The Cleistogamous Breeding System: A Review of Its Frequency, Evolution, and Ecology in Angiosperms

Abstract

Cleistogamy, a breeding system in which permanently closed, self-pollinated flowers are produced, has received increasing
attention in recent years, but the last comprehensive review of this system was over 20 years ago. The goal of this paper
is to clarify the different types of cleistogamy, quantify the number of families, genera, and species in which cleistogamy
occurs, and estimate the number of times and potential reasons why cleistogamy has evolved within angiosperms. Cleistogamous
species were identified through a literature survey using 13 online databases with references dating back to 1914; only those
species well-supported by floral descriptions or empirical data were included in the data set. On the basis of this survey,
we suggest the use of three different categories of cleistogamy in future studies: dimorphic, complete, and induced. Based
on these categories, cleistogamy in general is present in 693 angiosperm species, distributed over 228 genera and 50 families.
When analyzed on a family level across the angiosperms, the breeding system has evolved approximately 34 to 41 times. Theoretical
investigations indicate that the evolution of cleistogamy in taxa may be influenced by the presence of heterogeneous environments,
inbreeding depression and geitonogamy, and differential seed dispersal, as well as by various ecological factors and plant
size. Cleistogamy will undoubtedly be discovered in additional species as the reproductive biology of more taxa is examined
in the future. Such information will be invaluable for understanding the selective pressures and factors favoring the evolution
of cleistogamy as well as the evolutionary loss of this breeding system, a subject that has received little attention to date.

Full text

The Cleistogamous Breeding System: A Review of Its Frequency,
Evolution, and Ecology in Angiosperms
Theresa M. Culley
and
Matthew R. Klooster
Department of Biological Sciences
University of Cincinnati
614 Rieveschl Hall
Cincinnati 45221-0006, U.S.A.
Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
Types of Cleistogamy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Dimorphic Cleistogamy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
Complete Cleistogamy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
Induced Cleistogamy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
Occurrence of Cleistogamy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
Description of Survey . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
Survey Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
Evolution of Cleistogamy in Angiosperms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
Phylogenetic Implications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
Advantages and Disadvantages of CH and CL Flowers . . . . . . . . . . . . . . . . . . . . . . . . . 19
Selection for Cleistogamy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
Variable Environments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
Inbreeding Depression and Geitonogamy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
Differential Seed Dispersal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
Variable Ecological Factors and Plant Size . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
Implications for Future Research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
Literature Cited . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
Abstract
Cleistogamy, a breeding system in which permanently closed, self-pollinated flowers
are produced, has received increasing attention in recent years, but the last comprehensive
review of this system was over 20 years ago. The goal of this paper is to clarify the differ-
ent types of cleistogamy, quantify the number of families, genera, and species in which
cleistogamy occurs, and estimate the number of times and potential reasons why cleis-
The Botanical Review 73(1): 1–30
Copies of this issue [73(1)] may be purchased from the NYBG Press, The
New York Botanical Garden, Bronx, NY 10458-5125, U.S.A. Please in-
quire as to prices.
Issued 30 March 2007
© 2007 The New York Botanical Garden

togamy has evolved within angiosperms. Cleistogamous species were identified through a
literature survey using 13 online databases with references dating back to 1914; only those
species well-supported by floral descriptions or empirical data were included in the data
set. On the basis of this survey, we suggest the use of three different categories of cleis-
togamy in future studies: dimorphic, complete, and induced. Based on these categories,
cleistogamy in general is present in 693 angiosperm species, distributed over 228 genera
and 50 families. When analyzed on a family level across the angiosperms, the breeding
system has evolved approximately 34 to 41 times. Theoretical investigations indicate that
the evolution of cleistogamy in taxa may be influenced by the presence of heterogeneous
environments, inbreeding depression and geitonogamy, and differential seed dispersal, as
well as by various ecological factors and plant size. Cleistogamy will undoubtedly be dis-
covered in additional species as the reproductive biology of more taxa is examined in the
future. Such information will be invaluable for understanding the selective pressures and
factors favoring the evolution of cleistogamy as well as the evolutionary loss of this breed-
ing system, a subject that has received little attention to date.
Introduction
Cleistogamy, a sexual breeding system defined as the production of permanently
closed, self-pollinated flowers, has intrigued botanists for centuries and is now recog-
nized as an important system found in a variety of plant taxa. The term was first used by
Kuhn in 1867 to describe bud-like flowers that never opened but yet developed into
fruit. He called these cleistogamous flowers (literally, “closed marriage”). Darwin
(1877) noted that in a cleistogamous species, these flowers may be the only type pro-
duced or they may also appear together on the same plant along with open, typically
insect-pollinated flowers (known as chasmogamous or “open marriage” flowers). He de-
scribed cleistogamy in genera such as Impatiens,Oxalis, and Viola as evidence of natu-
ral selection. Cleistogamy was also discussed by Darwin’s contemporaries over subse-
quent decades (e.g., Kerner von Marilaun, 1902). Since then, many scientists have
investigated the ecological, developmental, and evolutionary aspects of cleistogamous
(CL) and chasmogamous (CH) flower production in a variety of plant species (e.g.,
Schemske, 1978; Waller, 1979; Mitchell-Olds & Waller, 1985; Antlfinger, 1986; Schmitt
& Ehrhardt, 1987, 1990; McCall et al., 1989; Bennington & McGraw, 1995; Culley,
2002). The production of CH and CL flowers was once thought to be directly analogous
to outcrossing and selfing, but CH flowers of some species are now known to occasion-
ally self-pollinate before anthesis via delayed self-pollination (Culley, 2000, 2002) as
well as geitonogamy (Stewart, 1994). The number of studies involving species with
cleistogamous flowers has risen dramatically in recent years (Fig. 1). Today, cleis-
togamy is a multifaceted term used to refer to this unique floral type and its subsequent
fruits and seeds, and also to the plant species that produce these closed flowers.
Despite the attention that cleistogamous species have received in the literature, the
extent of cleistogamy within angiosperms is still not fully understood. Cleistogamous
species and genera have been frequently listed in the literature (Darwin, 1877; Kerner
von Marilaun, 1902; Ritzerow, 1908; Rickett, 1932; Uphof, 1938; Camp & Gilly, 1943;
Maheshwari, 1962), but the most recent comprehensive reviews of cleistogamy (Con-
nor, 1979; Lord, 1981; Campbell et al., 1983) are well over 20 years old. Furthermore,
two of these reviews (Connor, 1979; Campbell et al., 1983) focus strictly on the
Poaceae, and a later paper (Plitmann, 1995) focuses on floral dimorphisms in general.
2 THE BOTANICAL REVIEW

There is also some discrepancy in the literature as to what constitutes a cleistogamous
species. Confusion may be due to the use of the term “cleistogamy” to refer to species
with only CL flowers and to those that produce both CL and CH flower types.
Relative to other breeding systems, the evolution of cleistogamy has received relatively
little attention, despite recent advances in molecular techniques that have spurred the de-
velopment of comprehensive angiosperm phylogenies (e.g., Soltis et al., 2000, 2005; An-
giosperm Phylogeny Group, 2003; Hilu et al., 2003). Mapping the cleistogamous trait
onto these phylogenies would provide an estimate of the number of times cleistogamy has
evolved. The purpose of this review is to (1) clarify the different types of cleistogamy that
exist, (2) quantify how often cleistogamy occurs within angiosperm genera and species,
(3) estimate the number of times that cleistogamy has evolved within angiosperms, and
(4) identify ecological factors that may promote the evolution of cleistogamy in plants.
Types of Cleistogamy
Cleistogamy is a sexual breeding system in which the necessity of floral visitation
for fertilization depends upon the type of flower produced. Fertilization within individ-
ual cleistogamous flowers occurs without pollinator intervention, either by direct trans-
fer of pollen grains from anther to stigma or through germination of pollen grains in the
anther and pollen tube growth into the adjoining style within the bud (Mayers & Lord,
1983b). Hence, cleistogamy differs from asexual systems such as apomixis, in which
double fertilization is not required for full seed set. Open chasmogamous flowers may
be pollinated by floral visitors or may sometimes self-pollinate through prior or delayed
selfing. We consider here three major types of cleistogamy that vary in terms of their de-
velopmental pathways (Fig. 2). In all cases, the floral forms are influenced to varying
CLEISTOGAMY IN ANGIOSPERMS 3
Fig. 1. Number of papers published in 23 botanical and ecological journals within a 120-year inter-
val (1879–1999) that cite either “cleistogamy” or “cleistogamous.” Data were obtained through the
JSTOR online search engine (http://www.jstor.org), which has a 5-year moving wall.

4 THE BOTANICAL REVIEW
Fig. 2. Sample pathways that result in the production of three different types of cleistogamy.
Shown are the original primordial buds and a period of environmental change (shaded boxes) occurring
during the developmental sequence. In dimorphic cleistogamy, the primordial bud is already predeter-
mined to develop into either a chasmogamous (CH) or cleistogamous (CL) flower. Actual production of
either flower type (indicated by lines) is dependent upon the environment. In induced cleistogamy, there
is no fixed developmental trajectory, and each bud will develop into a CH flower unless the environment
prevents its mechanical opening. Finally, all buds in species with complete cleistogamy develop into CL
flowers.

degrees by the environment in which they occur. CL flowers are often favored under
poor growth conditions, presumably because they are energetically less costly (Waller,
1979; but see Cheplick, 2005b). The types of cleistogamy described below are adapted
in part from Lord (1981) and Campbell et al. (1983), both of whom modified their defi-
nitions from Hackel (1906).
dimorphic cleistogamy
Dimorphic cleistogamy is the same as Lord’s (1981) category of true cleistogamy,in
which prominent differences in CL and CH floral morphology result from divergent de-
velopmental pathways. CL flowers are modified early in development and are character-
ized by a reduction in corolla size and stamen size and/or stamen number relative to CH
flowers. This category is most commonly associated with cleistogamy in general, as ex-
emplified by many Impatiens and Viola species (e.g., Darwin, 1877). On a given plant,
both types of flower can appear at the same time but in different positions (spatially sep-
arated; Lloyd’s [1984] multiple strategy), or they may be produced sequentially during
the season (temporally separated; Lloyd’s [1984] conditional strategy). Spatial floral sep-
aration is evident, for example, in Amphicarpaea bracteata (Trapp & Hendrix, 1988) and
Vigna minima (Gopinathan & Babu, 1987), in which aerial CH, aerial CL, and subter-
ranean CL flowers are all produced. In the case of temporal separation, the sequence of
CH and CL flower production depends upon the species. In the forest herb Viola pubes-
cens, CH flowers are produced only in the early spring, followed by the appearance of
CL flowers a few weeks later, once light levels fall as the forest canopy forms (Culley,
2002). In other species, CL flowers appear before CH flowers, as in Ceratocapnos hete-
rocarpa (Ruiz de Clavijo & Jimenez, 1993), in which CH flower production is eventually
triggered by a specific photoperiod and temperature in the spring. Dichanthelium clan-
destinum also produces CL flowers during spring, with CH flowers subsequently appear-
ing in late summer (T. Bell, pers. comm.). Multiple flower-type sequences can also occur
throughout the season, as in Viola canadensis (CH-CL-CH; Culley, 2000) and Centaurea
melitensis (CL-CH-CL; Porras & Muñoz Álvarez, 1999). Finally, some species are capa-
ble of both temporal and spatial separation of floral types. For example, in Impatiens,in-
dividual plants can produce CH and CL flowers simultaneously (Waller, 1979; Antlfin-
ger, 1986), but seasonal trends are more often detected in populations with either flower
type produced first (e.g., Simpson et al., 1985; Masuda & Yahara, 1994; Lu, 2002). This
may reflect populational differences in light availability, with high light or dense shade
triggering CH or CL floral production, respectively (Schemske, 1978).
Changes in floral type during the season in dimorphic cleistogamy are due to alter-
ations in the initial production of primordial buds, which then develop into either CH or
CL flowers. Individual buds are incapable of converting from one floral form into an-
other once the developmental pathway has been determined. For example, if the light
environment of Viola pubescens suddenly declines, all CH buds on individual plants
typically abort within a few days, and plants begin producing new CL buds soon there-
after (Culley, 2002). The ability to produce CH and CL flowers has a discrete genetic
basis, which is affected by a number of abiotic and biotic factors such as light levels,
nutrient availability, pollinator availability, and herbivory. We have renamed this cate-
gory dimorphic instead of true cleistogamy or facultative cleistogamy to emphasize the
CLEISTOGAMY IN ANGIOSPERMS 5

