What is the difference between lactic acid and acetic acid production in sourdough, and how does the fermentation protocol determine which dominates?

Fructilactobacillus sanfranciscensis is an obligate heterofermentor: it cannot produce lactic acid alone and always co-produces ethanol and CO₂ through the phosphoketolase (PK) pathway. The biochemical decision point between lactic acid dominance and acetic acid dominance is the NADH/NAD⁺ redox balance in the bacterial cell. The PK pathway produces acetyl phosphate as a key intermediate. Acetyl phosphate has two fates: (1) reduction to ethanol via acetaldehyde, consuming 2 mol NADH per mol ethanol and regenerating NAD⁺ — this is the route that pairs lactic acid production; (2) conversion to acetate via phosphotransacetylase and acetate kinase, generating ATP but not consuming NADH. When route 2 is taken, the cell needs an alternative NADH sink — often fructose reduction to mannitol, which regenerates NAD⁺ without producing ethanol. The result of route 2 is lactic acid + acetic acid + mannitol + CO₂. Practically, the lactic-to-acetic ratio is controlled by four fermentation variables. Temperature: warm fermentation (30–35°C) accelerates lactic dehydrogenase preferentially, producing more lactic acid relative to acetic; cool fermentation (8–12°C) slows lactic dehydrogenase more than it slows acetate kinase, shifting production toward acetic acid. Hydration: stiff doughs (below 65% hydration) have limited free water, which restricts molecular diffusion and enzymatic activity; under this water stress the pathway shifts toward acetic acid production. Duration: early in fermentation (first 4–8 hours), glucose is the primary substrate and lactic acid dominates; after glucose depletion, maltose becomes the primary substrate and, as the NADH/NAD⁺ ratio shifts, acetic acid production increases. Fructose availability: when fructose is available as an alternative electron acceptor (reduced to mannitol), F. sanfranciscensis channels more acetyl phosphate to acetic acid rather than ethanol, increasing the acetic acid fraction without increasing ethanol. In practice: a 68–72% hydration dough fermented for 4–6 hours at 28°C produces a mild, lactic-dominant (yogurt-like) sourness. A 62–65% hydration stiff dough fermented for 12–16 hours at 8–10°C produces a sharper, acetic-dominant (vinegar-forward) sourness. Both are authentic sourdough — the flavor direction is a deliberate technical choice, not randomness.

How does gluten network development work in sourdough dough, and what role does the acidic pH from organic acids play in disulfide bond formation and protease activity?

The gluten network is a viscoelastic protein matrix formed by two wheat storage protein families: glutenins (polymeric, providing cohesiveness and elasticity) and gliadins (monomeric, providing extensibility and plasticity). High-molecular-weight glutenin subunits (HMW-GS) are the primary determinants of dough strength. The most baking-relevant alleles are 1Dx5 + 1Dy10 (encoded by the Glu-D1 locus on chromosome 1D) and 1Bx7 + 1By8 (Glu-B1). HMW-GS 1Dx5 contains 16 cysteine residues: 14 form intra-chain disulfide bonds that stabilize the subunit's secondary structure, and 2 terminal cysteines are available for inter-chain intermolecular disulfide crosslinking. Strong flour (bread flour, W250–W380 alveograph) has glutenin polymers reaching molecular weights above 10 million Da through these inter-chain SS bonds. Mechanical work — kneading, stretching, folding — deforms the polymer network and aligns glutenin chains so that terminal SH groups from adjacent molecules come into proximity. In the presence of dissolved oxygen (the reason artisan bakers often autolyse in an uncovered bowl) and natural ascorbic acid from flour (milled wheat contains ~1–3 mg/100g ascorbic acid), the following redox chain promotes disulfide bond formation: ascorbic acid is oxidized by O₂ in the presence of ascorbic acid oxidase (naturally present in wheat) or by glutathione oxidase to dehydroascorbic acid (DHAA). DHAA then directly oxidizes free SH groups on glutenin cysteines via a thiol-disulfide exchange reaction, converting –SH + –SH pairs to –S–S– crosslinks. This is why commercial bakers add ascorbic acid (E300) as a dough improver — it accelerates gluten network crosslinking. The acidic pH of sourdough (3.8–4.5 by mid-bulk fermentation) has two important effects on the gluten network. First, gliadins approach their isoelectric point (pI approximately 4.0–4.5) at sourdough pH, causing increased gliadin self-association and tighter hydration into the gluten matrix — the dough feels stronger and more extensible at once. Second, and more counterintuitively, the low pH inhibits endogenous wheat cysteine protease (EC 3.4.22.x, a family of thiol-endopeptidases with optimum activity at pH 4.5–6.5). These proteases are native to the wheat grain (present in the flour) and their activity during long fermentation would normally degrade the glutenin polymer network, weakening gluten and eventually producing slack, sticky dough. At pH below 4.0, cysteine proteases are largely inactivated. The organic acids produced by F. sanfranciscensis — lactic and acetic acid — serve a dual role: they create the sour flavor that consumers associate with artisan sourdough, and simultaneously protect the gluten network from enzymatic degradation during the long cold proof. This is why sourdough doughs can be cold-proofed for 14–20 hours at 4°C without the structural collapse that would occur in commercial-yeast doughs at the same temperature and duration. The acidic inhibition of protease is the mechanistic basis for the superior dough handling properties that experienced sourdough bakers observe empirically but rarely quantify: a properly acidified sourdough holds its shape better on the bench (due to inhibited protease + low temperature slowing remaining enzymatic activity), and the oven spring is stronger because the gluten network has maintained its crosslink density through the long fermentation.

