Salt in sourdough cultures
Sometimes people here on The Fresh Loaf have wanted to favor yeast growth/activity in their sourdough cultures. A particular scientific study of the effects of process conditions such as temperature and salt concentrations on lactic acid bacteria and yeast has been discussed numerous times in this context and others. Gänzle’s 1998 publication reported on studies of two strains of Lactobacillus sanfranciscensis and one strain of Candida milleri yeast, each in isolation. These were in grain-free liquid culture. 4% NaCl inhibited the growth of the Lactobacilli, while the yeast was more tolerant at that concentration and up to 8% NaCl. These findings have been of interest to some here.
Can it be extrapolated to actual dough, particularly diverse sourdough cultures, and lactic acid bacteria (LAB) or yeast strains besides these three? Gänzle’s 2004 follow-up study (with Brandt and Hammes) on a combination of LAB and yeast in rye sourdough was largely consistent with their earlier data from varying salt concentration.
What I think may have escaped notice here is that 4% and 8% salt values are "w/vol" solution percentages (aka solution = solute/solvent = g/100ml = g/100g water), NOT baker's percentages.
The conversion factor from % solution to baker’s percentage is the hydration. A 4% solution is 4 x .75 = 3 in baker’s%. To convert the other way is the inverse: 2% salt (baker’s %) at 75% hyd is 2 / .75 = a 2.67% solution.
In practical terms, to reach the lactobacillus growth-inhibitory 4% solution, one needs 3% salt at 75% hydration or only 2% salt at 50% hydration. So a typical 2% salt in the total dough at 80% hydration is already a 2.5% solution.
For the visually inclined, I made this table of NaCl solution percentages at a given Baker’s percentage and hydration level. [color code 0-3.9, 4-7.9, and 8+ ranges]
As I posted elsewhere, I've been salting my starters for about a month now and haven't seen much of an effect on the speed to maturity, though this process has seemed to have improved the flavor of my sourdough breads since then. In their chapter in "Handbook on Sourdough Biotechnology," 2013, Gänzle and Gobbetti point out two other strains of lactobacilli that have been found to grow at up to 6% salt.
Wondering if I had selected for particularly salt-tolerant LAB, I did a short experiment to test whether my starter could now grow at higher salt concentrations. I set up small portions of dough using all-purpose flour at 75% hydration, with salt at 3%, 4%, 5%, and 6% solution. These equate to 2.25%, 3%, 3.75%, and 4.5% baker's percentages. Seed was 10% pre-fermented flour in a 70% hydration whole wheat starter kept at 2.2% baker's percentage salt, 3.25 solution percentage. I compensated for the seed water and salt, accordingly, so the above were final percentages.
What seems clear, then, is that salt has the potential to shift the balance between different strains of bacteria and yeasts. In my limited survey of the literature, the first statements that I found about whether that would actually be advantageous were in a section of Neysens and De Vuyst's 2005 publication. They cite at least a few known precedents of the use of NaCl in traditional sourdough processes. (They don't mention what I suppose is another precedent, the pâté fermentée / old dough method)
Neysens and De Vuyst refer to a sourdough patent from 1973, which states, "An important ingredient in the liquid starter is salt (NaCl) which we found has the critical effect of maintaining the yeast and bacteria growth rates in the proper relationship. Because the starter of the invention contains a much higher proportion of water than the conventional sponge, there is a tendency for bacterial growth to outstrip yeast growth. The addition of salt counteracts this tendency by preferentially accelerating yeast growth. To attain this desirable effect when the liquid starter of the invention is made up, one includes about 1 to 3% of salt, based on the amount of flour." https://patents.google.com/patent/US3734743A/en
Experiments by Simonson et al. showed skew towards yeast by decreased competition with LAB in certain ranges of NaCl, such that they identified dramatically different ratios of LAB to yeast after culture at different salt concentrations.
Salt-containing sourdough cultures are also reported to favor lactic acid production over acetic acid production at some concentrations.
All this to say, salt is another tool that we can use strategically to influence fermentation in our breads. I hope more of us will give it a try.
--Tom
P.S.