dual production of different floral forms and to avoid any confusion with the following
category.
complete cleistogamy
Defined as the production of only CL flowers on an individual, complete cleistogamy
has been reported in several species, especially in orchids and grasses. For example, the
Hawaiian endemic Schiedea trinervis self-pollinates in the bud, which never opens
(Wagner et al., 2005). This category has been maintained in previous reviews of cleis-
togamy (Hackel, 1906; Connor, 1979; Lord, 1981). Most indications of complete cleis-
togamy are based on observations of only a few individuals, often in artificial environ-
ments such as a greenhouse. This is especially true of orchid species, which are often
cultivated under highly artificial conditions, with only a few individuals per species.
Darwin (1877) was skeptical about reported cases of complete cleistogamy and recom-
mended further study under natural conditions to ensure that no CH flowers are ever
produced. Thus, to verify complete cleistogamy within a plant species, investigations
must include monitoring the flowering phenology of multiple individuals in the natural
environment.
induced cleistogamy
This is the same as Lord’s (1981) category of pseudocleistogamy and Uphof’s (1938)
ecological cleistogamy. In this case, the environment arrests the development of CH
flowers prior to anthesis and results in a mechanical failure of the flower to open, result-
ing in the production of a CL flower (Schoen & Lloyd, 1984). In contrast to dimorphic
cleistogamy, there are no morphological differences between CL and CH flowers other
than lack of floral expansion and anthesis in CL flowers (Lord, 1981). Shifts from CH to
CL flower production occur more quickly during the season in this category than in di-
morphic cleistogamy. In addition, there are no fixed developmental trajectories that dif-
fer for the two floral types (Fig. 2). In this category, unfavorable conditions such as
drought and low temperatures often promote CL flower production (Uphof, 1938). For
example, several Festuca species in the Far East Arctic that often generate masses of CH
flowers produce CL flowers only under conditions of low temperature and high relative
humidity (Connor, 1998). Portulaca species in Hawaii often produce CL flowers under
conditions of reduced light and temperature, resulting in a failure of buds to open (Kim
& Carr, 1990). We have renamed this category from Lord’s (1981) and Uphof’s (1938)
designations to emphasize the effect of the environment in inducing a change in flower
type.
Another reported category of cleistogamy is that of preanthesis cleistogamy,in
which self-pollination occurs first in the bud, followed by anthesis and opportunities for
outcrossing. For example, self-fertilization of Lacandonia schismatica occurs in flower
buds, as pollen grains germinate within the anthers and pollen tubes grow within the re-
ceptacle before the flower opens (Márquez-Guzmán et al., 1993). CH flowers of the in-
breeder Mimulus nasutus are also capable of self-fertilization prior to anthesis (Diaz &
Macnair, 1999). Because these flowers do not remain closed, we do not consider them
to be an example of cleistogamy, but rather of prior self-pollination in CH flowers
(Lloyd & Schoen, 1992). Consequently, examples of this floral morphology will not be
considered further in this review.
6 THE BOTANICAL REVIEW

The three types of cleistogamy presented here occur along a gradient of morphologi-
cal change influenced by the environment. That is, in response to different environmen-
tal conditions, individuals within a species may be able to alter their CH/CL flower pro-
duction within a short period of time (induced cleistogamy) or gradually in response to
seasonal changes (dimorphic cleistogamy), or they may produce only CL flowers re-
gardless of the environment (complete cleistogamy). It would not be unexpected to find
cases of cleistogamy that are difficult to partition into these categories, especially as
more species are examined in the future.
The type of cleistogamy can also vary among populations in some species, such as
the orchid Corallorhiza bentleyi, in which populations produce either CL flowers or
varying numbers of both CH and CL flowers (Freudenstein, 1999; J. Freudenstein, pers.
comm.). In populations of Vigna minima, individual plants produce only one type of
flower or a combination of aerial CH flowers, induced aerial CL flowers, and subter-
ranean CL flowers (Gopinathan & Babu, 1987). In some species, cleistogamy is not the
only method of self-pollination. For example, in addition to cleistogamy in Drosophyl-
lum lusitanicum (Droseraceae), selfing can occur simultaneously on an individual plant
in CH flowers via prior selfing, delayed selfing, or geitonogamy (Ortega-Olivencia et
al., 1998).
In some species, the ratio of CH to CL flowers may also fluctuate among individuals
and populations. In a population of Viola pubescens, for example, 74% of sampled indi-
viduals produced both types of flowers during a single season, 13% produced only CH
flowers, 9% produced only CL flowers, and the remaining 4% were vegetative; all of
these individuals had produced both CH and CL flowers the previous year (Culley,
2002). Similarly, individuals of Viola sororia may not always produce both flower types
every year (Solbrig, 1981). Percentage cleistogamy (defined as the number of CL buds
divided by total bud number) differed substantially among populations of Impatiens pal-
lida (range, 69–100%; Schemske, 1978) and I. biflora (80–98%; Schemske, 1978). In
addition, the proportion of flowering plants producing CL flowers varied within popula-
tions of Oxalis montana (range, 25–68%; Jasieniuk and Lechowicz, 1987) and Dantho-
nia spicata (0–94%; Clay, 1983b; Cheplick, 2005b).
To fully investigate the occurrence of cleistogamy within a species in future studies,
it will be necessary to monitor the flowering phenology and floral production in several
populations and individuals within populations over several seasons. Because the cate-
gories presented here are based on ecological and genetic factors, they can be very use-
ful in describing the breeding system of any given plant species.
Occurrence of Cleistogamy
In one of the most frequently cited reviews of angiosperms, Lord (1981) reported
that cleistogamous flowers were detected in 56 families and 287 species. These initial
figures included all types of cleistogamy (including preanthesis cleistogamy). Taxa that
exhibit only dimorphic cleistogamy are listed in her Table I, and consist of 29 families,
67 genera, and at least 148 species. In a later review, Plitmann (1995) reported cleis-
togamy in 112 genera and 202 species, with 81% of the latter possessing dimorphic
cleistogamy. To update Lord’s (1981) estimate and to include subsequent reviews and
papers (e.g., Campbell et al., 1983, for Poaceae), we conducted an extensive literature
search with a goal of classifying angiosperm species into the three categories of cleis-
togamy presented above. In contrast to previous reviews (e.g., Camp & Gilly, 1943; Ma-
CLEISTOGAMY IN ANGIOSPERMS 7

heshwari, 1962) that were based on species or genera lists derived from older anecdotal
sources (e.g., Rickett, 1932), our approach was to consider only those species as cleis-
togamous that were well-supported in the literature with floral descriptions or empirical
data. Thus, our study is more thorough and conservative than many previous reviews.
description of survey
The literature survey was performed using online databases and supplemented with
printed journals when appropriate. A number of different online search engines were
used to locate publications from 1914 (the earliest that many search engines can cur-
rently access) to the present that contained the keywords “cleistogamy,” “cleistoga-
mous,” or “autogamy.” The primary search engines used in this analysis were the fol-
lowing: Academic Search Premier, BioOne, BIOSIS, Blackwell-Synergy, Ingenta
Online Journals, Ingenta Select, JSTOR Ecology and Botany Collection, Ohiolink Elec-
tronic Journal Center, Oxford Journals Online, ProQuest Research Library, and Wilson
OmniFile-Full Text Mega Edition. Articles, including several Ph.D. dissertations, were
acquired in electronic or printed formats. A few cases of purported complete cleis-
togamy were verified by using online search engines and virtual herbaria to locate addi-
tional information or floral images of the species in question.
Based on descriptions of floral morphology and/or development in each article,
species were placed into one of the three categories of cleistogamy. In some cases, the
author’s classification of cleistogamy type was revised based upon the criteria presented
above. All grass species cited by Campbell et al. (1983) were considered dimorphic un-
less they were described as having complete cleistogamy. In our review, studies clearly
indicated that CL flowers were present in certain species, but these studies lacked suffi-
cient morphological or developmental descriptions of the CH flowers. These species
could not be assigned to one of the three categories with confidence and were therefore
placed in an “unclear” category of cleistogamy. In a few cases, the literature indicated
that certain species produced both CH and CL flowers, but the species could not be fur-
ther assigned to a dimorphic or induced category, so they were also left as “unclear.”
The taxonomy of species, genera, and families was also updated to reflect currently ac-
cepted nomenclature. Selected investigations were excluded from the analysis if (1) the
description of cleistogamy was ambiguous in that it could potentially include prior self-
ing, (2) cultivated species were genetically modified or bred to produce cleistogamous
flowers, (3) taxa consisted of cleistogamous hybrids or mutants, or (4) cleistogamy was
reported from only a few individuals of a given species grown in an artificial environ-
ment. Thus, this review is a conservative estimate of the total number of cleistogamous
species documented in literature.
survey results
We found that cleistogamy in general is present in 693 angiosperm species, distri-
buted over 228 genera and 50 families [[TR I]](Table I; species list available from T.
Culley upon request). This species estimate is more than twice the number of cleistoga-
mous species reported by Lord (1981). The number of families was lower in our survey
than indicated by Lord (1981) because of modern taxonomical rearrangements (An-
giosperm Phylogeny Group, 2003; Soltis et al., 2005) and the more conservative nature
of our review. This review includes five families (Aizoaceae, Aristolochiaceae, Bromeli-
8 THE BOTANICAL REVIEW

