especially common under montane climate condi-
tions, if potential pollinators are less active (or less
common) than at lower elevations. Low temperatures
and mist may constrain flowering periods or reduce
pollinator abundances or may alter the relative pro-
portions of visitor species that differ in their effective-
ness as pollinators. Altitudinal migrants or rare visi-
tors may be especially important as pollinators of
some plant species. Although related plant species
have not been compared over an elevational gradient,
some species in Monteverde are not well pollinated
(Bronstein 1988a). For 10 species of hummingbird-
pollinated'plants, stigmatic pollen loads were used
to calculate the number of compatible grains deliv-
ered per ovule. For only one species did pollinated
flowers collect a substantial number of grains per
ovule. For the other nine species, the majority of flow-
ers were poorly pollinated (W. Busby and S. Kinsman,
unpubl. data).
Even where pollinator activity is high, effective
pollination can be rare. For two common mass-flow-
ering self-incompatible Inga spp. (Fabaceae) trees at
Monteverde, experimental pollinations were com-
bined with detailed observations of pollinators and
flowers to determine the frequency of highly effec-
tive pollination. Visitors to these trees are strong fli-
ers (hummingbirds, hawkmoths, and skippers) and
are able to move pollen over substantial distances,
but they do so infrequently. Although visitation is
high and half of the flowers receive pollen, relatively
few receive the required outcross pollen (Koptur
1984a).
8.1.6. Research and Conservation:
Plants' Mates
Population genetic consequences of pollen dispersal.
Pollinators and seed dispersers help shape the mat-
ing patterns of plants (Hamrick et al. 1993). The poten-
tial for a plant's mates to be distant and diverse
through pollen exchange is strongly influenced by
pollinator populations and movement patterns. For
conservation, it is critical to learn how pollinators
influence genetic characteristics of plant populations.
Flying pollinators that can move pollen relatively long
distances are especially important in effecting gene
flow (Hamrick and Loveless 1989). Bees, hawkmoths,
and bats are strong fliers that can move many kilome-
ters while foraging (Murawski 1995); trap lining hum-
mingbirds and Lepidoptera may also effect substan-
tial gene flow. Knowing the role of these pollinators
for Monteverde plants will help determine conse-
quences of forest loss and fragmentation, declines in
plant and pollinator populations, and changes in re-
sources that support pollinators.Tropical evidence. Many lowland tree populations
appear to outcross frequently (Bawa et al. 1990), to
maintain high levels of genetic variation, and to expe-
rience substantial gene flow (Hamrick and Loveless
1989, Hall et al. 1994, Murawski 1995, Chase et al.
1996). At La Selva, the great majority of trees require
outcrossing (Kress and Beach 1994), and their animal
pollen vectors (usually bees) import the required out-
cross pollen grains (Bawa 1974). On Barro Colorado
Island, in Panama, 16 common species of shrubs and
trees representing a variety of pollination systems and
seed dispersal syndromes show high levels of genetic
variation and substantial genetic exchange among
populations (Hamrick and Loveless 1989). Less com-
mon trees at the same site show less genetic variation
(Murawski 1995). Similar variation in Asian tropical
forest trees are attributed in part to pollinator behav-
ior (Murawski 1995).Monteverde questions. We know little about gene
flow, genetic variation, and the effectiveness of pol-
lination for most plants in the Monteverde region
(Koptur 1984a, Linhart et al. 1987a). The single Monte-
verde tree species that has been examined (Gibson
and Wheelwright 1996; see Gibson, "Seed Dispersal,"
pp. 289-291) demonstrates gene flow. Effective plant
population sizes may be different from those in the low-
lands because the steep elevational gradient collapses
suitable habitat. Altitudinal migrants (e.g., certain hum-
mingbirds, butterflies, and hawkmoths) may provide
occasional long-distance pollen movement, although
studying the original patterns may now be impossible
because most Pacific slope forest has been eliminated.
Mating systems in reforestation windbreaks may not be
characteristic if nursery-grown saplings were derived
from seeds of one or a few parent trees.
It is critical to maintain the communities and eco-
systems that support pollinators' and seed dispersers'
travel over both short and long distances (see William-
son and Darling, "La Ventana," p. 438; Nielson and
DeRosier, "Windbreaks and Birds," pp. 448-450; and
Harvey, "Windbreaks and Trees," pp. 450-451 vol-
ume). Demonstration of plant gene flow patterns, es-
pecially over an elevational gradient, could provide
strong arguments for integrated systems of reserves
and corridors to support pollinator and seed disperser
populations (see Chap. 12, Conservation Biology).
Even individual plant species that occupy a restricted
elevational range may have pollinators and seed dis-
persers that depend on wide-ranging resources (Bron-
stein 1995). Plant species in forest fragments are es-
pecially important to study (Hall et al. 1994, 1996,
Boshier et al. 1995a, b, Chase et al. 1995), particularly
those on the largely deforested Pacific slope. Their
gene flow patterns have probably already been altered255 Plant-Animal Interactions