Local stability properties of complex, species-rich soil food webs with functional block structure.

Francisco De Castro,Diana H Wall,Aaron G Maule,Tancredi Caruso,Jacopo Grilli,Zoë Lindo,Thomas Fleming,Johnathan J Dalzell,Stefano Allesina,Stavros D Veresoglou,Mark Emmerson,Diego Garlaschelli,Franciska De Vries,Maarja Öpik,Silja Emilia Hannula,Thomas Bolger,Richard D Bardgett,Sina M Adl,Matthias C Rillig

doi:10.1002/ece3.8278

Abstract

Ecologists have long debated the properties that confer stability to complex, species‐rich ecological networks. Species‐level soil food webs are large and structured networks of central importance to ecosystem functioning. Here, we conducted an analysis of the stability properties of an up‐to‐date set of theoretical soil food web models that account both for realistic levels of species richness and the most recent views on the topological structure (who is connected to whom) of these food webs. The stability of the network was best explained by two factors: strong correlations between interaction strengths and the blocked, nonrandom trophic structure of the web. These two factors could stabilize our model food webs even at the high levels of species richness that are typically found in soil, and that would make random systems very unstable. Also, the stability of our soil food webs is well‐approximated by the cascade model. This result suggests that stability could emerge from the hierarchical structure of the functional organization of the web. Our study shows that under the assumption of equilibrium and small perturbations, theoretical soil food webs possess a topological structure that allows them to be complex yet more locally stable than their random counterpart. In particular, results strongly support the general hypothesis that the stability of rich and complex soil food webs is mostly driven by correlations in interaction strength and the organization of the soil food web into functional groups. The implication is that in real‐world food web, any force disrupting the functional structure and distribution pattern of interaction strengths (i.e., energy fluxes) of the soil food webs will destabilize the dynamics of the system, leading to species extinction and major changes in the relative abundances of species.

Highlights

The study of the stability of ecological communities has a long history (Allesina et al, 2015; Goodman, 1975; Grilli et al, 2017; Jacquet et al, 2016; May, 1972, 1973; Moore & Hunt, 1988; Rooney et al, 2006) but is a much debated topic (Donohue et al, 2013; Grimm & Wissel, 1997; Loreau & de Mazancourt, 2013). Much of this debate is focused on the relationship between stability and complexity: Ecological networks appear extremely complex in terms of species richness and connections, which are central to the topology of the web, and interaction strengths, which control energy fluxes between species
Our results show that both aspects of structure enhance the stability of our models of soil food webs compared with randomly assembled predator–prey networks, even when the soil food web is very large
The classic soil food web model emerged from the hypothesis that soil food webs consist of three main energy channels, with trophic interactions arranged mostly within them, and that stability and resilience properties depended on how energy flows through those channels

Summary

| INTRODUCTION

The study of the stability of ecological communities has a long history (Allesina et al, 2015; Goodman, 1975; Grilli et al, 2017; Jacquet et al, 2016; May, 1972, 1973; Moore & Hunt, 1988; Rooney et al, 2006) but is a much debated topic (Donohue et al, 2013; Grimm & Wissel, 1997; Loreau & de Mazancourt, 2013). We used random matrix theory (May, 1972; Tang et al, 2014) to generate the weights of the links and evaluate the implication of the constructed topology for the local stability properties of the population dynamics of these species-rich soil food web models. We tested these hypotheses with the overarching goal of assessing how plausible and generally large levels of species richness as well as the recent view of the topological properties of soil food webs impact the dynamic stability of belowground ecological networks, which are of vital importance for ecosystem functioning and represent an ideal model for large (i.e., very many species) and highly structured (i.e., nonrandom topologies) networks (Allesina et al, 2015; Allesina & Tang, 2012; Grilli et al, 2017)

| METHODS

| DISCUSSION