The Enigma of Food Chain Length: Absence of Theoretical Evidence for Dynamic Constraints

Abstract

The question of what, if anything, limits the length of food chains has been problematic ever since Elton originally asked it decades ago. Here, we reexamine the major evidence for one of the chief hypotheses: dynamical constraints. Reportedly, food chain stability decreases as chain length increases. However, the separate roles of self‐damping (where a species shows density dependence independent of the densities of other species in the food chain model) and chain length have not been adequately determined. We examined food webs of varying chain length while controlling for self‐damping. In contrast to the prevailing hypothesis, we find a weak but significant increase in stability in longer chains when we hold constant the number of species and the number of self‐damping terms. Thus, the conclusion that longer food chains are more dynamically fragile, thereby setting an upper limit to food chain length, has at best very limited theoretical support. Our results indicate that community stability should tend to lengthen, not shorten, food chains. Mechanisms limiting the length of food chains remain to be convincingly demonstrated

Omnivory and the stability of simple food webs

“…All of the signi®cant results were in agreement with theoretical predictions that more omnivorous species and their prey will show more stable dynamics (McCann and Hastings 1997), and that food webs containing more species are less stable (Saunders 1978;Sterner et al 1997).…”

“…They found that return times were longer for systems with more species and where fewer species had self-damping terms. Contrary to Pimm and Lawton (1977), Sterner et al (1997) found that adding extra trophic levels while controlling for the number of self-damping terms produced a weak but signi®cant increase in stability (the opposite of return time). Third, it has become clear that it is not straightforward to classify some of the protozoan species used by Lawler and Morin (1993) as omnivores or non-omnivores.…”

“…Sterner et al (1997) found that the number of self-damping Table 2. B shows the coecient of variation of density for predators (shaded bars), and Colpidium striatum (unshaded bars) when with the predator species indicated beneath the bar.…”

“…Another dierence is that McCann and Hastings (1977) used non-linear (type II) functional responses rather than linear functional responses as in Pimm and Lawton (1977). Second, Saunders (1978) and Sterner et al (1997) tested the eect of number of trophic levels on the return time of food webs identical to those used by Pimm and Lawton (1977) except they controlled for the number of species and number of self-damping terms (where species have their own density dependence). They found that return times were longer for systems with more species and where fewer species had self-damping terms.…”

“…2. Food webs containing more species will be less stable than food chains containing fewer species (Saunders 1978;Sterner et al 1997).…”