Content uploaded by Patricia Hoc
Author content
All content in this area was uploaded by Patricia Hoc on Jan 28, 2016
Content may be subject to copyright.
In: Pollination: Mechanisms, Ecology and Agricultural Advances ISBN 978-1-61209-634-6
Editors: Nichole D. Raskin and Patrick T. Vuturro, pp. © 2011 Nova Science Publishers, Inc.
Chapter 2
POLLINATION MECHANISMS IN
PASSIFLORA SPECIES: THE COMMON
AND THE RARE FLOWERS -
ECOLOGICAL ASPECTS AND
IMPLICATIONS FOR
HORTICULTURE
M. T. Amela García and P. S. Hoc
Depto. de Biodiversidad y Biología Experimental,
Facultad de Ciencias Exactas y Naturales,
Universidad de Buenos Aires,
PROPLAME-PHRIDEP (CONICET),
Argentina
ABSTRACT
Passionvines have flowers with the following basic architecture: 5 sepals, 5 petals, a
corona formed by concentric cycles (radii, pali, operculum, limen) and an
androgynophore that bears 5 anthers, the ovary, 3 styles and 3 stigmas.
Self-pollination may be achieved but some species are self-incompatible, so pollen
vectors are required. The different relative sizes and orientation of the floral pieces of the
various Passiflora species have implications on which visitors will pollinate; to perform
pollination, they must have the adequate size to contact both anthers and stigmas in the
same or in successive visits to different flowers. Pollen removal (from the anthers) and
deposition (in the stigmas) is carried out by means of different parts of the body of the
different pollinators, depending on their size and behaviour.
The anthers are dehiscent and the stigmas are receptive as soon as the flower opens
until it closes.
The styles move throughout anthesis: they tilt down to the anthers and uplift
afterwards. Thus, three floral stages occur: in the first and the third, only the anthers can
be contacted by the legitimate visitors while in the second, both the anthers and the
stigmas are placed in the way of the pollinators.
M. T. Amela García and P. S. Hoc
2
The style movements succeed in all the studied species. However, in some species,
in a proportion of the flowers the styles remain upright since the flowers open. These
flowers are not able to receive pollen, neither by the pollinators nor by themselves, so
they are functionally staminate. In fewer species, the dehiscence of the anthers does not
happen in some flowers, so they are functionally pistillate. Finally, the three types of
flowers may coexist in the same plant. This brings about the simultaneous occurrence of
pollen donor-receptor flowers and only pollen donors, pollen donor-receptor and only
receptor flowers or the three types of flowers in a single plant, respectively, leading to the
corresponding functionally andromonoecious, gynomonoecious or trimonoecious
systems. Certain floral traits seem to be associated with the absence of styles movements,
such as a less developed gynoecium, minor-sized and nectarless flowers.
In this chapter, an update of the recorded aspects at the moment as well as original
data are discussed, taking in account the ecological interpretations of style movements,
analysing the possible causes of the incidence of the less frequent flowers and
considering the implications for fruit production in this edible fruited genus, some species
of which are grown commercially.
INTRODUCTION
Passionvines have hermaphrodite flowers with the following basic architecture (Figure
1): 5 sepals, 5 petals, a corona formed by different kinds of pieces arranged in various cycles
(radii, pali, operculum, limen) and 5 anthers and a gynoecium with 3 styles and stigmas that
stand over an androgynophore. The architectural differences among Passiflora species
flowers lie on the total flower size and on the relative sizes of each piece (especially the
androgynophore length and the corona pieces) and the floral pieces orientation (especially the
radii and anthers), which have implications in the kind (involving size and behaviour) of
visitors that will carry out pollination.
Other floral traits that vary among species are flower colour, odour and anthesis period.
The floral cycles perform movements throughout anthesis situating the reproductive
cycles in different positions; based on those changes, three floral stages were defined as
follows. In the bud, the styles are upright, the staminal filaments are adnate to the styles and
the anthers are parallel to the filaments and facing inwards. As soon as the sterile cycles
separate and recurve, the anthers make a spin over their filaments while these bend until they
reach a perpendicular plane with respect to the floral axis. Thus, the anther face changes to an
outward position (extrorse), remaining in different orientations (Figs. 3 and 4, Table 1),
depending on the pollination syndrome. This constitutes the first floral stage (Figure 4, A, B),
along which the styles progressively bend, bringing the stigmas towards the level of the
anthers. Stage 2 is defined as the time during which the stigmas are at the same level of the
anthers, or even lower (Figure 4, C). During the third stage (Figure 4, D), the styles go back
to the former position, while the stamens descend and the corona, petals and sepals incurve,
closing the flower.
Anthers are dehiscent and stigmas are receptive throughout the whole anthesis; so the
flowers are not dicogamous. Due to the styles’ movements, the stigma surface and the pollen
are not available (in the way of the pollinators) at the same time during stages 1 and 3
(hercogamy), but they are both available during stage 2, so the hercogamy is temporal, or,
what is equivalent, there is functional dicogamy.
Pollination Mechanisms in Passiflora Species
3
To carry out pollination, a visitor must have the adequate size and/or behaviour to contact
both or either an anther and a stigma in the same or in successive visits to different flowers.
Pollen removal (from the anthers) and pollen deposition (on the stigmas) can only be
performed by a single pollinator when the stigmas had bent down to the level of the anthers.
Pollen removal and deposition is carried out by means of the dorsal thorax of hymenopterans
(in the melittophilous species), different parts of the head of hummingbirds (in the
ornitophilous ones), on the head and neck of bats (in the quiropterophilous ones), on the
mouth parts and antennae of sphingids (in the sphingophilous ones). The bees land on the
larger pieces of the corona (radii); the rest of the animals hover while they feed.
The style movements have been reported for all the species in which the floral biology
has been studied (Knuth, 1908; Janzen, 1968; Sazima and Sazima, 1978; Corbet and Willmer,
1980; Girón Vander-Huck, 1984; Kay,2001; AmelaGarcía, 1999 and others). In spite of this,
it has been detected that in certain species a proportion of the flowers do not bend the styles in
any moment of the anthesis, remaining straight as when the flowers open. These flowers are
not able to receive pollen from the legitimate visitors, i.e., the pollinators, so they become
functionally staminate. Other floral traits like a minor size of the gynoecium, minor sized and
nectarless flowers seem to be associated with the lack of styles movement. In some species,
the anthers do not open in a few flowers, so these flowers become functionally pistillate. The
ocurrence of non-bending styles flowers, non-dehiscent anthers flowers, or both of them
concomitantly with hermaphrodite flowers makes the sexual system functionally
andromonoecious, gynomonoecious or trimonoecious, respectively.
Figure 1. Floral architecture. Longitudinal section of flower of Passiflora mooreana showing the floral
parts. Drawing by Amela García, M. T. Scale bar = 10 mm. a, androgynophore; ann, annulus; ant,
anther; c, cup; ch, nectar chamber; f, staminal filament; l, limen; n, nectary; o, operculum; of,
operculum filaments; ov, ovary; p, pali; r, radii; st, stigma; sty, style; t, trochlea.
