
of adaptive zones - has phytophagy promoted insect diversification.
The American Naturalist, 132, 107–128.
Moore JC, Pannell JR (2011) Sexual selection in plants. Current
Biology, 21, R176–R182.
Muenchow GE (1987) Is dioecy associated with fleshy fruit.
American Journal of Botany, 74, 287–293.
Murase K, Shigenobu S, Fujii S, et al. (2017) MYB transcription
factor gene involved in sex determination in Asparagus officinalis.
Genes Cells., 22, 115–123.
Muyle A, Shearn R, Marais GAB (2017) The evolution of sex
chromosomes and dosage compensation in plants. Genome
Biology and Evolution, in press.
Norman JK, Weller SG, Sakai AK (1997) Pollination biology
and outcrossing rates in hermaphroditic Schiedea
lydgatei (Caryophyllaceae). American Journal of Botany, 84,
641–648.
Ornduff R (1966) The origin of dioecism from heterostyly in
Nymphoides (Menyanthaceae). Evolution, 20, 309–314.
Paland S, Lynch M (2006) Transitions to asexuality result in
excess amino acid substitutions. Science, 311, 990–992.
Pannell J (1997a) The maintenance of gynodioecy and androdioecy
in a metapopulation. Evolution, 51, 10–20.
Pannell J (1997b) Widespread functional androdioecy in Mercurialis
annua L (Euphorbiaceae). Biological Journal of the Linnean
Society, 61, 95–116.
Pannell JR (2000) A hypothesis for the evolution of androdioecy:
the joint influence of reproductive assurance and local
mate competition in a metapopulation. Evolutionary Ecology,
14, 195–211.
Pannell JR (2002) The evolution and maintenance of androdioecy.
Annual Review of Ecology and Systematics, 33, 397–425.
Pannell JR (2008) Consequences of inbreeding depression due
to sex-linked loci for the maintenance of males and outcrossing
in branchiopod crustaceans. Genetics Research, 90, 73–84.
Pannell JR (2015) Evolution of the mating system in colonizing
plants. Molecular Ecology, 24, 2018–2037.
Pannell JR (2017) The when, who, how and why of sex determination
in plants. Current Biology, in press.
Pannell JR, Barrett SCH (1998) Baker’s law revisited: reproductive
assurance in a metapopulation. Evolution, 52, 657–668.
Pannell JR, Korbecka G (2010) Mating-system evolution: rise of
the irresistible males. Current Biology, 20, R482–R484.
Pannell JR, Verdu M (2006) The evolution of gender specialization
from dimorphic hermaphroditism: paths from heterodichogamy
to gynodioecy and androdioecy. Evolution, 60, 660–
673.
Pannell JR, Dorken ME, Pujol B, Berjano R (2008) Gender variation
and transitions between sexual systems in Mercurialis
annua (Euphorbiaceae). International Journal of Plant Sciences,
169, 129–139.
Pannell JR, Eppley SM, Dorken ME, Berjano R (2014) Regional
variation in sex ratios and sex allocation in androdioecious
Mercurialis annua. Journal of Evolutionary Biology, 27, 1467–
1477.
Pannell JR, Auld JR, Brandvain Y et al. (2015) The scope of
Baker’s law. New Phytologist, 208, 656–667.
Peischl S, Dupanloup I, Kirkpatrick M, Excoffier L (2013) On
the accumulation of deleterious mutations during range
expansions. Molecular Ecology, 22, 5972–5982.
Prenner G, Rudall PJ (2007) Comparative ontogeny of the
cyathium in Euphorbia (Euphorbiaceae) and its allies:
exploring the organ-flower-inflorescence boundary. American
Journal of Botany, 94, 1612–1629.
Pujol B, Zhou SR, Sanchez Vilas J, Pannell JR (2009) Reduced
inbreeding depression after species range expansion. Proceedings
of the National Academy of Sciences of the United States of
America, 106, 15379–15383.
Rabosky DL (2009) Heritability of extinction rates links diversification
patterns in molecular phylogenies and fossils. Systematic
Biology, 58, 629–640.
Rabosky DL, Goldberg EE (2015) Model inadequacy and mistaken
inferences of trait-dependent speciation. Systematic
Biology, 64, 340–355.
Renner SS (2001) How common is heterodichogamy? Trends in
Ecology & Evolution, 16, 595–597.