Explainers · 2026-07-09

Patreon for sourdough creators: Fructilactobacillus sanfranciscensis phosphoketolase heterofermentation, Kazachstania humilis wild yeast symbiosis, lactic vs acetic acid ratio mechanics, gluten network disulfide crosslinking, organic acid protease inhibition, alpha-amylase starch hydrolysis, Maillard reaction and Strecker degradation, starch gelatinization physics, and the Apple Tax

Sourdough Patreon retention is built on the technical layer that process video cannot fully carry: the biochemistry of why your starter ratio and fermentation temperature produce one flavor profile and not another, what is actually happening in the gluten network during a 14-hour cold proof, why the pH that makes bread sour also prevents structure collapse, and the specific molecular events during oven spring that produce the open crumb. Sourdough audiences are heavily iOS—Instagram crumb-shot photography and TikTok scoring reveals are among the most iOS-concentrated content categories on Patreon. Apple Tax exposure begins November 1, 2026.

Wild yeast and bacterial ecology of a sourdough starter

A mature sourdough starter is a stable microbial consortium dominated by two organism classes: obligate heterofermentative lactic acid bacteria (LAB) and wild yeasts. The organisms that dominate a healthy, regularly fed starter are not random environmental microbes—they are specialists selected by the specific stresses of a flour-and-water environment: high osmotic pressure, periodic starvation and nutrient pulse, acidic pH, and competition from other microbes.

The bacterial dominant in most well-established starters is Fructilactobacillus sanfranciscensis—reclassified from Lactobacillus sanfranciscensis in 2020 following whole-genome phylogenetic analysis. It was first isolated from San Francisco sourdough in 1971 and named after the city. Despite the name, it is found in established starters worldwide wherever the ecological conditions are right: obligate heterofermentation, tolerance of pH as low as 3.5, and a specific nutritional niche based on maltose metabolism. It is not the only LAB present—other heterofermentative species (Ligilactobacillus pontis, Levilactobacillus brevis, Secundilactobacillus kimchiensis) contribute to the ecosystem, especially in young or variable starters—but F. sanfranciscensis is the dominant and best-characterized species in artisan bread fermentation science.

The primary wild yeast in San Francisco-type starters and most established European artisan starters is Kazachstania humilis, formerly classified as Candida humilis and also known as Saccharomyces cerevisiae var. hollandicus. K. humilis is osmotolerant, survives at pH 3.5–4.0, and grows in the presence of the concentrations of organic acids produced by F. sanfranciscensis—conditions that inhibit commercial baker’s yeast (Saccharomyces cerevisiae). Commercial S. cerevisiae struggles in established sourdough starters because of its low acid tolerance and is progressively outcompeted by K. humilis over successive feedings, which is why adding commercial yeast to a sourdough starter gradually changes its character over months of maintenance.

