Editors note: This publication contains the video of the talk from the Fermentology webinar series, as well as a lightly edited transcript of the lecture.

Abstract

How do microbial communities change as fermentation techniques move around the world? What happens when people mix far-flung traditions and local ingredients in new ways in new places for new flavours? Joshua Evans talks about experiments with novel misos he has conducted among chefs and fermenters in some of Copenhagen’s leading kitchens. He discusses the ideas behind the experiments, shares some results, and explores what these culinary fermentation experiments tell us about microbial biogeography and domestication histories. He also reflects on the social context of these experiments and what it means to share and remix cultures in today’s world.

Josh is a PhD candidate in Geography and the Environment at the University of Oxford, and a visiting PhD student at the University of Copenhagen. Previously he was Lead Researcher at Nordic Food Lab, a non-profit institute in Copenhagen that conducted open-source gastronomic research for chefs, academics, and the public.

Watch the talk

Image attribution: Experimental misos, by Joshua Evans

Introduction

I research microbes and humans and how they change each other in novel fermentation practices.  Before I tell you what this looks like in practice (spoiler: it involves miso) I want to tell you a little bit about how I came to this kind of research.

It began at a place called Nordic Food Lab in Copenhagen. It doesn't exist anymore, but it was a non-profit organization set up by a restaurant called Noma. Maybe some of you have heard of this restaurant. It's become relatively renowned in the last decade or so for rediscovering and championing a lot of Nordic ingredients and foodways that either had been forgotten about or had never been developed gastronomically. 

The team at Noma set up this place called Nordic Food Lab to investigate the edible potential of the Nordic region and to share these results in an open source way. We worked with neglected and underutilized foods like wild plants, seaweeds, insects, and also with fermentation. 

The approach we were pursuing involved combining different techniques

from around the world with ingredients from the Nordic region to create flavors that were reminiscent of the traditional product but also flavors that had never really existed before. By now this approach has become fairly common in restaurants and even many home kitchens around the world. At that time, it was maybe a bit less so. It's great to see that so much has changed in the last five or ten years, not just in fermentation but in cooking more generally. 

A really good example of this technique was one of the first experiments that Noma developed, which colloquially they came to call peaso—this was a miso made using yellow peas instead of soybeans, which are traditionally used in Japan. We can think of this as a kind of translation, as if we were translating a book from Japanese into Danish, for example. 

In all these fermentation experiments, the main goal for the chefs was flavor. But as I learned more about the microbes that were involved in these processes—how quickly they reproduce, how quickly they can adapt to new environments—the more it became clear to me that these experiments in translating fermentation techniques might not only be creating new flavors, but they might be having bigger effects. They might be also shaping microbial life without these chefs realizing it or willing it. And so that idea became the seed of my subsequent PhD research. 

The Ecology of Novel Misos

I really wanted to investigate whether the pursuit of these new flavors in these novel fermentations might also be changing the microbes and the humans involved, and if so, how. To do this, my project has combined fermentation experiments, DNA sequencing, and ethnography to get at how these humans and microbes are shaping each other in what feminist science studies scholar Donna Haraway calls “a dance of mutual becoming”, which I find an apt phrase for this process. 

I carried out three experiments. The first was with miso at Noma. I wanted to see if making misos with different ingredients changes their microbial ecology. The second experiment was with koji. Koji is a fungus that is used to make miso (more on this in a minute). I looked at whether growing koji over and over again with its own spores, sort of like saving the seed from a plant and then using it to plant the next crop, at different sites would lead to each population becoming different from each other. Perhaps they would evolve differently.

The third experiment was with kombucha made by a colleague of mine at another restaurant in Copenhagen called Amass. I wanted to see whether kombuchas that he had made by assembling them from scratch were different from kombuchas grown in a more traditional way by using a mother (also known as a SCOBY, which stands for "symbiotic culture of bacteria and yeast"). All of these experiments were based on existing practices of novel fermenters in Copenhagen. I'm only going to focus on the first experiment here, the one about misos. 

