Functional Responses Shape Node and Network Level Properties of a Simplified Boreal Food Web

• Published in: Frontiers in Ecology and Evolution
• Main Authors: Jenilee Gobin, Thomas J. Hossie, Rachael E. Derbyshire, Samuel Sonnega, Tucker W. Cambridge, Lee Scholl, Nicolas Diaz Kloch, Arthur Scully, Kiefer Thalen, Graeme Smith, Carly Scott, Francis Quinby, Jordan Reynolds, Hannah A. Miller, Holly Faithfull, Owen Lucas, Christopher Dennison, Jordan McDonald, Stan Boutin, Mark O’Donoghue, Charles J. Krebs, Rudy Boonstra, Dennis L. Murray
• Format: Article in Journal/Newspaper
• Language: English
• Published: Frontiers Media S.A. 2022
• Subjects: ecological communities; interaction strength; kill rate; Lepus americanus; Lynx canadensis; predator functional response; Evolution; QH359-425; Ecology; QH540-549.5; Canada; Yukon; Lynx
• Online Access: https://doi.org/10.3389/fevo.2022.898805; https://doaj.org/article/5c70d3dca03c45a48047c381bf137ef1

Ecological communities are fundamentally connected through a network of trophic interactions that are often complex and difficult to model. Substantial variation exists in the nature and magnitude of these interactions across various predators and prey and through time. However, the empirical data needed to characterize these relationships are difficult to obtain in natural systems, even for relatively simple food webs. Consequently, prey-dependent relationships and specifically the hyperbolic form (Holling’s Type II), in which prey consumption increases with prey density but ultimately becomes saturated or limited by the time spent handling prey, are most widely used albeit often without knowledge of their appropriateness. Here, we investigate the sensitivity of a simplified food web model for a natural, boreal system in the Kluane region of the Yukon, Canada to the type of functional response used. Intensive study of this community has permitted best-fit functional response relationships to be determined, which comprise linear (type I), hyperbolic (type II), sigmoidal (type III), prey- and ratio-dependent relationships, and inverse relationships where kill rates of alternate prey are driven by densities of the focal prey. We compare node- and network-level properties for a food web where interaction strengths are estimated using best-fit functional responses to one where interaction strengths are estimated exclusively using prey-dependent hyperbolic functional responses. We show that hyperbolic functional responses alone fail to capture important ecological interactions such as prey switching, surplus killing and caching, and predator interference, that in turn affect estimates of cumulative kill rates, vulnerability of prey, generality of predators, and connectance. Exclusive use of hyperbolic functional responses also affected trends observed in these metrics over time and underestimated annual variation in several metrics, which is important given that interaction strengths are typically estimated over ...