You love the blog, so subscribe to the Beervana Podcast on iTunes or Soundcloud today!

Wednesday, July 15, 2009

What Sours a Beer?

In yesterday's post about Devil's Kriek, Samurai Artist sparked a conversation about the souring properties of brettanomyces. Since sour beers are becoming more common, it's a timely discussion. Sourness may be tart and clean as in a Berliner Weisse, dry and austere as in some lambics, our face-puckeringly intense, as in some Flemish reds. These different qualities come from different microorganisms, and it's worth spending a post mentioning a few of the biggies.

My source material here is Jeff Sparrow's Wild Brews, which I recommend highly for anyone interested in a deep understanding of sour beers. Let's start with a pithy opening from the start of his fourth chapter, "Beer-Souring Microorganisms." Here, he describes the actors that create the funk:
"Four dominant types of microorganisms commonly ferment and acidify wild beers: brettanomyces, lactobacillus, pediococcus, and saccharomyces. Sever other important players also merit a mention, including acetobacter, enterobacter, and various oxidative yeasts."
Now, in this next pithy passage, he describes the particular nature of the funk those actors produce:
"The acids most important to wild beers include lactic and acetic acid. Acetic acid, present in copious amounts in vinegar, is sharp, pungent, and greatly increases the perception of sourness. Lactic acid, found in spoiled milk, is less objectionable and contributes a 'tangy' character, sometime perceived as 'sweet' by brewers in contrast to other acids."
(There are actually a host of other acids that contribute flavor like caproic acid, which Sparrow says gives a "goaty," "sweaty," or "zoolike" character. But you can read his book if you want the full monty.)

This wild yeast inspires the most awe and fear among brewers. It will eat anything, including dextrins and sugars that other yeasts find unpalatable, achieving nearly 100% attenuation. (Brewers joke that it will start eating the glass in a bottle if you leave it long enough.) Attenuation is the percentage of available sugars a yeast will eat. Wyeast's Northwest ale yeast, a non-brettanomyces yeast, attenuates at about 70%, for example. Brett will produce both acetic and lactic acids, but the former only under certain circumstances. There are at least five species of brettanomyces and many strains within each. The most common is brettanomyces bruxellensis, named for a strain from Brussels.

Lactobacillus is a type of wild yeast bacteria that gives Flanders beers (red and brown) their character, as it does to some German ales like Gose and Berliner Weisse. It is not a major player in lambics, however--the lactic there comes from pediococcus (see below). As the name suggests, lactobacillus produces lactic acid. Lactobacillus is far more finicky than brettanomyces, preferring warm temperatures, a low-oxygen environment, and low levels of hop acids.

As alluded to above, pediococcus is the beastie that gives lambics their lactic, not lactobacillus. This is mainly a function of the life cycle of a lambic. Pediococcus ferments in beer with little or no oxygen; likewise, it gives off no carbon dioxide. In a lambic, the pediococcus kicks in after 3-4 months when, fascinatingly, the wort is exceptionally sour as a result of early enterobacter production. The pediococcus begins when the lambic warms up, creating "long strands of slime" on top of the wort. You can drink the beer at this stage, but it's oily and known as the "sick" stage. But from the sickness comes the lactic, and eventually, the slime is reabsorbed as the brettanomyces begin gobbling up everything that's left.

The upshot? "Sour" isn't a fixed flavor. Different beers have different compounds and acids that contribute characteristics that define style. Brewers have very different attitudes to the kinds of sour beers produce. When I was at Allagash last year, Jason Perkins and Rob Tod described their efforts to cultivate native brettanomyces. On the other hand, Ron Gansberg doesn't want brett in his brewery; he's a lactobacillus man. Matt Swihart is a brett man, but is he only a brett man, or will future batches exhibit the character of other funky bacteria? I guess we'll have to wait and see.

