Revision 1

Accepted · first revision · record

Title: How predictable is the species composition of an assembling microbial community?
Plain: When a mixture of many microbes is grown in the same food again and again, the community that forms is similar at the level of broad families but differs in its exact species. How much of the final composition can be predicted in advance, and from what measurements, is not known.
Precise: Given a pool of S microbial strains and a defined medium with specified carbon sources, predict the relative abundances of the community reached after serial dilution, at family and species level, from measured monoculture traits, pairwise outcomes or metabolic capabilities. Quantify the predictability limit by comparing prediction error with replicate-to-replicate variation. An answer is the minimal data set and model that predict composition to within replicate variation, or a demonstration that species-level composition is intrinsically unpredictable (for example from multistability or historical contingency).
Settled by: Prospective predictions of community composition for new inocula and media, scored against replicated experiments.
Kind: phenomenon
Answer type: identification
References: (empty)
Literature status: partially-resolved
Status note: 2018 serial-dilution experiments found family-level composition reproducible across inocula and set by metabolic structure, with species-level variation unexplained.
Posed since: (empty)