our elected officials and their constituencies about the
costs of complacency. Local action is also important.
Without the work of the Tropical Science Center,
which manages the MCFP, and the Monteverde Con-
servation League, which manages the adjacent Inter-
national Children's Rain Forest (see Chap. 10, Con-
servation Institutions), the area's forests would be
reduced to a small island in a sea of agricultural land.
Deforestation would directly ravage biodiversity and,
by affecting local hydrology, would exacerbate the
effects of greenhouse warming.
We may be granted a reprieve from this global
threat. Temperatures have risen over the past century,
but in a series of upward and downward swings rather
than a steady progression (Houghton et al. 1996).
Cooling trends have characterized some decades.
Patterns suggest an inherent climate variability on
which the greenhouse effect is superimposed. Action
taken now may help to determine how much bio-
diversity survives until a possible reprieve comes.
Thus, applied conservation measures such as habitat
management, establishment of corridors for dispersal,
captive breeding and releases, and reintroduction of
species into previously occupied habitats may prove
valuable (see Pounds and Brenes, "The GTLC," p. 174).
Even if our only accomplishment is to provide another
generation of humans with the opportunity to know
the "optimism and inner meaning" of a glass frog's
song, the effort is worthwhile.
5.8 SummaryA sudden crash of amphibian populations in 1987 led
to the disappearance of the endemic Golden Toad and
many other species from seemingly undisturbed habi-
tats in the Monteverde region of Costa Rica's Cordil-
lera de Tilaran. This and similar cases in highland
areas of other continents have caused alarm and con-
troversy because of the suggestion that subtle agents
may be threatening biodiversity on a global scale.
Debate has focused on whether the observed patterns
differ from those expected from normal population
dynamics. Whereas arguments have hinged on stan-
dards of scientific proof and the scarcity of long-term
demographic data needed to judge whether a popu-
lation is in decline, diverse tropical faunas afford an
approach that does not rely on these data. A compari-
son of the number of disappearances at Monteverde
to the number that could be expected for an assem-
blage of demographically unstable populations sug-
gests that the declines go beyond natural fluctuations.
Although discussion has focused on amphibian
declines, reptiles have also been affected. Prey scar-
city may account for a decrease in the abundance ofcolubrid snakes; the species for which declines have
been documented eat anurans (or their eggs or larvae).
It is doubtful, however, that declines of anoline liz-
ards are a secondary consequence of amphibian de-
clines. The two may be components of a single phe-
nomenon. Both began in the late 1980s, have taken
place in seemingly undisturbed upland habitats, and
have led to disappearances of populations. A broad
taxonomic focus is needed to identify the causes of
the observed patterns and explore the implications.
We must rely on the available information on ecol-
ogy to guide future research. The spatial configuration
of the 1987 crash of anuran populations, viewed in re-
lation to local patterns of distribution and diversity,
helps explain why so many species disappeared from
the area. Much of the high overall species richness is a
consequence of a diverse array of climate and vegeta-
tion zones. The crash affected all of these zones and
thus all geographic components of diversity. Broad
patterns of distribution, examined in the context of
evolutionary history, suggest that Monteverde's am-
phibians are generally more vulnerable to regional or
global extinction than its reptiles. The former include
a higher proportion of upland species restricted to the
Cordilleras of Costa Rica and Western Panama. The
declines have mostly affected highland areas, and these
upland endemics have no potential for recolonizing
from nearby lowlands. The anuran species that have
disappeared represent a wide range of life histories and
reproductive behaviors. The diversity of oviposition
sites, including subterranean nests, casts doubt on the
hypothesis that UV radiation has caused declines by
reducing hatching success. Population structure as a
function of habitat use may help to explain why some
anuran species have disappeared while others have
persisted. Although virtually all anuran populations
crashed in 1987, species not dependent on bodies of
water were less likely to disappear than species asso-
ciated with aquatic habitats. In contrast to the island-
like population structure of the latter, the former tend
to be distributed more continuously over the land-
scape. The relatively large number of local populations
may reduce vulnerability to stochastic mechanisms of
extinction in the wake of deterministic declines.
It appears unlikely that a single cause can explain
the declines at Monteverde, but there may be a prin-
cipal underlying factor that has set the stage for
proximate causes of mortality. Epidemic disease may
have played a role. However, there is no evidence
that a single outbreak, spreading in a wavelike fash-
ion, has caused all the declines in Central America.
An alternative hypothesis is that a change in the
ecological context over a large area may have encour-
aged epidemics of the same, or different microparasites
at different times and places. Airborne pollution may170 Amphibians and Reptiles