Community Ecology Processes, Models, and Applications

Among the 15 food webs analysed, however, only

four are from terrestrial ecosystems (Broom, Coa-

chella, EcoWEB60 and Grasslands); three have rath-

er low connectance; and one, Coachella, has very

high connectance. This variation among terrestrial

webs is unlikely to be due to the nature of data

collection, and has more to do with the types of

species in the food webs (e.g. seed eaters versus

parasitoids).

As in competitive and mutualistic networks,

transitive structures play an interesting role also in

food webs, based on consumer–resource inter-

actions. Soil food webs have been shown to contain

trophic loops that are suggested to contribute to

their stability (Neutelet al.2002). Also known as

hypercycles, such interaction structures may even

have played an important role in prebiotic evolu-

tion (Boerlijst and Hogeweg 1991).

1.4.2 Mechanisms

What mechanisms determine connectance, how

does it differ between food webs and how does it

vary with species richness? Broadly, one can sepa-

rate the candidate mechanisms into two classes:

those related to foraging ecology and those affect-

ing overall network stability. Foraging-based

mechanisms focus on why consumers eat the num-

ber of resources that they do, and how this depends

on traits of the consumer (such as body size; e.g.

Woodwardet al.2005) and the diversity of available

resources. Stability-based mechanisms focus on the

effects of connectance and species richness on dy-

namic stability, with the idea that unstable combi-

nations of connectance and species richness will be

less common in natural systems than stable combi-

nations. Clearly these are not mutually exclusive

mechanisms, though from here on we will focus

on foraging-based mechanisms.

Qualitative foraging-based explanations of con-

nectance rely on the fact that connectance is derived

from the summed number of resources of each of

the consumers in a food web. Consequently, if there

are many generalist consumers connectance will be

higher than if there are many specialists. An exam-

ple of this type of explanation of connectance is

provided by a study of the natural enemies of

aphid species in a meadow (Van Veenet al.2008).

Aphids are attacked by parasitoids, pathogens and

predators, and the hypothesis was made that para-

sitoid webs would have lower connectance than

pathogen ones, which in turn would have lower

connectance than predator ones. The idea was that

the intimate and prolonged interactions between

Benguela pelagic

Broadstone stream

Broom

Capinteria

Caricaie lakes

Coachella

EcoWEB41

EcoWEB60

Grasslands

Mill stream

Sierra lakes

Skipwith pond

Small reef

Tuesday Lake

Ythan

100 120 140 160

Species richness

80604020

0.0

0.1

0.2

0.3

0.4

Connectance

0.5

0.6

Figure 1.6Connectance and species richness of 15 food webs from Petcheyet al.(2008). Note how connectance
declines with increasing species richness.

18 SHAPE AND STRUCTURE