Implications of scaled d15N fractionation for community predator–prey body mass ratio estimates in size-structured food webs

Reum, Jonathan, Jennings, Simon and Hunsicker, Mary (2015) Implications of scaled d15N fractionation for community predator–prey body mass ratio estimates in size-structured food webs. Journal of Animal Ecology, 84 (6). 1618–1627. ISSN 0021-8790

Full text not available from this repository. (Request a copy)

Abstract

Nitrogen stable isotope ratios (δ15N) may be used to estimate community-level relationships between trophic level (TL) and body size in size-structured food webs and hence the mean predator to prey body mass ratio (PPMR). In turn, PPMR is used to estimate mean food chain length, trophic transfer efficiency and rates of change in abundance with body mass (usually reported as slopes of size spectra) and to calibrate and validate food web models. When estimating TL, researchers had assumed that fractionation of δ15N (Δδ15N) did not change with TL. However, a recent meta-analysis indicated that this assumption was not as well supported by data as the assumption that Δδ15N scales negatively with the δ15N of prey. We collated existing fish community δ15N–body size data for the Northeast Atlantic and tropical Western Arabian Sea with new data from the Northeast Pacific. These data were used to estimate TL–body mass relationships and PPMR under constant and scaled Δδ15N assumptions, and to assess how the scaled Δδ15N assumption affects our understanding of the structure of these food webs. Adoption of the scaled Δδ15N approach markedly reduces the previously reported differences in TL at body mass among fish communities from different regions. With scaled Δδ15N, TL–body mass relationships became more positive and PPMR fell. Results implied that realized prey size in these size-structured fish communities are less variable than previously assumed and food chains potentially longer. The adoption of generic PPMR estimates for calibration and validation of size-based fish community models is better supported than hitherto assumed, but predicted slopes of community size spectra are more sensitive to a given change or error in realized PPMR when PPMR is small.

Item Type: Article
Faculty \ School: Faculty of Science > School of Environmental Sciences
Depositing User: Pure Connector
Date Deposited: 11 May 2016 13:00
Last Modified: 24 Jul 2019 22:10
URI: https://ueaeprints.uea.ac.uk/id/eprint/58646
DOI: 10.1111/1365-2656.12405

Actions (login required)

View Item View Item