The coexistence of K. humilis and F. sanfranciscensis is not accidental—it is a metabolic symbiosis structured by substrate niche separation. When flour is mixed with water, the primary carbohydrates available are sucrose (from the grain), glucose, fructose, and maltose (from amylase-mediated starch hydrolysis, which begins immediately upon hydration). K. humilis preferentially ferments maltose and glucose. F. sanfranciscensis has a constitutively expressed maltose phosphorylase and can also ferment maltose, but K. humilis outcompetes it for maltose when both are present. Crucially, K. humilis lacks the fructose transport system that F. sanfranciscensis has, leaving fructose largely accessible to the bacteria. The result is a stable two-species ecosystem: K. humilis drives CO&sub2; production (leavening) from maltose and glucose; F. sanfranciscensis acidifies the dough from glucose, fructose, and maltose metabolized via its phosphoketolase pathway; fructose acts as an electron acceptor for bacterial NAD&sup+ regeneration (via reduction to mannitol), influencing the lactic-to-acetic acid ratio.

Sourdough educators and recipe documentors who can explain the starter ecology layer—why the stable state emerges, what destabilizes it (irregular feeding, temperature swings, chlorinated water inhibiting bacteria but not yeast, or the reverse), and how to diagnose what’s wrong from the smell and texture—produce content that no book or YouTube algorithm-optimized video bothers to cover at this depth.

Fructilactobacillus sanfranciscensis phosphoketolase heterofermentation

F. sanfranciscensis is an obligate heterofermentor, meaning it cannot perform simple homofermentation (glucose → 2 lactic acid). It lacks phosphofructokinase, the enzyme required for the Embden–Meyerhof–Parnas (EMP) glycolysis pathway. Instead, it operates exclusively through the phosphoketolase (PK) pathway, also called the pentose phosphate-linked heterofermentative pathway.

In the PK pathway, glucose enters as glucose-6-phosphate (from glucose uptake via phosphoenolpyruvate phosphotransferase system, PEP-PTS) and is isomerized to fructose-6-phosphate. Fructose-6-phosphate is then oxidized and decarboxylated by fructose-6-phosphate phosphoketolase (F6PPK) to yield erythrose-4-phosphate + acetyl phosphate + CO&sub2;. Wait—that is incorrect. The PK pathway proceeds: glucose-6-phosphate → 6-phosphogluconate (via glucose-6-phosphate dehydrogenase, consuming NAD&sup+;, producing NADH) → ribulose-5-phosphate + CO&sub2; (via 6-phosphogluconate dehydrogenase, consuming another NAD&sup+;, producing another NADH) → xylulose-5-phosphate (via ribulose-5-phosphate epimerase) → xylulose-5-phosphate is cleaved by phosphoketolase to glyceraldehyde-3-phosphate (GAP) + acetyl phosphate. GAP follows the lower half of the EMP pathway to pyruvate, which is reduced to lactic acid (consuming 1 NADH, regenerating NAD&sup+;). Acetyl phosphate is reduced to ethanol via acetaldehyde (consuming 2 NADH, regenerating 2 NAD&sup+;). Net per mol glucose: 1 mol lactic acid + 1 mol ethanol + 1 mol CO&sub2;. Net ATP yield: 1 mol ATP (versus 2 mol from EMP homofermentation—heterofermentation is energetically less efficient). This obligate co-production of ethanol and CO&sub2; alongside lactic acid is the defining biochemical feature of F. sanfranciscensis and all obligate heterofermentors.

Maltose metabolism in F. sanfranciscensis differs from that in yeast. F. sanfranciscensis uses maltose phosphorylase (not the maltase that K. humilis uses) to cleave maltose: maltose + inorganic phosphate → α-glucose-1-phosphate + β-glucose. The α-glucose-1-phosphate is converted by phosphoglucomutase to glucose-6-phosphate and enters the PK pathway normally. The β-glucose (free glucose released) can be taken up by phosphotransferase and also enters as glucose-6-phosphate, or it can be taken up by K. humilis if present. This maltose phosphorylase reaction is the basis for the maltose symbiosis: one mol of maltose yields two mol of hexose-6-phosphate entering the PK pathway, and the free glucose released is available to K. humilis as an additional substrate.