CLEISTOGAMY IN ANGIOSPERMS 9
Table I
List of cleistogamous families and genera in angiosperms. Shown are the number of species within each genera classified as exhibiting
complete, induced, dimorphic, or unclear cleistogamy. Additional references for some species can be found in Lord (1981)
Family Genus Complete Induced Dimorphic Unclear References
Acanthaceae Aechmanthera . 1 . . Maheshwari,1962
Blechum . 1 Uphof, 1938
Dianthera . 1 . . Maheshwari, 1962
Dicliptera . 1 . . Uphof, 1938
Dipteracanthus . . 1 . Uphof, 1938
Eranthemum . 2 1 . Darwin, 1877; Uphof, 1938
Ruellia . . 9 . Lord, 1981; Sigrist & Sazima, 2002
Schaueria . . 1 . Lord, 1981
Stenandrium . 1 . . Uphof, 1938
Aizoaceae Sarcozona . . . 1 Keighery, 1988
Alismatacea Luronium . 1 . . Kerner von Marilaun, 1902
Aristolochiaceae Aristolochia . . 1 . Pfeifer, 1966
Asteraceae Ainsliaea . . 2 . Watanabe et al., 1992
Catanche . . 1 . Kaul et al., 2000
Centaurea . . 1 . Porras & Muñoz Álvarez, 1999
Gymnarrhena . . 1 . Lord, 1981
Balsaminaceae Impatiens . . 4 . Lord, 1981; Masuda & Yahara, 1994
Boraginaceae Cryptantha . . 3 . Calvino & Galetto, 2003; Grau, 1981
Lithospermum . . 6 . Lord, 1981; Gleason & Cronquist, 1991
Brassicaceae Cardamine . . 1 . Lord, 1981
Draba . . 1 . Cruden, 1977
Geococcus . . 1 . Kaul et al., 2000
Subularia . 1 . . Kerner von Marilaun, 1902
Thlaspi . 1 . . Uphof, 1938; Maheshwari, 1962
Bromeliaceae Tillandsia . . 2 . Gardner, 1982; Gilmartin & Brown, 1985
Campanulaceae Campanula . . 1 . Kerner von Marilaun, 1902
Githopsis . . 1 . Morin, 1983
Howellia . . 1 . Lesica et al., 1988
Triodanis . . 1 1 Lord, 1981; Ritzerow, 1908; Trent, 1939; Maheshwari, 1962
Caryophyllaceae Illecebrum . 1 . . Kerner von Marilaun, 1902
Polycarpon . . 1 . Maheshwari, 1962

10 THE BOTANICAL REVIEW
Table I (continued )
Family Genus Complete Induced Dimorphic Unclear References
Schiedea 3 . . . Wagner et al., 2005
Stellaria . 1 . . Maheshwari, 1962
Cistaceae Halimium . . 1 . Lord, 1981
Helianthemum . . 1 . Lord, 1981
Lechea 1 . . . Nandi, 1998
Tuberaria 1 . . . Herrera, 1992; Nandi, 1998
Commelinaceae Commelina . . 4 . Lord, 1981
Commelinantia . . 1 . Maheshwari, 1962
Murdannia . 1 . . Lord, 1981
Plowmanianthus 2 . 2 . Hardy & Faden, 2004
Tradescantia . . 1 . Maheshwari, 1962
Droseraceae Drosera . . 1 . Darwin, 1877
Drosophyllum . . . 1 Ortega-Olivencia et al., 1998
Elatinaceae Elatine 1 3 . . Maheshwari, 1962; Keighery, 1984
Ericaceae Vaccinium . . 1 . Vander Kloet, 1993
Fabaceae Amphicarpaea . . 3 . Lord, 1981; Trapp & Hendrix, 1988; Trapp, 1988
Astragalus . . 1 . Gallardo et al., 1993
Clitoria . . 2 . Gomez & Kalamani, 2003
Galactica . . 1 . Kaul et al., 2000
Glycine . . 4 1 Lord, 1981; Schoen & Brown, 1991; Takahashi et al., 2001;
Moyle et al., 2004
Lathyrus . . 3 . Lord, 1981; Kaul et al., 2000
Lespedeza . . 13 . Lord, 1981; Cole & Biesboer, 1992
Lotononis . 12 . . Van Wyk, 1990
Macroptilium . 1 . . Drewes & Hoc, 2000
Medicago . . 6 . Novosyelova, 1998
Ononis . . 3 . Lord, 1981
Phaseolus . . 3 . Lord & Kohorn, 1986; Delgado-Salinas, 2000;
Kaul et al., 2000
Pisum . . 1 . Kaul et al., 2000
Rhynchosia . . 1 . Lord, 1981
Tephrosia . . 2 . Lord, 1981; Kaul et al., 2000
Trifolium . . 1 . Kaul et al., 2000

CLEISTOGAMY IN ANGIOSPERMS 11
Vicia . . 1 . Lord, 1981
Vigna . . 1 . Gopinathan & Babu, 1987; Kaul et al., 2000
Voandzeia . . 1 . Lord, 1981
Gentianaceae Sebaea . . 1 . Lord, 1981
Gesneriaceae Streptocarpus . . 1 . Lord, 1981
Hydrocharitaceae Blyxa . 2 . . Jiang & Kadono, 2001
Ottelia . 1 1 . Lord, 1981; Jiang & Kadono, 2001
Juncaceae Juncus . . 1 . Lord, 1981
Lamiaceae Ajuga . . 1 . Ruiz de Clavijo, 1997
Lamium . 1 1 1 Lord, 1981; Trent, 1939; Maheshwari, 1962
Salvia 1 . 1 . Kerner von Marilaun, 1902; Uphof, 1938;
Plitmann, 1995
Scutellaria . . 1 . Sun, 1999
Lentibulariaceae Utricularia 1 . 4 . Lord, 1981; Gleason & Cronquist, 1991; Khosla et al.,
1998; Yamamoto & Kadono, 1990
Liliaceae Narthecium . 1 . . Jacquemart & Desloover, 1992
Lythraceae Ammannia 1 . . . Maheshwari, 1962
Lythrum . 1 . . Kerner von Marilaun, 1902
Malpighiaceae Aspicarpa . . 3 . Lord, 1981
Camarea . . 1 . Lord, 1981
Gaudichaudia . . 1 . Lord, 1981
Janusia . . 1 . Lord, 1981
Malvaceae Gossypium . . 4 . Lord, 1981
Malva . . . 1 Trent, 1939
Marantaceae Calathea . . 1 . Le Corff, 1993
Mayacaceae Mayaca . 1 . . Uphof, 1938; Maheshwari, 1962
Nyctaginaceae Acleisanthes . . 16 . Lord, 1981; Levin, 2000; Levin, 2002
Mirabilis . . 1 . Lord, 1981
Nymphaeaceae Euryale . . 1 . Kadono & Schneider, 1987
Orchidaceae Appendicula . . . 1 Uphof, 1938; Maheshwari, 1962; Catling, 1990
Bletia . 1 . . Catling, 1990
Bulbophyllum 1 . . 2 Uphof, 1938; Maheshwari, 1962
Calanthe 1 . . . Maheshwari, 1962
Caularthron . . . 1 Pupulin, 1998
Cheirostylis . . . 1 Jones, 1997
Chloraea . . . 1 Uphof, 1938; Maheshwari, 1962
Corallorhiza 2 . . . Catling, 1990; Freudenstein, 1994, 1999

12 THE BOTANICAL REVIEW
Table I (continued )
Family Genus Complete Induced Dimorphic Unclear References
Dendrobium 3 2 Uphof, 1938; Maheshwari, 1962; Catling, 1990
Liparis . 2 . . Uphof, 1938; Catling, 1990
Oberonia . . . 1 Uphof, 1938; Catling, 1990
Plocoglottis 1 . . . Maheshwari, 1962
Polystchya . 1 . . Catling, 1990
Spiranthes 1 . . . Catling, 1990
Thelasis . . . 1 Uphof, 1938; Maheshwari, 1962
Thelymitra 1 . . . Catling, 1990
Orobanchaceae Epifagus . . 1 . Howes, 1999
Oxalidaceae Oxalis . . 6 . Darwin, 1877; Lord, 1981; Jasieniuk & Lechowicz, 1987;
Berg, 2003
Papaveraceae Papaver . 3 . . Uphof, 1938
Ceratocapnos . . 1 . Ruiz de Clavijo & Jimenez, 1993
Corydalis . . 1 . Endress, 1999
Plantaginaceae Plantago 11 . 1 1 Primack, 1978; Lord, 1981
Poaceae Achnatherum . . 1 . Campbell et al., 1983
Aciachne 1 . . . Campbell et al., 1983
Acrachne 1 . . . Campbell et al., 1983; Connor, 1979
Agrostis . . 1 . Campbell et al., 1983; Connor, 1979
Amphibromus . 1 2 . Crozier & Thomas, 1993; Campbell et al., 1983; Cheplick
& Clay, 1989
Amphicarpum . . 2 . Lord, 1981; Cheplick 2005b
Andropogon . . 10 . Campbell, 1982; Campbell et al., 1983
Aristida . . 4 . Lord, 1981; Campbell et al., 1983
Astrebla . . 2 . Campbell et al., 1983
Austrodanthonia . . 1 . Lord, 1981
Avena . . 4 . Uphof, 1938; Lord, 1981; Campbell et al., 1983
Bothriochloa . . 6 . Lord, 1981; Campbell et al., 1983
Bouteloua 2 . 1 . Campbell et al., 1983; Columbus, 1998
Brachyachne 1 . 1 . Campbell et al., 1983
Briza 3 . 11 . Campbell et al., 1983
Bromus . . 8 . Lord, 1981; Campbell et al., 1983; Bartlett et al., 2002
Calamagrostis . . 2 . Campbell et al., 1983

CLEISTOGAMY IN ANGIOSPERMS 13
Calyptochloa . . 1 . Campbell et al., 1983
Catapodium . . 1 . Campbell et al., 1983
Chasmanthium . . 1 . Campbell et al., 1983
Chloris . . 1 . Campbell et al., 1983
Cleistochloa . . 2 . Campbell et al., 1983
Cleistogenes . . 3 . Lord, 1981; Campbell et al., 1983
Cottea . . 1 . Campbell et al., 1983
Dactyloctenium . . 1 . Campbell et al., 1983
Danthonia . . 17 . Lord, 1981; Campbell et al., 1983; Clay, 1983a;
Cheplick & Clay, 1989; Cheplick, 2005b
Deschampsia 7 . 1 . Campbell et al., 1983; Holderegger et al., 2003
Desmazeria . . 1 . Campbell et al., 1983
Dichanthelium . . 19 . Lord, 1981; Campbell et al., 1983; Bell & Quinn, 1985;
Cheplick, 2005b
Dichanthium . . 1 . Campbell et al., 1983
Dichelachne . . 1 . Campbell et al., 1983; Edgar & Connor, 1982
Digitaria . . 5 . Campbell et al., 1983
Dimorphochloa . . 1 . Campbell et al., 1983
Diplachne . . 3 . Campbell et al., 1983
Echinochloa . . 1 . Campbell et al., 1983
Ectrosia . . 4 . Campbell et al., 1983
Ehrharta . . 1 . Lord, 1981
Eleusine . . 1 . Campbell et al., 1983
Enneapogon . . 4 . Campbell et al., 1983
Enteropogon 1 . . . Campbell et al., 1983
Eragrostis 2 . 4 . Campbell et al., 1983; Judziewicz & Peterson, 1990
Eremitis . . 4 . Campbell et al., 1983
Eriachne . . 3 . Campbell et al., 1983
Erianthus . . 1 . Campbell et al., 1983
Erioneuron . . . 1 Campbell et al., 1983
Festuca . 4 2 1 Campbell et al., 1983; Connor, 1998
Garnotia 1 . . . Campbell et al., 1983
Gymnachne 1 . . . Campbell et al., 1983
Gymnopogon . . 2 . Campbell et al., 1983
Habrochloa 1 . . . Campbell et al., 1983
Helichtotrichon . . 1 . Campbell et al., 1983
Heterachne . . 3 . Campbell et al., 1983