PHOTO: Cantillon casks, Thom's Beer Blog / link

By the way, Sparrow reproduces a lot of very cool graphs he got from Raj Apte, and you can find those at his site. One I particularly enjoyed is a graph showing the waves of activity in lambic fermentation, particularly in the first year. Click on it to see an enlarged version. You'll find more cool stuff if you follow the link to his site, too.


  1. Sorry to nitpick, but Lactobacillus is a bacteria, not a wild yeast.

  2. Right you are, and a gram-positive bacteria at that.

    Nitpickers have your field day. I'm basically repeating Sparrow's wisdom here. Although I love my lambics, their cultures come, as with most of my yeasts, from a smack-pack of Wyeast's purest. I'm neither a scientist nor an advanced brewer of sour ales.

  3. Any idea on how many strains of lacto and pedio there are?

    I've heard that cultured Cantillon dregs are intense fermenters and pile on the acidity, so how are they different from my lacto/pedio/brett combo that takes ages to create even mild acidity?

    It always seems to simple reading about it, but gets very confusing in practice.

  4. Nice article, Jeff....

    @ DA

    There are many different strains of Bacteria and Wild Yeasts. Each one can add a different nuance. Belgian Sour beers have very complex COMBOS of Bacteria, yeast and funk that's developed over a series of years of conditioning, brewing and age or just plain ol' local micro flora.

    Most of our selections of Brett, Pedio, Lacto, etc. are isolated strains. Of course some of these strains are combined to increase complexity. This kind of explains a lot of one dimensional sourness we get in some American Sour beers. We have a long way to go, but some brewers are getting their faster than others. ;-}

    It's like making a Grilled Cheese Sandwich... You can use American cheese with Wonder Bread and call it good, or you can use 5 different cheeses and some homemade rustic or rye bread and create a total different animal.

  5. Nice topic Jeff, it's a sour geeks delight.

    I'm not sure how many beer friendly strains of lacto and pedio there are. My understanding about lacto is that primarily it effects the flavor by producing lactic acid and that it doesn't contribute as much to the aroma and taste profile. Pedio affects flavor more because of the production of diacetyl (and other compounds) as well as lactic acid.

    Brett adds considerable amounts of aroma/flavor compounds in addition to producing acetic acid. Contrary to Samurai Artist's comment in the Devil Kriek post, Brett does sour a beer without any need for pedio/lacto. Lacto and pedio will add acidity and make the beer more acidic/sour, but brett will sour beer on its own, particularly in the presence of oxygen.

    I find that a lot of US produced sour beers have fairly high levels of acetic acid and VA because of the use of relatively new small barrels. Most US producers seem to stick to 55 gallon wine barrels which are "used" but still fairly new. These smaller barrels have a higher oxygen transfer rate than larger barrels. (Jeff Sparrow has a table of oxygen transfer rates for varous containers in his book.) Studies from the wine industry (still some of the best sources for brett research) show that in the presence of higher amounts of oxygen, brett will produce higher levels of acetic acid (partly by oxidizing ethanol if I recall correctly). But even with low/no oxygen it will also produce acetic acid.

    Newer barrels have higher amounts of celliose which brett can use as a food source, so newer barrels (1-3 years old) can support higher amounts of brett than some older barrels can. So between younger barrels and higher oxygen transfer rates, it makes sense that these beers would have more acetic acid. Maybe that's why the US made beers in the flanders red style (which has more acetic acid) have generally been more successful than the lambic-styles, at least to me.

    In addition Brett produces a plethora of compounds that can increase the percpetion of sourness (in the aroma and somewhat in the flavor). But the major contribution of brett is the array of flavors and aromas: barnyard, horse blanket (both of which are described as "mousy" in the wine world), smokiness, spice, cloves, and a wide variety of fruit esters from peach and cherry to lemon, pineapple, and apple.

    Though both White Labs and Wyeast offer pure strains of brett, I've found that they can have different flavor profiles. The Wyeast ones seem to be more cherry/peach and somewhat more austere, and the White Labs ones seem to produce more tropical fruit esters and don't end up as dry and austere. I've had my best results using both and blending.

