Tropical Forest Community Ecology

280 Lee A. Dyer

TROPHIC CASCADES

Definitions

The term “trophic cascades” has been defined in
many ways, which has created problems (Hunter
2001), but the most restrictive definition is: a
measurable increase in primary productivity due
to negative effects of predators on herbivore
biomass (Paine 1980, Power 1990, Carpenter and
Kitchell 1993). Defined as such, the trophic cas-
cade hypothesis is also known as the “green world
hypothesis” (Polis 1999) and was first proposed by
Hairstonet al. (1960; HSS, for Hairston, Smith,
and Slobodkin). Here I use the HSS definition of a
trophic cascade, but there are many other types
of trophic cascades hypotheses that are poten-
tially important forces in terrestrial systems and
they fall under a more general definition provided
in theoretical and empirical studies (Hunter and
Price 1992, Carpenter and Kitchell 1993, Polis
1999, Halaj and Wise 2001, Dyer and Letourneau
2003, Letourneauet al. 2004, Schmitz 2004,
Schmitzet al. 2004) – indirect effects of one
trophic level on a non-adjacent level.This includes
indirect effects among individual species or entire
trophic levels, with the effects acting on densi-
ties, traits, or community parameters, such as
species richness (Figure 16.1).Two additional cas-
cades hypotheses that I consider here are the
“trait-mediated cascade” and the “diversity cas-
cade,” both of which could be important in
tropical communities. A trait-mediated trophic
cascade is a change in plant biomass caused by
modifications in herbivore foraging behavior in
the presence of predators (Schmitzet al. 2004).
A diversity cascade is an indirect effect of diver-
sity at one trophic level on a non-adjacent trophic
level (Dyer and Letourneau 2003). No trophic cas-
cade hypothesis has been fully tested in a tropical
system (Dyer and Coley 2001).
Trophic cascades hypotheses have been
extended to the ecosystem exploitation hypothe-
sis (EEH), which incorporates variation in primary
productivity and generalizes predictions for even
and odd numbers of trophic levels that might
result along a productivity gradient (Fretwell
1977, 1987, Oksanenet al. 1981, Oksanen 1991,
Hairston and Hairston 1997). The HSS and EEH

Enemies

Herbivores

Plant chemistry Plant biomass

Resources

C3 B1

B 3

B2

C2

A1 C1

A2

Figure 16.1 A simplified path diagram of selected

direct and indirect effects that are examined in tropical

food webs. Solid lines are direct effects, dashed lines are

indirect effects, arrowheads are positive effects, and

circle heads are negative effects. The letters next to the

lines could be path coefficients or any other statistic of

effect size, allowing comparisons between magnitudes

of direct and indirect effects. The top-down cascade

model predicts that B will be an important pathway.

Plant defense (Moenet al. 1993) and resource

availability (reviewed by Stam p2003) hy potheses

predict that A will be an important pathway. Resource

limitation or bottom-u pcascade models (Lindeman

1942, Slobodkin 1960, Hunter and Price 1992) predict

that C will be an important pathway. This chapter

focuses on pathways A and B.

models of multitrophic interactions have endured

numerous attacks. Some ecologists have dismissed

trophic cascades as one of many indirect effects

that are unlikely to be of great importance in ter-

restrial systems (Polis and Strong 1996, Menge

2000, Halaj and Wise 2001). The criticism most

relevant to tropical systems is that diverse terres-

trial systems are unlikely to contain linear trophic

levels, thus direct effects of one trophic level on

another are never likely to be strong enough

to cascade in any direction (Polis and Strong

1996). Omnivory, intra-guild predation, inter-

ference competition, spatial heterogeneity, prey

refugia, and other factors that putatively buffer