14 THE BOTANICAL REVIEW
Table I (continued )
Family Genus Complete Induced Dimorphic Unclear References
Hordeum . . 4 . Campbell et al., 1983
Hypseochloa . . 1 . Campbell et al., 1983
Leersia . . 2 . Lord, 1981; Campbell et al., 1983
Leptochloa 1 . 1 . Campbell et al., 1983
Melica . . 3 . Campbell et al., 1983
Microlaena . . 1 . Lord, 1981
Microstegium . . 1 . Ehrenfeld, 1999; Cheplick, 2005a, 2005b
Muhlenbergia . . 1 . Campbell et al., 1983
Nassella 3 . 2 . Lord, 1981; Campbell et al., 1983
Oryza . . 1 . Kerner von Marilaun, 1902
Panicum . . 2 . Campbell et al., 1983
Pappophorum . . 3 . Campbell et al., 1983
Paspalum . . 1 . Campbell et al., 1983
Pennisetum . . 3 . Campbell et al., 1983
Pheidochloa . . 1 . Campbell et al., 1983
Piptatherum . . 1 . Campbell et al., 1983
Piptochaetium 2 . 7 . Campbell et al., 1983
Poa 1 . 2 . Uphof, 1938; Maheshwari, 1962; Campbell et al., 1983
Puccinellia . . 1 . Campbell et al., 1983
Relchela . . 1 . Campbell et al., 1983
Rottboellia . . 2 . Campbell et al., 1983
Rytidosperma . . 1 . Campbell et al., 1983
Schizachyrium . . 17 . Campbell et al., 1983
Secale . . 1 . Campbell et al., 1983
Setaria . . 1 . Campbell et al., 1983
Sorghum . . 2 1 Campbell et al., 1983; Lazarides et al., 1991
Spathia . . 1 . Campbell et al., 1983
Sporobolus 1 . 10 . Lord, 1981; Campbell et al., 1983; Gleason & Cronquist,
1991
Stipa 2 . 39 . Campbell et al., 1983; Jacobs et al., 1989
Tetrapogon 1 . . . Uphof, 1938; Campbell et al., 1983
Thyridolepis 1 . 2 . Campbell et al., 1983
Tridens 2 . 1 . Campbell et al., 1983

CLEISTOGAMY IN ANGIOSPERMS 15
Triodia . . 1 . Lord, 1981
Triplasis . . 2 . Lord, 1981; Campbell et al., 1983; Cheplick, 1996
Trisetum 1 . 1 . Campbell et al., 1983
Vulpia . . 15 . Lord, 1981; Campbell et al., 1983
Podostemaceae Griffithella 1 . . . Khosla et al., 2001
Polemoniaceae Collomia . . 1 . Lord, 1981
Polygalaceae Polygala . . 2 . Lord, 1981
Polygonaceae Emex . . 1 . Kaul et al., 2000
Polygonum . 3 1 . Kerner von Marilaun, 1902; Lord, 1981
Pontederiaceae Heteranthera . 1 2 . Wylie, 1917; Uphof, 1938; Lord, 1981
Monochoria . . 1 . Lord, 1981
Portulacaceae Portulaca . 2 . . Kim & Carr, 1990
Ranunculaceae Ranunculus 1 3 . . Uphof, 1938; Maheshwari, 1962; Deyuan, 1990
Rosaceae Aremonia . . 1 . Kerner von Marilaun, 1902
Rubiaceae Houstonia . . 2 . Lord, 1981
Relbunium 1 . . . Freitas et al., 1995
Scrophulariaceae Antirrhinum . . . 4 Oyama & Baum, 2004
Glossostigma . . 1 . Beardsley & Olmstead, 2002
Limosella . 1 . . Kerner von Marilaun, 1902
Linaria . 1 . . Maheshwari, 1962
Mimulus . . . 1 Trent, 1939
Neogaerrhinum . . 1 . Lord, 1981
Nuttallanthus . . 1 . Lord, 1981
Sairocarpus . . 2 . Lord, 1981
Scrophularia . . 1 . Kaul et al., 2000
Triphysaria 1 . . . Jamison & Yoder, 2001
Vandellia . . 2 . Lord, 1981
Solanaceae Nicotiana . . . 1 Trent, 1939
Salpiglossis . . 1 . Lord, 1981
Solanum . . 1 . Lord, 1981
Urticaceae Fleurya . . 1 . Kaul et al., 2000
Violaceae Hybanthus . . 1 . Lord, 1981

16 THE BOTANICAL REVIEW
Table I (continued )
Family Genus Complete Induced Dimorphic Unclear References
Viola . 2 78 . Darwin, 1877; Kerner von Marilaun, 1902; Nieuwland,
1914, 1916; Uphof, 1938; Maheshwari, 1962; Beattie,
1969; Munz, 1974; Culver & Beattie, 1978; Lord, 1981;
Wagner et al., 1990; Kim et al., 1991; Masuda & Yahara,
1992; Ballard & Wujek, 1994; Ballard, 1994; Sakai &
Sakai, 1996; Gil-Ad, 1997; Elisafenko, 1998; Culley,
2000; Kaul et al., 2000; Dinc &Yildirimli, 2002; Berg,
2003; Dinc et al., 2003; Marcussen, 2003; Cortes-
Palomec, 2004; Oakley, 2004; Eckstein and Otte, 2005
Total 72 61 536 24

aceae, Orobanchaceae, and Urticaceae) not previously recognized as containing cleis-
togamy (Darwin, 1877; Kerner von Marilaun, 1902; Ritzerow, 1908; Rickett, 1932;
Uphof, 1938; Camp & Gilly, 1943; Maheshwari, 1962; Lord, 1981). We found that
cleistogamy was most often reported within the Poaceae (n=326 species), Violaceae
(80), Fabaceae (61), Orchidaceae (24), and Acanthaceae (19). This result differs slightly
from Darwin (1877), who noted that cleistogamy was most common in the Fabaceae,
Acanthaceae, and the Malpighiaceae. Within genera, cleistogamy was most commonly
reported in Viola (Violaceae; 80 species), Stipa (Poaceae; 41), Dichanthelium (Poaceae;
19), Danthonia (Poaceae; 17), Schizachyrium (Poaceae; 17), Acleisanthes (Nyctagi-
naceae; 16), Vulpia (Poaceae; 15), Briza (Poaceae; 14), Plantago (Plantaginaceae; 13),
and Lespedeza (Fabaceae; 13). In most cases, the type of cleistogamy reported was di-
morphic (536 of 693 species, or 77.3%), similar to Plitmann’s (1995) estimate of 81.7%
(165 of 202 species). Complete cleistogamy (72 species, 10.4%) and induced cleis-
togamy (61 species, 8.8%) were scattered throughout various genera and families (Table
I). Only 24 species (3.5%) that produced cleistogamous flowers could not be assigned
further to a specific category type because of lack of reported information. These counts
are undoubtedly underestimates and are biased toward species and genera that have
been more widely studied (e.g., those in Violaceae or Poaceae).
Evolution of Cleistogamy in Angiosperms
The widespread distribution of cleistogamy within angiosperms suggests that this
breeding system may have evolved repeatedly over time. Selective factors favoring its
evolution may differ, given its occurrence in various plant families and habitats. Both
the number of times that cleistogamy has evolved as well as the theory and potential se-
lective pressures behind such events are reviewed below.
phylogenetic implications
Recent advances in molecular genetics and taxonomy make it possible to estimate the
number of times cleistogamy has evolved within the angiosperms by mapping the trait
onto phylogenetic trees. We did this for three comprehensive, family-based angiosperm
phylogenies derived from 18S rDNA, rbcL and atpB sequences (Soltis et al., 2000),
matK sequences (Hilu et al., 2003), and a supertree generated from 27 source trees from
published and unpublished studies (Soltis et al., 2005). Of the 50 cleistogamous families
identified in our review, 44 were listed in the Soltis et al. (2000) phylogeny and 36 in the
Hilu et al. (2003) phylogeny, while all 50 were found in the Soltis et al. (2005) phy-
logeny. Cleistogamy was widespread across monocot and dicot families, having evolved
multiple times (Fig. 3). Analysis on a family level across the angiosperms suggested that
the breeding system evolved approximately 34 to 41 times (Table II). Cleistogamy
evolved at least six to eight times within monocots, but was much more common in di-
cots, where it evolved at least 24 to 31 different times (Table II).
These values are most likely underestimates because cleistogamy may also have
evolved repeatedly within certain families and genera. For example, complete cleis-
togamy has evolved three separate times within Schiedea (Caryophyllaceae), a Hawai-
ian endemic genus containing diverse breeding systems (Wagner et al., 2005). In the pri-
marily chasmogamous genus Acleisanthes (Nyctaginaceae), the joint production of CL
and CH flowers has evolved at least twice (Levin, 2000, 2002). Alternatively, the ability
CLEISTOGAMY IN ANGIOSPERMS 17

18 THE BOTANICAL REVIEW
Fig. 3. Summary tree for angiosperms, adapted from Soltis et al. (2005), with the monocot section
derived from Angiosperm Phylogeny Group (2003). Orders in italic boldface type contain cleistogamous
species, including all three types of cleistogamy (dimorphic, complete, and induced), as outlined in this
review. Black bars indicate the presence of both CH and CL flowers, gray bars indicate loss of CL flow-
ers, and equivocal decisions are indicated by dashed lines. Only nodes receiving bootstrap or jackknife
support above 50% are indicated. Actual values are shown only for nodes receiving less than 75% sup-
port.

