stressed ecosystems, if such species are also keystones and/or strong interactors,
as has been suggested (e.g. Scheffer & Carpenter, 2003 ; Solanet al., 2004). For
instance, in freshwaters pH has strong effects on the distribution of body mass,
biodiversity, ecosystem processes and food-web properties (Woodward &
Hildrew,2002a). At low pH fish populations are often replaced at the top of
the food web by much smaller invertebrate predators, which benefit from
mesopredator release (sensuCourchamp, Langlais & Sugihara, 1999 ). According
to food-web theory, this should reduce average interaction strength, as the size
disparity between consumers and resources shrinks and might account for
the apparently high stability of the communities of acidified freshwaters, and
their slow recovery following reversals of acidification (Woodward & Hildrew
2002a; Ledger & Hildrew, 2005 ).
Conclusions
Recent advances in both theory and empirical data have led to a tighter coupling
of the ecosystem and community approaches, and the implicit links between
structure (e.g. the distribution of body size) and process (e.g. nutrient cycling;
energy flux) are now starting to be addressed more rigorously. Metabolic theory
has been proposed as a suitable framework within which to develop an inte-
grated structural-functional approach, because basal metabolic rate, which is
negatively correlated with body size, affects pattern and process across multiple
levels of ecological organization, from individual organisms (e.g. growth and
survival rates) to entire ecosystems (e.g. nutrient cycling and productivity
(Woodwardet al., 2005a, b). In the context of these large-scale problems, and
the general theories proposed to address them, size seems likely to play a central
role, linking as it does individual-level processes with ecosystem structure.
Many of the component parts of this scaling process have been in place for
some time: the challenging thing now is to develop an understanding of how to
scale up to making the right connections.
It is interesting that for such a basic question we still have rather patchy and
heterogeneous information on the size relations of predators and their prey
(Warren, 2005 ; Humphries, this volume). More systematic data collation from a
wide range of ecosystems (see Broseet al., 2005) is obviously one key way
forward, but new data are also needed to allow the effects of scaling from the
individual to species to be examined. Size provides a way of building models that
take individual-level processes and explore their consequences across popula-
tions and communities, but at present much of the information we have about
how the mechanisms underpinning foraging and feeding dynamics of indivi-
duals comes from comparisons using intraspecific variation in body size. New,
individual-based, food-web models may provide a means of testing how to
generalize from individuals to species level aggregations and beyond. One
potential way forward is to construct food-web models that employ size-based
BODY SIZE AND PREDATORY INTERACTIONS IN FRESHWATERS 113