Friday, 21 June 2013

First Wild Yeast Hunt

I've been hush-hush on my wild yeast hunt, aside from a few posts on the methodologies I was planning on using to purify and identify wild strains of yeast. But I haven't been sitting still - the hunt's been on, for over a month. This post is a mid-hunt writeup; at this point I have a number of yeast strains, but I have not yet completed a full analysis of their fermentation properties, the flavour of the resulting beer, nor have I done any form of species identification aside from a bit of selective media and pictures taken under a microscope.

The isolation process is fairly simple. Six kernels of malt were dropped into a small tube containing 6 ml of 1.040 unhopped wort. At specific intervals a small amount of fermenting wort was sterilely removed and plated on a wort-agar plate containing penicillin and streptomycin. These antibiotics prevent the growth of bacteria, ensuring that only wort-fermentating yeasts (and other fungi) are purified.

My sampling strategy is simple; using information gleaned from two sources: Wild Brews: Culture and Craftsmanship in the Belgian Tradition by Jeff Sparrow and the paper Brewhouse-Resident Microbiota Are Responsible for Multi-Stage Fermentation of American Coolship Ale, I picked a series of time points in which specific organisms should dominate the ferment, but I am hoping shorter time points will suffice, under the expectation that the fermentation process will be accelerated due to the small volume and relatively large inoculum. By plating out at these time point I hoped to get enriched cultures of very specific yeast species, notably:

  • Day 4: Oxidative yeasts like Pichia & Kloeckera apiculata
  • Day 10: Fermentative yeasts like Saccharomyces
  • Day 20: Saccharomyces, some Brettanomyces
  • Day 30: Saccharomyces, some Brettanomyces
Progress of the ferment.
L:R: Day 1, Day 4, Day 10, Day 20 & Day 30

As always, the rest is below the fold...


Wort-Agar Plates

For this trial run, I am using wort-agar plates with penicillin and streptomycin to purify the yeasts at every stage. Aside from the antibiotics, these are made of easily-attainable materials and can be easily prepared at home. Simply put, they are 1.002 S.G. wort with enough agar to solidify them, plus the antibiotics (which are optional). Sterile petri dishes (or reusable glass petri dishes) are required.

  • 2g of Dry Malt Extract
  • 1.7g Agar
  • 100ml water
  • 100x antibiotic solutions (optional)
  1. In a beaker or jar dissolve DME into water, then mix in agar (agar will not dissolve). Cover opening of container with foil*.
  2. Sterilize using an autoclave or pressure-cooker; 20 minutes at 121C**.
  3. Let cool to ~50C, add antibiotics
  4. Pour into sterile paltes, using sterile techniques. 20-30ml is required per plate. Resulting gel should be 5-8mm in thickness.
  5. Let cool to solidify; will store in the fridge for upto 2 months.
* If using glass petri dishes and no anitbiotics, add the wort/agar directly to the petri dishes and sanitise them in the steam phase (e.g. do not immerse them in the water of the pressure cooker).
** If using glass petri dishes with antibiotics, the dishes should be sanitized at the same time.

Day 0 - Culture Set-up

Inoculated 6 ml of 1.040 wort with 6 uncrushed grains of malted 2-row barley. Wort was clear & sweet smelling at this point.

Day 4 - The First Yeasts

Day 4 yeasts - lots of Rhodotorula
On day 4 the wort appeared modestly cloudy and had notable sour and fruity aromas. Underlying those aromas was an unpleasant enteric odour (weakly vomit-like) - likely caused by the enteric bacteria known to be present early in most ferments. These enteric bacteria can be nasty - while most of them are harmless, occasionally food-poisoning causing strains of e. coli can be found in young worts. Within the ferment these bacteria typically are out-completed by other organisms - or they ruin the batch. But it is because of these that I'm using antibiotic-containing plates. Last thing I want is to brew a batch of EHEC ale...

I withdrew & plated 50ul of fermenting wort and cultured for 5 days at room temperature (May 2-4 long weekend got in the way). The picture to the right shows the plate a few days after plating. As you can see the colonies are rather dense, indicating I should have plated slightly less volume. The plate is dominated by red colonies, which are most likely Rhodotorula - an oxidative yeast common early in wild ferments, but which can also cause opportunistic infections in immune-suppressed individuals. While predominant, Rhodotorula is not the only thing growing on these plates.

