Community Ecology Processes, Models, and Applications

5.9 Adding food web interactions into the equation

All of the models discussed above consider species
interactions only within a single trophic level. Al-
though many of the empirical studies above implic-
itly considered both predator and prey species in
their analyses, few explicitly considered the role
that food web interactions might play. There is
increasing evidence that not only do predators
play an important role in structuring communities
(reviewed in Chaseet al. 2002), but that they are
expected to respond differentially to spatial factors
relative to prey communities, thus indirectly alter-
ing metacommunity structure (Holtet al. 1999; Holt
and Hoopes 2005).
Predators often are more sensitive to habitat area
than prey species (reviewed in Holt and Hoopes
2005; Ryall and Fahrig 2006), primarily because
predators typically need larger areas in order to
maintain viable populations. Smaller habitat
patches are often less likely to maintain large pre-
dators, and thus can have higher diversity and
abundance of prey species than larger habitat
patches (e.g. Terborghet al. 2001; Gotelli and Ellison
2006; Ostmanet al. 2007). Further, because preda-
tors are more likely to be present in larger habitats,
and because predators can strongly influence prey
community structure, this can have a strong influ-
ence on species-area curves (Holtet al. 1999; Holt
and Hoopes 2005). For example, on rocky outcrops
(glades) in the Missouri Ozark mountains, Ostman
et al. (2007) found no relationship between habitat
area and insect species richness. However, when
the occurrence of the voracious insectivorous lizard
Crotophytus collaris, primarily only found on larger
glades, was statistically controlled, significant spe-
cies-area relationships emerged, albeit with differ-
ent slopes with and without the predatory lizards
(Fig. 5.3).
Habitat isolation can also disproportionately in-
fluence predators, and thus alter prey metacommu-
nity structure (reviewed in Holt and Hoopes 2005).
Several empirical studies have shown that more
isolated habitats have lower rates of predation
(and parasitism), which in turn can alter the struc-
ture of the entire food web (Zabel and Tscharntke
1998; Kruess and Tscharntke 2000; Watts and

Didham 2006; Shulman and Chase 2007). In pond-

dwelling invertebrate communities, Shulman and

Chase (2007) showed that satellite ponds that were

more isolated from the source pond had lower

predator–prey species richness ratios relative to

ponds that were closer to the source pond. This

resulted primarily because predator richness was

highest near the source pond, but decreased with

distance, allowing prey species richness to increase

farther from the source.

However, predators may also have larger home

areas and be able to utilize a large area of a set of

patches that appear more isolated to prey species

(Oksanen 1990; Van de Koppelet al.2005). This may

result in different effects of spatial structure on

predator–prey interactions, as it basically decou-

ples predators from single prey populations and

may lead to local overexploitation (e.g. van de

Koppelet al. 2005). Hence predators may integrate

dynamics over several local ecosystems, which may

be similar or of varying quality (see below).

5.10 Cross-ecosystem boundaries

There is a growing body of evidence suggesting

that resource subsidies and predators (often termed

bottom-up and top-down regulation, respectively)

frequently cross traditionally defined ecosystem

boundaries (e.g. Poliset al. 2004). Such fluxes of

resources and consumers are not generally consid-

ered under traditional metacommunity ecology,

1.4

1.3

1.2

1.1

–1.5 –1 –0.5

z = 0.27

z = 0.35

Glade size (log 10 ha)

log

10

(arthropod sp.richness)

0 0.5

1

Figure 5.3Effects of patch area on insect species richness

in Ozark glades with and without insectivorous lizard

predators. Redrawn from Ostmanet al.(2007).

66 SPACE AND TIME