Flour and feeding schedule shape the sourdough microbiome

A sourdough starter is one of the few microbial ecosystems people have kept in their kitchens for thousands of years without knowing exactly what is in it. A paper published in Microbiology Spectrum by a team from North Carolina State University tested one specific variable: whether the type of flour affects which microorganisms win.

Three flours were compared — all-purpose, bread and whole wheat — with starters run in parallel. Differences in bacterial composition turned out to depend on the flour: Companilactobacillus were more abundant in whole wheat starters, and Levilactobacillus in those fed on bread flour.

The fungal result is the more interesting one, because it runs against intuition. The dominant yeast in all the starters was the genus Kazachstania, not Saccharomyces cerevisiae — that is, not the yeast sold in a block and popularly credited with making dough rise. A sourdough is therefore not a diluted version of baker's yeast but a distinct system.

The scale of the phenomenon is larger than one study suggests. More than sixty types of bacteria and more than eighty yeast species have been described in sourdoughs worldwide. That is a great deal for one food culture kept in a jar, and the diversity translates directly into taste: microbial composition governs the acid profile, the aroma and the pace of fermentation.

The practical conclusion is simple and testable at home: change the flour and you change the microorganisms, and changing the microorganisms changes the taste of the bread. The caveat is simple too — the study used starters established under controlled conditions, and a years-old starter in a home kitchen has a history the experiment does not reproduce.

Where those microorganisms come from is a separate question. They arrive in the jar from the flour, the air, hands and vessels, and each feeding then applies selection: the winners are those that tolerate low pH, the rate of dilution and the sugars available. A sourdough is therefore cultivated but not inoculated — nobody adds a starter culture, and yet a community of repeatable structure forms every time.

It is worth noting that sourdough yeasts and baker's yeast do different jobs. Saccharomyces cerevisiae in a block has been selected on one criterion: producing carbon dioxide quickly. A sourdough rises more slowly because its microorganisms are doing something else as well — making organic acids that change the flavour, delay staling and alter the availability of some minerals in the flour.

This is a good place to recall how widely fungi sit in an everyday kitchen. Bread, beer, wine, blue cheese, soy sauce, tempeh and kōji are all products of fungal work — and in most of them not a single hypha is visible.

Date and scope — 29 August 2026

The article appeared online on 25 November 2025 and in the January 2026 issue. Controlled starters compared three flour types and two feeding schedules. Both factors changed bacterial and fungal composition, so flour alone does not “decide” everything; inoculum, temperature, water, ratios and the surrounding environment also matter. Sequencing describes the relative abundance of microorganisms, but it does not translate automatically into flavour, rise or bread safety. The result is useful because it identifies variables a baker can control deliberately, not because it establishes one recipe for a perfect starter.