lasted for more than ten years). It is unclear whether experiments of the type ‘a
wolf in the sheep field and the grass grows longer’ tell us anything about
community assembly other than that wolves eat sheep.
Complexity
Although habitat architecture is likely to influence species interactions, its
effects are confounded by species richness. Architecturally complex environ-
ments offer a wider variety of niches for species to exploit (Jeffries,1993).
Species-poor communities present few opportunities for prey specialization
and an increased likelihood of strong interactions between predators and
prey, with both of these conditions thought to favour trophic cascades (Polis &
Strong,1996).
The physical complexity of a habitat also affects how the organisms within
it interact, particularly by providing refugia. This, coupled with a wider array
of prey, increases the number of pathways through the web, increasing com-
partmentalization and imparting stability to the system (Neutelet al., 2002).
However, many types of spatial structure do not provide complete refugia for
prey but merely restrict movement, thereby reducing the encounter rate.
Models indicate that encounter rate is a function of the fractal dimension of
the environment, with movement and predation being constrained in more
complex environments (Johnson, Hatfield & Milne,1995; Cuddington & Yodzis,
2002 ). Due to reduced encounter rate, the foraging success of predatory insects
(Heck & Thoman,1981; Grevstad & Klepetka,1992) and fish (Crowder & Cooper,
1982 ; Winfield, 1986 ; Diehl, 1988 ; Nelson & Bensdorff, 1990 ) depends upon the
structural complexity of the environment (e.g. plant density). In lakes, zoo-
plankton (Timms & Moss,1984) and fish (Persson & Eklo ̈v, 1995) use the com-
plex structure of plant stands as refugia from predators, even though these
refugia do not absolutely exclude predators, presumably because encounter
rate and predation risk is reduced (Nelson & Bensdorff,1990).
The behaviour of both predators and prey may change as a consequence of
structural complexity, such that foraging, and hence encounter and predation
rates, are affected (Romareet al., 2003). In oyster beds, even though complexity
both reduces predation on mud crabs and enhances their activity, the presence
of toadfish affects their behaviour more, such that a cascade is still apparent in
the presence of complexity (Grabowski,2004). The cascade described in New
in large body-sized primary producers, to which the consumers have little
access, the passage of organic matter through the web is relatively small and a
triangular trophic pyramid produced. Furthermore, the presenceof large body--
sized primary producers adds structural complexity to the habitat, which
results in increased species richness and reduced interaction strength.BODY SIZE AND TROPHIC CASCADES IN LAKES 131