The acetyl phosphate branch of the PK pathway is the fork that determines lactic vs acetic acid output. Acetyl phosphate can go to ethanol (via phosphotransacetylase → acetaldehyde dehydrogenase → alcohol dehydrogenase, consuming 2 NADH total, regenerating 2 NAD&sup+;) or to acetic acid (via phosphotransacetylase → acetate kinase, generating 1 additional ATP but consuming no NADH). The choice is governed by the cellular NADH/NAD&sup+ ratio. When NADH is in excess (pyruvate reduction to lactate alone cannot consume all the NADH produced by the oxidative decarboxylation steps), more acetyl phosphate is channeled to ethanol to regenerate NAD&sup+;. When an alternative NADH sink is available—specifically fructose reduction to mannitol via mannitol-1-phosphate dehydrogenase—the cell can regenerate NAD&sup+ via mannitol rather than ethanol, channeling acetyl phosphate to acetic acid instead, producing acetic acid + mannitol rather than ethanol.

The documentation opportunity for Patreon: most sourdough guides describe the lactic/acetic ratio outcome (stiff/cold/long = more acetic) without explaining the biochemical why. The why—that it is a cellular redox balance decision, that the presence of fructose as an electron acceptor is the proximate cause, and that the same flour composition on different days (different fructose-to-maltose ratio depending on flour freshness and amylase activity) will produce different flavor outcomes even with identical protocol—is the layer that builds patron understanding beyond recipe following.

Lactic vs acetic acid: the four fermentation variables

Lactic acid (2-hydroxypropanoic acid, pKa 3.86) and acetic acid (ethanoic acid, pKa 4.76) have different sensory thresholds and different flavor profiles. At equal molar concentration in dough, lactic acid produces a mild, creamy, yogurt-like acidity—its pKa of 3.86 means it is 82% dissociated (ionized) at dough pH 4.2, weakening its perceived sharpness because the dissociated carboxylate anion is less volatile and less directly sensory-active on the palate than the undissociated acid. Acetic acid at pKa 4.76 is only 28% dissociated at pH 4.2, meaning 72% remains as undissociated acetic acid—volatile, sharp, and directly sensory-active. The characteristic “sharp” vinegar note of San Francisco sourdough versus the milder tang of a French levain represents this pKa difference played out in the fermentation conditions of each tradition.

The four variables that experienced sourdough educators document in their fermentation protocols:

Temperature: Warm bulk fermentation (28–34°C) accelerates F. sanfranciscensis growth rate and preferentially elevates lactic dehydrogenase activity relative to the acetate kinase branch. At 28°C, lactic-to-acetic acid ratios (mass ratio) in the range of 3:1 to 5:1 are typical. At 8–12°C (cold-proofed overnight), the ratio shifts toward 1.5:1 to 2:1, with a perceptibly sharper crumb. The temperature response is not symmetric—acetic acid production is relatively cold-tolerant, while lactic acid production is temperature-sensitive in a way that makes it disproportionately suppressed at lower temperatures.

Hydration: Stiff doughs (below 65% hydration) have reduced water activity, which restricts enzyme mobility and diffusion within the dough matrix. The lactic dehydrogenase (LDH) reaction—cytoplasmic, requiring direct substrate diffusion—is more sensitive to water limitation than the acetate kinase reaction. Stiff doughs therefore shift toward higher acetic acid output. This is the mechanistic basis for the traditional baguette de tradition française, which uses a very stiff starter (levain dur, 45–50% hydration) to produce a milder, lactic-dominant sourness despite long fermentation times. Contrast with a 100%-hydration liquid levain fed at 1:5:5 (starter:flour:water) producing a bolder acetic character.