Renner SS (2014) The relative and absolute frequencies of
angiosperm sexual systems: dioecy, monoecy, gynodioecy,
and an updated online database. American Journal of Botany,
101, 1588–1596.
Renner SS (2016) Pathways for making unisexual flowers and
unisexual plants: Moving beyond the “two mutations linked
on one chromosome” model. American Journal of Botany, 103,
587–589.
Renner SS, Ricklefs RE (1995) Dioecy and its correlates in the
flowering plants. American Journal of Botany, 82, 596–606.
© 2017 John Wiley & Sons Ltd
RARITY OF DIOECY IN PLANTS 1239
Renner SS, Won H (2001) Repeated evolution of dioecy from
monoecy in Siparunaceae (Laurales). Systematic Biology, 50,
700–712.
Ricklefs RE (2007) Estimating diversification rates from phylogenetic
information. Trends in Ecology & Evolution, 22, 601–610.
Rivkin LR, Case AL, Caruso CM (2016) Why is gynodioecy a
rare but widely distributed sexual system? Lessons from the
Lamiaceae. New Phytologist, 211, 688–696.
Ross MD (1982) 5 evolutionary pathways to subdioecy. The
American Naturalist, 119, 297–318.
Russell JR, Pannell JR (2015) Sex determination in dioecious
Mercurialis annua and its close diploid and polyploid relatives.
Heredity, 114, 262–271.
Sabath N, Goldberg EE, Glick L et al. (2015) Dioecy does not consistently
accelerate or slow lineage diversification across multiple
genera of angiosperms. New Phytologist, 209, 1290–1300.
Sakai AK, Weller SG (1999) Gender and sexual dimorphism in
flowering plants: a review of terminology, biogeographic
patterns, ecological correlates, and phylogenetic approaches
In: Gender and Sexual Dimorphism in Flowering Plants (eds. M.
A. Geber, T. E. Dawson and L. F. Delph) pp. 1–32. SpringerVerlag,
Berlin, Heidelberg.
Saumitou-Laprade P, Vernet P, Vassiliadis C et al. (2010) A
self-incompatibility system explains high male frequencies in
an androdioecious plant. Science, 327, 1648–1650.
Savolainen V, Heard SB, Powell MP, Davies TJ, Mooers AO (2002)
Is cladogenesis heritable? Systematic Biology, 51, 835–843.
Schaefer H, Renner SS (2010) A three-genome phylogeny of
Momordica (Cucurbitaceae) suggests seven returns from
dioecy to monoecy and recent long-distance dispersal to
Asia. Molecular Phylogenetics and Evolution, 54, 553–560.
Schaefer H, Renner SS (2011) Phylogenetic relationships in the
order Cucurbitales and a new classification of the gourd
family (Cucurbitaceae). Taxon, 60, 122–138.
Schlessman MA (2010) Major events in the evolution of sexual
systems in apiales: ancestral andromonoecy abandoned.
Plant Diversity and Evolution, 128, 233–245.
Schlessman MA, Vary LB, Munzinger J, Lowry PP II (2014)
Incidence, correlates, and origins of dioecy in the island flora
of New Caledonia. International Journal of Plant Sciences, 175,
271–286.
Schultz ST (1994) Nucleo-cytoplasmic male sterility and alternative
routes to dioecy. Evolution, 48, 1933–1945.
Schwander T, Crespi BJ (2009) Twigs on the tree of life? Neutral
and selective models for integrating macroevolutionary
patterns with microevolutionary processes in the analysis of
asexuality. Molecular Ecology, 18, 28–42.
Sinclair JP, Maxwell GD, Freeman DC (2013) Consanguineous
mating, specialization, and the environment: how multiple
variable interactions affect the evolution of dioecy. American
Journal of Botany, 100, 1038–1049.
Spigler RB, Ashman TL (2012) Gynodioecy to dioecy, are we
there yet? Annals of Botany, 109, 531–543.
Stebbins GL (1950) Variation and Evolution in Plants. Columbia
University Press, New York.
Stevens P (2001 onwards) Angiosperm Phylogeny Website.
http://www.mobot.org/MOBOT/research/APweb/.
Tanurdzic M, Banks JA (2004) Sex-determining mechanisms in
land plants. Plant Cell, 16, S61–S71.
Taylor PD, Bulmer MG (1980) Local mate competition and the
sex-ratio. Journal of Theoretical Biology, 86, 409–419.
