We have good news. From the 568 sourdough samples our participants sent us we have completed the first of many stages of identification of the life therein. Most sourdough starters contain both bacteria and fungi. The fungi produce carbon dioxide, the bacteria the acid (usually, we actually think that some of the fungi in sourdoughs are producing acid too). We haven’t identified the fungi yet. Soon though.

As for the bacteria, when we first looked at the results, we were in for several surprises. Though it took some steps to get there. In our analyses the first data we see are in the form of a file that contains the individual decoded DNA of the organisms present in the sourdough. Not the entirety of their DNA, but instead one gene that happens to be particularly good at telling us about the bacteria. Think of that gene as the title of the bacteria’s book. You can’t judge a book by its cover or title but it is a good place to start. We carry out this step using algorithms. There are too many individual decoded pieces of DNA for us to read  and organize on our own. We use the computer algorithms (or rather, Angela Oliverio did) to sort the bacteria into something akin to species (what microbiologists call “OTUs,” or operational taxonomic units. Way to go microbiologists in making something marvelous, the names of life forms, boring. Just to make discussion a little easier I’ll refer to the OTUs as species from here on out but know that they are a little bit different from species). Once the computer algorithm lumps all of the different decoded sequences into their species (akin to stacking books with the same name together), we get a sense both of how many kinds of bacteria we found and in which samples we found them.

The answer? How many kinds of bacteria are in sourdough? Probably you want to know. But first let me give you some context.

The Diverse World Hypothesis

The study of microbes around the world has shown us that nearly every habitat that is studied is astonishingly diverse. In part this diversity reflects the reality that to microbes each tiny bit of dirt or flesh is a whole world, a world sufficient for millions, billions or even trillions of individuals.  More individual bacteria live in your colon, for example, than there are individual trees on Earth (one hundred times more!).  Even a tiny, tiny sample of microbes can contain many resources, habitats, individuals and hence species. In a thimble of soil one can find ten thousand species. In a sample of feces, thousands of species. In a swab of microbes from a belly button, hundreds of species. On the basis of these results, one might expect many, many, species in sourdough starters and also expect sourdough starters from different regions of the world to be very different.  The world is big, especially to a bacterium.

The Simple Fermentation Hypothesis

But we also had reason to believe the exact opposite.  The many hundreds of studies of sourdoughs have tended to find just a small handful of bacteria and often a single (or no) fungus species in the average starter. This is what one finds anyway when one cultures (grows) the microbes from those sourdoughs on petri dishes as isolated single species all on their own. When we talked to our friends in the baking industry, this is what most expected. Each individual starter, they hypothesized, would tend to have a couple of bacteria species and most often a fungus. Sampling more sourdough starters, they hypothesized, would reveal different species because although each starter might have few species, collectively there might be many.

What did we find? Wait one more second. Let me spell out the expectations a little more clearly in terms of the units that biologists like to use. We talk about alpha diversity (the number of species found in a sample) and gamma diversity (the number of species found when all of the samples are considered together). We talk about these measures of diversity when thinking about birds or fish and we can talk about them with sourdough. The diverse world hypothesis suggests that the alpha diversity present in any particular sourdough should be high and that the gamma diversity of all sourdoughs should also be high. Plot the diversity of sourdoughs against the number of samples and it should continue to increase as more samples are taken. This is what soil samples look like. High alpha diversity, high and ever increasing gamma diversity. The simple fermentation hypothesis, on the other hand, predicts a low alpha diversity but agrees with the diverse world hypothesis in predicting a high gamma diversity (high and ever increasing).

So what do we find already?

So far our results present the most interesting possible scenario. They don’t fit with either of our hypotheses. The alpha diversity of samples is high, which fits with the diverse world hypothesis. The average sourdough had tens (and often many tens) of species, dominated by species of Lactobacillus and their relatives. You can see this in the figure at left (it is a sketch for good reason, we are just beginning to explore). Some of these species we know a lot about. Among the most common species were Lactobacillus plantarum, Lactobacillus brevis, Lactobacillus paralimentarius and three other species of the Lactobacillaceae. Together these bacteria accounted for a large part of the microbial abundance we encountered. But, they were almost always accompanied by many other species, each of which we anticipate contributes something, be it conspicuous or subtle, to the abilities of the individual sourdough starter. We think these other species were missed in classic studies of sourdough starters because although they grow readily in starters, they don’t grow as easily isolated on their own in petri dishes. 

But when we look across the starters, the results no longer match the diverse world hypothesis. The sourdough starters we were sent had been fed many different ingredients and made in many different regions (from Alaska to Australia with many samples in between). Some were hundreds of years old, others brand new. Yet, while the alpha diversity of sourdoughs was high, the gamma diversity was (relatively) low. Lower than we expected. Far lower. Overall, we found around three hundred species of bacteria in the sourdoughs, which superficially seems like a lot. But it is almost a hundred times fewer species of bacteria than would be found in a thimble full of soil. It is fewer kinds of bacteria than live in some belly buttons. And some individual starters had nearly one hundred species of bacteria in them, which is to say nearly a third of what we have encountered in sampling the starters of the world. The take home. We found high alpha diversity, relatively low gamma diversity and very low beta diversity (a new term, beta diversity is a measure of how much the species present in one sample differ from those in the next). Finally, if we plot a curve showing the number of new species we encountered as we considered more samples we find that the curve levels off; Adding more samples does not seem to yield more kinds of sourdough bacteria.

There are many caveats to what we have figured out so far. We expect to be building on the insights of this first analysis for a decade, or more. But we thought you would like to be in on the ground floor, to see and think about what we are seeing now even as we continue to explore. Pardon our dust, we are doing science. 

The biggest caveat is that we have so far only looked at bacteria, not fungi. The fungal data will be forthcoming (thank you Angela and Jessica!). What do we expect for the fungi? We now have three hypotheses for the fungi. Either the simple fermentation hypothesis or the diverse world hypothesis might still hold for the fungi. Or the fungi might be like the bacteria and show a pattern that doesn’t yet have a name (high alpha diversity, low gamma diversity). One thing we do know is that in our other studies, particularly those of homes, that while the bacteria in one site tend to be similar to those in another (and influenced more by human behaviors than by climate or geography), that the fungi show the opposite pattern. The fungi present in any particular house seem to be most influenced by climate, by soil type, and by region. Might the same be true for starters? We’ll know soon.

In the meantime, keep sending us your ideas, notes, stories and photos about your sourdough starters and bread (even if we haven’t sampled your starter). We love them. We can’t always respond quickly to every one of your emails, but we always read them. And, if you did send in your starter (and if we were able to identity its microbes) we will soon be sharing with you, on a map, the list of the species of bacteria, and then later fungi, we found in the starter. Some starters were shy and wouldn’t share their secrets but most did. We would like to share those species lists with the name of your starter and its story (but without your name). If you would prefer to not have your starter’s story shared, please let us know.

 

Image Credits:

Header Image: Lea Shell

Figures 1 and 2: Lauren Nichols