Duration: In the first 4–6 hours of bulk fermentation, glucose and fructose are the primary substrates. Later, maltose predominates as amylases release it from starch. The shift from glucose to maltose substrate correlates with higher acetic acid production, partly because of the maltose phosphorylase route and partly because glucose depletion changes the cellular redox environment. Longer fermentations (12–24 hours total including retard) accumulate more acetic acid cumulatively than shorter ones at the same temperature.

Inoculation rate: A higher starter percentage (20–30% preferment) produces faster acidification but with less time for the acetate branch to accumulate—resulting in a milder, more lactic profile. A lower inoculation rate (5–10% preferment) produces a longer lag before acidification, during which amylase activity and yeast fermentation proceed longer before organic acid inhibition sets in—producing more complex flavor but with more time for acetic accumulation. Sourdough educators who publish their inoculation rate alongside temperature and hydration in a structured variable table provide the documentation that enables patron replication without guesswork.

Gluten network development and disulfide crosslinking

The gluten network forms from two wheat storage protein families present in flour: glutenins (polymeric, linked by inter-chain disulfide bonds) providing elasticity and cohesiveness, and gliadins (monomeric, hydrogen-bonded and hydrophobic within the network) providing extensibility and plasticity. The viscoelastic balance between these two determines dough handling character: too much glutenin crosslinking produces a tough, elastic dough that resists shaping and tears; too little produces a slack, extensible dough that spreads without holding shape.

High-molecular-weight glutenin subunits (HMW-GS) are the molecular scaffolding. They are encoded by the Glu-A1, Glu-B1, and Glu-D1 loci on the group-1 homeologous chromosomes of hexaploid bread wheat. The most technologically important alleles: Glu-D1d (encoding subunits 1Dx5 + 1Dy10), the highest-quality allele for baking strength; and Glu-B1b (encoding 1Bx7 + 1By8). HMW-GS 1Dx5 has 16 cysteine residues. Fourteen form intra-chain disulfide bonds, stabilizing the subunit’s repetitive β-spiral secondary structure. Two terminal cysteines are available for inter-chain crosslinking. Glutenin macropolymers (GMP) reach molecular weights above 10 million Da in strong bread flour—among the largest natural polymers in common food ingredients. These are the molecules that make bread dough elastic in a way that pasta dough, cookie dough, and cake batter are not.

Gluten development via mechanical work (kneading, stretch-and-fold, coil folding) operates by bringing reactive cysteine SH groups on adjacent glutenin chains into proximity. In the oxidizing environment of a hydrated dough, thiol-disulfide exchange reactions progressively increase the inter-chain SS bond density. The ascorbic acid naturally present in flour (~1–3 mg/100g in whole wheat, lower in white flour) participates via the ascorbic acid oxidase/glutathione oxidase cascade: ascorbic acid → dehydroascorbic acid (DHAA) under enzymatic oxidation; DHAA oxidizes free SH groups to SS bonds via thiol-disulfide exchange (DHAA + 2 R–SH → ascorbic acid + R–S–S–R). This is why commercial dough improvers contain ascorbic acid (E300) and why kneading in the presence of atmospheric oxygen builds gluten faster than kneading in an inert atmosphere.

In sourdough, the stretch-and-fold technique (four sets of four cardinal-direction folds over the first 2–3 hours of bulk fermentation at 30-minute intervals, without mechanical kneading) achieves gluten development through deformation rather than sustained mechanical work. Each fold aligns polymer chains, increases SS bond formation probability at chain termini, and traps the CO&sub2; produced by K. humilis in an increasingly reinforced gluten matrix. An underfolded dough at the same bulk fermentation time will have lower GMP mass (fewer inter-chain SS crosslinks) and less gas retention, producing a flat, dense crumb. An overfolded dough becomes so elastic that it tears during shaping rather than extending smoothly—the SS bond density has overshot the optimal range. Documenting the dough feel at each fold stage—slack and extensible after fold 1, increasingly elastic through fold 4, and the specific tactile change that indicates adequate development—is proprietary knowledge that patrons cannot find in a recipe card and that is poorly communicated by video alone without the matching written description.

