Inferring ecosystem networks as information flows

Jie Li,Matteo Convertino

doi:10.1038/s41598-021-86476-9

Abstract

The detection of causal interactions is of great importance when inferring complex ecosystem functional and structural networks for basic and applied research. Convergent cross mapping (CCM) based on nonlinear state-space reconstruction made substantial progress about network inference by measuring how well historical values of one variable can reliably estimate states of other variables. Here we investigate the ability of a developed optimal information flow (OIF) ecosystem model to infer bidirectional causality and compare that to CCM. Results from synthetic datasets generated by a simple predator-prey model, data of a real-world sardine-anchovy-temperature system and of a multispecies fish ecosystem highlight that the proposed OIF performs better than CCM to predict population and community patterns. Specifically, OIF provides a larger gradient of inferred interactions, higher point-value accuracy and smaller fluctuations of interactions and alpha-diversity including their characteristic time delays. We propose an optimal threshold on inferred interactions that maximize accuracy in predicting fluctuations of effective alpha-diversity, defined as the count of model-inferred interacting species. Overall OIF outperforms all other models in assessing predictive causality (also in terms of computational complexity) due to the explicit consideration of synchronization, divergence and diversity of events that define model sensitivity, uncertainty and complexity. Thus, OIF offers a broad ecological information by extracting predictive causal networks of complex ecosystems from time-series data in the space-time continuum. The accurate inference of species interactions at any biological scale of organization is highly valuable because it allows to predict biodiversity changes, for instance as a function of climate and other anthropogenic stressors. This has practical implications for defining optimal ecosystem management and design, such as fish stock prioritization and delineation of marine protected areas based on derived collective multispecies assembly. OIF can be applied to any complex system and used for model evaluation and design where causality should be considered as non-linear predictability of diverse events of populations or communities.

Highlights

The detection of causal interactions is of great importance when inferring complex ecosystem functional and structural networks for basic and applied research
A plethora of conceptual approaches, frameworks and algorithmic tools including but not limited to Pearson’s correlation coefficient (PCC)[13, 14], Bayesian networks (BNs) and dynamic Bayesian networks (DBNs)[15,16,17,18,19], neural networks, graphical Gaussian models (GGMs)[20, 21], Wiener–Granger causality (GC) m odel[22], structural equation modeling (SEM)[23,24,25,26,27,28], convergent cross mapping (CCM)[29] and information-theoretic models[30,31,32,33] for instance, to tackle causal interactions and infer complex networks in terms of correlation, predictability and probability have been well established; most tools are solely tested on low-dimensional systems and some are even untested on ecosystems at different levels of complexity or simulated ones
In this paper we propose the Optimal Information Flow (OIF) model and assess its validity and performance in causality inference by comparing optimal information flow (OIF) to well-documented Convergent cross mapping (CCM) and correlation model

Summary

Introduction

The detection of causal interactions is of great importance when inferring complex ecosystem functional and structural networks for basic and applied research. The accurate inference of species interactions at any biological scale of organization is highly valuable because it allows to predict biodiversity changes, for instance as a function of climate and other anthropogenic stressors This has practical implications for defining optimal ecosystem management and design, such as fish stock prioritization and delineation of marine protected areas based on derived collective multispecies assembly. The flourishing development of complexity science[1, 2] has shed light on research questions and applications in many interdisciplinary fields, for instance, climate c hange3–5, epidemiology[6, 7] and ecosystem sciences at multiple scales[8,9,10] In this burgeoning science, complex network models play a central role in the quantitative analysis, synthesis and design (including predictions) of ecosystems and their visual representation. Ushio et al.[37] applied CCM to a complex fish ecosystems with 15 species after removing seasonality from abundance data in order to assess “true” or biological interactions

Methods

Results

Discussion

Conclusion