I’m not sure how to make use of this. I feel confident that wild yeast-driven bread can be made this way, but because we don’t want the salt to high in the final product, it takes dilution by additional water or dough. A high PFF bread like ciabatta seems the obvious application, since the low hydration allows higher salt concentration. Two big questions in my mind are: 1) How quickly will the LAB retake the yeast during BF and proof? Dunno the LAB growth rates but E. coli can grow with a doubling time of 20 minutes at their preferred 37 deg C. It calls for another try. 2) Can a yeast-dominated 5-6% (solution) salt starter be maintained long-term without encroachment of bad bugs? If it’s not getting sour, seems like other bacteria and molds should be able to grow. The salt would help, at least.
REFERENCES:
Ganzle et al., "Modeling of Growth of Lactobacillus sanfranciscensis and Candida milleri in Response to Process Parameters of Sourdough Fermentation," Applied and Environmental Microbiology, 1998.
Brandt et al., "Effects of process parameters on growth and metabolism of Lactobacillus sanfranciscensis and Candida humilis during rye sourdough fermentation," 2004.
Gänzle and Gobbetti (editors), "Handbook on Sourdough Biotechnology," 2013.
Neysens and De Vuyst, "Kinetics and modelling of sourdough lactic acid bacteria," Trends in Food Science & Technology, 2005. Simonson et al., “Response of wheat sourdough parameters two temperature, NaCl and sucrose variations,” Food Microbiology, 2003.
.
Tom, thanks for posting your literature search and findings along with your experiments and explanations. This is all extremely interesting. I have long wondered why my starter seems to get sluggish after a while and this may offer a good explanation. I would surmise from your post that my starter’s LAB population over time starts to outstrip the yeast population. When I notice this slowing of the starter I typically start doing feedings with whole rye and that seems to perk it up. Based on your findings it might seem logical to add some salt, perhaps 2% by baker’s math and maintain that in my starter.
Have you converted your starter to a salt starter? If you are, at what % by baker’s math are you keeping it. How long have you kept it a salt starter? Have you noticed any change in behavior over a long period of time?
I’m very interested to learn more about this, thanks for this detailed post Tom.
Benny
Hi Benny,
So glad you found this interesting! As for my starter, I keep it at 2.2% baker's percentage and usually 70% hydration now. When I first made the change, I suppose almost 2 months ago, it was 100% hydration, so the same as 2.2% solution. I didn't notice any change in behavior. Gradually, I started shifting to lower hydration, I think 80% at the beginning, 75%, 70%, then 60%. As I mentioned, this progressively increased the salt concentration. Again, I didn't see much difference, maybe some slowdown at 60%, but there were two variables changing and I didn't think much of it.
Only after Doc directed my attention to the Gantzle paper did I realize that the 4% figure I'd heard about inhibiting LAB was not baker's percentage, and that I had already gotten very close to it (2.2%BP@60% hydration-->3.7% solution). I interpret my experiment (done just yesterday) as indicating that the LAB in my starter are still inhibited around 4% solution as in the first paper, even with my treatment applying a selective pressure.
After I dropped the hydration to 80%, I started getting big smiles and thumbs up from the two sourdough eaters in my family, so I think that's when we first discerned an improvement. Still, I was also changing to a stiff starter; it could have been due to one or both factors.
By the way, I saw two reports that LAB are actually boosted over their baseline (0 salt) growth rate at 0.7-1% NaCl, before the rate dips down at higher concentrations. Yeast, in contrast, don't show the same boost at low salt (relative to no salt).
--Tom
.
.
.
It stands to reason that starters should be fine at 2% salt (baker's %), since they continue along happily in our final dough. Bakers who want more sour may want to favor LAB in their preferment, in which case I interpret the studies as suggesting 0.5-1% (same as baker’s %, at 100% hydration) of that concentration may be ideal.
.
Shame Rus Brot only writes in Russian, but seems like adding salt to a preferment, at least, is a known German technique (Monheim salt sourdough). Here is Yippee's write up I found though: http://www.thefreshloaf.com/node/58095/20181121-rus-brots-german-rye-using-monheims-sourdough
http://brotgost.blogspot.com/2016/01/salzsauer.html
Google Translate to the rescue:
” Monheim's sourdough is one of the most interesting ways to use German sourdoughs, and a feature of this method is the addition of 1.5% -2% salt to the sourdough by weight of the sourdough flour.”
I wondered if Yippee’s CLAS is made with any salt. 0.7% NaCl was optimal for LAB.
Let me know if you want any details from his blog that Google doesn't translate well.
.
.
.
.
.
