of frequency- dependent selection, in which the intensity
of selection (in the form of predation) depends directly
on the abundance of the prey (such that rare species
have a selective advantage). The result of predator
pressure is to preserve diversity by preventing extinction
by competition (plate 9- 25).
Note that the predation hypothesis is basically
opposite to the interspecific competition hypothesis.
The competition hypothesis suggests that competition
among species promotes diversity by leading to
specialization, narrower niches, and tighter species
packing. The predation hypothesis suggests that
predators actually reduce interspecific competition, thus
permitting coexistence among competing species. The
predation hypothesis does not predict specialization.
Indeed, specialization would seem less likely to occur
because predators keep competition levels low. Likewise,
the predator hypothesis allows for wide niche overlap
among similar species.
Tropical forests contain impressive predator richness,
circumstantial evidence for the possible importance of
predator effects. In a study conducted by Roman Dial
and Joan Roughgarden in a tropical moist forest in
Puerto Rico, lizards in the genus Anolis were shown
to have strong effects on the arthropod community of
the rain forest canopy (plate 9-26). When lizards were
experimentally excluded, large arthropods such as orb-
weaver spiders, cockroaches, beetles, and katydids all
significantly increased.
What if predators are nonselective, even when prey
densities vary? John Terborgh suggested an intriguing
model of how predators, specifically cats, might be
inadvertently structuring rain forest communities.
Citing Louise Emmons’s work, Terborgh began by
stating that the Jaguar, Puma, and Ocelot all forage
nonselectively, taking whatever they encounter and
can catch. The populations of prey species correlate
directly with the frequency upon which each prey
species is taken by a cat. If agoutis represent 40% of prey
species, agouti remains show up in cat scat 40% of the
time. Predators are taking what they find as they find it.
Terborgh goes on to argue that since prey species differ
in their fecundities (rates of reproduction), predation by
nonselective cats could significantly reduce certain low-
fecundity prey populations. In other words, peccaries
can produce more offspring annually than pacas, so if
cats do not ever discriminate between peccaries and
pacas, pacas must decline more than peccaries, since
they cannot replenish their losses as quickly. Because
pacas and peccaries both eat many of the same things,
an ecologist might be tempted to conclude that paca
reduction was due to its losing the competition for food
with peccaries, never guessing that cat predation was
the real reason. Terborgh’s argument demonstrates that
predation can result in a loss of diversity, as well as act
to maintain it, if predation is, in fact, nonselective rather
than frequency dependent (plates 9- 27– 28).Some Broad Patterns of Tropical
Species RichnessIt should come as no surprise that animal groups such
as amphibians and reptiles, unable to physiologically
regulate body temperature, decline in diversity and
abundance with latitude. The Arctic and Antarctic are
inhospitable to crocodiles, lizards, snakes, and turtles.
These animals are ectothermic and thus unable to obtain
sufficient body heat during times of extreme cold. Such
harsh conditions do not occur in subtropical or tropical
latitudes, and thus reptiles thrive in such regions.
Amphibians are similar in that they too are ectothermic,
but they differ from reptiles in being dependent on an
aquatic or semiaquatic environment, especially for
reproduction, so they are even more limited by harsh
climate, and their species richness pattern is closely
correlated with water availability as well as warm
temperature. Thus, like reptiles, amphibians exhibit
highest species richness in low latitudes. Physiological
constraint undoubtedly is a major factor in the latitudinal
diversity gradient (LDG) for reptiles and amphibians.
Insects and other arthropods are also ectothermic,
but they are small and often able to overwinter as eggs
or pupae. Insect numbers may be impressive during
the short Arctic growing season, but arthropod species
richness is nonetheless far greater at lower latitudes. Like
various vertebrate groups, insects (among arthropods)
exhibit their highest species richness within rain forests.
LDGs show complex and nonrandom patterns. For
example, most of the increase in mammalian diversity
in the tropics is due largely to bats (Chiroptera) and
rodents (Rodentia). But monkeys, sloths, anteaters,
and various marsupials all contribute to enhanced
diversity. Mammalian diversity in Neotropical lowland
rain forests correlates with both productivity and
habitat characteristics. Louise Emmons showed that
the density and number of species of Amazonian
mammals (excluding bats) correlate positively with150 chapter 9 why are there so many species?