Monteverde : Ecology and Conservation of a Tropical Cloud Forest

(やまだぃちぅ) #1

be large. With a few exceptions (Bronstein 1991; Weiss,
"A Fly Larva," pp. 278-279), their effects in Monte-
verde are poorly known. They offer outstanding op-
portunities for research on constraints on the evolu-
tion of defense (Baker et al. 1983), and on the effects
of flower-chewing, nectar robbery, and pollen feed-
ing (McDade and Kinsman 1980, Dayanandan et al.
1990). The role of third parties may be particularly
important in forest fragments because fragmentation
may affect populations of pollinators and third par-
ties in different ways.


8.1.2. Floral Sexuality and Pollinators


Floral sexual systems. Most flowering plants in the
world and in Monteverde are hermaphroditic: each
flower has functional female and male reproductive
organs. In monoecious species, separate female and
male flowers occur on each plant; in dioecious
species, individual plants have either female flow-
ers or male flowers but not both. There are many
variations of these three main sexual systems (Wyatt
1983, Richards 1986).


Correspondences between pollination and sexual sys-
tems. In tropical forests, each floral sexual system
seems to be associated with particular pollination
systems. These patterns were first determined from
examination of 276 plant species at La Selva (Kress
and Beach 1994), where beetle pollination is associ-
ated with monoecy and pollinators of hermaphroditic
plants are more diverse. Similar patterns occur in
Monteverde. There is no single explanation for cor-
respondences between floral sex and pollinator. Some
could be a simple result of phylogeny. For example,
beetles could be associated with monoecy simply
because over evolutionary time, they have retained
associations with the plant families that happen to be
monoecious (e.g., Araceae and Cyclanthaceae; Schatz
1990). Alternatively, being associated with beetles
might have fostered the evolution of monoecy in
these plant families. Another association—one between
hummingbirds and understory, self-compatible, her-
maphroditic plants—has been explained in terms of
outcrossing (Kress and Beach 1994). A comprehensive
approach to understanding the pollinator-floral sex
correspondences must consider pollen movement
patterns and resulting plant mating structures (Bawa
1994; see Wheelwright, "Sex Ratios," pp. 87-88).


Dioecy. The pollination methods of dioecious an-
giosperms in tropical forests are especially intriguing.
Two estimates of the global incidence of dioecy are
6% and 4% of all angiosperm species (but see Bawa
1980a, Richards 1986, Renner and Ricklefs 1995).

Comparatively, dioecy is overrepresented in tropical
forests (Bawa 1980a), occurring in around 15% of the
Monteverde flora (Table 8.1), 16% of species in a
montane forest in Venezuela (Sobrevilla and Arroyo
1982), and 17% of 507 species surveyed at La Selva
(Kress and Beach 1994), with a similar frequency in
Peninsular Malaysia (Bawa 1992). In two seasonally
dry neotropical forests (Jalisco, Mexico, and Guana-
caste, Costa Rica), 12% and 22% of angiosperms are
dioecious, respectively (Kress and Beach 1994). The
frequency of dioecy is even higher when only trees
are considered. At two lowland sites, La Selva and Pen-
insular Malaysia, 23% and 26% of tree species are dio-
ecious (of 334 and 711 tree species, respectively; Bawa
1992). In two montane forests, Monteverde (Table 8.1)
and Venezuela (Sobrevilla and Arroyo 1982), 28% and
31% of tree species are dioecious (of 500 medium-large
Monteverde trees, and of 36 tree species in the Venezu-
elan forest). In Monteverde, the frequency of dioecy
declines with the forest strata: dioecy occurs in 19% of
small trees, 15% of vines, 14% of epiphytes, 9% of
shrubs and treelets, and 0.6% of herbs.
Pollination modes of these dioecious species are
of interest (Bawa 1994). At La Selva and in a drier
forest in Costa Rica, the small, unspecialized, pale or
white, open flowers of dioecious species are polli-
nated predominantly by small bees and small diverse
insects (Bawa 1994). Monteverde also reflects this
pattern, as dioecy is especially frequent in plants
pollinated by small bees (19%) and by small diverse
insects (21%); 64% of Monteverde's dioecious spe-
cies exhibit these pollination systems (Table 8.1).
Monteverde differs from La Selva in showing an ad-
ditional association between dioecy and wind polli-
nation. Dioecy occurs in more than 40% of Monte-
verde's 103 wind-pollinated species (Table 8.1), but
in only 15% of Monteverde's flowering plants over-
all. Wind pollination occurs in 5% of Monteverde's
flowering plants but in 13% of dioecious canopy trees
(Table 8.1). In contrast, at La Selva, wind pollination
is rarer overall (< 3%) and is not reported for trees
(Kress and Beach 1994). Why these particular polli-
nation systems (small bee, small diverse insects, and
wind) are especially frequent for dioecious plants is
not clear (Bawa 1994).

Pollen delivery. Floral sexual systems, like the compo-
sition of nectar and the timing of its availability, can
influence pollinator movement and effectiveness. Stud-
ies of rare sexual systems provide clues about the ways
that evolution can mold floral traits in ways that influ-
ence pollinator behavior. Three of these studies are from
Monteverde: one a case of intersexual floral mimicry
(see Agren and Schemske, "Deceit Pollination," pp. 279—
281), one of andromonoecy, and one of androdioecy.

251 Plant-Animal Interactions
Free download pdf