M. T. Amela García and P. S. Hoc
4
Figure 2. Flowers of Passiflora exhibiting different pollination syndromes. A, C, ornitophily; B,
melittophily; D, sphingophily. A and C, photographs leg. by J. P. Torreta and O. R. Di Iorio,
respectively.
All these related features have not been studied in detail and future research should take
them into account, as well as others, such as the possibility of underdeveloped ovules
associated with non-bending styles flowers, in which case, the gynoecium would be a
pistillode and the system would be addressed as truly andromonoecious.
Here, a compilation of original data is presented and it is discussed with the reported
aspects at the moment. The ecological significance of the styles’ movements are interpreted.
Furthermore, the traditional concepts of hercogamy and dicogamy are discussed.
MATERIALS AND METHODS
In this chapter, the terminology employed and revised by Tillet (1988) is used for flower
morphology, without taking into account their origin, but rather their function. The terms used
for sex expression in flowers is that of Wyatt (1983); although sexuality is a property of the
gametophytes and not of the flowers that bear them, we prefer to call the flowers
hermaphrodites when they bear the structures in which both gametophytes form, as it is a
conservative denomination.
Pollination Mechanisms in Passiflora Species
5
Figure 3. Flower opening and anther orientation. A, sepals and petals separating, the anthers face the
floral axis except one already spinning; B, 3 anthers spinning (right), 2 anthers facing downwards (left);
C, anthers facing outwards; D, anthers facing downwards; a bee landed on the radii and licking. A-B,
P. caerulea; C, P. misera; D, P. foetida.
Most of the observations described here refer to species studied in Argentina (Amela
García, 1999 and unpubl. data), from subgenus Passiflora, supersect. Passiflora: P. caerulea,
P. mooreana (sect. Stipulata), P. palmatisecta (sect. Passiflora), P. foetida, P. chrysophylla
(sect. Dysosmia) and subg. Decaloba, supersect. Decaloba: P. misera, P. urnaefolia (sect.
Decaloba), P. capsularis (sect. Xerogona) and supersect. Cieca: P. suberosa (sect. Cieca),
according to the classification of MacDougal and Feuillet (2004).
The field and laboratory methodology applied is described in detail in Amela García
(1999), and in Amela García and Hoc (1997, 1998).
All the photographs were taken by Amela García, M. T., except when quoted as a
different author.
M. T. Amela García and P. S. Hoc
6
RESULTS
1. Floral Architecture and Function
The base of the flower is constituted by the hypanthium; this is formed by the floral cup
and the floral tube. The first may acquire, mainly, the form of a plate, cup or urn, of varying
depth, depending on the species; it is the region compressed between the apex of the pedicel
and the base of the androgynophore, up to the insertion of the operculum (excluding it)
(Figure 1); there, in some species, the tube originates (Figure 2, A), which ends in the
insertion of the perianth; the tube may be not developed or reduced only to an edge in the cup.
The sepals and petals are inserted on the outer edge. In contrast with most Angiosperms, the
sepals contribute to the visual attraction when the flower is open, as they form together with
the petals the outline of a star when the flower is observed from the front or above (Figure 2,
B), at least in the species in which they are not deflexed (Figs. 2C, 3C, 5C). The corona is
found more internally, comprising the pieces that are between the perianth and the
androgynophore. The corona is formed by the radii (sing. radius), the pali (sing. palus), the
operculum, the nectar chamber, the annulus, the nectary, the limen and the trochlea. The radii
are filaments of varying width, length and position (alone and in group), depending on the
species; they are arranged in one or more cycles; the conjunct is called nimbus. It has
different functions, depending on the pollination syndrome (see section “Floral architecture
variants”). The pali are shorter filaments, frequently with a wider apex, arranged in one or
more cycles, forming a fence (sepimentum) that apparently deters the access to the nectar to
non-pollinating visitors. The operculum is a membrane, smooth or with radial fan-like
foldings, curved towards the androgynophore, that reaches the upper edge of the limen,
constituting an impediment for access to the nectar by nectar thieves and, in the pendant
flowers, for the nectar to drop; the smooth operculum fits the limen by means of a crease,
constituting a barrier more difficult to transpose than the folded operculum (which only leans
on the limen), reinforced sometimes with filaments that originate from its edge. The nectar
chamber consists in the cavity formed by the floral cup, closed by the operculum and the
limen. The annulus is a ring that emerges from the inner surface of the floral cup that may
divide the nectar chamber in an antechamber and an inner chamber. The nectary covers the
inner surface of the nectar chamber, not necessarily the annulus. The limen is a membrane,
generally placed in the base of the androgynophore or a bit up, over this, where the operculum
sits. The trochlea is an annular bulk in the base of the androgynophore, over the limen; it is
supposed to restrict the access to the nectar to non legitimate visitors and reinforce the
androgynophore base against the pushing of the bees when they lick (Figure 3, D). The
androgynophore separates the androecium and the gynoecium from the reward and, in that
way, reduces the damages that the visitors could provoke.
2. Floral Pieces Movements, Floral Stages and Implications in Pollination
Among other authors, the floral movements were described by Masters (1871) and
MacDougal (1994), although not completely. The movements described in this section
correspond to most Passiflora spp. flowers; variants of them are detailed in section 4 (“Floral
Pollination Mechanisms in Passiflora Species
7
architecture variants and their relation to pollinators size and behavior”). In the bud, the
anthers filaments are parallel to the floral axis and the anthers are extrorse. When anthesis
begins (Figure 3, A), the sepals, the petals and the nimbus expand until each cycle remains
campanulate, perpendicular to the floral axis or deflexed, depending on the species (Table 1).
The anthers spin on their insertion in their filaments while these filaments, originally parallel
to the floral axis, descend backwards and place forming an acute angle with the floral axis; in
this way, the anthers remain facing outwards (Figure 3, C); in other species, the filaments
descend even more, till they rest almost perpendicular to the floral axis; in these species, a
spin of 90o takes part between the filament and the connective, changing the anthers from a
radial position to a tangential one; thus, the dehiscent face rests facing towards the corona,
forming a parallel plane to this one (Figure 3, D). In some species, the anthers rest in an
intermediate position between the two ones described (Table 1). The styles, which are parallel
to the floral axis in the bud, tilt progressively to reach the same level of the anthers (Figure 4,
A-C).
The styles’ movements are useful to define three floral stages (Figure 4).
Stage 1 (Figure 4, A-B). The anthers spin over their filaments and these bend from their
vertical position to a horizontal one. Progressive bending of the styles until they reach the
level of the anthers.
Stage 2 (Figure 4, C). The stigmas remain just over the level of the anthers, among or
below these (depending on the species). The styles do not always move synchronously, so it
must be considered that the flower is in this stage when at least one of the stigmas is in the
described position, i.e., plausible of receiving pollen from a legitimate pollinator.
Stage 3 (Figure 4, D).Progressive lifting of the styles up to their former position in the
bud, parallel to the floral axis. The anthers’ filaments curve downwards. The radii, the corolla
and the calix incurve, thus closing the flower, which does not open again. Although this stage
consists in the progressive closing of the flower, it must be taken into account as a third stage,
as it is visited and sometimes there is pollen left in the anthers, depending on the visits
received in the former stages. The described movements have been observed in P. caerulea,
P. mooreana, P. palmatisecta, P. foetida, P. chrysopylla, P. capsularis and P. suberosa. In a
few species (P. misera and P. urnaefolia), the lifting of the styles or the bending of the anther
filaments does not always occur (Figure 6, C).