Thomson JD, Barrett SCH (1981) Selection for outcrossing, sexual
selection, and the evolution of dioecy in plants. The
American Naturalist, 118, 443–449.
Thomson JD, Brunet J (1990) Hypotheses for the evolution of
dioecy in seed plants. Trends in Ecology & Evolution, 5, 11–16.
Torices R, Mendez M, Gomez JM (2011) Where do monomorphic
sexual systems fit in the evolution of dioecy? Insights
from the largest family of angiosperms. New Phytologist, 190,
234–248.
Vamosi JC, Otto SP (2002) When looks can kill: the evolution
of sexually dimorphic floral display and the extinction of
dioecious plants. Proceedings of the Royal Society of London B:
Biological Sciences, 269, 1187–1194.
Vamosi JC, Vamosi SM (2004) The role of diversification in
causing the correlates of dioecy. Evolution, 58, 723–731.
Vamosi SM, Vamosi JC (2005) Endless tests: guidelines for analysing
non-nested sister-group comparisons. Evolutionary
Ecology Research, 7, 567–579.
Vamosi JC, Otto SP, Barrett SC (2003) Phylogenetic analysis of
the ecological correlates of dioecy in angiosperms. Journal of
Evolutionary Biology, 16, 1006–1018.
Vamosi JC, Zhang Y, Wilson WG (2007) Animal dispersal
dynamics promoting dioecy over hermaphroditism. The
American Naturalist, 170, 485–491.
Volz SM, Renner SS (2008) Hybridization, polyploidy, and evolutionary
transitions between monoecy and dioecy in Bryonia
(Cucurbitaceae). American Journal of Botany, 95, 1297–
1306.
Waelti MO, Page PA, Widmer A, Schiestl FP (2009) How to be
an attractive male: floral dimorphism and attractiveness to
pollinators in a dioecious plant. BMC Evolutionary Biology, 9,
190.
Webb CJ (1999) Empirical studies: evolution and maintenance
of dimorphic breeding systems In: Gender and Sexual Dimorphism
in Flowering Plants (eds. M. A. Geber, T. E. Dawson and L.
F. Delph) pp. 61–95. Springer-Verlag, Berlin, Heidelberg.
Weiblen GD, Oyama RK, Donoghue MJ (2000) Phylogenetic
analysis of dioecy in monocotyledons. The American Naturalist,
155, 46–58.
Weller SG, Sakai A, Rankin A, Golonka A, Kutcher B, Ashby K
(1998) Dioecy and the evolution of pollination systems in
Schiedea and Alsinidendron (Caryophyllaceae:Alsinoideae) in
the Hawaiian Islands. American Journal of Botany, 85, 1377–
1388.
West SA (2009) Sex Allocation. Princeton University Press,
Princeton.
Westergaard M (1958) The mechanism of sex determination in
dioecious flowering plants. Advances in Genetics, 9, 217–281.
Willson MF (1979) Sexual selection in plants. The American Naturalist,
113, 777–790.
Wolf DE, Takebayashi N (2004) Pollen limitation and the evolution
of androdioecy from dioecy. The American Naturalist,
163, 122–137.
Wright SI, Barrett SC (2010) Evolution. The long-term benefits
of self-rejection. Science, 330, 459–460.
Yakimowski SB, Barrett SCH (2016) The role of hybridization
in the evolution of sexual system diversity in a clonal, aquatic
plant. Evolution, 70, 1200–1211.
Yampolsky C (1930) Induced alteration of sex in the male plant
of Mercurialis annua. Bulletin of the Torrey Botanical Club, 57,
51–58.
© 2017 John Wiley & Sons Ltd
1240 J. K €AFER, G. A. B. MARAIS and J. R. PANNELL
Yampolsky C, Yampolsky H (1922) Distribution of sex forms in
the phanerogamic flora. Bibliotheca Genetica, 3, 1–62.
Zhang LB, Simmons MP, Kocyan A, Renner SS (2006) Phylogeny
of the Cucurbitales based on DNA sequences of nine
loci from three genomes: implications for morphological and
sexual system evolution. Molecular Phylogenetics and Evolution,
39, 305–322.
J.K. and G.A.B.M. initiated the review, and J.K.,
G.A.B.M. and J.R.P. conceived the outline. J.K. wrote
the first draft and prepared figures; J.K., J.R.P. and
G.A.B.M. wrote the final version.
© 2017 John Wiley & Sons Ltd
RARITY OF DIOECY IN PLANTS 1241