Low-molecular-weight glutenin subunits (LMW-GS) contribute to the network as shorter crosslinking elements. ω-gliadins are entirely non-crosslinking (no cysteine residues) and contribute extensibility; α-, β-, and γ-gliadins have cysteine residues that form intra-chain SS bonds only and interact with the network via disulfide bonds at specific positions. The ratio of HMW-GS to LMW-GS to gliadin varies with wheat variety and dramatically affects the final dough character—a key reason why variety selection in grain-to-loaf sourdough content is technically justified, not just marketing.

Organic acid protease inhibition during bulk fermentation and cold proof

Wheat grain contains a suite of endogenous proteases, including cysteine proteases (asparaginyl endopeptidase family, Cys protease EC 3.4.22.x), serine proteases, and aspartic proteases. These are present in the flour and become active upon hydration. Their natural role is to mobilize storage proteins during germination for seedling nutrition. In bread dough, their activity degrades glutenin polymers, reducing gluten network strength and dough stability over time. This is why commercial yeast doughs left to over-ferment become sticky and difficult to handle: prolonged yeast activity maintains dough pH near neutral (5.5–6.0), where wheat cysteine proteases are fully active, degrading the gluten continuously.

Sourdough fermentation suppresses this protease degradation via pH. As F. sanfranciscensis produces lactic and acetic acids, dough pH falls from an initial 6.0–6.5 toward 3.8–4.5 by mid-bulk-fermentation. Wheat cysteine proteases have their activity optimum at pH 4.5–6.5; below pH 4.0, they are effectively inhibited. The organic acid accumulation that produces sourness therefore simultaneously inhibits the enzymatic mechanism that would otherwise degrade the gluten network. This is the mechanistic basis for the paradox observed in practice: a 14–20 hour cold-proofed sourdough loaf often has better structure and stronger oven spring than a 2-hour room-temperature commercial-yeast loaf, despite far longer total fermentation time. The acidity that makes sourdough identifiable as sourdough is also the chemical signal that suppresses protease and preserves gluten structure.

Temperature adds a second protection layer. Cold retard at 4°C reduces the diffusion coefficients of proteases and their substrates. Proteolytic degradation rates at 4°C are roughly 10–20-fold lower than at 25°C (Q⊂10; ≈ 2 per 10°C). The combination of low pH (chemical inhibition) and low temperature (kinetic suppression) allows cold-proofed sourdough to maintain excellent dough structure for 14–24 hours in the refrigerator. Documenting both the pH drop during bulk fermentation (measurable with a narrow-range pH strip calibrated for acidic dough, or with a probe pH meter) and the role of cold temperature in the structural equation gives patrons a scientifically grounded understanding of cold retard that elevates far above “more flavor from the fridge.”

Glutathione, the tripeptide gamma-glutamyl-cysteinyl-glycine (GSH) present in flour at ~100–150 mg/kg, also affects gluten network stability. GSH is a reducing agent that cleaves SS bonds (GSH + R–S–S–R → GSSG + 2 R–SH), softening dough. At low pH, GSH is more protonated (pKa of thiol group is ~8.7, so this is less about pH directly and more about the competing oxidation by organic acid fermentation environment). The net effect at sourdough pH is that GSH-mediated SS cleavage is outpaced by the ascorbic acid/DHAA-mediated SS formation when conditions are correct—but high-activity starters that ferment too fast (before the gluten network has been adequately developed by folding) can produce GSH-dominant conditions in the first hour, explaining why underdeveloped sourdough doughs feel sticky even before obvious over-fermentation.

Alpha-amylase and beta-amylase: starch hydrolysis during fermentation and baking

Starch is the primary carbohydrate in flour, comprising 65–75% of flour dry weight. The amylases in flour—both endogenous to the grain and added by millers when grain amylase activity is low—begin hydrolyzing starch the moment flour is hydrated. This hydrolysis provides sugars for both K. humilis (leavening) and F. sanfranciscensis (acidification). The amylase system also directly affects final crumb texture, crust color, and shelf life.