to produce cleistogamous flowers can be lost over time, as is evident in Viola (Ballard et
al., 1999), where species that produce only CH flowers, (e.g. V. pedata, V. maviensis, V.
helenae) are embedded within clades containing species with both flower types (Fig. 4).
Interestingly, most of the seven Hawaiian Viola species appear to possess only CH flow-
ers (Wagner et al., 1990), with one exception, V. kauensis (Ballard et al., 1999). The
most parsimonious explanation is that the ancestral colonist to Hawaii may have been
completely chasmogamous, with dimorphic cleistogamy subsequently evolving in situ.
Alternatively, dimorphic cleistogamy could have been the ancestral condition, with the
loss of CL flowers occurring shortly after radiation to the different islands. Within Viola,
species lacking CL flowers often inhabit relatively stable environments, such as the
shaded understory of tropical mesic forests (V. chamissoniana), or the high-light envi-
ronments of open bogs (V. maviensis) or savannah (V. pedata, V. pedunculata), where in-
sect pollinators are common. The relationship between loss of CL flowers and stable
abiotic environments and/or pollinator activity is an area of promising research for fu-
ture phylogenetic or ecological investigations.
advantages and disadvantages of ch and cl flowers
In order for cleistogamy to evolve, CL flowers must exhibit fitness advantages dis-
tinct from those of CH flowers. A variety of these fitness advantages have already been
identified. First, CL flowers may offer reproductive assurance when pollinators are rare
or absent (Mitchell-Olds & Waller, 1985; but see Cheplick, 2005b). Rather than acting
solely as a fail-safe mechanism for back-up reproduction, CL flowers of some species
may actually increase seed production in the event that CH flowers are left unpollinated
(Redbo-Torstensson & Berg, 1995; Berg & Redbo-Torstensson, 1998). Second, CL
flowers in some species may be energetically less costly to produce, resulting in more
resources available for seed production (Waller, 1984) or for larger CL seeds with
higher fitness. Third, CL flowers possess an inherent automatic selfing advantage be-
cause both maternal sets of genes can be passed on to the progeny (Mitchell-Olds &
Waller, 1985), in contrast to only one set of genes passed on to outcrossed offspring.
Fourth, CL selfing prevents disruption of locally adapted gene complexes by avoiding
the recombination that frequently accompanies outcrossing (Waller, 1984). In addition,
CL seeds do not disperse very far in some species, further preserving locally adapted
CLEISTOGAMY IN ANGIOSPERMS 19
Table II
Estimate of the number of times cleistogamy has evolved within the angiosperms,
based on three phylogenies generated by Soltis et al. (2000), Hilu et al. (2003),
and Soltis et al. (2005). Shown are the total number of angiosperm families within
each phylogeny that was identified in the current review as containing cleistogamy
Soltis et al. Hilu et al. Soltis et al.
(2000) (2003) (2005)
Number of families 44 36 50
Eumagnolids 2 2 2
Monocots 8 6 8
Dicots 24 26 31
Total 34 34 41

complexes but also potentially leading to sibling competition (Schmitt et al., 1985).
Fifth, consistent selfing through CL flowers can eliminate deleterious recessive alleles
within populations (Clay & Antonovics, 1985); over time, this could lead to a decrease
in the level of inbreeding depression, especially in those species with complete CL. Dis-
advantages to CL reproduction include decreased genetic variation and increased ge-
netic drift, high levels of inbreeding depression (if caused by expression of several dele-
terious alleles of small effect that cannot be purged), and increased sibling competition
among CL seeds that are dispersed within the immediate vicinity of the maternal plant.
20 THE BOTANICAL REVIEW
Fig. 4. Phylogeny of Viola Section Melanium (V. arvensis and V. calcarata) and Section Viola (all
other species) showing the presence of both CH and CL flowers (black bars) and loss of cleistogamy
(gray bars). Equivocal decisions are indicated by the dashed lines. Species listed in bold lack CL flowers.
The figure is adapted from the strict consensus cladogram of Ballard et al. (1999) and retains branches
with at least 50% bootstrap or jackknife support.

In order for CH flowers to be maintained in species with dimorphic cleistogamy, they
must offer a selective advantage different from that of CL flowers. First, CH seeds may
exhibit heterosis (Mitchell-Olds & Waller, 1985; Schmitt & Gamble, 1990) if they result
from outcrossing events between genetically distinct individuals. Second, genetically
variable progeny produced by CH flowers (Antlfinger, 1986) would be favored in spa-
tially or temporally heterogeneous habitats (Mitchell-Olds & Waller, 1985; Holsinger,
1986). Finally, CH seeds are often dispersed farther from the maternal plant, thus avoid-
ing sibling competition (Schmitt et al., 1985). Disadvantages to CH flower production
include high energetic cost of production in some species as well as pollinator reliance
for fertilization.
selection for cleistogamy
The evolutionary pathway to cleistogamy remains relatively understudied compared
with more widely investigated breeding systems such as dioecy (e.g., Charlesworth &
Charlesworth, 1978). A number of different ideas, however, have been proposed to ac-
count for the production of one or both flower types within populations, and these theo-
ries consolidate several of the ideas presented above (see also Oakley et al., in prep).
Variable Environments
Two related sets of models (Lloyd, 1984; Schoen & Lloyd, 1984) demonstrate that
dimorphic cleistogamy can be selected for in two types of heterogeneous environments
based on the relative fitness and cost of CH and CL flowers. First, environments can
vary simultaneously within a given plant (the fine-grained environment of Lloyd, 1984
and Schoen & Lloyd, 1984), as when flowers at different spatial positions experience
varying numbers of pollinator visits or levels of herbivory. In this case, a phenotype that
produces both CH and CL flowers at a single time would exhibit the highest fitness if
each flower type is produced in the environment for which it is best suited. More often,
however, environments vary across an area or season (a coarse-grained environment;
Lloyd, 1984; Schoen & Lloyd, 1984), and the fitness of a phenotype would be opti-
mized by the production of a different flower type within each environment to maximize
reproductive success.
Flowering phenology in several angiosperm species is consistent with these models.
In Impatiens, CH flowers are typically produced in areas of high light intensity, as along
edges of populations (Schemske 1984), while CL flowers are produced in more shaded
areas (Schemske 1978). CL seed production in Impatiens is also higher in more stressful
(Bennington and McGraw, 1995) or unpredictable environments (Waller, 1979) because
CL flowers may be less costly, even though their fitness may still be considerably less
than that of CH flowers. Fitness may also increase for CH flowers produced when polli-
nators are most active, while less costly CL flowers only appear when pollinators are
less frequent and as CH flowers decline in fertility. Thus, the ability of plants to assess
and respond appropriately in a heterogeneous environment can lead to selection for a
phenotype that produces both flower types, either sequentially or simultaneously, de-
pending upon the nature of the environment (Lloyd, 1984). If a plant cannot properly
assess the environment, the least costly flower type would be favored (i.e., complete
cleistogamy in most species) as long as the fitness benefits outweigh the cost, assuming
no inbreeding depression (see below). These environmental effects were also invoked by
CLEISTOGAMY IN ANGIOSPERMS 21

Darwin (1877), who suggested that dimorphic and complete cleistogamy may have
evolved from induced cleistogamy as environmental conditions became more pre-
dictable and plants were able to track environmental changes.
Inbreeding Depression and Geitonogamy
The role of inbreeding depression has been invoked in several evolutionary models.
Masuda et al. (2001) expanded on Schoen and Lloyd’s (1984) multiple strategy model
by incorporating inbreeding depression and CH geitonogamy. Their results reinforce the
importance of CH fertility in CL flower production; they showed that both flower types
will be favored when some geitonogamy occurs in CH flowers and inbreeding depres-
sion is severe. Similarly, Stewart (1994) suggested that increased allocation to CH re-
production may increase geitonogamy, resulting in diminished fitness gains because
cross-fertilization is less likely. This could constrain selection for greater CH reproduc-
tion, while inbreeding depression would constrain selection for greater CL reproduction.
Consequently, the dual production of CH and CL flowers may be an evolutionarily
stable strategy even if the level of inbreeding depression is substantial, as long as there
is a positive correlation between geitonogamy and CH flower production. The level of
inbreeding depression, however, has generally been low in dimorphic cleistogamous
plant species, ranging from 0–0.10 (Wilken, 1982; Culley, 2000; Lu, 2002; Eckstein and
Otte 2005). Inbreeding depression may be reduced in these populations because of a
past history of CL selfing and purging of deleterious alleles. Even if CH progeny exhibit
a high overall level of inbreeding, CH flowers may still be favored because they serve to
maintain gene flow and thus significantly reduce gene fixation in populations (Knight
and Waller, 1987).
Differential Seed Dispersal
CH and CL seeds of many angiosperms, especially grasses, exhibit heteromorphism
with discrete seed types often produced concurrently on individual plants (e.g., Chep-
lick & Clay, 1989; Cheplick, 1996). These seeds may be differentially dispersed, with
CL seeds distributed around the natal site and CH seeds dispersed further away (Schmitt
et al., 1985). Such differential seed dispersal may explain in part the maintenance of di-
morphic cleistogamy. According to Holsinger (1986), intermediate selfing rates will be
stable if less-fit selfed progeny are less successful migrants than outcrossed progeny, es-
pecially in areas with spatial environmental heterogeneity (Schmitt et al., 1985) in
which local adaptation occurs. Although CH progeny are not always outcrossed, these
ideas can be expanded to explain the maintenance of dimorphic cleistogamy in cases
where CH progeny are primarily outcrossed.
In such species, dimorphic or induced cleistogamy will be maintained within popula-
tions if the dispersal of selfed CL seeds locally and outcrossed CH seeds more distantly
results in higher fitness gains for both seed types. CL seeds would have highest fitness
in the maternal environment, since both sets of chromosomes come from the maternal
plant, which may already be adapted to its local environment. In contrast, recombination
and the possible appearance of new alleles in outcrossed CH progeny would disrupt lo-
cally adapted gene complexes, resulting in depressed CH fitness within the maternal en-
vironment. Hence, different allelic combinations in CH seeds may favor adaptation to a
wider array of heterogeneous habitats, as well as reduced sibling competition. In one of
22 THE BOTANICAL REVIEW

the few investigations of differential seed dispersal, CL progeny of Impatiens capensis
exhibited higher fitness within the maternal location than they did 12 m away, but, para-
doxically, CH flowers also had higher fitness at the maternal site as well (Schmitt &
Gamble, 1990). Considered within a metapopulational context, differential dispersal of
CH seeds may also serve to increase the distribution of a species, while CL seeds serve
to ensure populational survival within a given site (Koller & Roth, 1964).
Variable Ecological Factors and Plant Size
Cheplick (2005b) suggested that variation in the expression of cleistogamy in some
species may reflect the influence of environmental factors such as density, soil moisture
and nutrition, light levels, and plant size. In a review of several grass species, he found
that the CH/CL ratio declined with increasing density in Amphicarpum purshii because
plant size was reduced by intraspecific competition. In Microstegium vimineum,the
CH/CL ratio was lowest in greenhouse plants exposed to a sunny environment but was
highest in small plants from the shady interior. In contrast, the CH/CL ratio of Dichan-
thelium clandestinum was not affected by light levels (Cheplick, 2005b), although an
earlier study found significant effects among different populations (Bell & Quinn,
1987). Relative amounts of CH and CL flowering is also known to vary in other an-
giosperm species in relation to day length and temperature (Evans, 1956; Mayers &
Lord, 1983a) or to light and nutrient availability (Le Corff, 1993).
Plant size also affects reproductive output of certain species exhibiting dimorphic
cleistogamy. In some taxa, individual plants must attain a certain size before reproduc-
tion of any type can occur, as in Impatiens capensis (Waller, 1979). In general, plant
size affects either CH reproduction or a combination of CH and CL reproduction. For
example, greater plant size was associated with an increase in percentage chasmogamy
in Mimulus nasutus (Diaz & Macnair, 1999), and only larger plants produced CH flow-
ers in Viola sororia (Solbrig, 1981) and Danthonia spicata (Clay, 1982). In the grass
Amphicarpaea bracteata, light intensity influenced plant size, which was associated
with increased aerial CH flower production (Trapp & Hendrix, 1988). Greater plant size
was related to both CH and CL flower production in Oxalis montana, but size explained
a significantly greater proportion of variation in CL flower number (Jasieniuk & Le-
chowicz, 1987). CH and CL flower production also increased substantially with plant
weight and density in Collomia grandiflora (Wilken, 1982). Although plant size is gen-
erally less likely to directly affect CL reproduction, more cleistogamous plant species
need to be examined in future studies.
Implications for Future Research
Continued research on cleistogamous plant species will offer further insight into the
evolutionary history and prevalence of this breeding system in angiosperms. Cleis-
togamy will undoubtedly be discovered in additional species as the reproductive biology
of more species is examined in the future. New discoveries of cleistogamy are espe-
cially likely in families related to those in which cleistogamy has already been docu-
mented (Table I), as in the Surianaceae which is sister to the Polygalaceae and Fabaceae
(Soltis et al., 2000). As more cleistogamous species are identified, they can be used to
test evolutionary theories and origins of cleistogamous taxa, especially as phylogeny
resolutions increase with greater taxonomic sampling. Empirical data are also needed to
CLEISTOGAMY IN ANGIOSPERMS 23

