Geoscience Reference
In-Depth Information
the same community (or pattern), but few studies to date have been able to
use multiple approaches in one system, mostly due to the limited availability
of suitably taxonomically resolved, individual-level data that is needed as the
basis of applying both size- and species-based perspectives simultaneously
(but see
Reuman et al., 2008
).
Studies of trophic interactions have recently begun to recognise the bene-
fits offered by such a plurality of views, which when used to examine the same
question can provide a more complete answer. A recent example is that of
Woodward et al. (2010)
, who constructed food webs in which nodes were size
classes, rather than the more traditional representation of trophic species.
The size- and taxonomic-based food webs they constructed exhibited very
different structures, and the former were also much more amenable to
predictive modelling of network structure.
In traditional food webs, there are unexpected cases reported of smaller
species that appear to be feeding upon much larger ones (
Woodward and
Warren, 2007
). Often this is a consequence of the practice of using species
averages. If the size distributions of the two species overlapped this pattern
could result if large individuals of the small species fed on small individuals of
the larger species, creating a mirage that obscures the true extent of size
structuring within the web (
Ings et al., 2009; Woodward and Warren, 2007
).
Similarly, averaging might underestimate true predator-prey mass ratios (i.e.
among those individuals that are actually interacting within a feeding link),
by up to two orders of magnitude (
Woodward and Warren, 2007
). However,
these artefactual patterns often disappear once we adopt the size-class-based
food webs approach (
Woodward and Warren, 2007; Woodward et al., 2010
).
C. Individual-Based Food Webs: An Emerging Field
To view food webs in terms of both size and taxonomy, datasets examining
trophic interactions must ultimately be based at the individual level (or at
least contain some intraspecific information). The above difficulties with past
studies of trophic relations and attempts to bridge the terrestrial-aquatic
divide have led to a call for empirical data to be recorded on the sizes of
interacting individuals within local communities, to allow the analysis of size
structure with as little aggregation as possible (
Ings et al., 2009; Woodward
and Warren, 2007; Woodward et al., 2010
). Data collected at the individual
level also enable us to create a firm foundation to scale across levels of
ecological organisation. Such datasets allow examination of patterns of size
structure at the individual, population and community level and could be
used to infer how these patterns are linked, while avoiding some of the pitfalls
of averaging which have befallen past studies. To date, such individual-based
data have been used to examine some of these patterns in a very small
