COS 117-6
A mechanistic theory for food chain length

Thursday, August 14, 2014: 3:20 PM
Regency Blrm F, Hyatt Regency Hotel
Colette L. Ward, Integrative Biology, University of Guelph, Guelph, ON, Canada
Kevin S. McCann, Integrative Biology, University of Guelph, Guelph, ON, Canada
Background/Question/Methods

Multiple hypotheses propose an ostensibly disparate array of drivers of food chain length (FCL), including resource availability, dynamic stability, and aspects of ecosystem size. These hypotheses are well supported in experimental in vitro and in silico systems, however patterns in natural settings are contradictory. Here, we posit that these hypotheses effectively argue for 2 underlying mechanistic drivers of FCL: diversity and food web topology.

Results/Conclusions

We show that, for the case where diversity is constant, FCL is readily predicted by a simple conceptual framework for the effect of changes in vertical energy flux within food webs. Specifically, using a tri-trophic Lotka-Volterra model, we show that rising energy flux (e.g. increasing attack rates, resource carrying capacity) gives rise to increasingly top-heavy (wasp-waisted) biomass pyramids, which render omnivory energetically beneficial, in turn reducing FCL. We consider this theory in terms of empirically documented drivers of food chain length (productivity and ecosystem size) and demonstrate that these drivers are context-dependent. We then test the predictions of our theory using empirical data from lake and marine food webs. We show that the diversity mechanism is not supported in natural settings, and instead that ecosystem size is the most important driver of FCL in low-productivity systems, while productivity is most important in large and high-productivity systems. These results stand in contrast to classical hypotheses which predict a positive effect of productivity on FCL, and provide a mechanistic explanation for the Dynamic Stability hypothesis. Moreover, these results provide a unifying framework for drivers of FCL and reconcile the seemingly contradictory nature of published results for its drivers.