Friday, 12 August 2016

La Pasta Madre: Italian GF sourdough starter experiment, Day 1

Sourdough or natural leaven is the natural fermentation of starchy material by a mixture of many types of wild yeasts and bacteria that are adapted to coexistence. This complex fermentation results in a diverse assortment of molecules contributing to the flavor, nutrition, and even the texture of bread differently than the single strain of yeast we buy at the store. The difference is even more pronounced in GF bread. Based on a wide body of research, I believe that the traits selected for store-bought yeast are simply not the best match for the properties of our GF flours; wild yeasts and associated bacteria are much better for turning these flours into bread - real bread, with only a few simple ingredients and no additives needed to compensate. In fact, one of the ways certain sourdough bacteria benefit bread texture and reduce staling is by producing a molecule that functions similarly to a gum! 

I recently acquired an Italian gluten-free book, A Tavola Senza Glutine, which was published by Slow Food Editore - between that fact and the amazing-looking pictures, I knew it had to be good! Unsurprisingly, all of the leavened breads use a natural sourdough starter, which I was eager to make. However, I found the method of making the starter quite surprising indeed. Let me explain:

Conventional sourdough starter methods usually advise refreshing (discarding a portion and feeding fresh flour) frequently, up to 2 or even 3 times a day until the microbial community is established, and this approach is often used in GF sourdough tutorials as well. Some recent artisan-bread experts feel the starter grows more robustly when fed/divided just once a day and at the other 12 hour intervals simply stirred to ensure enough oxygen for aerobic fermentation throughout the mixture. I’ve found this latter strategy to work well in the past with my own sweet-potato-based GF starters also. Certain traditional methods, on the other hand, involve a straight batch fermentation - nothing is discarded or divided, the flour-water mixture is just left alone until it balances itself out! It is more often done in situations where the entire batter is used (as opposed to maintaining a continuous starter for future batches) but I have also seen examples of long-lived starters created in this way. This last method is of course the least reliable and the success or failure of it seems highly dependent on temperature, ingredients, and multiple other conditions, because without intervention it is harder for the desirable bacteria and yeasts to gain a stronghold. (I personally have never gotten it to work right.) But the simplicity of it is intriguing - and tradition hints our ingredients’ natural course of fermentation may be suited to this approach, as seen in recipes for various Asian rice- and bean-based batters and Ethiopian teff flatbread as well as some of the traditional American methods of making potato yeast. 

So I was very interested (albeit a little nervous) to try the starter method in A Tavola Senza Glutine, which seems to strike a nice balance between the high-turnover and the hands-off approach: the starter is left alone for 2-3 days at a time, and at each interval, 20% of the mixture is discarded and replaced with fresh flour and water to maintain a constant quantity. Simple and unfussy, right? addition to the low-maintenance schedule, this is a much lower discard ratio than with a typical sourdough starter recipe. So much lower, in fact, that I began to question whether I was reading the recipe correctly, and if it in fact might be the other way around. The word used, sostituitelo, means replace - does this mean replace as in “swap out” or could it mean replace in the sense of “return / put back with”? Am I supposed to add the fresh flour to the original jar (unusual ratio, but seemed to be what was instructed) or the removed portion? Had it been the reverse - removing a small portion to combine with fresh flour and discarding the rest, then also discarding and replacing half of the mixture at each subsequent feeding - this procedure would be far more typical, though less economical. To clarify, I went looking for some other instances of the same wording in other types of recipes where the intended procedure is more obvious, and it seems pretty clear that the usage of “replace” here does indeed mean the fresh flour is to be added to the main mixture, not combined with the small removed portion. (If you’re wondering why this difference matters so much, don’t worry - I will explain further about the role of discarding during the establishing phase in the Day 4 update.)
Regardless, that wasn’t the only odd thing about this starter recipe. It is also a lower hydration percentage - it starts out at just 60% hydration (not including the small amount of honey to get things going). The subsequent feedings/waterings add liquid at 67% of the flour weight, meaning the starter does become a little less dry with each discard, but even 67% remains a much lower hydration than the starters I’m most familiar with. 

But! In spite of all these doubts, I still want to try it the way it's (I think?) written. I know, I should have asked someone about the confusing wording, and I will feel very silly if I misunderstood the procedure. At the very least, it will be interesting, right? Let’s see what happens!

