Statistical Inference for Food Webs

• Activity Number: 605
• Type: Invited
• Date/Time: Thursday, August 5, 2010 : 8:30 AM to 10:20 AM
• Sponsor: Section on Statistics and the Environment
• Author(s): Grace Chiu*+ and Joshua Gould and Anton H. Westveld
• Companies: CSIRO and Dalhousie University and University of Nevada at Las Vegas
• Keywords: Bayesian melding ; deterministic models ; food webs ; ecological networks ; mass balance ; social networks

Abstract

Quantifying aspects of dependence within a food web has largely been descriptive or deterministic; existing schemes to assess reliability of such quantification are often non-statistical. In this paper, we attempt to make formal quantitative inference for within-web dependence via statistical models that characterize (a) trophic relations, and (b) inter-compartmental transfer. For (a), we propose the use of statistical modeling for social-network data to identify trophic compartments and other relevant characteristics of the web. For (b), Bayesian melding is employed to incorporate traditional deterministic views on ecological network mass balance with statistical inference of physical quantities involved. Together, (a) and (b) can provide a more honest representation of the true state of nature, a formal quantitative assessment of food web characteristics, and new ecological insight.