Where kveik comes from
Posted in Beer on 2018-09-12 16:27
Yeast ready for use, Sõru, Estonia
I've written before about the kveik research paper by Preiss, Tyrawa, and van der Merwe. That paper was not accepted by the reviewers, who had a number of complaints. In addition to the serious complaints one reviewer, amusingly, had difficulty believing that farmhouse brewers ferment at 30-40C. Anyway, to get the paper accepted Richard and co went to work and expanded the research quite a lot, with help from Kristoffer Krogerus. That second round of research uncovered enough new information that I think it's worth doing a second blog post on the new, second edition of the paper. (Which now has been peer-reviewed and accepted.)
The first paper showed very clearly that kveik is a domesticated yeast. That is, a yeast that has been reused by humans for so long that it has changed into something that works better for humans and is adapted to the environment humans have created for it. So kveik is definitely not wild yeast. It's beer yeast. But what kind of beer yeast?
The family tree in the first version of the paper was based on DNA fingerprinting of a small part of the genome, and comparison with just a few other strains. That doesn't definitively answer whether kveik is a distinct subgroup of yeast or not. For that, you need full genome sequencing and a larger data set, which is expensive. However, they managed to get funding for it, so Kristoffer Krogerus sequenced 6 kveik strains and placed them into the big family tree for yeast.
Substructure inside "Beer 1"
The results were quite surprising: kveik belongs to Beer 1. This is the group that has Belgian/German strains on one side, and UK/US ones on the other. Kveik slots in near the root of Beer 1. You see the tree above: the basic division inside Beer 1 is between the US/UK/Bel/Ger yeasts on the one side, and kveik + 3 non-kveik yeasts on the other. (More on those below.)
So what does this tell us?
Obviously, kveik shares an origin with the biggest group of beer strains from commercial brewing. Almost certainly, Beer 1 started in continental Europe, and some of the strains were then taken across the channel to the UK, while others came to Norway. How and when that happened we don't know, but it's clearly a long time ago. Gallone et al calculated that the yeasts in Beer 1 began diverging around 1600 CE, but personally I consider that a low end estimate. That is, it probably can't have happened much more recently, but it may well have happened earlier. So I think we can confidently say kveik split off from the rest of Beer 1 centuries ago.
From the genetic signature of kveik (high heterozygosity) the researchers conclude that it is probably a mosaic (a kind of hybrid). This needs to be explained, I guess. Wild yeast can produce offspring by having two parents mate to make a child by mixing their DNA, like humans do. They can also produce new cells by simply "budding," letting a new cell form out of an existing cell. Budding is something humans (happily!) can't do. Brewer's yeast can usually only reproduce by budding, but there are some exceptions.
The family tree with the two halves of the genome (haplotypes) inserted separately. Look for the two sets of strain names in red around the outermost circle.
What the researchers think is that at some point in the evolution of brewer's yeast, a yeast from Beer 1 mated with some other yeast from outside Beer 1, and that all the kveiks descend from this child. They used an algorithm to divide the genome into the genes they think came from one parent, and those they think came from the other. Making the tree again, but placing the two halves of the genomes into it separately, one half ends up in the same place in Beer 1, while another ends up in a kind of no-man's land not really close to anything. (See the two groups of red names in the diagram above.)
The researchers think this means that kveik was formed by a parent from very early Beer 1 mating with a wild yeast. This is not a firm research conclusion, however, just what looks most likely from the available data. Since all six kveiks show the same signature, very likely this hybridization happened a long time ago. Whether it happened on the continent or in Norway we have no idea.
(This hybridization event, by the way, if it happened, is similar to how lager yeast formed, by normal brewer's yeast mating with a wild yeast from a different, cold-tolerant species. In the case of kveik, both parents were Saccharomyces cerevisiae, but from very different populations within the species.)
But what about those three non-kveik yeasts that are their closest neighbours? Those are three hefeweizen yeasts, intriguingly. Unfortunately, that result is probably because both are mosaics (within-species hybrids) and that similarity is what causes them to cluster together, rather than any historical relationship. Probably. It's a tantalizing fact, though, that the two most aromatic types of beer yeast end up right next to each other. (I'm not counting Brettanomyces, because it's a completely different genus, group of species.)
Idun Blå, Norwegian bread yeast
One thing the reviewers questioned was whether kveik might be bread yeast rather than real farmhouse yeast, since the sahti and gotlandsdricke brewers are all using bread yeast. The researchers fingerprinted the Norwegian baking yeast Idun Blå, and it ends up in a completely different place in the tree. So kveik is definitely not related to Idun Blå.
To summarize, kveik is definitely domesticated brewer's yeast, relatively closely related to continental beer yeast. It's probably a hybrid between those and a wild yeast that was then domesticated further over centuries.
The big lesson we learned from the Gallone et al 2016 paper was that most brewing yeasts are related, which shows that properly domesticated yeast must have been valuable thing that brewers took good care of, and shared with each other. That's how a single family of yeast managed to spread out from Belgium and Germany to the UK, and from the UK to the US. Other, similar, family tree research projects have found similar results, so there's every reason to be believe this is correct.
Kveik getting started, Norwegian farmhouse ale festival, Hornindal
This paper brings that out even more clearly. Even in Western Norway it turns out brewers are still using descendants of this same Beer 1 family. The fact that all the kveiks, from Hardanger in the south to Sykkylven in the north, are related to each other shows both that brewers have been sharing their yeasts, and also that they have been very careful to preserve their domesticated yeast. One also gets the impression that there weren't that many options: this single type of yeast was apparently preferred over all alternatives. So maybe domestication of yeast didn't happen that many times?
It's worth repeating that the one non-Norwegian farmhouse yeast that was included (#16 Simonaitis, from Lithuania) is not related to the kveiks. So there are farmhouse yeasts that are not kveik. This leads us to the still open questions. We've now collected farmhouse yeast from eastern Norway, more from Lithuania, from Latvia, and from Russia. Where do those fit? Are the eastern farmhouse yeasts a separate group that ferments hot and fast? Is eastern Norwegian farmhouse yeast kveik, or something else? We don't know yet.
But since two of the six known Lithuanian yeasts belong to completely different species my guess is that the non-kveiks don't all belong to a single family. But we won't know for certain until the researchers do some more work.
Big thanks to Richard Preiss and Kristoffer Krogerus for taking the time to answer my questions on how to understand the results.
In 2016 I was contacted by Canadian researcher Richard Preiss
Read | 2017-10-06 10:02
Read | 2017-08-08 09:09
Robert Manktelow - 2018-09-12 21:31:51
Great to see this Lars. Could you advise if the kveik as per the registry was used in the study, and if so which ones. So for example, is Voss 1 Sigmund Gjernes kveik, or another Voss Kveik? Is the Stordal Ebbegarden that of Jens Aage Øvrebust?
Lars Marius Garshol - 2018-09-13 06:32:49
@Robert: Good point. In the supplementary material, page 10/14, there is a table showing more detail on the cultures. The numbers, like Voss 1 and Voss 2, refer to specific strains inside a culture. The registry only has the mixed cultures (which would be Voss 1, Voss 2, and whatever else is in the culture).
Marco - 2018-09-13 15:21:58
Out of curiosity, are there any well known beer 2 strains?
Lars Marius Garshol - 2018-09-16 09:43:04
@Marco: Kristoffer Krogerus and the anonymous qq have done a fantastic piece of detective work, and produced a version of the tree with meaningful names: http://beer.suregork.com/?p=3919
The White Labs Belgian and French saison yeasts all seem to be Beer 2, and also some British strains. See the diagrams there and you can figure it out.