The stigmas are receptive during the whole anthesis in most of the species in which this
state was measured (Amela García, 1999), but in some of them they react more intensely
during stage 2, which is in accordance with the stigma’s position in this instance (capable of
receiving pollen from the pollinators). Similar results were obtained by Akamine and
Girolami (1959) in P. edulis, who employed different techniques, and by Souza et al. (2004),
who measured a decrease of the percentage of receptivity at the beginning of the evening,
when flowers must have been changing to stage 3 (although the stage’s changes were not
recorded by these authors).
Considering the possibility of contact with the legitimate pollinators in their way to the
reward, floral stages 1, 2 and 3 are functionally staminate (pollen donor), hermaphrodite and
functionally staminate, respectively. Nevertheless, the second stage could be deemed mainly
functionally pistillate (pollen receptor); the pollen offered in this stage depends on the
removal done by the visitors during stage 1.
If fertilization occurs, all the floral pieces persist (on the contrary, only the bracts) and the
ovary begins to increase its size generally at the following day of anthesis.
M. T. Amela García and P. S. Hoc
8
The lifting of the stigmas and the descent of the anthers during stage 3 would reduce the
possibility of self-pollination, except in the species of subg. Passiflora, supersect. Passiflora,
sect. Dysosmia, in which the styles, after lifting, curve upwards the upper third of their length
(Figure 6, D); nevertheless, fruits from self-pollination only were obtained from P. foetida but
not from P. chrysophylla (Amela García, 1999). The lifting of the styles, besides, must
diminish the possibility of damage or pollen clogging to the stigmas once they have been
pollinated, especially by some of the pollen robbers or thieves that remain near the anthers or
even inside the flowers while these close or when they are already closed.
Figure 4. Styles movements throughout anthesis. A-B, stage 1; C, stage 2; D, stage 3; A,C, D, anthers
facing downwards; B, anthers facing outwards. Drawing by P. S. Hoc based on P. caerulea (A, C, D)
and P. capsularis (B).
The unique moments when a flower can self-pollinate is when it opens (as the anthers are
dehiscent and facing the stigmas and they turn outwards on their filaments afterwards) or
when the stigmas descend to the level of the anthers and place among them. In both instances,
only the edges of the stigmas can be impregnated with self-pollen. In the cases when during
stage 3 the stamens do not descend, there could be a possibility of self-pollination if the
anthers contact the edges of the stigmas when these uplift, but the lack of stamens descending
in this stage occurs in presumptive self-incompatible species (AmelaGarcía, 1999); besides in
Pollination Mechanisms in Passiflora Species
9
the third stage little pollen is left on the anthers if pollinators have visited the flowers
assiduously, even fewer in the anthers edges.
In the species of Passiflora, the transfer of autogamous and geitonogamous pollen is
restricted to the removal that the visitors have performed during stage 1 (for both cases), and,
in the case of geitonogamy, to the synchrony in the change of stages between the flowers of a
single plant (that is not total) and the quantity of flowers that are visited in each arrival to a
plant. According to Janzen (1968), the pollinators would mainly transfer pollen that has been
deposited on their bodies during the previous visits to flowers in stage 1 of co-specific plants;
although some visitors remove it, in the species of Xylocopa a line of pollen remains in the
center of the thorax dorsum (Amela García, 1999), a fact that was also observed by Corbet
and Willmer (1980).
Webb and Lloyd (1986) interpreted erroneously which happens in Passiflora, as they
included the genus in a subclass of hercogamy that does not consider the styles’ movements,
nor how the contact with the pollinators occurs; but the genus neither can be placed in another
of the subclasses that they defined. As in all of them, they consider that the pollinators contact
both reproductive cycles in the same visit. Due to the spatial separation between the anthers
and the stigmas during stage 1 (hercogamy), and that the stigmas (although they are
receptive) remain out of the way of the pollinators that visit the flowers in this stage
(contacting only the anthers), both reproductive cycles are not presented effectively at the
same time. With the ulterior movement of the styles, and the location of the stigmas at the
same level of the anthers, hercogamy disappears, so, there would be a “temporal” hercogamy
(only during stage 1, and again in stage 3). Although during the whole anthesis the stigmas
are receptive and the anthers expose the pollen, considering that during stage 2 the stigma
receptivity is greater, that in some cases the stigmas descend bellow the level of the anthers
and that these are almost empty if the flower has been visited, stage 2 is mainly pistillate;
there would be, then, a tendency to dicogamy (stage 1 = staminate; stage 2 = pistillate). The
proper Webb and Lloyd (1986) stated that, in general, dicogamy and hercogamy are not
equivalent alternatives, and frequently occur together.
Lloyd and Webb (1986) commented that, in the flowers of some species that present the
pollen first, the stigmas may situate nearer the way of the pollinator than where the anthers
have presented, diminishing in that mode the interference with autogamous pollen; this occurs
sometimes in Passiflora when the stigmas place below the level of the anthers (Janzen, 1968;
Amela García, 1999).
3. The Approach to the Flowers: Floral Orientation
Flowers of most Passiflora species are solitary, derived from a higher numbered
inflorescence (Ulmer and MacDougal, 2004). Reduction of the peduncle has led to a unique
stalk in the majority of the species. The peduncle and pedicel from a continuous unit only
interrupted by the abscission layer and the bracts, usually three. What matters for pollination
biology is the length of the whole unit, which contributes to the exposition of the flower
outside the foliage, so peduncle plus pedicel will be addressed with a unique term (“stalk”)
from here and on.
The flowers may appear erect, patent or pendant (Figure 5, A-C). Each species has a
typical gravitational orientation, apparently measured by geotropic responses of the buds
M. T. Amela García and P. S. Hoc
10
(MacDougal, 1994). The floral stalk curves with respect to the eventual orientation of the vine
carrying branch (more or less horizontal or vertical), in such a way that the flower maintains
its proper position. For example, in erect flowers, when the branches are situated more or less
perpendicular to the ground, an upwards curvature of the apex of the stalk is observed (Figure
5, D); this orientation is readily observed in the buds (Figure 5, D). When the ovary begins to
grow, the stalk changes the curvature and the fruit becomes pendant (Figure 5, D).
Figure 5. Orientation of Passiflora flowers. A, erect flower; B, pendant flower; C, patent flower; D,
progressive change of the curvature of the stalk from the bud to the fruit in an erect flower. A, P.
capsularis; B, P. alata; C, P. palmatisecta; D, P. urnaefolia.
Pollination Mechanisms in Passiflora Species
11
4. Floral Architecture Variants and their Relation with Pollinator Size and
Behaviour
The comparisons with out-groups on the base of the floral morphology and the classic
theories of pollination syndromes suggest that the ancestor of the genus was pollinated by
insects, probably hymenopterans (Faegri and Pijl, 1976, cited in MacDougal, 1994). Among
the genus, melittophily is the most common syndrome (MacDougal, 1994). Ornithophily
appeared independently in various lines and the records of pollination by wasps and bats are
isolated (MacDougal, 1994). Vogel (l990) suggested pollination by lepidopterans for one
species and it is suspected for 2 more (Amela García, 1999).
