Alternative stable states of microbiome structure and soil ecosystem functions

Theory predicts that biological communities can have multiple stable states in terms of their species/taxonomic compositions. The presence of such alternative stable states has been examined in classic ecological studies on the communities of macro-organisms (e.g., distinction between forest and savanna vegetation types). Nonetheless, it remains an essential challenge to extend the target of the discussion on multistability from macro-organismal systems to highly species-rich microbial systems. Identifying alternative stable states of taxonomically diverse microbial communities is a crucial step for predicting and controlling microbiome processes in light of classic ecological studies on community stability.

By targeting soil microbiomes, we inferred the stability landscapes of community structure based on a mathematical framework of statistical physics. We compiled a dataset involving 11 archaeal, 332 bacterial, and 240 fungal families detected from > 1,500 agroecosystem soil samples and applied the energy landscape analysis to estimate the stability/instability of observed taxonomic compositions. The statistical analysis suggested that both prokaryotic and fungal community structure could be classified into several stable states. We also found that the inferred alternative stable states differed greatly in their associations with crop disease prevalence in agroecosystems. We further inferred "tipping points", through which transitions between alternative stable states could occur.

Our results suggest that the structure of complex soil microbiomes can be categorized into alternative stable states, which potentially differ in ecosystem-level functioning. Such insights into the relationship between structure, stability, and functions of ecological communities will provide a basis for ecosystem restoration and the sustainable management of agroecosystems.