test evolutionary theories previously developed (Lloyd, 1984; Schoen & Lloyd, 1984;
Masuda et al., 2001). This information will be invaluable for understanding the selective
pressures and factors favoring the evolution of cleistogamy as well as the evolutionary
loss of this breeding system, a subject that has received little attention to date.
Although the expression of cleistogamy is affected by the environment in many
species, for most species, the necessary reproductive and ecological information is often
not reported. For example, variation of floral types within and between individuals and
populations needs to be quantified. In many species, there is little indication of the de-
gree to which the expression of cleistogamy is a phenotypic trait influenced by the envi-
ronment (but see Cheplick, 2005b), or how it is affected by the interaction of genetic
and environmental factors. Investigations would also greatly benefit from the applica-
tion of developmental and genetic techniques to identify the genes responsible for the
production of CH and CL flowers in different species.
We encourage future investigators to classify the cleistogamous breeding system
using the three primary categories as outlined in this paper to facilitate future compar-
isons among taxa. In particular, reports of cleistogamy in the literature need to include a
detailed description of the reproductive ecology and developmental biology of the vari-
ous floral types, preferably in multiple populations and over several seasons. Ultimately,
by further investigating the role of the environment and its influence on floral gene ex-
pression, we may come to better understand the factors directly influencing the transi-
tion from CH to CL flower production and the evolution of cleistogamy in angiosperms.
Acknowledgments
We thank Timothy Bell, who kindly shared his original notes from his cleistogamy
survey of the grasses, and Harvey Ballard and his group at Ohio University for their
suggestions. Anne Wick and MaryAnn Paul assisted in portions of the JSTOR literature
review for Figure 1, and the librarians at the Chemistry-Biology library at the University
of Cincinnati and the Lloyd Library in Cincinnati were indispensable in locating certain
references. Harvey Ballard kindly supplied the information for Figure 4. Valuable com-
ments on the manuscript and data set were provided by Timothy Bell and Gregory
Cheplick. This study was made possible through funding from the University Research
Council at the University of Cincinnati.
Literature Cited
Angiosperm Phylogeny Group. 2003. An update of the Angiosperm Phylogeny Group classification for
the orders and families of flowering plants: APG II. Bot. J. Linn. Soc. 141: 399–436.
Antlfinger, A. E. 1986. Field germination and seedling growth of CH and CL progeny of Impatiens
capensis (Balsaminaceae). Amer. J. Bot. 73:1267–1273.
Ballard, H. E. 1994. Violets of Michigan. Michigan Botanist 33:131–199.
——— & D. E. Wujek. 1994. Evidence for the recognition of Viola appalachiensis. Syst. Bot. 19(4):
523–538.
———, K. J. Sytsma & R. R. Kowal. 1999. Shrinking the violets: phylogenetic relationships of infra-
generic groups in Viola (Violaceae) based on internal transcribed spacer DNA sequences. Syst.
Bot. 23: 439–458.
Bartlett, E., S. J. Novak & R. N. Mack. 2002. Genetic variation in Bromus tectorum (Poaceae): differ-
entiation in the eastern United States. Amer. J. Bot. 89(4): 602–612.
24 THE BOTANICAL REVIEW

Beardsley, P. M. & R. G. Olmstead. 2002. Redefining Phrymaceae: the placement of Mimulus, tribe
Mimuleae, and Phryma. Amer. J. Bot. 89(7): 1093–1102.
Beattie, A. J. 1969. Studies in the pollination ecology of Viola. I. The pollen-content of stigmatic cavi-
ties. Watsonia 7:142–156.
Bell, T. J. & J. A. Quinn. 1985. Relative importance of chasmogamously and cleistogamously derived
seeds of Dichanthelium clandestinum (L.) Gould. Bot. Gaz. 146(2): 252–258.
——— &———.1987. Effects of soil moisture and light intensity on the chasmogamous and cleis-
togamous components of reproductive effort of Dichanthelium clandestinum populations. Canad. J.
Bot. 65: 2243–2249.
Bennington, C. C. & J.B. McGraw. 1995. Natural selection and ecotypic differentiation in Impatiens
pallida. Ecol. Monogr. 65: 303–323.
Berg, H. 2003. Factors influencing seed: ovule ratios and reproductive success in four cleistogamous
species: a comparison between two flower types. Pl. Biol. 5: 194–202.
——— & P. Redbo-Torstensson. 1998. Cleistogamy as a bet-hedging strategy in Oxalis acetosella,a
perennial herb. J. Ecol. 86: 491–500.
Calvino, A. & L. Galetto. 2003. Cleistogamy in the rare high Andean perennial herb Cryptantha capit-
uliflora (Boraginaceae). Pl. Syst. Evol. 237(1-2): 41–50.
Camp, W. H. & C. L. Gilly. 1943. The structure and origin of species with a discussion of intraspecific
variability and related nomenclatural problems. Brittonia 4: 323–385.
Campbell, C. S. 1982. Cleistogamy in Andropogon L. (Gramineae). Amer. J. Bot. 69(10): 1625–1635.
———,J. A. Quinn, G. P. Cheplick & T. J. Bell. 1983. Cleistogamy in grasses. Annual. Rev. Ecol.
Syst. 14: 411–441.
Catling, P. M. 1990. Auto-pollination in the Orchidaceae. Pp. 121–158 in J. Arditti (ed.), Orchid biol-
ogy, reviews and perspectives. Timber Press, Portland, Oregon.
Charlesworth, B. & D. Charlesworth. 1978. A model for the evolution of dioecy and gynodioecy.
Amer. Naturalist 112: 975–997.
Cheplick, G. P. 1996. Cleistogamy and seed heteromorphism in Triplasis purpurea (Poaceae). Bull. Tor-
rey Bot. Club 123: 25–33.
———. 2005a. Biomass partitioning and reproductive allocation in the invasive, cleistogamous grass
Microstegium vimineum: influence of the light environment. J. Torrey Bot. Soc. 132: 214–224.
———. 2005b. Plasticity of chasmogamous and cleistogamous reproductive allocation in grasses. Pp.
145–155 in J. T. Columbus, E. A. Friar, C. W. Hamilton, J M. Porter, L. M. Price & M. G. Simpson
(eds.), Monocots: comparative biology and evolution. Rancho Santa Ana Botanical Garden, Clare-
mont, California.
——— & K. Clay. 1989. Convergent evolution of cleistogamy and seed heteromorphism in two peren-
nial grasses. Evol. Trends Pl. 3: 127–136.
Clay, K. 1982. Environmental and genetic determinants of cleistogamy in a natural population of the
grass Danthonia spicata. Evolution 36: 734–741.
———. 1983a. Variation in the degree of cleistogamy within and among species of the grass Dantho-
nia. Amer. J. Bot. 70(6): 835–843.
———. 1983b. The differential establishment of seedlings from chasmogamous and cleistogamous
flowers in natural populations of the grass Danthonia spicata (L.) Beauv. Oecologia 57: 183–188.
——— & J. Antonovics. 1985. Demographic genetics of the grass Danthonia spicata: success of prog-
eny from chasmogamous and cleistogamous flowers. Evolution 39: 205–210.
Cole, C. T. & D. D. Biesboer. 1992. Monomorphism, reduced gene flow, and cleistogamy in rare and
common species of Lespedeza (Fabaceae). Amer. J. Bot. 79: 567–575.
Columbus, T. J. 1998. Morphology and leaf blade anatomy suggest a close relationship between
Bouteloua aristidoides and B. (Chondrosium)eriopoda (Gramineae: Chloroideae). Syst. Bot.
23(4): 467–478.
Connor H. E. 1979. Breeding systems in the grasses: a survey. New Zealand J. Bot. 17: 547–574.
———. 1998. Breeding systems in New Zealand grasses. XII. Cleistogamy in Festuca. New Zealand J.
Bot. 36: 471–476.
Cortes-Palomec, A. C. 2005. Ecological factors, mixed breeding system and population genetic struc-
ture in a subtropical and a temperate violet species. Ph.D. diss., Ohio Univ.,Athens, Ohio.
CLEISTOGAMY IN ANGIOSPERMS 25

