dragonflies are voracious predators on a variety of
terrestrial insects, including species that are impor-
tant pollinators of terrestrial plants. Thus, fish, by
reducing the number of larval and consequently
adult dragonflies near ponds, facilitated higher le-
vels of pollination to plants in the surrounding
terrestrial community (Fig. 5.4).
The environmental conditions and interactions
that organisms face in one part of the landscape
can strongly influence their interactions in the
other. When these sorts of movements are frequent
and important, we need to expand our view of the
metacommunity. In doing so, we will need to de-
velop a more general theory that can explicitly ac-
count for cross-ecosystem flows of organisms. This
more general theory will have to consider interac-
tions among organisms that move between ecosys-
tem types in a landscape context. An organism will
be constrained by the availability of the ecosystem
type that is most limiting to the population. For
example, for anadramous fishes, the availability of
pelagic habitat for the adult stage will have little
influence on abundance if breeding and juvenile
habitats (e.g. streams) are degraded (Allendorf
et al. 1997). For many amphibians, the quality of
the pond in which larvae will develop is largely
irrelevant in the absence of adequate terrestrial
habitat (Semlitsch 1998). We can consider the avail-
ability of ecosystem types in a similar manner to the
way we consider resources in more traditional con-
sumer–resource models (Chase and Leibold 2003).
Specifically, if one ecosystem type that is needed for
an organism to persist is in limited supply, it is that
availability that will govern the dynamics of the
population, and thus the interactions of that organ-
ism in the other ecosystem types that it uses.
5.11 Conclusions
We have provided a broad overview of metacom-
munity ecology. By our definition, metacommunity
ecology encompasses a large number of spatio-tem-
poral processes that occur above the level of the
local community, and thus our treatment is a bit
broader than previous synthetic treatments of me-
tacommunities (Leiboldet al. 2004; Holyoaket al.
2005). In our view, metacommunity ecology encom-
passes the ‘mesoscale’ (sensuHolt 1993) of commu-
nity ecology, above the level of the local
community, but below that of large-scale biogeo-
graphic studies. Thus, there is a rather large spatial
scale and span of questions that can be addressed
under the metacommunity umbrella. These in-
clude: (1) patterns of metacommunity structure,
including the relationships betweena-,b- andg-
diversity; (2) patterns of interspecific interactions,
relative abundances and coexistence in local com-
munities that are linked through dispersal; (3) SARs
and their variation; (4) interactions among species
across traditional ecosystem boundaries.
Metacommunity ecology is a rapidly growing
field, but one in which considerable new insights,
synthesis, theories and empirical studies are needed.
We have discussed a number of recent advances
in metacommunity ecology, including diversity par-
titioning at different spatial scales, the interaction
between stochastic and deterministic factors, food
web interactions and cross-ecosystem boundaries.
Additionally, although beyond the scope of the
current chapter, evolutionary and historical factors
can also play a significant role in the structuring of
metacommunities, and this aspect is just beginning
to be considered (e.g. Urban and Skelly 2006;
Kraftet al. 2007). Thus, there is considerably more
work needed in these and other areas to gain a
better understanding of the processes that maintain
community diversity and structure at different
spatial scales. In addition, this understanding will
be critical in order to develop the concepts and tools
necessary to manage, conserve and restore ecosys-
tems in this increasingly human-dominated world
(see Chapter 9).Acknowledgements
We thank Jens A ̊stro ̈m, Wade Ryberg, Herman Ver-
hoef and Peter Morin for comments and discus-
sions that helped us to improve this chapter.68 SPACE AND TIME