Inferring Ecosystem Networks as Information Flows

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 α-diversity including their characteristic time delays. We propose an optimal threshold on inferred interactions that maximize accuracy in predicting fluctuations of effective α-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.

The Eco-Evo Mandala: Simplifying Bacterioplankton Complexity into Ecohealth Signatures

“…This invariance across scales is associated with stable or scale-free distribution, manifesting the relative optimality of a community, and the exponent of the distribution is likely habitat-specific (modulated by environmental dynamics) rather than blueprinted by biology. Thus, it is much more appropriate to focus on the collective distribution of abundance and interactions (as information flow due to the data-driven approach of ecosystem monitoring [29]) revealing ecosystem organization, rather than treating any specific bacterium and its abundance in isolation. By decoupling and decoding the genetic, structural, and functional signals emanating from aquatic microbiomes (here, focusing on the bacterioplankton but extendable to eukaryotic and viral components), we demonstrated a novel means of evaluating ecosystem health (as community organization).…”

“…The biomass conversion process of the system involves the interspecies abundance variation resulting from competitive (or cooperative) interactions, making information theoretic approaches, such as transfer entropy [39], suitable for signal detection. Transfer entropy (TE) is an assumption-free and probabilistic means of maximizing uncertainty reduction about species interactions [29,40]. It is ideal because species interactions are usually non-normal, non-linear, bidirectional, highly unpredictable, and asynchronous: methods such as TE have been found to be superior to correlational approaches [41,42], and other models for functional network inference [29].…”

“…Transfer entropy (TE) is an assumption-free and probabilistic means of maximizing uncertainty reduction about species interactions [29,40]. It is ideal because species interactions are usually non-normal, non-linear, bidirectional, highly unpredictable, and asynchronous: methods such as TE have been found to be superior to correlational approaches [41,42], and other models for functional network inference [29]. Moreover, there is growing evidence of law-like behavior within communities, such as Taylor's law and the abundance-size spectrum [19,43], which are highly robust and seem to shift only in the presence of significant environmental pressure.…”

“…The community dynamics of the marine microbiome contain many signals informative of environmental pressure. The color of this "noise", in a dynamical sense, is extremely meaningful for the network topology underlying the observed ecological phenomena over time [27,29]. Traditional bioinformatics considers the dynamics of phylogenetic traits such as taxonomy or species abundance; for example, the diversity and abundances of populations within these communities change with habitat geomorphology and physicochemistry [30][31][32][33][34][35].…”

In.To. COVID-19 socio-epidemiological co-causality

“…Gradients in positivity as a function of cases or hospitalizations define risk perception patterns on which predictive models are calibrated to produce forecasts. Linear predictive models (selected upon linear socio-epidemiological relationships inferred on weekly data) are used to perform infection case and hospitalization forecasts whose predictive power is tested via non-linear predictability indicators (i.e., Transfer Entropy, TE, measuring the time-delayed uncertainty reduction between positivity and epidemiological information (see Li and Convertino 28 for details of TE as information flow), as discussed in the Material, Methods and Implementation section). These indicators are based on probability distribution functions of the variables of interest and yet they implicitly consider uncertainty distributions that are also attributable to other unexplained uncertainty sources.…”