communities, dispersal allowed species to recol-
onize habitats where they were driven extinct,
increasing overallb-andg-diversity.
5.7 Local–regional richness relationships
It is commonly observed that the relationship be-
tweeng-diversity anda-diversity – called local–
regional richness (LRR) relationships – is positive
(e.g. Ricklefs 1987, 2004; Cornell 1993; Rosenzweig
and Ziv 1999; Shurin and Srivastava 2005). That is,
as the number of species in the regional pool in-
creases, so does the number of species that coexist
in any given locality. Originally, the shape of this
relationship was thought to be able to discern
whether communities were saturated with species
(in which case the LRR should be asymptotic) or
whether they were unsaturated (in which case the
LRR should be linear) (e.g. Terborgh and Faaborg
1980; Cornell 1993). However, more recent analyses
have suggested that LRR relationships can be linear
even when communities are saturated with species,
or asymptotic even when communities are not
saturated (reviewed in Shurin and Srivastava 2005).
Despite the potential problems with inferring
causation from LRR patterns, they can provide
valuable information on metacommunity structure.
Using a dataset of severalDaphniaspecies co-occur-
ring in Baltic Sea rockpools (Bengtsson 1989, 1991),
combined with a modelling approach, Hugueny
et al. (2007) were able to discern between the patch
dynamic (Levins’ metapopulation) and source sink
models. Since these models have different under-
lying colonization and competition structures,
Hugeneyet al. were able to gain a deeper under-
standing of the processes structuring this commu-
nity. Additionally, multivariate analyses exploring
the effects of environmental conditions on patterns
of diversity allow a more direct comparison of how
these factors directly and indirectly influence local
and regional richness when the LRR is explicitly
considered (Harrisonet al. 2006b).
5.8 A synthesis of metacommunity models
In any metacommunity, a variety of processes, in-
cluding those related to neutral models (e.g. sto-
chasticity) and those related to niche models (e.g.
determinism), are operating simultaneously (Chase
2005, 2007; Gravelet al. 2006; Leibold and McPeek
2006; Adleret al. 2007; Clarket al. 2007). In analogy
to population genetics, where genetic drift is only
part of the equation and is tempered by the impor-
tance of natural selection, in metacommunity ecol-
ogy, the ecological drift invoked by neutral theory
(Hubbell 2001) is only part of the equation, and can
be tempered by deterministic factors, which Chase
(2007) referred to as ‘niche selection’.
When niche selection is strong, such as in low
productivity or highly disturbed communities, the
stochasticity associated with ecological drift should
be lower than when niche selection is weaker
(Booth and Larson 1999; Chase 2003, 2007). Indeed,
in an experimental study on the assembly of small
pond communities, Chase (2007) found that pat-
terns ofb-diversity (site-to-site differences in spe-
cies composition) were not different from what
would be expected from purely stochastic (neutral)
assembly when ponds were permanent. However,
when drought was experimentally imposed on one-
half of those ponds, community structure became
more similar between ponds, and patterns of
b-diversity were better predicted by expectations
based on models that incorporate species niches
(Chase 2007).
Another way to depict the same idea is to consider
the dual mechanisms that can influence coexis-
tence among species in a metacommunity. Co-
existence results from a balance between stabilizing
and equalizing forces (Chesson 2000; Adleret al.
2007). Equalizing factors serve to make species re-
sponses to variation in the environment similar, en-
abling them to persist in the same patches, while
stabilizing factors serve to make species responses
to the environment more different, i.e. niche differ-
entiation and partitioning (e.g. Chesson 2000). The
equalizing factors represent the components of spe-
cies coexistence that are neutral, whereas the stabi-
lizing factors represent the components of species
coexistence that are niche related (Adleret al.2007).
It is easy to see here how both niche and neutral
perspectives can (and indeed must) coexist in the
same conceptual space, since the neutral theory is
simply a special case of a more complete theoretical
construct.INCREASING SPATIO-TEMPORAL SCALES 65