Sourdough microbiome revealed

How the diversity of microbes in sourdough starters can affect the flavour and rise of the bread
23 February 2021

Interview with 

Elizabeth Landis, Tufts University and Angela Oliverio, The University of Colorado, Boulder


Sourdough bread


During the pandemic a lot of us took up baking, especially bread, and in particular sourdough bread. This is the oldest form of breadmaking and requires the use of what’s called a sourdough starter - essentially water mixed with flour and allowed to pick up yeasts and bacteria from the air so it ferments. The sourdough starter is what makes the bread rise. And now, scientists have collected over 600 sourdough starter samples from people’s kitchens across the USA and other parts of the world to find out which microbes are making it into people’s bread, and the effects different combinations of yeasts and bacteria might have on the final taste. Eva Higginbotham spoke with Angela Oliverio from The University of Colorado, Boulder, and Elizabeth Landis, who’s at Tufts…

Elizabeth - The first thing that we wanted to know was mainly just what microbes are in sourdough, because though people are fermenting flour and water in homes and bakeries all over the world, we really didn't have an idea of what microbes were in home fermentations of sourdough.

Eva - So what did you do?

Elizabeth - So the first thing that we did is we partnered with Rob Dunn's lab at North Carolina State University. They asked people to fill out a survey where they told us about their sourdoughs - so everything from where they lived and what type of home they had to whether they kept their sourdough starter in the fridge or at room temperature, and even down to whether or not they had pets in their homes. So they were telling us about the parameters around their sourdough starter. And then we asked people to send us samples of their sourdough so that we could sequence the microbes that are there and also culture some of those microbes to later do experiments with.

Eva - Angela, you also worked on this project. What did you do once you received the samples?

Angela - Yeah, so we actually extracted the DNA in those samples and used gene sequencing to look at what bacteria and fungi, and particularly yeast, were present in those samples. So we found a lot of different microbes, probably not surprising to any bakers out there we found a lot of Saccharomyces cerevisiae - it was definitely a dominate yeast. But we also found other more interesting or niche yeasts that are known to be associated with sourdough samples that you wouldn't be able to go out and buy at a store. And then we also found a lot of different species of lactic acid bacteria, Lactobacillus sanfranciscensis - which I bet that you can probably guess how that was named! -was the dominant bacteria in most sourdoughs where it occurred. And we found that within starters, there were often only one dominant species or bacterial species.

Eva - So there wasn't that much diversity within each individual sourdough starter, but there was quite a bit of diversity between different ones?

Angela - Exactly, yeah.

Eva - Did you find any patterns? Was it like some bits of the US have this dominant species versus other bits, or across the world were there differences?

Angela - Yeah, so we did look into this because some quintessential sourdough lore is that the sourdough is going to be different depending on where you are geographically, but we didn't find that taxonomic composition was correlated with geographic distance in the US at all. So it was overall a pretty poor predictor explaining any of the variation that we found across sourdough starters.

Eva - So if it isn't geography that's making a difference, what is making a difference between the different starters?

Angela - Great question! So we actually tested 33 types of metadata. It included things like the age of the starter, what sort of flour you were using in your starter - so like rye or wheat - and then a bunch of other sort of home characteristics. And then we also collected climatic factors, but altogether those predictors really only accounted for a little bit less than 10% of the overall variation in community composition.

Eva - I see, so is it still something of a mystery then how the diversity is generated?

Angela - Yeah, so we really started with more of the abiotic factors. But we also know that interactions amongst taxa also shape outcomes. And when we were looking at the data, we found some interesting correlative trends between different starter species. So for example, there were a couple of lactic acid bacteria that seemed to co-occur often together. Then Liz designed a series of experiments to look at the interactions between different yeast and bacterial taxa.

Eva - So Liz, what did you do to study the interaction between the different microbes in each starter?

Elizabeth - So we first autoclaved flour and tried to reduce the microbial load that was in the flour itself and mixed it with water and made these really controlled sourdoughs, where we had just one yeast and just one bacteria. And we tried that with all of the most abundant bacteria and yeast that we cultured from sourdough and found that generally we see the same patterns of microbes being able to grow together in the lab as we see in the world.

Eva - Did each of those different starters that you created behave differently in some way?

Elizabeth - We took a subset of 40 starters and looked for differences in the aromas of those doughs. So we partnered with an expert sensory panel and they did both sensory analysis and chemistry on those samples to determine differences in aroma, and then we looked at also the way that the dough rises over time.

Angela - We detected 123 volatile organic compounds. So those are basically the different compounds that sourdough starters are releasing, and those are thought as a way to capture what's contributing to the flavour profile of sourdough. So the sensory analysis yielded 14 dominant notes across the subset of 40 starters that were selected. And we're talking pretty expected notes such as yeasty, vinegary, but some pretty interesting notes as well, including acetic sour, green apple, fermented sour. We might've had one that was like toasted corn chips.

Eva - Are these like specialists smellers who smell the different starters and can say, "Oh, this is definitely apple"?

Angela - Exactly.

Eva - Okay. And were you able to match up then different combinations of microbes with different aromas and ,ultimately I guess, with the different tastes that the bread would have?

Angela - Right, yeah. So we didn't go as far as actually baking any bread, but we were able to link the functional outputs to what microbes were found in particular starters. I think it's a really cool proof of concept to show that these particular microbes really do yield a very specific sensory output. And this group of microbes was also directly linked to dough rise rates, which is just another functional output.


Add a comment