Firstly, what is a miso? Miso is a traditional fermented product from Japan that usually combines cooked soybeans and a fungus called koji that is usually grown on some grain (often rice, but sometimes barley or wheat), and mixing them with salt. This mixture is fermented for a few months, or even a few years. This fermentation yields an umami-rich paste that can be used in all sorts of preparations, from soups and sauces to glazes and marinades. I'm sure that many of you are familiar with miso. Maybe you cook with it in your kitchen. But how does miso go from being an unappetizing mash of beans and grains to this delicious, amazing, umami rich paste? Well, a lot of it has to do with the koji. The scientific name for this fungus is Aspergillus oryzae. The fungus makes a lot of enzymes which break down some of the large molecules found in foods into smaller molecules. They turn proteins into amino acids, they turn starches into sugars, and turn fats into fatty acid chains. 

In the context of miso, these enzymes help create the amino acids that produce the umami taste, and the sugars created by the koji breaking down starches help bacteria and yeasts grow in the miso, which then produce acidity and other flavors. This is where miso gets a lot of its rich flavor from. Miso also has analogs in Korea and China, and it's found in a variety of forms throughout Japan. Some of them are more sweet, some are more savory; some are more salty, some are less salty; some are very young, some are much older. There are hundreds of different kinds. 

The experiment I conducted was based on Noma's standard miso recipe, which had been developed over eight or so years. It involves three parts of bean to two parts koji, plus 4% of that total weight in salt. It's quite a standard ratio, except for the lower salt concentration compared to traditional misos. We made eight different kinds of miso by varying the proteinous substrate. Instead of using soybeans, we used yellow peas, fava beans, and Gotland lentils, a specific type of lentil from an island in Sweden called Gotland.

We also cooked each of these three substrates normally in water and nixtamalized them. Nixtamalization is a technique from Mexico and Central America that's traditionally used to process corn. Many corns are still very tough even after you boil them in water and it's almost impossible to make a dough out of them. If you want to make masa out of corn, which is what you need to make tortillas and tamales and all of the different amazing Mexican preparations with corn, you have to nixtamalize it first. This involves cooking it in an alkaline solution (i.e. one that has a high pH – the opposite of acidic). Traditionally this might have been done with ash from a fire or with calcium hydroxide. Nowadays it's mainly done with chemicals like calcium hydroxide that are found naturally. This process breaks down some of the molecules, similarly to what the enzymes do in the koji, which then makes it possible to emulsify it into a paste. 

This is an example of a technique that Noma had developed and incorporated as part of its effort to  make “translated fermentations”. They were not only drawing on Japanese traditions and bringing them to Denmark, but they were then mixing this Japanese technique with a Mexican technique to yield a new and unprecedented fermented product. 

In addition to our experimental proteins, we also made misos with soybean as a control. This was what would be considered a normal, traditional miso. We also made a miso with rye bread. Rye bread is a protein-rich Danish staple that people eat all the time. You can make miso out of it. 

We made each of these eight misos (yellow pea, fava bean, Gotland lentil, soybean, and rye) in three replicates. In other words, we made three batches of each type of miso. The reason for this was to ensure that we had consistent data across batches. Often even if you use the same recipe with the same ingredients, it can still turn out very differently. This is because even very small differences in the starting conditions of a fermentation can yield bigger differences further down the road. By making three separate batches we could control for this possible batch variation. We fermented everything for three months at 28 degrees and ambient humidity. We sampled each one three times: at the start, in the middle of the fermentation, and at the end.

Once sampled, I brought them into the lab and extracted the DNA, purified the DNA, prepared it for sequencing, and then sequenced their DNA.

Along with the samples from the Noma experiment, I also received samples from six misos made using the same recipe in Japan. These were made at a restaurant in Tokyo called Inua that was affiliated with Noma. It was founded by a former R&D chef that used to work at Noma. Sadly, it had to close last year because of COVID, but I managed to get these samples before it closed. Even though there were no batch replicated with these (there's only one sample per type), we could still  compare them to the Noma misos. 

The first thing we found is that miso ecology does in fact vary according to the ingredients. But interestingly, some were more different than others. For example, the fava and soy misos had a very clear effect, whereas the pea and lentil misos had some effect but it was a bit less clear. And there are a few reasons why this could be.

The second result is that there is a really clear effect of the nixtamalization process. So the three nixtamalized misos that we made had a very similar ecology to each other even though the three ingredients that they were made from differed. Which is notable. 

We also found that when we compare the peaso made in Copenhagen to the peaso made in Japan with the same yellow peas and the same barley, we find that they had comparable compositions, but in different proportions. This is particularly interesting when we consider this finding in relation to the one about the impacts of nixtamalization, but more on that later. 

