Skip to main content
Are We Mapping Ecosystems or Models? How Reconstruction Choices Shape Food Web Inference

Are We Mapping Ecosystems or Models? How Reconstruction Choices Shape Food Web Inference

This is a Preprint and has not been peer reviewed. This is version 2 of this Preprint.

Add a Comment

You must log in to post a comment.


Comments

There are no comments or no comments have been made public for this article.

Downloads

Download Preprint

Supplementary Files

Authors

Tanya Strydom, Baran Karapunar , Jennifer A Dunne, Pincelli Hull, Crispin T.S. Little, Catalina Pimiento, Andy Ridgwell, Andrew P Beckerman , Alexander M Dunhill

Abstract

Ecological networks are increasingly used to compare community organisation, quantify interaction turnover, and infer responses to environmental change. However, interaction data are often incomplete, requiring researchers to predict network structure using approaches that encode different ecological assumptions. The extent to which these choices influence macroecological inference remains under explored. We investigate this problem using four successive marine assemblages spanning the Early Toarcian Extinction Event and subsequent recovery. We compare six food-web models representing feasibility-based, realised, and structural approaches to network construction. By quantifying network topology, pairwise interaction turnover, and extinction dynamics while holding species composition constant across models, we assess how model choice influences inferences about ecological change through time. Model choice explained substantially more variation in network topology than temporal turnover across extinction and recovery phases. Models also predicted markedly different sets of pairwise interactions, even when their overall network structures were similar. Although species-level vulnerability rankings were comparatively robust across models, link-dependent extinction outcomes were more model-dependent, indicating that conclusions about ecological dynamics can depend on how interactions are reconstructed. Our findings demonstrate that ecological networks are not neutral representations of communities, but model-dependent representations that can shape macroecological inference. Network reconstruction should therefore be treated as part of hypothesis specification rather than a technical step preceding analysis. Matching models to inferential goals and testing conclusions across alternative network representations are essential for distinguishing robust ecological patterns from those contingent on modelling assumptions.

DOI

https://doi.org/10.32942/X2XQ1M

Subjects

Ecology and Evolutionary Biology, Life Sciences, Other Ecology and Evolutionary Biology

Keywords

Network reconstruction, Macroecology, Ecological networks, Interaction turnover, Trophic organisation, Palaeoecology, Ecosystem resilience

Dates

Published: 2026-06-17 04:22

Last Updated: 2026-10-28 15:28

Older Versions

License

CC BY Attribution 4.0 International

Additional Metadata

Conflict of interest statement:
None

Data and Code Availability Statement:
The empirical data, derived network datasets, and all analysis scripts are archived in a public Zenodo repository ([DOI: 10.5281/zenodo.20444969](https://doi.org/10.5281/zenodo.20444969)), which includes a complete README file describing software dependencies, execution order, and reproduction instructions.

Language:
English

Metrics

Views: 463

Downloads: 158