As the insert to the left shows, scattered among the orange/red Rhodotorula colonies are three distinct types of unpigmented (white/grey) colonies. The larger are round-to-irregular shaped, elevated and mucoid (moist-looking) colonies. The medium colonies are circular, convex and powdery (dry-looking) colonies. The smallest are circular, convex and mucoid. While these appear to be three discrete strains, each colony within a single "morphology group" may represent unique yeasts with similar colony structures.

These early yeasts are likely oxidative yeasts which require oxygen to grow, meaning they are unlikely to be suitable for a pure fermentation. On the surface this would suggest that purifying them further is useless, but the wine literature suggests that they may work well as an early co-fermenter, to both add unique flavour compounds as well as for their anti-bacterial properties. I streaked out one Rhodotorula colony, just to take its picture, but didn't store a culture as they don't appear to be important to wild ferments and are potentially dangerous. I did, however, select three of the large/mucoid, medium/powdery and small/mucoid colonies, for a total of 9 colonies. These colonies were cultured in 1ml of 1.020 wort and frozen for future work, as well as streaked out for high-magnification (90x) imaging:

Large Mucoid yeasts 1-3. Yeasts 1&3
have a similar spherical morphology with
multiple small vacuoles, yeast
2 may be ovoid and has few vacuoles.

Medium Powdery yeasts 1-3.
All three have a similar ovoidmorphology with 2-3 small
Small Mucoid yeasts 1-3. All three have a
similar ovoid morphology with multiple
small vacuoles.
Rhodotorula, with their highly
elongated morphology and
poorly defined vaculoles

Day 10 - Looking for Saccharomyces

I drew another sample on day 10, which was checked for gravity and then plated. The wort was slightly turbid - about the same as on day 4, and had a gravity of 1.045 - 5 points higher than the OG! I suspect this increase in gravity may be due to extraction of materials from the malt used to initiate the ferment. The wort aroma had changed significantly; from fruity/sour to an aroma more like fermenting beer. 35ul of fermenting wort was plated out on a beer-agar plate and grown at room temperature for 4 days, producing a plate with nicely separated colonies.

As you can see, about 30% of colonies are still the red (Rhodotorula) colonies, but the remainder are unpigmented (white) medium-sized colonies which are glossy, but not outright mucoid. There were no obviously equivalent colonies on the day 4 plate, although the medium/powdery were close. Five of the white colonies were selected for imaging & further experimentation.
Day 10 Colonies. All appear similar; ovoid, most with a single large
vacuole. Day 10 yeasts also appear larger than the Day 4 yeasts.

Day 20 - Late Bloomers

First plate - too many yeast!
The wort has become quite cloudy, and has begun to developed a nice layer of kraussen, meaning that we probably have significant Saccharomyces activity. Due to the turbidity of the wort, I decreased the amount plated to 20ul. Despite this precaution I ended up with a lawn of yeast - literally, thousands of colonies all merging together into a single layer of yeast. That's the bad news; the good news is the plate smelled like fermenting beer! I'm certain that I'm on the right track, and that I have some Saccharomyces growing.

That's Better
But a lawn of yeast, of course, doesn't work for me - I want single pure colonies. So I scrapped a bunch of the yeast into 1ml of pure water, vortexed it to get a single-cell suspension, and then streaked out 20ul of that solution, using the conventional streak-plate method to maximize my chance of getting single colonies.

This second attempt was successful. As you can see too the right, on the 2nd streak I got single colonies. The colonies all appeared the same - small, circular, slightly raised and moist in appearance. I purified 5 of these, imaged them, and then froze them down. The uniformity of the colonies gives me hope that I am looking at Saccharomyces - they should be dominating the ferment at this point, and these colonies are consistent with the description of Saccharomyces colonies I've read elsewhere.
Day 20 Yeasts

To the right are micrographs of the 5 isolated strains. I'm in the process of upgrading my microscope, so the image quality here isn't fantastic. Regardless, we can see the morphology and the morphology of all 5 is similar - ovoid, with a central vacuole. The fact they are ovoid and appear to bud at the ends (see image at bottom-left) makes it unlikely these are Saccharomyces cerevisiae. That said, other saccharomyces, such as S. boulardii, have a more ovoid morphology. So we may still have sacc. The morphology is also consistent with Brettanomyces. Unfortunately, these samples were scrapped off plates instead of being recovered from a liquid culture, so we cannot see if they form elongated chains of end-link yeast cells; a characteristic of Brettanomyces but not Saccharomyces.

