Coexistence holes characterize the assembly and disassembly of multispecies systems

Abstract

A central goal of life science has been to understand the limits of species coexistence. However, we know surprisingly little about the structure of species coexistence below such limits, and how it affects the assembly and disassembly of ecological systems. Here we introduce a novel hypergraph-based formalism that fully captures the structure of coexistence in multispecies systems. Our formalism uncovers that, below its limits, coexistence in ecological systems has ubiquitous discontinuities that we call "coexistence holes." These coexistence holes do not occur arbitrarily but tend to obey patterns that make them predictable. We provide direct evidence showing that the biotic and abiotic constraints of empirical systems produce an over-representation of coexistence holes. By highlighting discontinuities in the form of coexistence holes, our work provides a new platform to uncover the order and structure of the assembly and disassembly of ecological systems.

Untangling the complexity of priority effects in multispecies communities

“…Though not as widely recognized, three other outcomes are also possible in 2-species competition: (IV) two species eventually coexist despite requiring a specific arrival order; (V) only one species survives or two species coexist depending on who arrives first; and (VI) the species that arrives late always replaces the species that arrives early. The presence of these other possibilities has been supported by empirical evidence (Drake, 1991;Warren et al, 2003;Amor et al, 2020;Angulo et al, 2020). Of these six scenarios, (III), (V) and (VI) all represent priority effects.…”

“…For example, let us focus on the persistent species combinations (i.e., knowing the nodes in assembly graphs). While a community with S species has 2 S ≠ 1 potential species combinations, it has been shown that only a small fraction of such combinations can persist (Angulo et al, 2020). Because of the sparsity of these persistent combinations, it can be possible to infer all persistent combinations using a minimum of S + 1 experiments in a community with S species (Maynard et al, 2020).…”

“…To explain the work required, we first focus on identifying the persistent species combinations (i.e., knowing the nodes in assembly graphs). While a community with S species has 2 S − 1 potential species combinations, it has been shown that only a small fraction of such combinations can persist (Angulo et al, 2020). Because of the sparsity of these persistent combinations, recently developed computational tools based on Bayesian inference (Maynard et al, 2020) are only interested in bottom-up assembly (i.e., starting with no species, and then adding species one by one), then persistent combinations that are unreachable via introduction of single species do not need to be mapped to construct the corresponding assembly graph.…”

“…For example, arrival order dictates one of the three well-studied outcomes of 2-species competition in the Lotka-Volterra model (Figure 1A) (Case, 2000): (I) coexistence (two species coexist regardless of who arrives first); (II) deterministic exclusion (the same species always excludes the other regardless of who arrives first); and (III) history-dependent exclusion (the species that arrives first excludes the other). Though not as widely recognized, three other outcomes are also possible in 2-species competition: (IV) two species eventually coexist even though it requires a specific arrival order; (V) only one species survives or two species coexist depending on who arrives first; and (VI) only the species that arrives late survives (Warren et al, 2003;Amor et al, 2020;Angulo et al, 2020). Of these six scenarios, (III), (V) and (VI) all represent history-dependent priority e ects.…”

Multi-Replicated Enrichment Communities as a Model System in Microbial Ecology

“…Communities from two soil inocula were assembled in minimal media supplemented with one, two, or three carbon sources for 10 transfers (48 h each; Estrela et al, 2020a). Shown is the total number of families (left) and exact sequence variants (ESVs; right) in each of the self-assembled communities at transfer 10. sequential invasion of species (Angulo et al, 2020;Serván and Allesina, 2020). These theoretical studies have posed fascinating questions, such as whether top-down and bottom-up community assembly are fundamentally different from one another and posited the existence of coexistence holes and forbidden assembly paths that, violating the empirical assembly rule described above, could be present when communities are assembled from the bottom-up.…”