Figure 5.5. Montane Anole (Norops altae) displaying his dewlap in a territorial dispute with another
male. This species has declined at Monteverde. Photograph by Michael and Patricia Fogden.A patch of upland tropical forest ordinarily con-
tains fewer species of amphibians and reptiles than a
similar-sized patch of lowland rain forest (Scott 1976,
Duellman 1988). The number of leaf-litter species
found per plot at Monteverde was similar to that of
mid-elevation sites in Costa Rica's Braulio Carillo
National Park but was smaller than that of nearby
lowland sites such as Tortuguero (Fauth et al. 1989).
Seventy-nine species of amphibians and reptiles are
known from the highest elevations at Monteverde
(Zones 3 and 4; Appendix 8), compared to 135 in low-
land rain forest at La Selva (Donnelly 1994, Guyer
1994), despite the smaller sampling area of the latter.
These 79 species, however, are fewer than half of
those known from the total Monteverde area (161
species; Appendix 8). Much of the high overall spe-
cies richness is a consequence of the diverse array of
environments. As one travels up the rain-shadowed
Pacific slope, over the continental divide, and down
the wet Caribbean slope (Fig. 1.5), one encounters
different species of amphibians and reptiles in dif-
ferent areas. The maximum number of species known
from a single distributional zone is 52% of the over-
all total (Table 5.1). Differences in species composi-
tion between the two slopes contribute strongly to this
total; 102 species (64%) are restricted to one slope or
the other, extending in some cases to the vicinity of
the continental divide (Zones 3 and 4; Table 5.2;
Appendix 8). Of these species, 60% of the reptiles in-
habit the relatively dry Pacific side, compared to 30%
of the amphibians (Fisher exact test, P = .004). The
pattern is similar when lizards and snakes are com-
pared to anurans (60% vs. 27%; P = .002).
The percentage of species inhabiting the highest
elevations (Zones 3 and 4) is greater for amphibians
than it is for reptiles (63% vs. 41%; P = .004; Table
5.2). The same is true when anurans are compared to
lizards and snakes (60% vs. 41%; P= .015). Few spe-
cies, however, occur only in these uppermost life
zones—13% of the amphibians versus 3% of the rep-
tiles (P= .015). The difference is not significant when
anurans are compared to lizards and snakes (9% vs.
3%; P = .094). Although areas of greatest change in
species composition along climatic gradients may
correspond to boundaries between vegetation zones
(Heyer 1967), the above patterns illustrate that spe-
cies do not appear and drop out in perfect synchrony
at these boundaries. Communities are not discrete but
instead intergrade in a complex fashion.
The spatial pattern of the anuran declines at Monte-
verde helps to explain why so many species disap-
peared from the area. The 1987 crash affected all cli-
mate and vegetation zones in the region (Pounds et al.
1997) and thus all geographic components of species
richness. I compared the observed patterns of pres-
ence and absence to what they might have been if the
declines had affected only the Caribbean slope, only
the Pacific slope, or only the highest elevations (Zones154 Amphibians and Reptiles