Figure 6. Rare flowers or unusual position of floral parts. A, twisted styles (right); B, non-dehiscent
anthers; C, non-lifting styles during stage 3; D, lifted styles in stage 3 with the upper third of their
length curved downwards. A, P. caerulea; B-C, P. misera; D, P. foetida.
These syndromes usually manifest in the genus as follows:
Melitophily (Figure 2, B). Anthesis is diurnal. The flower’s size ranges from small to
large. The visual and olfactory attraction, as well as the surface where the bees land (Fig. 3,
D, Table 1), are provided by the radii, which are blue or purplish with white stripes, scented
and long. The dehiscent side of the anthers face the radii (Figure 3, D; Figure 4, A, C); the
hypantium is dish or cup shaped (constituted only by the cup, the floral tube is not
M. T. Amela García and P. S. Hoc
12
developed). Generally speaking, two variants can be considered: 1) erect or patent flowers,
with the radii extended and the nimbus forming a platform perpendicular to the floral axis; 2)
pendant flowers, with the radii fringed in their distal part and the nimbus campanulate (from
where the hymenopterans climb). In the erect flowers, the bees land on the outer edge of the
radii and walk a bit till they place their mouth parts between the limen and the operculum,
separating them; opening in that way allows access to the nectar. After liking, they walk
backwards and fly away. In the pendant flowers, the bees may visit legitimately in two ways:
1) entrance and exit of the flower in the same position, 2) the bee changes the body position
inside the flower (Varassin and Gomes da Silva, 1998). The different sized melittophilous
flowers are pollinated by different sized-bees (Amela García and Hoc, 2001). The length and
disposition of the radii conditions the bees of what size could land (Table 1).
Table 1. Structural characteristics of the hypanthium,
radii and other corona pieces involved in the access to the nectar
in Passiflora spp. with different pollination systems
Passiflora sp.
(pollinators)
Anther
orientation
Hypanthium
shape
Nectar barrier
Radii orientation
operculum
pali
P. caerulea
(large bees)
tangential
campanulate
smooth, with
filaments
short
P. mooreana
(large bees)
tangential
campanulate
smooth, with
filaments
short
P. palmatisecta
(sphingids)
tangential
tubular
smooth, without
filaments
absent
P. foetida
(medium sized-
bees)
tangential
campanulate
smooth, without
filaments
short
P. chrysophylla
(medium sized-
bees)
tangential
dish-shaped
smooth, without
filaments
short
P. misera
(medium sized-
bees)
radial
campanulate
folded
Long
P. capsularis
(butterflies)
radial
campanulate
folded
absent
Pollination Mechanisms in Passiflora Species
13
P. suberosa
(wasps and bees)
intermediate
dish-shaped,
shallow
folded
short
P. urnaefolia
(large bees)
intermediate
campanulate
folded
short
Ornitophily. Anthesis is diurnal. The flowers are large. The orientation is erect, patent or
pendant. The visual attraction is given by the scarlet sepals and petals (Figure 2, A, C), the
radii are short and parallel to the floral axis (thus preventing access to the abundant nectar
against the thieves) or reduced (tubercular) without odour (Tillet, 1988). The anthers face
outwards (Figure 2, A), the androgynophore is longer than in the melittophilous flowers
(Figure 2, C) and the hypanthium is tubular or urceolate (constituted by the cup and the floral
tube, of varying length, depending on the species). The length of the hypanthium conditions if
the flowers will be visited by short or long-billed hummingbirds; flowers with short
hypanthiums may be visited either by short or long-billed hummingbirds but flowers with
long hypanthiums must be visited by long-billed hummingbirds (Christensen, 1998). These
birds contact the reproductive organs with different parts of the head (throat, cheeks,
forehead), not the bill (Girón Vander-Huck, 1984; Christensen, 1998).
Quiropterophily. Few species show this syndrome (Sazima and Sazima, 1978; Kay,
2001). Most of the anthesis takes part at night. The long stalks make the flowers stand out
from the foliage. The flowers are large. The radii are short and completely (P. penduliflora)
or more or less (P. mucronata) parallel to the floral axis. The faint odour is fruity (P.
mucronata) or musty (P. penduliflora). Most flower parts are greenish (P. penduliflora) or
whitish (P. mucronata). P. mucronata erect flowers become asymetric during their aperture
(the anthers and the stigmas remain grouped in semicircle and facing outwards of the plant);
instead, P. penduliflora pendant flowers bend the anthers and stigmas a bit, facing the ground.
The firm and flexible stalk supports the bats visits (Sazima and Sazima, 1978).
Wasp and wasp and bee-pollinated flowers. Anthesis is diurnal. The flowers pollinated
by wasps are more or less erect, the radii are not conspicuously colour-banded and the odour
is similar to musk or escatol (MacDougal, 1994). In the small P. suberosa flowers, pollinated
by wasps and bees (Kosnitchke and Sazima, 1997), the scant radii are incurved at their
middle, resting as spider legs seen from the side (Table 1) and the sepals might contribute to
support the pollinating insects while they lick (Amela García, 2008).
Psicophily. It is suspected in P. jorullensis (Vogel, com. pers.) and in P. capsularis
(Amela García, 1999). The suspicion is based on the size of the flowers, the length of the
androgynophore, the width of the access to the nectar and the orange coloration (Vogel, com.
pers.), the outward orientation of the anthers (Figure 5, A; Table 1), the campanulate nimbus
(Figure 5, A; Table 1), diurnal anthesis and the presence of butterfly scales on the flowers
(Amela García, 1999). The syndrome could not be confirmed by the observations of
pollinators for any of the species by now; only thieves have been recorded for P. capsularis
(Amela García, 1999).
M. T. Amela García and P. S. Hoc
14
Phalaenophyly. Pollination by moths is suspected in P. palmatisecta (Amela García,
1999) taking into account the following traits: anthesis before dawn till mid-morning, patent
or erect flowers (Figure 5, C), developed floral tube, sepals and petals curved backwards
(Figure 5, C), radii apparently not forming a landing surface but instead a highly dissected
image in front view with a darker centre that might function as a nectar guide (Figure 2, D),
narrow access to the nectar, the stigmas never descend below the anthers (Figure 5, C),
whitish colour, intense citric odour. In spite of this, no pollinators could be registered until
now.
5. Structural Flower Class and Pollen Transfer
Faegri and Pijl (1979) included Passiflora among the "dish-bowl" flower type, with the
reproductive organs elevated over the plane formed by the perianth and the nimbus. Endress
(1996) included the genus in a more precise classification: among the flowers with canalized
access to the reward, which is arranged in a circle (“roundabout flowers”). Thus, Endress
(1996) considered the species of Passiflora in which the visitors land on the nimbus, placing
the body radially oriented to the flower and moving on it around the androgynophore,
walking laterally, describing a circle, separating the operculum and the limen at different
points when they probe for nectar; while doing so, the pollination is carried out nototribically
(Figure 3, D). So do the hymenopterans. This “several testes from different points from a
circle” mode of visit could be applied for hummingbirds, who in fact sometimes lick in a
similar way (moving apart from the flower and returning again to a different point), although
it is not so frequent and exhaustive, as it depends on the relative position of the flower on the
plant, i.e., if the flower is accessible from different sides for these hovering, licking animals.
