Geoscience Reference
In-Depth Information
body mass ratio for small predators and overestimate the ratio for large
predators (see above). Estimates of the strength of per capita interactions
between predator and prey species to be used in dynamic models should be
based on the body mass ratios of interacting prey and predators individuals,
(though this would also require reliable estimates of the size distributions
of the species in the whole environment not just those used to determine
diets). These are, in a sense, the true PPMRs, while body mass ratios
calculated from species averaging (mean mass of predator species divided
by mean mass of prey species) might lead to erroneous conclusions about the
dynamics and stability of ecological communities.
Many ecologists working with marine pelagic ecosystems have developed a
purely size-oriented modelling approach, in which individual organisms are
only classified by their size and taxonomic identity is not considered. Here, it
is the dynamics of the community size spectra itself that is of primary interest
(e.g.
Benoit and Rochet, 2004; Law et al., 2009; Silvert and Platt, 1978
).
A problem with this approach is that it cannot be used to explore the effects
of perturbations like species losses. To do that, size spectra need to be
disaggregated into component species, as pointed out by
Law et al. (2009)
.
Accounting for size as well as taxonomic identity of individuals makes it
possible to assess the risk and extent of secondary extinctions following the
loss of a species. However, so far, most models of this kind have been
relatively simple, dealing with webs consisting of only a few species (
De
Roos et al., 2003, 2008; Persson and De Roos, 2007; Persson et al., 2007
).
D. Caveats
The main rate-limiting step to conducting an individual-based comparative
analysis of species and size-based food webs has been the scarcity of suitable
data, and although we now have access to seven study systems, all of them
are aquatic and all bar the Celtic Sea are from running waters. It has often
been proposed that aquatic systems are inherently more strongly size
structured than terrestrial systems (
Shurin et al., 2006; Yvon-Durocher
et al., 2011
), and it would therefore be instructive to carry out similar studies
in the latter. Unfortunately, such data are not yet available, and this repre-
sents a considerable research gap that clearly needs to be addressed.
Another caveat that should be borne in mind is that, in common with all
other published food webs, our seven study systems are to, differing degrees,
partial descriptions of real networks, largely due to logistic constraints. For
instance, the Celtic Sea contains only fishes, whereas the Afon Hirnant, the
three Chilean Streams and Broadstone Stream contain only invertebrates. In
the latter case, this is because the stream is naturally acid and hence fishless
(although it has since been invaded by brown trout at pH has risen;
Layer
