differences in the feeding biology of the organisms involved (Aljetlawiet al.,
2004 ). Nonetheless, such experiments have, at least in some cases, provided
evidence to support the idea that size-selective predation is directed towards the
most profitable prey (e.g. Kislalioglu & Gibson, 1976 ; Pastorok,1981; Turesson
et al., 2002). Although optimal foraging theory predicts that consumers will feed
on prey of a certain size range, for the reasons outlined above, this selectivity
can manifest itself at a variety of temporal scales, from instantaneous decisions
about feeding on individual prey, which can respond to dynamic variables such
as hunger level and predation risk, through to the evolution of feeding adapta-
tions, such as gape size and mouthpart morphology. The latter are essentially
fixed traits for an individual at any given time and provide the size selection
‘envelope’. This fundamental food-size niche sets the size limits to predation for
an individual. Within these limits, prey may or may not be taken, or may be
taken with greater or lesser frequency, as a result of more dynamic foraging
decisions (e.g. Turessonet al., 2002), or through variation in the efficiency of the
foraging processes within those limits (e.g. Pastorok,1981).
This set of processes emphasizes the predators’ view of size selection, and the
detail of these processes comes from studies of single species of prey with single
species of predators (e.g. Woodward & Hildrew,2002c). If we are to extrapolate
to the broader food web these ideas have to be generalized across many prey and
predator species, and the simplest way to do this is to assume that prey are
equivalent in all respects except size. However, since the precise scalings
between body size and the costs and benefits will not be exactly the same for
each species (e.g. proportions of digestible tissue to total weight, mechanisms
for avoiding capture, etc.) species will have their own profit functions on the
size axis. In reality, however, many of these more subtle differences might not
be perceived by aquatic predators, at least not during the initial stages of the
predation process. Predators do not have any inherent reason to recognize
taxonomic distinctions unless they correspond to relevant functional differ-
ences from the predator’s perspective (i.e. those that affect foraging). This raises
the question of what proportion of the variation in the occurrence of feeding
interactions between a predator and a suite of prey is determined by size, as
opposed to other factors associated with prey type? If size predominates, then
modelling the patterns and determinants of trophic links in multispecies sys-
tems becomes easier, as feeding can essentially be modelled as contiguous on a
single niche axis (e.g. Williams & Martinez,2000).
Evidence from the Broadstone Stream food web supports the view that body
size is the primary determinant of diet: the first axis of an ordination of pred-
ators’ diets represented a gradient in relative body size between predators and
their prey, whilst other effects, such as foraging strategy and spatial overlap
were less important (Fig.6.2, redrawn after Woodward & Hildrew,2002b). The
relative contribution of taxonomic (or rather the morphological, physiological102 G. WOODWARD AND P. WARREN