Not speaking from experience or specific knowledge, but I suppose different temperatures, or salt for that matter, modify the metabolism of yeast and LABs, so that could affect flavour? Any known facts about that?
.
Ilya, the only studies of taste components affected by salt that I saw in my literature search were measuring lactic and acetic acids. There was another study (not looking at salt) that focused on the effects of shifting starters from firm to liquid culture. It seems they found that it causes differences in many compounds, but they didn’t include qualitative sensory assessments so there wasn’t a conclusion that one hydration was better than the other.
My first experiment (above) confirmed that lactic acid bacteria could be disproportionately inhibited in the range of 4-6% solution, in favor of yeast. Since the high salt preferment must be diluted to achieve a normal salt level in the total dough, this method would be implemented as some combination of low enough % prefermented flour (PFF) and/or reduced hydration in the preferment.
This raised the question: Once the salt concentration is returned to normal, will the LAB quickly catch up by the time bulk fermentation and proof are complete, in actual practice?
Ciabatta seemed an ideal formula to try this out, with its use of a high %PFF biga. I decided on a two-stage levain build from my 2.2% (baker’s percentage, BP) salted, 66% hydration whole wheat starter. Stage 1 was all purpose flour (APF) seeded with 20% PFF at 3.9% (BP) salt (6.1% solution at 65% hydration). This took 12 hours to double at room temperature. Stage 2 biga was APF seeded with the entire Stage 1 levain for 20% PFF at 2.4% (BP) salt (5% solution at 47% hydration). I thought I should give this more fermentation time as a salt retard than commercial yeast (CY), but I was away from home and let this go too long at room temperature, 8.5 hours, then refrigerated overnight. The biga made up 80% PFF. To make the final dough, I added 14% sprouted hard white wheat mash (at 80% absorbed and added water) along with 6% APF, 4.2% vital wheat gluten for 14% total protein, and enough water (21%) for 70% total hydration.
Stage 1 levain after 12 hours:
This turned out to be a very slack and sticky dough. I didn’t get good gluten development on the Kitchenaid stand mixer, but let it rest 30 min, gave a set of stretch and folds and took a portion for the aliquot jar. After another hour, another set, then divided the 2kg dough in half. One half was stretched and letter folded both ways, then placed in a 9x5” loaf pan, and received no more handling. The other half got coil folds instead, and other set 50 min later. At 3 hours in the aliquot jar, rise was measured at 50%. I loosely shaped and covered the remaining dough and let it proof until the aliquot showed an 85% rise at 6 hours in the jar. I baked the freeform ciabatta on a baking stone, next to the loaf pan, at 450 deg F with a pan of boiling water for steam 20 min, then 20 min with the water pan removed. The ciabatta baked up with a hard crust, little oven spring, and not a great crumb. I think overall fermentation was way too long, so most of the gas had been knocked out along the way. The pan loaf was pretty dense and a bit chewy, but had a mostly even crumb and a nice ratio of moist crumb to crispy crust. Maybe too much gluten? The flavor of each was not appreciably sour, much like a CY ciabatta.
Ok, the process needs some work. My takeaway from the experiment is that the wild yeasts in salt-retarded sourdough culture are indeed able to leaven bread before the LAB can catch up and turn it sour.
The bread may not have worked out to your liking but the theories have been supported by the resulting flavour of the bread so that is good.
Benny
This is a very interesting post. I appreciate the work that has gone into the experiments and asking questions rather than just following the way it was done in the past.
I am going to try a culture with salt more the 80%/2% variation and see what the bread turns out like. I won't tell my family and see what they say without knowing there is any change in method
were motivated by the need to reduce the dough dissolution.
Actually it worked! I refreshed my started at 45% hydration containing 2% salt with respect to water. I also refreshed with high gluten flour mixing in the spiral mixer.
It turned out to be a much more consistent starter than all previous ones, with less acidity and more rising strength.
Curious what you mean by "dough dissolution". Can you explain?
Also, is your 45% hydration starter maintained like a lievito/pasta madre? I'm interested to know if you still do this; does adding salt actually help in some way?
after several hours the dough became a sticky mess. After one day it was almost liquid.
I haven't baked for many years, but at the time a 1% salt (to flour) in my refreshes did wonders.
Any sources on what these concentrations are? I'm interested to know (interest mainly stems from what I've read regarding the ideal lactic to acetic ratio is being somewhere around 3:1 or 4:1.)