This allows us to differentiate between the separate aspects of species (habitat
breadth through beta diversity, and the regional heterogeneity through the number
of available habitats). These values allow a mechanism for comparing different com-
munities (Schluter and Ricklefs 1993).
If a species lives in a local area or habitat independent of any other species pre-
sent, then one would expect that the more species there are in the surrounding region
the more there will be in any particular local area. Thus, there should be a linear
relationship between species richness locally and that on the larger regional scale.
Such an assumption, at face value, may seem unlikely since we know that there are
many interactions between species. However, a linear relationship could occur where
a young community that is still evolving may not yet have achieved its full comple-
ment of species, so that local area richness reflects that in the wider region – biotic
interactions may not have come fully into play. Alternatively, a local patch that receives
a large number of immigrants (such as one with many sink populations) relative to
the competitive abilities of the residents will also reflect the richness of the region
that produces the dispersers – dispersing species overwhelm the residents. This
linear relationship has been termed an unsaturatedpattern.
In contrast, if there are strong biotic interactions between species locally such that
many species are excluded from the community, then one would expect that after
an initial colonization period a limit to the number of species would occur locally
irrespective of the number available in the region. A plot of local versus regional species
richness would produce a curve with an upper limit. This has been called the satur-
atedpattern because no more species can be added to the local area (Srivastava 1999;
Hillebrand and Blenckner 2002).
Processes such as dispersal and interspecific competition for space underlie these
patterns. Initially, these patterns were used to infer the mechanism. However, several
different processes can produce either of the above patterns (Chave et al. 2002; Shurin
and Srivastava 2005). For example, facilitation through a keystone predator can increase
local richness, the amount dependent on the regional pool of species. This process
could override the saturating pattern of strong interspecific competition.
Another problem arises when one tries to identify local and regional scales. Shurin
and Srivastava (2005) showed that the pattern changes from saturated to unsaturated
as the ratio of local to region increases. Thus, the pattern depends on the scale of
the study. Many saturated communities have been overlooked because the local scale
was too large. In general, such patterns of biodiversity provide only weak evidence
for the underlying mechanisms structuring that diversity.
MacArthur and Wilson (1967) and MacArthur (1972) proposed that the diversity of
species in discrete ecosystems can be considered as a dynamic equilibrium, a balance
between the rate of immigrating species and the rate of local extinction of species.
They chose islands to illustrate this principle (Fig. 21.3). Starting with a newly formed
island, the rate of immigration of new species declines as the full complement of species
from the source (the mainland) is achieved. However, as species accumulate on the
island some of them are going extinct, and the probability of extinction increases as
more species arrive. Thus, where the two rates are equal we achieve the equilibrium
of species number, S, for that island.
The immigration rate must be determined by the distancethe island lies from the
mainland, so a distant island should receive a lower immigration rate than a near
ECOSYSTEM MANAGEMENT AND CONSERVATION 379
21.13 Island biogeography and dynamic processes of diversity
