The "Pyler says" series explores excerpts from Baking Science & Technology, a textbook that teaches readers a range of baking and equipment concepts. The following passage is from Chapter 2: Minor Ingredients — San Francisco sourdough.

In the case of San Francisco sourdough French bread, as described by Sugihara et al. (1970, 1971), the microorganisms involved were first identified in the 1970s by researchers at University of California–Davis and the Western Regional Research Laboratory of the US Department of Agriculture. Looking at five mother sponges used commercially by San Francisco Bay Area bakers and some 200 isolates from these sponges, they learned that the yeast involved was Saccharomyces exiguus, a spore-former, and its non-spore forming equivalent, Torulopsis holmii. This yeast did not ferment maltose, but they also found a maltose-consuming yeast identified as S. inusitatus, but no S. cerevisiae, or conventional bakers yeast, was identified in any of the samples.

(S. exiguus has since been reclassified as Candida milleri sp. nov. Additionally, T. holmii is now identified as C. humili and found to be the principal yeast in Italian sweet sourdough items such as panettone and in sourdoughs made from whole durum flours. In this text, the names for these yeasts match those used by the researcher being quoted.)

Scientists were surprised that these yeasts generally did not ferment maltose because it is the principal fermentable sugar in doughs made without added sweeteners. This inability to use maltose turns out to be highly significant to the San Francisco sourdough matrix.

Typically, starter sponges contain 10 million to 25 million yeast cells per g, and when the count sinks below 4 million, poor proofing power and low loaf volume result. Going above that level, the dough balloons undesirably. In the dough, the starter is diluted by a factor of 7 to 10, and the initial yeast cell count is 2 million to 4 million per g, about one-fiftieth to one-hundredth of that in conventional bread dough. This low population is one reason for the typically long proof time of sourdough bread; however, it is the yeast that is responsible for the leavening of San Francisco sourdough products. The acid-generating bacteria contribute the sour flavor and acid functionality.

In experiments comparing sourdough yeasts with regular bakers yeast, Sugihara and co-workers (1970, 1971) demonstrated the unusual vigor of the sourdough yeasts in the very acid environment of the starter (pH 3.8 to 4.5). At higher pH (5.3, for example), the bakers yeast was superior in proofing power to the sourdough yeast. In these experiments, they also found the viability of the bakers yeast to diminish quickly. Just 6 hours after adding bakers yeast to the starter, its cell count dropped to less than 0.1% of original levels and sourdough yeast multiplied 5-fold.

Now bacteria enter the picture. Just isolating the bacteria of San Francisco sourdough bread involved creation of an unusual culture medium, according to the researchers (Sugihara et al. 1970, Kline and Sugihara 1971). The sourdough bacteria would grow on or ferment only maltose and not any of the usual sugars such as glucose, sucrose or fructose. This finding contradicted what is seen in rye sours and French bread sours, where those bacteria (typically heterofermative Lactobacillus types) consume most common mono- and disaccharide carbohydrates. In other words, San Francisco sourdough bacteria consume maltose and San Francisco sourdough yeasts do not.

The bacteria in San Francisco starters are gram-positive when young, catalase-negative, non-motile and appear as short rods but also as unusual involuted, pleomorphic and filamentous forms. They also require unsaturated fatty acids, fresh yeast extractives, carbon dioxide and an environment with pH of less than 6.0. Sugihara and his colleagues (1970, 1971) speculated that these sourdough bacteria were mutants derived from a known Lactobacillus that has lost its ability to ferment anything but maltose. They had discovered a new species, later named L. sanfranciscensis.

There also appears to be a symbiosis between the yeast and bacteria in sourdough. Studies that followed the work of Kline and Sugihara built on the knowledge that C. milleri cannot use maltose while L. sanfranciscensis can. From this body of work emerged findings that the bacterial enzyme phosphorylase, which acts on maltose, also releases glucose into the dough to stimulate the yeast. In addition, the bacteria secretes cycloheximide, an antibiotic that kills off competing microorganisms yet does not affect C. milleri.

The total acidity of sourdough, both dough and finished bread, is a good 10 times that of bread made by conventional methods. Kline et al. (1970, 1971) reported that baking produced little or no loss of total acidity or acetic acid.

It is remarkable that a product made out of the essentially flavorless ingredients flour, water, salt and yeast can have such rich aromas, flavors and tastes when consumed. In sourdough products, lactic and acetic acid provide a unique contribution, but bread flavor is also a function of the amino acids generated in the dough during fermentation and baking. Thiele et al. (2002) probed how sourdough lactobacilli, yeast and cereal enzymes added to amino acids that influence bread flavor. They learned that generating sufficient amino acid levels to positively affect flavor production depended primarily on pH (3.0 to 4.0), fermentation time and the consumption of those acids by the microflora present in the sourdough. The microbial formation of ornithine specifically enhanced the roasty note of bread crust odor. The amino acid ornithine is not metabolized by S. cerevisiae but is converted to 2-acetylpyrroline during baking, yielding the roasty, cracker-like aroma noted.

The pioneering work done by Kline, Sugihara and their colleagues to identify the how fermentation proceeds in San Francisco sourdough enabled other researchers around the world to better understand the sourdough fermentations native to their regions. For example, a group of Italian microbiolgists (Foschino et al. 1999) confirmed the presence of the L. sanfranciscensis bacteria in traditional sourdoughs used commercially to make panettone, pandoro, colomba, brioche and similar sweet products.

Readers interested in more detail about sourdough yeast and bacteria can consult Hansen (2006), Kulp (2003) and Sugihara (2003) for in-depth coverage of this topic. Volume II provides additional description of the processes and equipment for making baked foods using sourdough and starters.


Hansen, Å. 2006. Sourdough bread. In: Handbook of Food Science, Technology and Engineering, Vol. 4. Y.H. Hui, ed. CRC Press, Marcel Dekker: New York, NY.

Kline, L., and Sugihara, T.F. 1971. Microorganisms of the San Francisco sour dough bread process. II. Isolation and characterization of undescribed bacterial species responsible for the souring activity. Appl. Microbiol. 21 (3): 459.

Kline, L., Sugihara, T.F., and McCready, L.B. 1970. Nature of the San Francisco sour dough French bread process. I. Mechanics of the process. Bakers Digest 44 (2): 48.

Kulp, K. 2003. Bakers yeast and sourdough technologies in the production of US bread products. In: Handbook of Dough Fermentations. K. Kulp and J. Lorenz, eds. CRC Press: New York, NY.

Sugihara, T.F. 2003. Commercial starters in the US. In: Handbook of Dough Fermentations. K. Kulp and J. Lorenz, eds. CRC Press: New York, NY.

Sugihara, T.F., Kline, L., and McCready, L.B. 1970. Nature of the San Francisco sour dough French bread process. II. Microbiological aspects. Bakers Digest 44 (2): 48.

Sugihara, T.F., Kline, L., and Miller, M.W. 1971. Microorganisms of the San Francisco sour dough bread process. I. Yeasts responsible for the leavening action. Appl. Microbiol. 21 (3): 456.

Thiele, C., Gänzle, M.G., and Vogel, R.F. 2002. Contribution of sourdough lactobacilli, yeast and cereal enzymes to the generation of amino acids in dough relevant for bread flavor. Cereal Chem. 79 (1): 45.