Crozier, T. S. & J. F. Thomas. 1993. Normal floral ontogeny and cool temperature-induced aberrant flo-
ral development in Glycine max (Fabaceae). Amer. J. Bot. 80(4): 429–448.
Cruden, R. W. 1977. Pollen-ovule ratios: a conservative indicator of breeding systems in flowering
plants. Evolution 31(1): 32–46.
Culley, T. M. 2000. Inbreeding depression and floral type differences in Viola canadensis (Violaceae), a
perennial herb with chasmogamous and cleistogamous flowers. Canad. J. Bot. 78(11): 1420–1429.
———. 2002. Reproductive biology and delayed selfing in Viola pubescens (Violaceae), an understory
herb with chasmogamous and cleistogamous flowers. Int. J. Pl. Sci. 163: 113–122.
Culver, D. C. & A. J. Beattie. 1978. Myrmecochory in Viola: dynamics of seed-ant interactions in some
West Virginia species. J. Ecol. 66: 53–72.
Darwin, C. 1877. The different forms of flowers on plants of the same species. John Murray, London.
Reprint, 1986. University of Chicago Press, Chicago, Illinois.
Delgado-Salinas, A. 2000. New species of Mexican Phaseolus (Fabaceae). Syst. Bot. 25(3): 414–436.
De-yuan, H. 1990. A biosystematic study on Ranunculus subgenus Batrachium in S. Sweden. Nordic J.
Bot. 11: 41–59.
Diaz, A, & M. R. Macnair. 1999. Pollen tube competition as a mechanism of prezygotic reproductive
isolation between Mimulus nasutus and its presumed progenitor M. buttatus. New Phytol. 144:
471–478.
Dinc, M. & S. Yildirimli. 2002. A new species of Viola (Violaceae) from Turkey. Bot. J. Linn. Soc. 138:
483–487.
———, Y. Bag˘ci & S. Yildirimli. 2003. A new species of Viola L. (Violaceae) from South Anatolia.
Bot. J. Linn. Soc. 141: 477–482.
Drewes, S. I. & P. S. Hoc. 2000. Morphology and development of cleistogamous flowers in Macroptil-
ium fraternum (Fabaceae). Kurtziana 28(2): 229–238.
Eckstein, R. L. & A. Otte. 2005. Effects of cleistogamy and pollen source on seed production and off-
spring performance in three endangered violets. Basic Appl. Ecol. 6: 339–350.
Edgar, E. & H. E. Connor. 1982. Dichelachne (Gramineae) in New Zealand. New Zealand J. Bot. 20:
303–309.
Ehrenfeld, J. G. 1999. A rhizomatous, perennial form of Microstegium vimineum (Trin.) A. Camus in
New Jersey. J. Torrey Bot. Soc. 126(4): 352–358.
Elisafenko, T. V. 1998. The two types of flowering in the rare Siberian species of Viola (Violaceae). Bot.
Zhurn. (St. Petersburg) 83(6): 66–72.
Endress, P. K. 1999. Symmetry in flowers: diversity and evolution. Int. J. Pl. Sci. 160 (6 Suppl.): S3–
S23.
Evans, L. T. 1956. Chasmogamous flowering in Viola palustris L. Nature 178: 1301.
Freitas, B. M., J. E. Alves, G. F. Brandão & Z. B. Araujo. 1995. Pollination requirements of West In-
dian cherry (Malpighia emarginata) and its putative pollinators, Centris bees, in NE Brazil. J.
Agric. Sci.133: 303–311.
Freudenstein, J. V. 1994. Character transformation and relationships in Corallorhiza (Orchidaceae:
Epidendroideae). II. Morphological variation and phylogenetic analysis. Amer. J. Bot. 81: 1458–
1467.
———. 1999. A new species of Corallorhiza (Orchidaceae) from West Virginia, U.S.A. Novon 9: 511–
513.
Gallardo, R., E. Dominguez & J. M. Muñoz. 1993. The heterochronic origin of the cleistogamous
flower in Astragalus cymbicarpos (Fabaceae). Amer. J. Bot. 80(7): 814–823.
Gardner, C. S. 1982. Systematic study of Tillandsia subgenus Tillandsia. Ph.D. diss., Texas A&M
Univ., College Station, Texas.
Gil-Ad, N. L. 1997. Systematics of Viola subsection Boreali-Americanae. Boissiera 53: 1–130.
Gilmartin, A. J. & G. K. Brown. 1985. Cleistogamy in Tillandsia capillaries (Bromeliaceae). Biotrop-
ica 17(3): 256–259.
Gleason, H. A. & A. Cronquist. 1991. Manual of vascular plants of northeastern United States and ad-
jacent Canada. Ed. 2. New York Botanical Garden, New York.
Gomez, S. M. & A. Kalamani. 2003. Butterfly pea (Clitoria teratea): a nutritive multipurpose forage
legume for the tropics—an overview. Pakistan J. Nutr. 2(6): 374–379.
26 THE BOTANICAL REVIEW

Gopinathan, M. C. & C. R. Babu. 1987. Breeding systems and pollination in Vigna minima (Legumi-
nosae, Papilionoideae). Pl. Syst. Evol. 156: 117–126.
Grau, J. 1981. Two new species of the genus Cryptantha (Boraginaceae) and their taxonomic position.
Mitt. Bot. Staatssamml. München 17: 511–526.
Hackel, E. 1906. Über Cleristogamie bei den Grasern. Oesterr. Bot. Z. 56: 81–88.
Hardy, C. R. & R. B. Faden. 2004. Plowmanianthus, a new genus of Commelinaceae with five new
species from tropical America. Syst. Bot. 29: 316–333.
Herrera, J. 1992. Flowering variation and breeding systems in the Cistaceae. Pl. Syst. Evol. 179(3-4):
245–255.
Hilu, K. W., T. Borsch, K. Müller, D. E. Soltis, P. S. Soltis, V. Savolainen, M. W. Chase, M. P. Pow-
ell, L. A. Alice, R. Evans, H. Sauquet, C. Neinhuis, T. A. B. Slotta, J. G. Rohwer, C. S. Camp-
bell & L. W. Chatrou. 2003. Angiosperm phylogeny based on matK sequence information. Amer.
J. Bot. 90: 1758–1776.
Holderegger, R., I. Stehlik, R. I. Lewis Smith & R. J. Abbott. 2003. Populations of Antarctic hair-
grass (Deschampsia antarctica) show low genetic diversity. Arctic Antarc. Alpine Res. 35(2): 214–
217.
Holsinger, K. E. 1986. Dispersal and plant mating systems: the evolution of self-fertilization in subdi-
vided populations. Evolution 40: 405–413.
Howes, M. 1999. Beechdrops. Newslett. Ohio Div. Nat. Areas Preserves 21(1): 5.
Jacobs, S. W. L., J. Everett, H. E. Connor & E. Edgar. 1989. Stipoid grasses in New Zealand. New
Zealand J. Bot. 27: 569–582.
Jacquemart, A. L. & J. R. Desloover. 1992. The part taken by cleistogamy in Narthecium ossifragum
reproductive strategy. Flora 187 (1–2): 67–72.
Jamison, D. S. & J. I. Yoder. 2001. Heritable variation in quinine-induced haustorium development in
the parasitic plant Triphysaria. Pl. Physiol. 125: 1870–1880.
Jasieniuk, M. & M. J. Lechowicz. 1987. Spatial and temporal variation in chasmogamy and cleis-
togamy in Oxalis montana (Oxalidaceae). Amer. J. Bot. 74:1672–1680.
Jiang, M. & Y. Kadono. 2001. Growth and reproduction characteristics of an aquatic macrophyte Ot-
telia alismoides (L.) Pers. (Hydrocharitaceae). Ecol. Res. 16(4): 687–695.
Jones, D. L. 1997. A taxonomic revision of Cheirostylis (Orchidaceae) in Australia. Muelleria 10: 75–
83.
Judziewicz, E. J. & P. M. Peterson. 1990. A new cleistogamous South American species of Eragrostis
(Poaceae Chloridoideae). Brittonia 42: 47–49.
Kadono, Y. & E. L. Schneider. 1987. The life history of the Euryale ferox Salisb. in southwestern Japan
with special reference to reproductive ecology. Pl. Spec. Biol. 2: 109–115.
Kaul, V., A. K. Koul & M. C. Sharma. 2000. The underground flower. Curr. Sci. 78: 39–44.
Keighery, G. J. 1984. Cleistogamy in Elatine gratioloides A. Cunn. (Elatinaceae). W. Austral. Naturalist
16(1): 22.
———. 1988. Breeding systems of the Western Australian flora. W. Austral. Naturalist 17(5): 121–
125.
Kerner von Marilaun, A. 1902. The natural history of plants. Vol. II. Translated by F.W. Oliver, Blackie
and Son, London.
Khosla, C., K. R. Shivanna & H. Y. Mohan Ram. 1998. Pollination in the aquatic insectivore Utricu-
laria inflexa var. stellaris. Phytomorphology 48(4): 417–425.
———,——— &———. 2001. Cleistogamy in Griffithella hookeriana (Podostemaceae). S. African
J. Bot. 67(2): 320–324.
Kim, I. & G. D. Carr. 1990. Reproductive biology and uniform culture of Portulaca in Hawaii. Pacific
Sci. 44: 123–129.
Kim, K. S., B. Y. Sun, S. S. Whang & G. H. Chung. 1991. Biosystematic study on the genus Viola in
Korea: comparative morphology of the Viola albida complex. Korean J. Bot. 34(3): 229–238.
Knight, S. E. & D. M. Waller. 1987. Genetic consequences of outcrossing in the cleistogamous annual
Impatiens capensis. I Population genetic structure. Evolution 41: 969–978.
Koller, D. & N. Roth. 1964.Studies on the ecological and physiological significance of amphicarpy in
Gymnarrhena micrantha (Compositae). Amer. J. Bot. 51: 26–35.
CLEISTOGAMY IN ANGIOSPERMS 27

Kuhn, M. 1867. Einige Bermerkungen über Vandellia und den Blüten Dimorphismus. Bot. Zeitung 25:
65–67.
Lazarides, M., J. B. Hacker & M. H. Andrew. 1991. Taxonomy, cytology and ecology of indigenous
Australian sorgums (Sorghum Moench: Andropogoneae: Poaceae). Austral. Syst. Bot. 4(4): 591–
635.
Le Corff, J. 1993. Effects of light and nutrient availability on chasmogamy and cleistogamy in an un-
derstory tropical herb, Calathea micans (Marantaceae). Amer. J. Bot. 80:1392–1399.
Lesica, P., R. F. Leary, F. W. Allendorf & D. E. Bilderback. 1988. Lack of genic diversity within and
among populations of an endangered plant, Howellia aquatilis. Conservation Biol. 2: 275–282.
Levin, R. A. 2000. Phylogenetic relationships within Nyctaginaceae tribe Nyctagineae: evidence from
nuclear and chloroplast genomes. Syst. Bot. 25: 738–750.
———. 2002. Taxonomic status of Acleisanthes,Selinocarpus, and Ammocodon (Nyctaginaceae).
Novon 12: 58–63.
Lloyd, D. G. 1984. Variation strategies of plants in heterogeneous environments. Biol. J. Linn. Soc. 21:
357–385.
———. and D. J. Schoen. 1992. Self- and cross-fertilization in plants. I. Functional dimensions. Int. J.
Pl. Sci. 153: 358–369.
Lord, E. M. 1981. Cleistogamy: a tool for the study of floral morphogenesis, function and evolution.
Bot. Rev. 47: 421–449.
——— & L. U. Kohorn. 1986. Gynoecial development, pollination, and the path of pollen tube growth
in the Tepary bean, Phaseolus acutifolius. Amer. J. Bot. 73(1): 70–78.
Lu, Y. 2002. Why is cleistogamy a selected reproductive strategy in Impatiens capensis (Balsami-
naceae)? Biol. J. Linn. Soc. 75: 543–553.
Mayers, A. M. & E. M. Lord. 1983a. Comparative flower development in the cleistogamous species
Viola odorata. I. A growth rate study. Amer. J. Bot. 70: 1548–1555.
——— &———. 1983b. Comparative flower development in the cleistogamous species Viola odor-
ata. II. An organographic study. Amer. J. Bot. 70: 1556–1563.
Maheshwari, J. K. 1962. Cleistogamy in angiosperms. Pp. 145–155 in P. Maheshwari, B. M. Johri & I.
K. Vasil, (eds.), Proceedings of the Summer School of Botany June 2–15, 1960. Ministry of Scien-
tific Research and Cultural Affairs, New Delhi, India.
Marcussen, T. 2003. A new violet species (Violaceae) from the south-west Alps. Bot. J. Linn. Soc. 142:
119–123.
Márquez-Guzmán, J., S. Vázquez-Santana, E. M. Engleman, A. Martínez-Mena & E. Martínez.
Pollen development and fertilization in Lacandonia schismatica (Lacandoniaceae). Ann. Missouri
Bot. Gard. 80: 891–897.
Masuda, M. & T. Yahara. 1994. Reproductive ecology of a cleistogamous annual, Impatiens noli-
tangere L. occurring under different environmental conditions. Ecol. Res. 9: 67–75.
———,——— & M. Maki. 2001. An ESS model for the mixed production of cleistogamous and
chasmogamous flowers in a facultative cleistogamous plant. Evol. Ecol. Res. 3: 429–439.
McCall, C., T. Mitchell-Olds & D. M. Waller. 1989. Fitness consequences of outcrossing in Impatiens
capensis: tests of the frequency-dependent and sib-competition models. Evolution 43: 1075–1084.
Mitchell-Olds, T. & D. M. Waller. 1985. Relative performance of selfed and outcrossed progeny in Im-
patiens capensis. Evolution 39: 533–544.
Morin, N. 1983. Systematics of Githopsis (Campanulaceae). Syst. Bot. 8(4): 436–468.
Moyle, L. C., M. S. Olson & P. Tiffin. 2004. Patterns of reproductive isolation in three angiosperm gen-
era. Evolution 58: 1195–1208.
Munz, P. A. 1974. A flora of Southern California. Univ. of California Press, Berkeley, California.
Nandi, O. I. 1998. Floral development and systematics of Cistaceae. Pl. Syst. Evol. 212(1–2): 107–134.
Nieuwland, J. A. 1914. Notes on cleistogamous flowers of violets. II. Amer. Midl. Naturalist 3(7): 198–
200.
———. 1916. Cleistogamous flowers in the Pansies. Amer. Midl. Naturalist 4(1): 464–466.
Novosyelova, L. V. 1998. The bud cleistogamy of some annual species of the genus Medicago
(Fabaceae). Bot. Zhurn. (St. Petersburg) 83: 82–87.
Oakley, C. G. 2004. Inbreeding depression and mating system evolution in the perennial herb Viola
28 THE BOTANICAL REVIEW