The bees, instead, may fly over the flower and go to the reward from any side, regardless the
proximity of the foliage.
To get all the nectar in a single bout, the researcher must separate the operculum from the
limen in different points. In a equivalent manner, the hymenopterans that access the nectar
must not be able to extract all the nectar from a single place. The circular location of the
reward compels the pollinator to approach the flower from different points, thus contacting
more than an anther and/or a bent stigma.
6. Anthesis, Flowering Period, Inter-Plant Distance: Attraction and Resource
Availability
Each individual flower usually lasts from a morning to a day, depending on the species
(Amela García, 1999). During the course of the anthesis, the three floral stages succeed, and
each single flower does not open again. Meteorological conditions influence the duration of
the anthesis, delaying it when it is cloudy, cold, wet, and/or rainy or when the flowers are in
the shade (Amela García and Hoc, 1997; Amela García, 1999). In this way, the probability of
receiving visits from pollinators increases, as the frequency of them diminishes during such
conditions.
The passion vines have a long flowering period but the daily production of flowers is
scarce (monotonous blooming or steady state). Faegri and Pijl (1979) interpreted that the
Pollination Mechanisms in Passiflora Species
15
genus had gregarious flowering. This happens in the species with bigger specimens, such as
P. mooreana, P. caerulea and P. urnaefolia (Amela García, 1999), but the opening of flowers
per day is minor in the species with smaller specimens, such as P. lutea (Neff and Rozen,
1995) and P. palmatisecta, P. foetida, P. chrysophylla, P. capsularis, P. misera and P.
suberosa (Amela García, 1999). In some opportunities, species of the second group can
develop specimens of greater size (like P. misera), so they should be included in the first
group. The frequency of visits was higher in the species of the first group (Amela García,
1999), surely conditioned, at least in part, by the greater attraction and resource availability.
Passion vines grow rather sparsely. Nonetheless, sometimes, a conspicuous group is
found. This depends mainly on the distribution performed by the dispersal animals of the
zoochorous seeds and the proper conditions for the establishment of the seedlings.
The duration of the anthesis being less than 24 hours is common in species with flowers
pollinated by trapliners (Endress, 1996). The tropical taxa grow rather sparsely, so they
require pollinators capable of long distance travels; these perform daily travels that include
predictable resources, previously localized, that utilize regularly during an extended period
(Endress, 1996); this coincides with the monotonous flowering of Passiflora. Among the
trap-liners, species of Ptiloglossa, Bombus, Centris, Xylocopa (Janzen, 1971), sphingids,
birds and bats are encountered (Endress, 1996). The pollination by this type of animals
constitutes an effective mode of xenogamy. Many self-incompatible species of other genera
are pollinated by this type of flyers (Endress, 1996). Passiflora is an eminently tropical genus,
so it is probable that co-adaptation has occurred with this type of pollinators, mainly in the
self-incompatible species, the majority of which are pollinated by Xylocopa: P. mooreana and
P. caerulea (Amela García, 1999), P. edulis f. flavicarpa (Akamine and Girolami, 1959;
Corbet and Willmer, 1980; Sazima and Sazima, 1989), by hummingbirds: P. vitifolia (Snow,
1982) and by bats: P. mucronata (Sazima and Sazima, 1978; Kay, 2001). P. foetida, P.
chrysophylla, P. misera and P. urnaefolia also receive pollinators that exhibit this behavior
(Amela García, 1999).
In general, species of the minor sized plants group are partially or totally self-compatible
(Amela García, 1999) while species of the larger sized plants group are self-incompatible
(Amela García, 1999).
In Argentina, Passiflora species bloom usually between the middle of August till March;
depending on the weather, flowers may be encountered some months later, even in May
(Amela García, 1999). The maximum production of flowers occurs between September and
December. At least in the more conspicuous vines (P. caerulea, P. mooreana), the production
is slow at the beginning but it increases soon and remains stable for 3 or 4 months, then it
decreases. This would function as an attention grabber for the trap-liners, so they remain
attracted the rest of the season.
7. Rare Flower Types
7.1. Non-Bending Styles’Flowers
Non-bending styles’ flowers have been reported in Passiflora from time to time (Harms,
1925, Snow and Gross, 1980, La Rosa, 1984, cited in MacDougal, 1994). This last author
suspects that this sexual system might be more frequent in this genus, which includes more
than 525 species (MacDougal and Feuillet, 2004), but lacks detailed observation.
M. T. Amela García and P. S. Hoc
16
Lack of styles’ movements has been detected in P. quadrangularis and P. pinnatistipula
(Knuth, 1904 cited in Gottsbergeret al., 1988), P. edulis (Akamine and Girolami, 1957; Free,
1970), P. foetida, P. incarnata and P. alata (Gottsberger et al., 1988; May and Spears, 1988
and Varassin et al., 2001), respectively, in P. exsudans, P. oxacensis and P. sicyoides
(MacDougal, 1994), in P. mooreana, P. caerulea and P. chrysophylla (Amela García, 1999),
in P. cincinnata (Aponte and Jáuregui, 2002) and in P. misera (this chapter).
The relative amount of these flowers is minor than the normal ones in most of the species
in which they were counted (Akamine and Girolami, 1959; Gottsberger et al., 1988; Amela
García, 1999; Aponte and Jáuregui, 2002). This ratio varies between plants (Ruggiero et al.,
1976), along the shoot (MacDougal, 1994) and throughout the flowering season (May and
Spears, 1988).
Andromonoic species may have constant or labile proportions of male and hermaphrodite
flowers, depending on the predictable and efficient service of pollinators or on the uncertainty
of these at the beginning of the flowering season, correspondingly, and the optimal allocation
of limited reproductive resources (May and Spears, 1988).
The upright styles of most of the functionally andromonoic species discovered do not
move at all from their initial position, but the “upright” ones of P. cincinnata bend a bit and
return to the previous position or diminish the degree of curvature afterwards (Aponte and
Jáuregui, 2002). This intermediate type of flower was also observed by Akamine and
Girolami (1957), Free (1970), Ruggiero et al. (1976) and Aponte and Jáuregui (2002) in P.
edulis, in which the styles bend until they form an angle of 45º with the floral axis, not the 90º
one of the common flowers. This might be a misinterpretation, at least of the last author for
this species, as she registered the styles’ movements only for 3 hours and the anthesis lasts
more (Corbet and Willmer, 1980), so perhaps those flowers would have bent the styles
completely later. Moreover, styles’ movements’ speed is influenced by meteorological
conditions, experiencing a lag when it is cloudy, cold and/or rainy (Amela García and Hoc,
1997). Anyway, partial stylar movements have also been registered in P. mooreana (Amela
García and Hoc, 1998).
These functionally staminate flowers produce fertile pollen, as evidenced when it was
used to artificially pollinate hermaphrodite flowers (Akamine and Girolami, 1957).
7.1.1. Case Study 1: P. caerulea
The occurrence of upright flowers and their fate was studied from October to December
1997, in the natural population of P. caerulea growing in the campus of Buenos Aires
University. Sixty-six flowers of 7 plants were tagged, the styles movements registered during
the day of anthesis and the product of open pollination (flower or fruit) recorded.
