A 2024 metagenomic time-series study of spontaneous sourdough fermentations — 18 independent starters tracked over 14 days with daily 16S and ITS sequencing — found that every starter experienced a predictable stall between days 4 and 6 where pH stabilized, gas production dropped to near zero, and the microbial community underwent a sharp compositional transition. During this stall, Enterobacteriaceae populations that had dominated the first three days collapsed from above 70% relative abundance to below 5%, replaced by Lactobacillus species that would go on to dominate the mature culture. The stall was not a sign of failure — it was the transition itself, the interval during which the old community was dying and the new one had not yet reached functional density.
What makes this finding instructive is that the stall looked identical across all 18 starters despite wide variation in flour source, ambient temperature, and initial microbial inoculum. The trajectory was convergent: different starting communities, different environments, same bottleneck. The mechanism appears to be pH-driven — early Enterobacteriaceae metabolism produces enough organic acid to drop pH below their own tolerance threshold, creating a niche they cannot occupy but that lactobacilli exploit. The dominant community engineers its own extinction.
This is a pattern with resonance far beyond microbiology. Any system where early actors modify their environment in ways that favor their successors will exhibit a stall — a period of apparent stagnation that is actually a regime transition in progress. The temptation during the stall is to intervene, to diagnose failure and attempt correction. The sourdough study suggests that the wiser response is to recognize the stall for what it is — not a breakdown but a handoff, the quiet interval between one regime's collapse and another's rise.