The Covid-19 pandemic has tightened our social circles and narrowed the scope of our lives in ways no one imagined just a year ago. But the lockdown also brought an unexpected solace: Everyone, it seems, turned to baking sourdough bread. Social media has been overflowing with photos of frothy sourdough starters — many of them named, like a family pet — and the fresh-baked loaves that result. And though peak sourdough may have passed, many a fridge still contains that jar of starter.

Most home sourdough bakers know that their starter contains a vibrant herd of microbes, which leaven and flavor their bread. But where conventional breads rely on a single species of baker’s yeast — the microbial equivalent of a cattle ranch — sourdough is more like the Serengeti, a diverse ecosystem of interacting yeasts and bacteria. The nature of that ecosystem, and hence the flavor of the bread, is a profound expression of a particular time and place. Scientists are beginning to discover that the microbes in a sourdough depend not just on the native microbial flora of the baker’s house and hands, but also on other factors like the choice of flour, the temperature of the kitchen, and when and how often the starter is fed.

“When we study sourdough science, we learn that we know remarkably little for a technology that’s — what? — 12,000 years old,” says Anne Madden, a microbiologist at North Carolina State University. But even that limited knowledge is enough to cast light on a diverse, tumultuous microbial world — and provide a few hints to home bakers hoping to up their game. (We probably still have plenty of time to be baking, alas.)

Under the microscope, a sourdough starter is full of yeast cells, much smaller bacteria and bubbles of carbon dioxide produced by fermentation.


Rise of the microbes

Mix flour and water in a bowl and you have papier-mâché paste. But almost immediately, yeasts and bacteria from the environment and the flour itself begin feeding on the sugars in the flour, explains Erin McKenney, a microbial ecologist also at North Carolina State who has studied the events that unfurl as a sourdough starter forms. At first, just about any microbe can grow on this rich, new energy source, including spoilage bacteria. (That’s why brand-new sourdough starters often go through a black, putrid-smelling phase.)

But soon, conditions begin to change. One group of those early colonists begins to acidify the starter. By Day 3, these so-called lactic acid bacteria — named for one of the main acids they produce, which is also found in yogurt, cheese and other fermented milk products — have made the starter so acidic that many of the early colonists can’t survive, leaving only the lactic acid bacteria and a few acid-tolerant yeasts. This lactic acid, together with vinegary-smelling acetic acid, gives sourdough its characteristic tang.

It may also improve the nutritional quality of the bread, says microbiologist Guylaine Lacaze of the Belgian bakers’ supply and consulting company Puratos (she is aware of four as-yet-unpublished studies on the topic). The increased acidity activates an enzyme, phytase, that makes minerals like calcium and phosphate more available, she says.

By Days 10 to 14, the starter has settled into a stable state where yeasts and lactic acid bacteria grow vigorously, the yeasts producing enough carbon dioxide to leaven a loaf of bread. The starter is ready to use.

But the fact that new starters settle down within a couple of weeks doesn’t mean that they all end up with the same set of microbes. In one recent study, Madden and her colleagues shipped bags of the same flour to 18 professional bakers around the world, who then used the flour to create starters in their own kitchens using identical techniques. About a month later, the bakers and their starters convened in Belgium, where researchers used DNA sequencing to identify the microbes in each starter.

Even though all the bakers started with the same flour, their starters were all different. Most contained various strains of common baker’s yeast, Saccharomyces cerevisiae, along with a host of other yeasts in varying proportions, they found. The starters also contained a wide range of lactic acid bacteria, mostly in the genus Lactobacillus — though once again, the details varied widely from one starter to the next. Most microbes appeared to have come from the flour — a different draw each time­ — though a few also originated with the baker’s hands or kitchen.

 Bar graphs showing the diversity of yeasts and lactic acid bacteria that make up 18 different sourdough starters.
This chart shows yeasts (left) and lactic acid bacteria (right) that make up 18 different sourdough starters created by professional bakers around the world. The bakers began their starters with identical flour, yet ended up with dramatically different collections of sourdough microbes. Each genetically distinct species or strain is shown in a different color; note that the white space at the top of each bar represents microbes too rare to track separately.

Other research groups in Europe have seen similar diversity. “My conclusion is that every sourdough is different,” says Marco Gobbetti, a microbiologist at the Free University of Bozen-Bolzano in Italy. Indeed, he suspects, a constant flux of species may be the norm for any given sourdough over time, though what little evidence is available is still equivocal. This casts some doubt on treasured heirloom sourdoughs, some of which have been passed down for generations. While their owners may like to think that they’re baking with the same microbes their ancestors used, Gobbetti is skeptical.

But even if every sourdough is different, might they fall into several distinct groups based on the microbes that are present, in much the same way that terrestrial plant communities can be grouped into grasslands and forests despite a changing mix of species? The answer to that question might be coming soon. Elizabeth Landis, a microbiologist at Tufts University in Boston, and her colleagues (including Madden and McKenney) identified the microbes in 560 starters submitted by bakers around the world, then looked for recurrent groupings of microbes. Some species do appear to co-occur frequently, they found, perhaps because they specialize in feeding on distinct sugars. The yeast Kazachstania humilis, for example, can’t use the sugar maltose, which is therefore available for the lactic acid bacteria. (The paper describing these results is still under review, so Landis isn’t sharing details just yet.)