Most of the styles that did not descend remained upright, parallel to the floral axis, as
previously described; but in some flowers, the styles that did not descend were twisted
(Figure 6, A). This may be a teratogenic phenomenon, or a strategy that prevents the bending
of the styles.
Bending-styles’ and upright-styles’ flowers coexisted in a same plant, but not all the
plants bore upright-styles’ flowers.
Although the movements of the styles could be registered for all the flowers tagged, the
product of some flowers were lost, so the set of data are presented separated in Table 2. As it
can be seen, the proportion of flowers that bended the styles was substantially greater than the
ones with upright styles (Table 2, a); if the flowers in which the products could not be
Pollination Mechanisms in Passiflora Species
17
recorded are discarded, the minor sample gives a minor percentage of upright styles’ flowers
(Table 2, b). This low percentage is constituted by 1 flower that set fruit (Table 2, b), which
reached maturity. The unique possibility of pollination in the upright-styles’ flowers is at the
flower opening, when it can be impregnated with the pollen of its dehiscent anthers (as
described above). P. caerulea is highly self-incompatible, and produces self-fruits in very low
percentages, which are mainly seedless (Amela García and Hoc, 1997). Nevertheless, the
origin of the pollen on the stigma might have been hardly from another source (exogenous
pollen dropped from the body of a pollinator flying over the flower while passing pollen to
the hind legs or from pollen deposited by means of unusual behavior of a floral visitor). Other
flowers that did not bend the styles, although the product could not be recorded, had pollen on
their stigmas.
Even based on a low number to make conclusions, the unique flower of P. caerulea that
did not bend the stigmas had pollen on them and set fruit, so it was not female sterile. May
and Spears (1988), by means of artificial pollinations, detected two physiological flower types
among morphologically normal flowers that did not bend the styles: ones that set fruit
(although in less proportion than hermaphrodites) and ones that did not set fruit, depending on
the period of the flowering season during which pollination took place. They attributed this
phenomenon to a seasonal component. On the other hand, the upright styles’ flowers of P.
edulis were reported as female sterile by Free (1970) and Ruggiero et al. (1976) when hand-
pollinated, and the upright styles’ flowers of P. foetida drop after flowering (Gottsbergeret al.
1988). These observations do not contradict those of May and Spears (1988) but they
evidence that studies involving the whole flowering season are needed.
Table 2. Fruit production of bending and upright styles’ flowers
in a population of P. caerulea in Ciudad Universitaria, Buenos Aires
Number of tagged flowers (number of plants)
66 (7)
Bending styles (% flowers)
92
Upright styles (% flowers)
8
a. Occurrence of flowers with upright and bending styles.
Number of tagged flowers with recorded products (number of plants)
54 (7)
Bending styles flowers (%)
98
Fruits from bending styles flowers (%)
92
Upright styles flowers (%)
2
Fruits from upright styles flowers (%)
100
b. Fruit set in upright and bending styles’ flowers.
7.2. Non-Dehiscent Anthers Flowers
In some species, a proportion of the flowers do not open the anthers (Amela García,
1999). In each single flower, no anther opens (P. capsularis) or not all the anthers open at the
same time (P. urnaefolia and P. suberosa) (Amela García, 1999).
The amount of these flowers is less than the functionally hermaphrodite ones.
The late and asynchronous pollen exposition in some flowers of P. urnaefolia and P.
suberosa probably are due to the high humidity and long exposition to shade in the forest
environments where they were studied, as the dehiscence is produced when the endothecium
M. T. Amela García and P. S. Hoc
18
cell walls dehydrate (Valla, 1999). In P. suberosa, the anthers were not always dehiscent
when the flowers opened but they finally did; sometimes they dehisced at the end of stage 1
or during stage 2, when humidity was lower than 67 %, temperature was higher than 17-25 °C
and the flowers began to be exposed to the sun (Amela García, 2008). In P. urnaefolia, the
anthers were not always dehiscent when the flowers began to open, even in some flowers in
stage 2 they were still closed if the flowers were not exposed to the sun, and not all the
anthers open simultaneously. But in P. capsularis, in some flowers (more frequently in a
patch than in another of two studied ones) the anthers did not open at all (although they had
pollen), in spite of all the flowers were exposed to the same microclimatic conditions; in these
cases, a physiological factor must be implied. In some of those non-dehiscent anthers’
flowers, odor was not perceived (Amela García, 1999).
In these flowers, with the non-dehiscent anthers until stage 2, there would not be a
tendency to dicogamy, i.e., there would not be functional dicogamy, as the stigmas and the
pollen are exposed at the same time. In this case, the probability that the pollinators transfer
autogamous pollen is greater because this pollen has not been removed in the previous stage.
On the contrary, the anthers of P. misera, P. palmatisecta, P. caerulea, P. mooreana, P.
foetida and P. chrysophylla, species that were studied in sites where they receive sun during
the entire day, the anthers were dehiscent as soon as the flowers began to open (Amela
García, 1999). Besides, the ambient temperature where P. urnaefolia was growing was rather
lower than in the rest of the sites. It is probable that the high relative ambient humidity where
the plants of P. capsularis were had also influenced the lack of anthers’ aperture during the
whole anthesis, although it rests to be determined why the aperture occurred only in some of
the flowers of a single day.
Anyway, regardless of the cause that prevents the anthers’ dehiscence, the lack of this
makes the flowers functionally pistillate. This provokes the concomitantly occurrence of
pollen donor-receptor and only receptor flowers in a single plant, in which case the system is
functionally gynomonoecious.
7.3. Case Study 2: Non-Bending Styles’Flowers and Non-Dehiscent Anthers’Flowers:
P. misera
The occurrence of flower types in this species was recorded in a specimen growing in a
garden in Buenos Aires province during January and February 2010. The specimen occupied
a space of approximately 2 m high and 5 m width. Daily, 5 ± 2 flowers open (n = 12 days).
The flower types that were produced are reported in Table 3. Aberrant androecium
implies:
anthers do not spin over their filament (partially or completely)
anthers do not open: no anther opens (partially or at all) or only some anthers open
stamens do not bend down to a horizontal plane
stamens’ filaments shorter than normal ones
stamens’ filaments lack anther
This fact may occur in some of the stamens of a flower, not in all of them, and aberrant
androecium may occur in combination with lack of style movement. For example, in one of
the flowers that the anthers did not open, they did not spin over their filament, and a style
Pollination Mechanisms in Passiflora Species
19
(together with its corresponding stigma) was contracted like a blower (this style bent a bit but
not as much as the rest). Some flowers (in which the androecium was aberrant) were smaller
than the normal ones.
The teratology consisted in a flower with the ovary laterally bent, one of the stamens’
filaments shorter than the rest and one of the styles shorter that the other two; surprisingly,
this was the unique style of the three that bent.
The occurrence of these flower types, especially the teratologic ones, may be due to their
belonging from a single isolated specimen without genetic flow of this self-incompatible
species (Amela García, unpubl. data), which flowers regularly every year but does not
produce fruits.
When the three types of flowers coexist in the same plant, the system becomes
functionally trimonoecious.