    Another note on lacto: I was at a lecture at Hop Union some years ago and was told that a lot of strains of lacto can't live in beer. Apparently it's only certain strains that are beer friendly and aren't killed off by the alcohol. I can't remember if it was someone form Cargill, White Labs or Vinnie Cilurzo who said it. But the implication was that there were still plenty of strains that could survive in and sour beer.

  6. Dr. Wort does bring up a valid point about strains. The commercially available strains here won't completetly replicate the flaovrs of spontaneously fermented beers in Belgium. Adding the dregs of bottles of spontaneously fermented sours will get you some bugs sometimes. It depends on the age and condition of the bottle and even then it can be a crap shoot because not all of the bugs will necessarilly grow. But I've noticed that carboys where I've added dregs of bottles often do get secondary flavors that the pure pitched ones don't.

    There are other things going on besides brett/pedio/lacot in real lambic (like certain strains of candida and other oxidative yeats). Some people say that enteric bacteria also add a lot to lambic, but other sources like Sparrow, point out that many producers go to great lengths to eliminate them because they consider the contribution from enteric bacteria to be off-flavors (and potentially harmful as well).

  7. Average Bill,

    Thanks for the info. Harmful to humans or the beer? I was under the impression that harmful stuff couldn't grow in beer...

  8. @DA - I haven't researched enteric bacteria enough to say, but my understading is harmful to people, at least in some cases (beaches have been closed because of high levels of enteric bacteria from feces in the water). But at the same time they're considered useful gut microflora, just like some kinds of lactobacillus. Jim Liddil has said that enteric bacteria in lambic aren't pathogenic, so it could be a matter of what type of enteric bacteria it is and that not all enteric bacteria are harmful.

    But I have seen people cautioning against the idea of leaving fermenters open in the kitchen for 24 hours (as Liddil and some others suggest) in order to pick up enteric bacteria, because you may pick up pathogenic enteric bacteria. And enteric bacteria can produce a lot of undesirable things like diacetyl, slime/ropiness and fecal aromas.

    Again, I haven't researched it thoroughly, and I hate to give out misinformation, so don't quote me unless you research it more. But there are a lot of things in Goolge about food spoilage and enteric bacteria though (and the interwebs never lie, right?).

  9. The Doctor thinks that graph looks familiar.... ;-}

  10. I hope I never encounter "fecal aromas" in a beer. Uggh.

    One thing Sparrow seems to suggest indirectly is that if you let a lambic go long enough, it has a self-regulating quality. The byproducts of enteric bacteria diminish in time.

  11. So you are saying, you want a little bit of poop in your sour beer, just not too much.

  12. @DA

    Gives a whole new meaning to "killer beer." ;-]

    I've read similar info on open fermenting creating enteric bacteria strains. The Doctor has commented and cringed in the past when others want to brag about their OPEN SOUR MASHING.... Oh well, you can always brag about brewing a beer that works like a high colonic! :-O

  13. @DA

    See! It can be a compliment when people say your beer tastes like SH*T!


  14. @Jeff - They do seem self-regulating. But, hey, they've already gone bad (at least in "normal" beer terms)so what can really happen? I've had air locks dry out on some lambic carboys and have gotten some white mold but racked the underlying beer off with no issues. It's like in sausage making: white mold is normally fine, it's green fuzzy mold you need to be afraid of. (Remember, red meat isn't bad for you, blue and green meat is.)

    One fun thing about making lambic at home is that it's ever changing. We had a blend from a few years ago that just seemed too soft and pedestrian and, well, uninteresting. My brewing partner brought one over last weekend and after 2+ years in the bottle it's remarkable: complex, more acidic, and with real depth of flavors. I coulnd't believe it was the same beer (I'd drunk most of my half up because it didn't seem that special).

    The single biggest mistake people seem to make in brewing them at home is that they don't give them enough time to develop. We use about 20-25% one year old in most of our blends, but I wouldn't dream of using most of them before they had at least 2-3 years of age. Even then, the blend will change wih time.