Each microbial community seems to produce its own unique flavor profile, too, McKenney says. Some produce more lactic acid, which gives a yogurty flavor; others yield a sharper, more vinegary note from lots of acetic acid. And because each species of microbe has slightly different metabolic pathways, each is likely to add other flavorful metabolic byproducts to the mix — a big reason sourdough tends to have a subtler, more complex flavor than ordinary bread. “You could compare it to one single flower compared to a nice bouquet of different flowers. The complexity of all these different compounds is what you find in a sourdough bread,” says Karl De Smedt, who maintains a library of sourdough starters at Puratos.

Care and feeding

Not everyone agrees that sourdough microbial communities are so variable. In commercial bakers’ sourdoughs, which are fed daily or even more often, the microbes always have plenty of food. That creates a race, with the fastest-reproducing microbes dominating over time, says Michael Gänzle, a food microbiologist at the University of Alberta, Canada. In the long run, he says, the winners are the yeast Kazachstania and the lactic acid bacterium Lactobacillus sanfranciscensis (recently renamed Fructilactobacillus sanfranciscensis).

That’s not necessarily good news for the resulting bread: L. sanfranciscensis grows fastest because it has one of the smallest genomes among lactic acid bacteria, which means it has fewer metabolic pathways and thus fewer flavor-producing by-products than other bacteria, Gänzle says. (Score one for home sourdoughs, which Landis says might be more diverse.)

But the flavor of a sourdough bread depends on more than just the species of microbes present in the starter. “You can have really different sourdoughs even if the microflora is the same,” says Lacaze. “It depends also on the recipe of the sourdough, the parameters of the culture.” Stiffer starters — that is, those made with a lower proportion of water — trap more oxygen within the dough, and this encourages lactic acid bacteria to produce sharper-tasting acetic acid; in runnier starters, the same bacteria produce softer-tasting lactic acid.

Photo of two sourdough starters dripping off a spoon. The one on the left is less liquid than the one on the right.
Little details make a big difference to a starter’s flavor. The starter on the left is made with less water. This makes a stiffer starter that traps more air, leading the bacteria in the starter to make more acetic acid, which gives a sharp flavor. The runnier starter on the right yields a milder-tasting loaf. 


Temperature matters, too. Lactic acid bacteria do best in relatively warm conditions, for example, so fermenting in a warm kitchen makes for a sourer dough, while cooler conditions lead to more of the fruity flavors produced by the yeast. Moreover, lactic acid bacteria, despite what you’d think, aren’t fond of highly acid environments. Home bakers who leave an acidic starter in a cold fridge for weeks between bakings can find they end up with a blander bread that lacks the distinctive tang contributed by the bacteria. (Pro tip: If you’re going to leave your starter in the fridge for longer than a week, make sure to refrigerate it immediately after adding fresh flour, when it’s least acidic. That, says Lacaze, will help the lactic acid bacteria survive the prolonged cold to acidify the rising dough.)

One of the biggest ways that bakers can influence the flavor of their sourdough bread is through their choice of flours for the starter. To demonstrate this, McKenney and her team made four starters each from 10 different grains. Because grains differ in the mix of sugars they make available to sourdough microbes — corn, for example, lacks a starch-digesting enzyme that creates maltose — they might lead to different sets of microbes and, hence, different flavors. And that’s exactly what McKenney found (again, the results are not yet published). Starters made from amaranth flour tended toward meaty, toasty aromas. Those made from teff (an African grain) and sorghum gave fermented smells, while emmer and buckwheat gave more vinegary starters.

So far, McKenney and other sourdough researchers have taken only baby steps toward designer sourdoughs: Their science has not yet caught up to folk wisdom. “People would like to know step by step: ‘How do I make the end product I desire?’” McKenney says. “We can’t begin to offer anything that’s better than common baking knowledge or the best practices you learn from blogs or talking to friends.”

More answers could be coming soon, thanks to new citizen-science initiatives. McKenney, Madden and their colleagues run the Wild Sourdough Project, which invites home bakers to experiment with flours and growing conditions and report their results. Similarly, Puratos has launched the Quest for Sourdough, where anyone from newbies to professionals can register their sourdough. Those with particularly interesting or unique starters might be invited to submit them to Puratos’s sourdough library for further analysis.

But sourdough is interesting to more than just bakers. Sourdough and other food fermentations such as those that give us cheese, sauerkraut and kimchi provide relatively simple, easy-to-handle model ecosystems for studying microbial ecology more generally. “There’s lots of insights you can gain from studying fermented foods that you can then transfer to more complex microbial communities as well,” says Paul Cotter, a microbiologist at the Teagasc Food Research Centre in Ireland, and co-author of an article on food microbiology in the Annual Review of Food Science and Technology.

Sourdough offers an additional benefit, especially relevant in pandemic times when the microbial world seems so full of threat. “Sourdough is this one space where we all agree, as a society, that microbes are helping us do wonderful things,” Madden says. “If you love sourdough, you love wild microbes in our lives.”

Editor's note: This story was updated on August 13, 2020, to correct an error in the authorship of a study. The study, on the microbial diversity of sourdough starters originating from identical bags of flour, should have been attributed to Anne Madden and her colleagues, not Erin McKenney and her colleagues, as was originally stated.

Editor’s note: This story was updated on August 18, 2020, to correct an error. The original stated that baker’s yeast could not use the sugar maltose. It can, although several other yeasts common in sourdough, such as Kazachstania humilis, cannot.

This article is part of Reset: The Science of Crisis & Recovery, an ongoing series exploring how the world is navigating the coronavirus pandemic, its consequences and the way forward. Reset is supported by a grant from the Alfred P. Sloan Foundation.