Table 3. Flower types in a specimen of P. misera growing
in a garden in Merlo, Buenos Aires province
Tagged flowers (n)
30
Bending styles (% flowers)
64
Upright styles (% flowers)
10
Aberrant androecium (% flowers)
23
Teratologic flowers (% flowers)
3
7.4. Floral Traits Associated with Lack of Styles Movements
The gynoecia of the upright-styles’ flowers in certain species from section
Pseudodysosmia (subgenus Decaloba) are not fully developed, yellowish and smaller
(MacDougal (1994).
In P. caerulea, the stigma colour may be yellowish-green, violet, deep yellowish brown
(Amela García, 1999) or creamy green to murrey (Deginani, 2001). Some flowers that have
the stigmas both small and green are suspected of not descending the styles, but some of these
stigmas descended; flowers with upright styles may have the stigmas of “normal” colour, i.e.,
brown. So the stigma colour seems to be more associated with intra-plant variation rather than
with style movement capacity, although it seems more frequent that non-descending stigmas
are green and smaller.
The same was observed in P. mooreana (Amela García and Hoc, 1998). Some of the
flowers in which the styles did not descend had the stigmas small and of different colour than
the usual flowers, but other non-bending styles’ flowers had them of normal appearance.
Moreover, these differences have no relation with intrafloral colour change, as the stigma
colour remains constant throughout anthesis; it only had been observed to change in one
species, and when the flowers had closed (Amela García and Hoc, 1998).
In P. caerulea, not always the three stigmas of a single flower have the same
morphology, i.e., one stigma may be normal and the rest atrophied.
The lack of complete correlation between stigma size and/or colour with the lack of stylar
movements reinforces the suspicion of May and Spears (1988) that a physiological factor is
involved.
M. T. Amela García and P. S. Hoc
20
Some of the flowers of P. foetida that did not bend the styles had smaller ovaries than
hermaphrodites, some apparently atrophied (Gottsberger et al., 1988). P. incarnata produces
upright styles’ flowers with atrophied ovaries and styles at the beginning of the flowering
season (May and Spears, 1988).
In the species in which the anatomy of the styles was compared between upright and
descending styles’ flowers, the flowers that bend the styles had more aerenchyma in the
medium and upper part of the styles; this tissue was mainly localized in the bending “face” of
the styles and these styles were of greater diameter than the ones of upright styles’ flowers
(Aponte and Jáuregui, 2002); the larger amount of this kind of tissue and its distribution is
supposed to increase the flexibility and to favor the nastic-like movement (Aponte and
Jáuregui, 2002).
In the course of a study of the floral biology of P. chrysophylla at El Palmar National
Park in Entre Ríos province, Argentina (Amela García, 1999), a few smaller flowers were
recorded among the population, with smaller nectar chamber, nectarless, although scented,
whose styles did not descend but the stigmas were receptive (Amela García, 1999). The
nectarless but scented flowers might contribute to long distance attraction with a minor cost
for the plant, as they are deceptive. The less proportion of those flowers with respect to the
nectar-producing ones would constitute an effective imitative system, because the imitator
does not surpass the number of the model (Faegri and Pijl, 1979). In monoic species,
unisexual flowers are small (Jong et al., 2008), and among unisexual flowers, female flowers
are larger than male flowers (Delph, 1996). The minor sized and upright styles’ flowers of P.
chrysophylla function as males.
Flowers of smaller size than the normal one have also been found among a population of
P. caerulea growing naturally at El Palmar National Park (Amela García, unpubl. data). It
remains to be determined if these flowers bend the styles or not.
7.5. Pollination and Fertilization Problems. Implications for Horticulture
The economic importance of the passion fruit stands on the production of seeds, as it is
the pulp that covers them (arile), the main product consumed. The arile is only developed
when the ovule is fertilized. So it is important that the plant set fruits and that those fruits
contain seeds.
Low fruit set has been recorded more than once (Akamine and Girolami, 1957; May and
Spears, 1988 and cites therein; Sazima and Sazima, 1989).
Sometimes the ovary grows but few or no seeds are developed inside (Akamine and
Girolami, 1957, 1959; Amela García and Hoc, 1997; 1998). This may be due to reception of
incompatible pollen, as this fact was observed in self-incompatible species (Akamine and
Girolami, op. cit.; Amela García and Hoc, op. cit.) or pollen limitation, such as low frequency
of pollinators due to lack of nesting sites or competition by nectar robbers (Sazima and
Sazima, 1989).
Increasing non-bending styles’ flowers during the flowering season decreases the
potential number of flowers to set fruit. If artificial pollination is desired, at least a portion of
the upright flowers would set fruit but they are not morphologically recognizable. So effort in
hand pollination of all the flowers, regardless their stylar movements, would be employed if
that is the decision to be made. Manual work on passion fruit culture seems common. Each
fruit is covered by a wax film after collection in order to increase preservation (Vilela, com.
pers.).
Pollination Mechanisms in Passiflora Species
21
CONCLUSION
Although some of the species in the genus are partially or totally self-compatible (Amela
García, 1999), the temporal hercogamy and the tendency to the functional dicogamy
generated by the styles’ movements favor the genetic material exchange in the population.
Different types of flowers occurring along the flowering season have been reported for
Phaseolus (Hoc, unpubl. data), but in this case the difference stands in that the flowers are
chasmogamous or cleistogamous at the beginning and at the end of the flowering period.
The following considerations must be taken in account for future research concerning the
different flower types:
in some bending-styles’ flowers the styles are left twisted when they lift; this must
not be mistaken with non-bending-styles’ flowers that have the styles twisted as soon
as the flowers open. So observations throughout the whole anthesis are necessary to
study this phenomenon
more exhaustive studies, dealing with the whole flowering season and in natural
populations, are necessary to assess the occurrence of these “abnormal” flowers and
their physiological performance
more species must be studied in this aspect
artificial pollinations are necessary to discover the physiological capacity of the
completely and partially non-bending styles flowers
stigma receptivity must be tested
ACKNOWLEDGEMENTS
To the Consejo Nacional de Investigaciones Científicas y Técnicas, for the doctoral
fellowships and the grant BID-Conicet Nº 548 with which this study (part of the Ph. D. thesis
of M. T. Amela García), was financed.
REFERENCES
Akamine, E. K.and Girolami, G. (1957). Problems in fruit set in yellow passion fruit. Hawaii
Farm Sci., April, 3-5.
Akamine, E. K. and Girolami, G. (1959). Pollination and fruit set in the yellow passion fruit.
Hawai Agricultural Experiment Station, Tech. Bull., 39, 1-44.
Amela García, M.T. (1999). Biología floral y sistema reproductivo de especies nativas de
Passiflora (Passifloraceae) de la Argentina: Ph. D. thesis. Buenos Aires University.
Amela García, M. T. and Hoc, P. S. (1997). Floral biology and reproductive system of
Passiflora caerulea (Passifloraceae). Beitr. Biol. Pflanzen, 70, 1-20.
Amela García, M. T. and Hoc, P. S. (1998). Aspectos de la biología floral y el sistema
reproductivo de Passiflora mooreana (Passifloraceae). Darwiniana, 35, 9-27.