septemloba; and the evolutionary maintenance of cleistogamy. M.S. thesis, Florida State Univ., Tal-
lahassee.
Oakley, C. G., K. S. Moriuchi & A. A. Winn. In prep. Stable mixed mating in cleistogamous species: a
review of the ideas and evidence. Annual Rev. Ecol. Syst.
Ortega-Olivencia, A., J. A. López-Paredes, T. Rodríguez-Riaño & J. A. Devesa. 1998. Modes of self-
pollination and absence of cryptic self-incompatibility in Drosophyllum lusitanicum (Droseraceae).
Bot. Acta 111: 474–480.
Oyama, R. K. & D. A. Baum. 2004. Phylogenetic relationships of North American Antirrhinum (Veron-
icaceae). Amer. J. Bot. 91(6): 918–925.
Pfeifer, H. W. 1966. Revision of the North and Central American hexandrous species of Aristolochia
(Aristolochiaceae). Ann. Missouri Bot. Gard. 53: 116–196.
Plitmann, U. 1995. Distribution of dimorphic flowers as related to other elements of the reproductive
strategy. Pl. Spec. Biol. 10: 53–60.
Porras, R. & J. M. Muñoz Álvarez. 1999. Breeding system in the cleistogamous species Centaurea
melitensis (Asteraceae). Canad. J. Bot. 77:1632–1640.
Primack, R. B. 1978. Evolutionary aspects of wind pollination in the genus Plantago (Plantaginaceae).
New Phytol. 81(2): 449–458.
Pupulin, F. 1998. Orchid florula of Parque Nacional Manuel Antonio. Revista Biol. Trop. 46(4): 961–
10317.
Redbo-Torstensson, P. & H. Berg. 1995. Seasonal cleistogamy: a conditional strategy to provide repro-
ductive assurance. Acta Bot. Neerl. 44: 247–256.
Rickett, H. W. 1932. Distribution of cleistogamy among flowering plants. Typewritten notes available at
the New York Botanical Garden Library.
Ritzerow, H. 1908. Über Bau und Befruchtung kleistogamer Blüten. Bot. Zeitung (Flora) 98: 163–212.
Ruiz de Clavijo, E. 1997. A comparative study of the reproductive biology of two Ajuga species (Lami-
aceae) in the southwest of the Iberian Peninsula. Int. J. Pl. Sci. 158:282–291.
——— & M. J. Jimenez. 1993. Cleistogamy and chasmogamy in Ceratocapnos heterocarpa (Fumari-
aceae). Int. J. Pl. Sci. 154: 325–333.
Sakai, S. & A. Sakai. 1996. Why is there variation in mean seed size among plants within single popu-
lations? Test of the fertilization efficiency hypothesis. Amer. J. Bot. 83(11): 1454–1457.
Schemske, D. W. 1978. Evolution of reproductive characteristics in Impatiens (Balsaminaceae): the sig-
nificance of cleistogamy and chasmogamy. Ecology 59: 596–613.
———. 1984. Population structure and local selection in Impatiens pallida (Balsaminaceae), a selfing
annual. Evolution 38: 817–832.
Schmitt, J. & D. W. Ehrhardt. 1987. A test of the sib-competition hypothesis for outcrossing advantage
in Impatiens capensis. Evolution 41: 579–590.
——— &———. 1990. Enhancement of inbreeding depression by dominance and suppression in Im-
patiens capensis. Evolution 44: 269–278.
——— & S. E. Gamble. 1990. The effect of distance from the parental site on offspring performance
and inbreeding depression in Impatiens capensis: a test of the local adaptation hypothesis. Evolu-
tion 44: 202–230.
———, D. Ehrhardt & D. Swartz. 1985. Differential dispersal of self-fertilized and outcrossed prog-
eny in Jewelweed (Impatiens capensis). Amer. Naturalist 126: 570–575.
Schoen, D. J. & A. H. D. Brown. 1991. Whole- and part-flower self-pollination in Glycine clandestina
and G. argyrea and the evolution of autogamy. Evolution 45:1651–1664.
——— & D. G. Lloyd. 1984. The selection of cleistogamy and heteromorphic diaspores. Biol. J. Linn.
Soc. 23: 303–322.
Sigrist, M. R. & M. Sazima. 2002. Ruellia brevifolia (Pohl) Ezcurra (Acanthaceae): flowering phenol-
ogy, pollination biology and reproduction. Revista Brasil. Bot. 25: 35–42.
Simpson, R. L., M. A. Leck & V. T. Parker. 1985. The comparative ecology of Impatiens capensis
Meerb. (Balsaminaceae) in central New Jersey. Bull. Torrey Bot. Club 112: 295–311.
Solbrig, O. T. 1981. Studies on the population biology of the genus Viola. II. Effects of plant size on fit-
ness in Viola sororia. Evolution 35: 1080–1093.
Soltis, D. E., P. S. Soltis, M. W. Chase, M. E. Mort, D. C. Albach, M. Zanis, V. Savolainen, W. H.
CLEISTOGAMY IN ANGIOSPERMS 29

Hahn, S. B. Hoot, M. F. Fay, M. Axtell, S. M. Swensen, L. M. Prince, W. J. Kress, K. C. Nixon
& J. S. Farris. 2000. Angiosperm phylogeny inferred from 18S rDNA, rbcL, and atpB sequences.
Bot. J. Linn. Soc. 133: 381–461.
———,———,P. K. Endress & M.W. Chase. 2005. Phylogeny and evolution of angiosperms. Sin-
auer Associates, Sunderland, Massachusetts.
Stewart, S. C. 1994. Genetic constraints on mating system evolution in the cleistogamous annual Impa-
tiens pallida: inbreeding in chasmogamous flowers. Heredity 73:265–274.
Sun, M. 1999. Cleistogamy in Scutellaria indica (Labiatae): effective mating system and population ge-
netic structure. Molec. Ecol. 8: 1285–1295.
Takahashi, R., H. Kurosaki., S. Yumoto, O. K. Han & J. Abe. 2001. Genetic and linkage analysis of
cleistogamy in Soybean. J. Heredity 92(1): 89–92.
Trapp, E. J. 1988. Dispersal of heteromorphic seeds in Amphicarpaea bracteata (Fabaceae). Amer. J.
Bot. 75: 1535–1539.
——— & S. D. Hendrix. 1988. Consequences of a mixed reproductive system in the hog peanut, Am-
phicarpaea bracteata (Fabaceae). Oecologia 75: 285–290.
Trent, J. A. 1939. Studies pertaining to the life history of Specularia perfoliata (L.) A. DC., with special
reference to cleistogamy. Ph.D. diss., Ohio State University, Columbus.
Uphof, J. C. Th. 1938. Cleistogamic flowers. Bot. Rev. 4: 21–49.
Vander Kloet, S. P. 1993. Biosystematic studies of Vaccinium section Macropelma (Ericaceae) in
Hawaii. Pacific Sci. 47(1): 76–85.
Van Wyk, B.-E. 1990. Studies in the genus Lotononis (Crotalaria, Fabaceae). 13. Two new species and
notes on the occurrence of cleistogamy in the section Leptis. Bothalia 20(1): 17–22.
Wagner, W. L., D. R. Herbst & S. H. Sohmer. 1990. Manual of the flowering plants of Hawai’i. Uni-
versity of Hawaii Press, Honolulu.
———, S. G. Weller & A. Sakai. 2005. Monograph of Schiedea (Caryophyllaceae-Alsinoideae). Syst.
Bot. Monogr. 72: 1–169.
Waller, D. M. 1979. The relative costs of self- and cross-fertilized seeds in Impatiens capensis (Balsam-
inaceae). Amer. J. Bot. 66: 313–320.
———. 1984. Differences in fitness between seedlings derived from cleistogamous and chasmogamous
flowers in Impatiens capensis. Evolution 38: 427–440.
Watanabe, K., T. Yahara & H. Kadota. 1992. Natural hybrid populations between chasmogamous and
cleistogamous species, Ainsliaea fauieana and Ainsliaea apiculata (Asteraceae; Mutisiae): mor-
phology, cytology, reproductive mode and allozyme variation. Pl. Spec. Biol. 7(1): 49–59.
Wilken, D. H. 1982. The balance between chasmogamy and cleistogamy in Collomia grandiflora (Pole-
moniaceae). Amer. J. Bot. 69: 1326–1333.
Wylie, R. B. 1917. Cleistogamy in Heteranthera dubia. Stud. Nat. Hist. Iowa Univ. 7: 48–58.
Yamamoto, I. & Y. Kadono. 1990. A study of the reproductive biology of aquatic Utricularia species in
Southwestern Japan. Acta Phytotax. Geobot. 41(4-6): 189–200.
30 THE BOTANICAL REVIEW