Overall the novel misos exhibit more diverse ecologies than “standard” misos. I use “standard” in quotes because we are comparing our data to misos that are described in the scientific literature. I've been searching for scientific papers on miso ecology for many months, and I've only come up with one or two. The papers that do exist are quite old and don't use sequencing techniques. They are based only on culturing.

There must be literature that has focused on the ecology of miso in Japan, possibly written in Japanese. But there seems to be very, very little written about miso in English. Hopefully, as more studies are published, we'll have more results to compare ours to. So far, it seems that, somewhat unsurprisingly, our novel misos, which mix and match ingredients, have more diverse ecologies. 

Taken together, our findings strongly suggest that how one makes the miso tends to matter more than geography, i.e. where the miso happens to be made.

Why it matters

But why, you might ask, does any of this matter? Aside from the fact that miso ecology is fascinating and everyone should just intrinsically care about it of course, one way to answer this question is to see how these findings inform the concept of microbial terroir. Some of you might have heard of this idea. It's an idea that became common in the food world around 2012 as fermentation was growing in popularity. It draws on the French notion of terroir in which specific food products emerge from within specific geographic regions.

The idea of microbial terroir imagines that specific fermented products are shaped by a unique microbiome associated with the location in which they were made. This seems intuitively true. Often this idea of microbial terroir, at least in English, can imply that the terroir is naturally occurring, that it exists prior to people, and that it doesn't change over time. 

However, a more complex understanding of microbial terroir has emerged more recently. Microbiologist Ben Wolfe and his colleagues carried out a well known study of microbial ecology of cheese rinds. They found that the main factor contributing to different community compositions in these cheese rinds was not geographic location but the specific human practices used to create the cheese. For example, how much salt was added to the cheese, how the curd was initially cut, and how the cheese was treated during aging. 

Other studies have similarly illustrated how different practices and arrangements of space in food production can shape this microbial terroir, such as in wineries and breweries and bakeries. One example is the large sourdough starter survey published by Rob Dunn’s lab.

My miso experiment contributes to this emerging field of research by showing how human interventions oriented towards producing flavor have effects on the microbiomes of the misos. These data add to our newer understanding of microbial terroir which suggests that it’s not fixed, but it's something we continue to shape through our ongoing practices. This is true whether or not we're trying to, and whether or not we're realizing it. In other words, we are always part of our microbial terroir.

There are plenty of directions for fascinating further research. One possibility would be to create the replicate batches of the Japanese misos in order to control for possible batch variation. That would make the comparison between the Japanese misos and the Copenhagen misos much more robust.

Another option would be to carry out deeper sequencing on these samples to try and resolve the microbiomes more specifically. In my study I was only able to identify the microbial community to the genus level. There are probably some important differences happening at the species and even strain level that my data weren’t able to resolve. In the future, and with a bit more funding, I’d love to  gather data on the species, and even particular strains, to see if these more detailed data suggest new or different conclusions.

Space miso

One very new extension of this miso experiment is one I’ve started in collaboration with Maggie Coblentz, a colleague at MIT Media Lab. Maggie and I sent a sample of my soybean miso up to the International Space Station to ferment for one month. We wanted to see how the environment of microgravity and being constantly bombarded with cosmic radiation might affect the miso ecology compared with earthbound samples made using the same recipe. 

We sent the miso up to space last spring in March and April and just recently got the sample back.  After a long delay due to COVID, we are finally going to start preparing them for sequencing and generating some data. If all goes well, we'll be able to compare the space miso with some of the existing misos to see whether space makes a difference. Hopefully this comparison can generate some further insights into how, in fermentation and especially in these novel fermentation practices that I've been describing, we're constantly participating in this dance of mutual becoming that Haraway talks about, being shaped and shaping the microbial world, whether we know it or will it or not. 

Acknowledgements

Big thanks to the Noma Fermentation Lab team for hosting me and for helping me in this experiment. To my lab group here at the University of Copenhagen, the Section for Evolutionary Genomics, for helping me with all the lab work and the sequencing. I have a great colleague, he's a biogeographer named Florent Mazel in Switzerland who has been helping me with all of the analysis and the number crunching. And of course, big thanks to Fermentology for inviting me to speak.