FIRST, some notes on what you’ll need: 

Rice - I might be overthinking this, but I figure I ought to at least bring it up in case it turns out to be relevant: The rice varieties grown in Italy are mostly medium-short-grain japonica varieties (think risotto), some of which are available made into flour. However, I can’t figure out if the regular rice flour available in Italy is made from one of these or from a more common rice type, as most do not specify. Likewise, most rice flour sold in the US does not specify the rice variety, but I’d guess it’s mostly made from medium-long grain types - a different amylose/amylopectin ratio than shorter grains. Anyway, I’m working under the assumption that standard rice flour here is similar enough to the standard rice flour there to suit the purposes of this experiment. Let’s hope I’m right!

Corn - Italian corn meals/flours can be divided into 3 categories: bramata (coarse, like for polenta), fioretto (fine, like US stone-ground corn flour), and finally fumetto (superfine and consisting only of the softer endosperm, or inner part of the corn kernel). This last grade is what the book calls for, but it isn’t really a thing in the US, so I had to do the best I could with my finest sieve and hope the tiny bits of harder outer layer remaining wouldn’t be too disruptive. 

Water - Please do NOT use tap water - the chlorine compounds and other things in municipal tap water will interfere with the growth of the more fragile members of the sourdough microbial community. Use spring water, or filtered water may be OK if your water filter is very very efficient.

Honey - I suspect the recipe specifies acacia honey for only a few particular reasons. For one, it has a higher fructose content than many other honeys, meaning it won’t crystallize and, being fairly thin, will dissolve evenly in the mixture. The other reason is that it has a very bright unintrusive flavor - it is a light, cleanly sweet spring honey, without the waxy, earthy, or tangy notes found in some other varietals or wildflower blends. 
With these criteria in mind I was all set to use another kind of honey I already had. But combined with the already unusual nature of this formula, I decided at the last minute to err on the side of caution, worried that there is something special about acacia that I wasn’t aware of. (Perhaps it may be particularly conducive to yeast fermentation, considering its popularity among mead producers?) If it doesn’t work right, I didn’t want to be stuck wondering if it was because I used the wrong honey. Besides, I was already having to change one thing with the cornmeal - changing two things just wouldn’t be scientific, after all. 

Jar - OK, it’s not exactly an ingredient, but it’s definitely necessary and I want to bring up one point: The book recommends covering the jar with a cloth, which is indeed the traditional way. However, newer science indicates that contrary to popular (and traditional) belief, the vast majority of the characteristic yeasts and bacteria in a sourdough starter come not from the surrounding air, but are already present on the surface of the grains. So, I just use a loose-fitting lid because it’s less cumbersome. You shouldn’t have an airtight seal, and you may still want to use a cloth if you are having issues with condensation in the jar, but really, any cover that keeps out the fruit flies and such should be fine. 

The starter, day 1: 100 g rice flour, 50 g fine corn flour, 180 g water, 1 tablespoon (~16 g) acacia honey
Combine in jar and set aside in a cool place for 3 days. The book specifies 18º C (about 65º F) - it’s nowhere near that cold in my house at this time of year, so I put it near the AC vent and hope it averages out! (By the way, the quantities above are half of the recipe as written in the book, because this quantity seemed more manageable. If you want to do it exactly as written, double all the measurements above.)

Check back on Day 4 - that’s Monday - to see what happens next!


  1. I've been experimenting with GF sourdough here in Greensboro lately with some really good success flavor wise. I haven't gone yeast free yet though. I'm very interested to see how yours goes.

    1. What flours do you use in your GF sourdough? This particular starter is giving me some trouble (see my latest update!), but I've had great results in the past with a sweet potato/buckwheat/cornmeal combo.

    2. This comment has been removed by the author.

  2. Sorry to take so long to reply! I use brown rice, oat, and a bit of buckwheat. The buckwheat really seems to give it a kick. I've also tried adding a bit of Kombucha in place of some of the water and that was a spectacular success. My problem is how to substitute in the starter in my recipes. I read to add 1.5 cups of starter and remove 3/4 cup each of my dry and wet ingredients. That made the dough way too wet. So I've been playing around with that some.