Day 30 - Maybe Some Brett?

Day 30 plates: suspended yeast (left) and from the
top of the tube (right)
By day 22 there was a thick layer of yeast at the bottom of the tube, plus a ring of yeast around the top, and all fermentation activity appeared to have ceased. The wort remains cloudy, perhaps due to a poorly-sedimenting yeast, or due to on-going fermentation. I took two samples for day 30 - some of the wort, plus some of the yeast that formed a ring around the top of the tube. It is my hope that this ring represents more flocculant yeasts - and attractive feature in a yeast strain.

I also learned my lesson last time - I diluted the collected yeast 1:10 before plating 20ul; despite this I ended up with way too many colonies. Moreover, the plate had a strong funk to it - a funk that I hope represents the presence of Brettanomyces. Furthering these hopes is the high attenuation of the wort, which has dropped from a S.G. of 1.040 to 1.005. Between this high attenuation, ongoing fermentation, and the persistence of suspended yeast, I have high hopes that some Brettanomyces is present.

Despite the multitude of small, poorly separated colonies, I was able to pick 6 single colonies; three from each plate, and grow them up. Their micrograph can be seen above. The top row are the colonies form the suspended plate, and the bottom row is from the 'tube-top' plate. The morphology of these yeasts is largely the same as those from day 20 - ovoid, end-budding, with a large vacuole or two in each yeast. Two of the 'tube-top' yeasts (left & centre, bottom row) may be less avoid than the other day 30 yeasts, giving them a higher likelyhood of being Saccharomyces cerevisiae.

What's Next?

The ferment is continuing, and I will try and extract additional yeasts at later time points. At this point I have 25 strains of yeast - the next step with these is to perform mini-ferments; small (100ml) batches of mildly-hopped beer to determine the flavour and aroma profile of these yeasts, as well as some characteristics like flocculation & attenuation.

I hope to start this in 2-3 weeks, with a report following mid-August.  Stay Tuned!

Interesting Webpage:


  1. Sounds like a fun project. I've been wanting to do something similar, and considered using acidified, sterile wort, perhaps around 4.5 pH, to try to skip at least part of the enteric bacteria growth. This is a common technique these days for creating sourdough starters, since it skips or at least minimizes the godawful stench of the early stages.

    Which brings me to another point. You mention the vomit-like smell being from the enteric bacteria. Though this has always been my suspicion, brewers seem to have latched on to the idea that this smell is butyric acid produced by Clostridia (never read about it being present in these fermentations), and sourdough people have attributed this to the growth of the LAB Leuconostoc (which is used in non-stinky fermentations, too). Do you have any insight into the matter?

    I briefly raised this with Sam from Eureka Brewing and he thought while butyric acid smelled bad, it was not the source of the vomit-like smell from these fermentations.

    1. The lower pH should help; that, along with increasing alcohol content, is what kills off the organisms during a conventional lambic ferment. I thought about those kinds of approaches, but I was concerned about potential negative effects on the oxidative yeasts.

      As for buteric acid, it is likely part of the aroma, but not the whole thing. Our noses are not highly specific chemical detectors, and instead we perceive the combined effect of numerous compounds all being sensed simultaneously. Buteric acid is part of what make shit (excuse me, I meant 'faeces'), body odour, vomit, rancid butter, and many other stinky biological things smell. And yet, each of those substances smells radically different; those differences are due to the presence of other odorants along side the buteric acid.

      Personally, I'd be hesitant to claim any one bacteria as the source of the aroma. Enteric bacteria make enteric odors; the exact chemicals and species involved are variable, and will be different for every wild ferment. Moreover, the final aroma is likely due to a mixture of odorants from a number of individual species.

      BTW, clostridium & lactic acid bacteria like Leuconostoc are generally classified among the enteric bacteria.

    2. I see, thanks for your clarification. I didn't know those other bacteria were considered "enteric" though I knew they were often found in the intestines. I thought the term was used for things like E. coli, Salmonella, etc.

    3. It may be different in the brewing world, but in the microbiology/immunology world (where I work), any normal component of the intestines would be considered 'enteric'.