Amela García, M. T. and Hoc, P. S. (2001). Pollination of Passiflora: do different pollinators
serve species belonging to different subgenera? Acta Hort., 561, 71-74.
M. T. Amela García and P. S. Hoc
22
Amela García, M. T. (2008). Breeding system and related floral features of Passiflora
suberosa (Passifloraceae) under natural and experimental conditions in Argentina. Bol.
Soc. Argent. Bot., 43, 83-93.
Aponte, Y. and Jáuregui, D. (2002). Estructura morfoanatómica de los estilos y estigmas en
Passiflora edulis f. flavicarpa Degener y Passiflora cincinnata Mast. Anales de Botánica
Agrícola, 9, 35-47.
Christensen, A. B. (1998). Passionflowers and their birds, bees and butterflies. The ecology
and evolution of pollination and herbivory in Passiflora. Ph. D. thesis. Univ. of Aarhus,
Denmark.
Corbet, S. A. and Willmer, P. G. (1980). Pollination of the yellow passionfruit: nectar, pollen
and carpenter bees. J. Agric. Sci., 95, 655-666.
Deginani, N. G. (2001). Las especies argentinas del género Passiflora (Passifloraceae).
Darwiniana, 39, 43-129.
Delph, L. F. (1996). Flower size dimorphism in plants with unisexual flowers. In: D. G.
Lloyd, and S. C. H. Barret (eds.), Floral biology: Studies on floral evolution in animal-
pollinated plants (1st. ed., pp. 217-237). NY: Chapman and Hall.
Endress, P. K. (1996). Diversity and evolutionary biology of tropical flowers (2nd. ed.).
Cambridge: Cambridge Univ. Press.
Faegri, K. and Pijl, L.van der. (1979). The principles of pollination ecology (3rd. ed). Oxford:
Pergamon Press.
Free, J. B. (1970). Insect pollination of crops. London: Academic Press.
Girón Vander-Huck, M. (1984). Biología floral de tres especies de Passiflora: Graduation
thesis. Medellín, Univ. de Antioquía.
Gottsberger, G., J. M. F. Camargo and I. Silberbauer-Gottsberger. (1988). A bee-pollinated
tropical community: the beach dune vegetation of Ilha de Sao Luís, Maranhao, Brazil.
Bot. Jahrb. Syst., 109, 469-500.
Janzen, D. H. (1968). Reproductive behaviour in the Passifloraceae and some of its
pollinators in Central America. Behaviour, 32, 33-48.
Janzen, D. H. (1971). Euglossine bees as long-distance pollinators of tropical plants. Science,
171, 203-205.
Jong de Tom J., Avi S. and Frank, T. (2008). Sex allocation in plants and the evolution of
monoecy. Ecology Research, 10, 1087-1109.
Kay, E. (2001). Observations on the pollination of Passiflora penduliflora. Biotropica, 33,
709-713.
Koschnitzke, C. and Sazima, M. (1997). Biologia floral de cinco espécies de Passiflora L.
(Passifloraceae) em mata semidecídua. Revta. Brasil. Bot., 20, 119-126.
Knuth, P. (1908). Handbook of flower pollination. 2. Oxford: Clarendon Press.
Lloyd, D. G. and Webb, C. J. (1986). The avoidance of interference between the presentation
of pollen and stigmas in angiosperms. I. Dichogamy. New Zealand Journal of Botany, 24,
135-162.
MacDougal, J. M. (1994). Revision of Passiflora Subgenus Decaloba Section
Pseudodysosmia (Passifloraceae). Ann. Arbor. Mich. American Soc. Plant Taxon. Syst.
Bot. Mon., 41.
MacDougal, J. M. and Feuillet, C. (2004). Systematics. In: T. Ulmer and J. M. MacDougal
(Eds.), Passiflora: passionflowers of the world (1st. ed., pp. 27-31). Cambridge, Timber
Press.
Pollination Mechanisms in Passiflora Species
23
Masters, M. T. (1871). Contributions to the natural history of the Passifloraceae. Trans. Linn.
Soc., 27, 593-645.
May, P. G. and Spears, E. E. (1988). Andromonoecy and variation in phenotypic gender of
Passiflora incarnata (Passifloraceae). Amer. J. Bot., 75, 1830-1841.
Neff, J. L. and J. G. Rozen. (1995). Foraging and nesting biology of the bee Anthemurgus
passiflorae (Hymenoptera: Apoidea), descriptions of its immature stages, and
observations on its floral host (Passifloraceae). Amer. Mus. Novitates, 3138, 1-19.
Ruggiero, C., Banzatto, D. A. and Sánchez-Lam, A. (1976). Studies on natural and controlled
pollination un yellow passion fruit (Passiflora edulis f. flavicarpa Deg.). Acta Hort., 57,
121-124.
Sazima, M. and Sazima, I. (1978).Bat pollination of the Passion flower, Passiflora mucronata
in Southeastern Brazil. Biotropica, 10, 100-109.
Sazima, I. and Sazima, M. (1989).Mamangavas e irapuás (Hymenoptera, Apoidea): visitas,
interacoes e consequencias para polinizaçao do maracujá (Passifloraceae). Rvta. Bras.
Entom., 33, 109-118.
Snow, A. A. (1982). Pollination intensity and potential seed set in Passiflora vitifolia.
Oecologia, 55, 231-237.
Souza, M. M., Pereira, T.N.S., Viana, A.P., Pereira, M.G., Texeira do Amaral Júnior, A. and
Madureira, H. C. (2004). Flower receptivity and fruit characteristics associated to time of
pollination in the yellow passion fruit Passiflora edulis Sims f. flavicarpa Degener
(Passifloraceae). Scientia Horticulturae, 101, 373-385.
Tillet, S. S. (1988). Passionis Passifloris II. Terminología. Ernstia, 48, 1-40.
Ulmer, T. and MacDougal, J. M. (Eds.) (2004). Passiflora: passionflowers of the world (1st.
ed.). Cambridge, Timber Press.
Valla, J. J. (1999). Botánica: Morfología de las plantas superiores (13 ed.). Bs. As.:
Hemisferio Sur.
Varassin, I. G. and Gomes da Silva, A. (1998). A melitofilia em Passiflora alata Dryander
(Passifloraceae) em vegetação de Restinga. Rodriguesia, 50, 5-18.
Varassin, I. G.. Trigo, J. R. and Sazima, M. (2001). The role of nectar production, flower
pigments and odour in the pollination of four species of Passiflora (Passifloraceae) in
south-eastern Brazil. Bot. J. Linn. Soc., 136, 139-152.
Vogel, S. (1990). Radiación adaptativa del síndrome floral en las familias neotropicales. Bol.
Acad. Nac. Cs. Córdoba, Argentina, 59, 5-29.
Webb, C. J. and Lloyd, D. G. (1986). The avoidance of interference between the presentation
of pollen and stigmas in angiosperms. II. Herkogamy. New Zealand J. Bot., 24, 163-178.
Wyatt, R. (1983). Pollinator-plant interactions and the evolution of breeding systems. In: L.
Real (ed.), Pollination biology (1st. ed., pp. 51-95). N.Y.: Academic Press.
Reviewed by Gabriel Bernardello
Universidad Nacional de Córdoba
