Community Ecology Processes, Models, and Applications

persistence, for many species by first approaching a
patch dynamics perspective, but as land use intensi-
fication proceeds the end result will be low-diversity
communities ofr-selected good dispersers.
Taken together, human alterations of habitats
and disturbance regimes will likely result in sub-
stantial decreases in habitat and landscape quality
for many species. However, for some species, the
landscape created by humans is of exceptionally
high quality. These species are admittedly a minor-
ity but often abundant and obnoxious ones, with
good dispersal ability and either generalistic re-
quirements or being specialists on monocultures.
Some general expectations from metacommu-
nity theory on the effects of land use intensification
are summarized in Box 9.1b and Fig. 9.2.
First, in human-dominated landscapes, increased
heterogeneity and connectivity is likely to lead to
higher diversity and a larger regional species pool
(Bentonet al. 2003). However, high dispersal and
high connectivity may homogenize landscapes and

decrease diversity. This is because species in the

region then will behave as one population (patchy

populations,sensuHarrison 1991), decreasing both

local and, if dispersal dominates over sorting, re-

gional diversity. A larger regional species pool in-

creases the possibility for local ecosystem services to

be maintained. Niche theory suggests that increased

diversity increases utilization of available resources,

at least at low levels of diversity. Also, a diverse

species pool increases the probability that at least

some species survive more variable environmental

conditions (Elmqvistet al. 2003; Loreauet al. 2003).

Second, dispersal ability and intensity influence

the effect of landscape context (connectivity) on

metacommunity dynamics. It is predicted that low

dispersal increases the importance of local sorting

in patches, while at intermediate dispersal landscape

factors are more important. At high dispersal, local

conditions in patches may again seem more impor-

tant, because then sorting can be expressed in all

patches.

Decline if dispersal

dominates over

sorting

High connectivity

allows local sorting

of best competitors

Only good dispersers

persist (assuming local

extinctions occur)

Mass effects of

good dispersers

decrease diversity

Local Regional

Local diversity

increased by

mass effects

Local

diversity

Local “ecosystem

functioning”

(b) (d)

(a) (c)

Average “ecosystem

functioning” in region

Effect on β-diversity

depends on variation

among patches and

on dominance of

dispersal vs. sorting

Good competitors can disperse

to all suitable patches

Note:

Regional diversity

(andβ-diversity)

Good dispersers

assumed to be less

efficient (trade-off)

Stochastic extinctions

and environmental

variation added

Only competitive extinctions

(species sorting)

Landscape connectivity Landscape connectivity

Dispersal to patch Low High

Figure 9.2Relationships between landscape connectivity and (a) local diversity, (b) local ecosystem functioning,
(c) regional diversity and (d) regional average ecosystem functioning. Suggested relationships are based on
metacommunity theory but may not always hold (see text).

122 APPLICATIONS