fruiting in one plant family, Lauraceae, over the last
17 years (N. Wheelwright, unpubl. data). One advan-
tage of studying plant reproductive phenologies is that
fruit set and development reflect environmental con-
ditions integrated over periods as long as two years.
So far, there is no indication that reproduction in the
Lauraceae has changed over the last decade and a
half. Whether lauraceous trees indicate anything
about other plant species in Monteverde is unknown,
but the responses of trees to environmental changes
is likely to be quite different than that of animals.
For animals, the most useful—and alarming—data on
population trends in Monteverde come from the field
notes and casual observations of resident and visit-
ing herpetologists, ornithologists, and mammalogists.
Another important concept in conservation biol-
ogy is that of the "umbrella species" (Launer and
Murphy 1994). If we protect species that have expan-
sive and sensitive habitat requirements, we automat-
ically confer protection on species occupying the
same habitats. The Bare-necked Umbrellabird is a
large fruit-eating bird found on Monteverde's Atlan-
tic slope. The bird's common name was inspired by
the parasol of ebony feathers adorning its head, but it
is an umbrella in more than one way. Umbrellabirds
occupy large home ranges and exist at low popula-
tion densities. They migrate altitudinally in response
to seasonally changing fruit availability. If enough
forest along an elevational gradient were set aside to
ensure the survival of umbrellabirds, it would safe-
guard many other Atlantic slope animals and plants.
Even though some of the most basic aspects of the
biology of tapirs and Three-wattled Bellbirds are un-
known, protecting sufficient habitat to ensure their
persistence will protect thousands of species of plants,
invertebrates, fungi, and microbes whose biology will
remain unknown for decades.
In ecosystems throughout Central America, top
carnivores have been eliminated, populations have
become inbred, and ecosystem functions have been
disrupted. Although Monteverde remains pristine by
comparison, more aggressive population-centered
management may someday be necessary, including
transplants of individuals between populations to
increase local genetic diversity, supplementation of
food resources, addition or protection of nesting sites,
culling of excess individuals or of pest species, and
habitat manipulations such as prescribed burning.
The effort has begun, with the provision of nest boxes
for quetzals (M. Fogden, pers. comm.), devices to
guard quetzal nests from predators (G. Powell, pers.
comm.), captive breeding of amphibians at the Golden
Toad Research Laboratory (A. Pounds, pers. comm.;
see Pounds and Brenes, "Golden Toad Laboratory,"
pp. 171-172), and the Monteverde Conservation
League's reforestation projects (see the essays by Harvey,
p. 450, and Nielsen and DeRosier, p. 448). Understand-
ing subtle aspects of animal behavior will be important
in the management of critical species (Curio 1996).
Increasingly, conservation practice favors ecosystem
management over critical species- or population-cen-
tered management (Grumbine 1994; see Chap. 10, Con-
servation Institutions). It is widely accepted that eco-
system management is effective for many of the reasons
that make the protection of umbrella species attractive.
Ecosystem management relies less on labor-intensive
and costly species-specific data collection and inter-
vention. By preserving intact ecosystems of sufficient
size, normal ecological processes can take place. Nu-
merous species depend on disturbances to reduce
populations of competitors or predators, or to provide
opportunities for reproduction or open new sites for
colonization. For example, in Monteverde landslides
and wind storms are essential perturbations. Unless
ecosystems are large enough, such processes may not
take place or may produce different effects, and dis-
turbance-dependent species may be lost.
Monteverde's protected areas should apply land-
scape principles and reflect a design that has become
standard in conservation biology. First, core zones
should be established, centered on unique natural fea-
tures or imperiled populations and expansive enough
to support sufficiently large populations to avoid exces-
sive inbreeding or demographic stochasticity. Within
core zones, ecotourism, hunting, harvesting, and other
potentially disruptive activities should be prohibited or
tightly regulated. Second, buffer zones, where tourism,
forestry, and wildlife management are concentrated,
should surround core areas. Third, transition zones,
where more intensive activities such as agriculture and
residences occur, should be located outside of buffer
areas (Meffe and Carroll 1997). Land-use practices such
as leaving isolated fruiting trees in pastures, which serve
as "stepping stones" for dispersing forest inhabitants,
or retaining or planting corridors and windbreaks to fa-
cilitate dispersal by animals between habitat patches can
be effective for conservation. Birds (including avian
seed dispersers) use agricultural windbreaks as corri-
dors in Monteverde (see Nielsen and DeRosier, "Wind-
breaks," p. 448), and as a consequence, windbreaks that
are connected to forest or that have remnant forest
trees harbor more species and higher densities of trees,
(Harvey 1999; see Harvey, "Windbreaks and Trees,"
pp. 450-451). Ironically, windbreaks were planted in
Monteverde to improve milk production and crop
yields and to sustain the agricultural activities that frag-
mented the forest in the first place. In the future, wind-
breaks should be situated more intentionally in a land-
scape context, using appropriate native plant species,
to provide more effective habitat corridors.430 Conservation Biology