Do Sourdough Starters Change When You Relocate Them?
A Tale of Two Starters
By Tom Cucuzza, © The Sourdough Journey, October 2026
My wife and I enjoy traveling and we often sample sourdough breads in different bakeries and restaurants. My wife tells me she loves my sourdough, but she often asked whether we could bring home some of a bakery’s starter and keep it the same.
I typically recited the conventional wisdom: “If you bring someone else’s starter into your home and feed it your flour, in your environment, it will quickly change and become ‘your starter,” losing the uniqueness of the source.”
I had heard this repeated many times, but never quite believed it. In February 2025, I set out to test it.
I obtained a mature sourdough starter from a commercial source. Over the next fourteen months I maintained it in my kitchen using different feeding schedules, four different types of flour, local tap water, storage at different temperatures, and normal exposure to my kitchen environment.
In April 2026, I obtained a fresh sample of the original starter from the same baker and sent samples of both starters to a laboratory for DNA sequencing.
How much did the starters diverge over 14 months? The results surprised me, and challenged some widely held assumptions about how sourdough starters change over time.
How I Intentionally Mistreated My Starter
The baker had fed his starter twice a day since 2020 and never refrigerated it. This was a very well-maintained starter.
I suspected that following the same maintenance routine might produce little change, so I chose a different path. I intentionally mistreated my starter to try to push it to a breaking point. I drew the line at refrigerator storage, which some studies have shown can change a starter’s microbiology. Instead, I:
- Fed it 4 different types of supermarket bread flour
- Used local tap water from Cleveland and New York City
- Fed it twice a day about 50% of the time, once a day about 40% of the time, and skipped feedings entirely about 10% of the time.
- On six occasions, I left the starter unfed for 1-3 weeks at 50°F (10°C) while I traveled.
Compared to the routine the starter was accustomed to, I was a bad starter parent.
The DNA Testing
I had learned about a commercially available service in Bolzano, Italy where you can send samples of your sourdough starter for DNA testing – similar to 23andMe for sourdough starters. Testing costs approximately $400.
After 14 months of separate maintenance, I sent my sample and the fresh sample from the baker to Italy for testing.
Note: These tests compare two “endpoints.” I did not test the shared original starer from February 2025. I still have a frozen sample plan to test it. This test shows the differences between the two branches after 14 months, but it cannot tell us specifically which branch changed, or whether both changed.
The DNA Report
Sourdough starters typically have 1 or 2 dominant yeast species and 1 to 3 dominant bacteria species. The report identifies yeast and bacteria separately, showing each species’ share of its respective population. In the report, “dominant species” means the species that together account for at least 95% of the yeast or bacterial population.
The “identity” of a sourdough starter is typically defined by its dominant species. The report also identifies other detected “low abundance” species in each starter.
Scroll sideways to see all columns.
| Yeast Species | Source Starter | My Starter | Observations |
|---|---|---|---|
| Kazachstania humilis (Maudiozyma humilis) | 57.3% | 17.8% | Same two dominant yeasts; relative dominance reversed |
| Saccharomyces cerevisiae | 40.7% | 78.0% | |
| Dominant Yeast Subtotal | 98.0% | 95.8% | |
| Candida solani | <1% | 2.1% | Higher in my starter |
| Hanseniaspora uvarum | <1% | <1% | Both <1% |
| Candida saitoana | <1% | <1% | Both <1% |
| Candida anglica | — | 1.2% | Only in my starter |
| Bacterial Species | |||
| Fructilactobacillus sanfranciscensis | 99.7% | 99.8% | Essentially the same |
| Dominant Bacteria Subtotal | 99.7% | 99.8% | |
| Leuconostoc citreum/holzapfelii | <1% | <1% | Both <1% |
| Leuconostoc falkenbergense | <1% | — | Only in source starter |
| Acetobacter orientalis | — | <1% | Only in my starter |
| Acetobacter malorum/cerevisiae | — | <1% | Only in my starter |
| Lactiplantibacillus sp. | — | <1% | Only in my starter |
— = Not detected.
What does the report say?
After 14 months of separate maintenance, both starters shared the same two dominant yeast species and the same single dominant bacterial species. No new dominant yeast or bacteria species were identified.
Four low abundance species were also detected in both starters.
One new low abundance species was identified in the source starter and four in my starter.
The biggest difference between the starters is the population mix of dominant yeast species. In my starter, S. cerevisiae accounted for 78.0% of the yeast population and K. humilis for 17.8%. In the source starter, K. humilis led at 57.3% followed by S. cerevisiae at 40.7%.
On the bacteria side, the two starters were virtually identical in their dominant species: F. sanfranciscensis accounting for 99.7% and 99.8% of the bacterial population.
Are these two starters more similar or different?
The yeast proportions clearly differed. But how meaningful is it that both starters shared the same three dominant species?
To put that finding in context:
- More than 60 yeast species have been identified across two large sourdough studies. Both starters shared the same two dominant yeast species.
- Nearly 100 types of bacteria have been identified in those studies. Both starters shared the same single dominant bacterium.
- My analysis of over 500 cataloged starter samples identified more than 10,000 theoretical dominant-species combinations. Both starters shared the same dominant species combination.
- Across more than 500 samples analyzed, this particular combination of dominant species occurred in approximately 1% of samples. The species profile is uncommon.
If a researcher randomly collected sourdough starters, finding two starters with this same uncommon combination of dominant species would be like finding two matching needles in different haystacks. The shared low-abundance species make the resemblance even more striking.
Data source: Landis et al. (2021) study and dataset, von Gastrow et al. (2023) study and dataset, and author’s analysis.
Can you relocate a starter to another kitchen and “keep it the same?”
Yes, quite possibly.
Even under intentionally different maintenance routines the starters in this experiment had remarkably similar microbial profiles. Similar maintenance routines might have resulted in fewer differences.
Are the two starters exactly the same? No. Are they very similar? Yes. They are not “identical twins,” but I would describe them as “close siblings.”
These results challenged the conventional wisdom that had been repeated for years. I wanted to understand how these starters could be so similar. That question sent me on a months-long investigation through scientific research, experiments, and sourdough DNA datasets.
You can read about what I found in my new research paper, “Inside the Invisible World of Sourdough Starters.”
Support the Sourdough Journey
Donations help fund my independent research. If you would like to see more sourdough DNA studies like this, please consider making a donation at thesourdoughjourney.com/donate
Thank you for supporting independent sourdough research.