Alpha-amylase (EC 3.2.1.1) is an endoamylase that randomly cleaves α-1,4 glycosidic bonds anywhere along the amylose or amylopectin chain interior. It produces a mixture of maltose, maltotriose, glucose, and soluble dextrins—larger starch fragments. It has a temperature optimum around 60–70°C and is thermostable until approximately 80°C, meaning it remains active throughout most of the baking process, continuing to hydrolyze starch granules as they gelatinize, until the internal crumb temperature exceeds ~80°C. High alpha-amylase activity (from sprout damage—starch from field-germinated grain that has elevated amylase from the germination process, quantified by a low Falling Number score, typically below 200 seconds) produces a sticky, gummy crumb because gelatinized starch is continuously hydrolyzed by the still-active alpha-amylase during baking before the crumb structure can set. This is the flour quality defect that makes certain harvest years technically difficult for bakers.

Beta-amylase (EC 3.2.1.2) is an exoamylase that cleaves maltose units from the non-reducing ends of amylose and amylopectin chains. It cannot bypass α-1,6 branch points, so it produces maltose plus limit dextrins. Its temperature optimum is 50–60°C and it is thermolabile, denatured at approximately 65°C. Beta-amylase is the primary maltose-generating enzyme during bulk fermentation and the early oven rise phase below 65°C. The maltose it generates is the primary fuel for F. sanfranciscensis in the later stages of fermentation and for K. humilis throughout.

Damaged starch (starch granules physically damaged by milling rollers, increasing their surface area and accessibility to water and enzymes) provides a more accessible substrate for both amylases. White flour from modern roller mills has 6–10% damaged starch; stone-ground whole wheat has 3–5% because stone grinding generates less mechanical damage. High damaged starch accelerates sugar release and therefore acidification rate, which is one reason high-extraction and whole wheat sourdoughs ferment faster than white flour sourdoughs at the same temperature and inoculation rate.

The Falling Number (FN) test (ISO 3093) measures alpha-amylase activity in a standardized way: a flour-water slurry is heated and stirred, and the time (in seconds) for a standardized needle to fall through the gelatinizing slurry under gravity is recorded. High alpha-amylase activity liquefies the gelatinizing starch, giving a low FN (sticky crumb risk). Normal wheat flour should have FN > 300 seconds. Millers and bakers blend flour lots from different harvests to target a FN in the range of 250–350 seconds. Sourdough educators teaching grain-to-loaf workflows, or who work with heritage grain varieties and small-mill flour, document the FN of their flour lots as a key variable affecting fermentation speed and crumb texture outcomes.

Maillard reaction and Strecker degradation during baking

Crust flavor in sourdough is not simple caramelization. The brown color and complex aroma of a properly baked loaf is primarily the result of the Maillard reaction—a cascade of non-enzymatic amino acid–reducing sugar reactions—and a sub-pathway called Strecker degradation that produces the specific volatile aldehydes most responsible for the characteristic smell of freshly baked bread.

The Maillard reaction initiates at surface temperatures above approximately 120°C. In bread baking, the crust surface reaches 165–200°C; the Maillard cascade proceeds throughout the baking period at the crust surface. The reaction begins with condensation of a free amino group (from the ε-amino group of lysine residues in gluten proteins, or from free amino acids released by protease activity during fermentation) with a reducing sugar (glucose, fructose, maltose, or other reducing carbohydrates produced by amylase activity). The condensation product is a glycosylamine (an N-substituted amino sugar), which undergoes an Amadori rearrangement to produce 1-amino-1-deoxy-2-ketose (the Amadori product). The Amadori product is the key branch point: it can undergo enolization, dehydration, retro-aldolization, or elimination reactions. Multiple pathways from the Amadori product produce: α-dicarbonyl compounds (methylglyoxal, diacetyl, glyoxal, acetoin); furanyl compounds (5-hydroxymethylfurfural from hexose-derived Amadori products, furfural from pentose-derived products via arabinoxylans in wheat); and eventually brown polymeric melanoidins that provide crust color.

Strecker degradation is a specific reaction of α-amino acids with the α-dicarbonyl compounds produced from Amadori product degradation. The mechanism is a transaminase-like reaction: the α-amino acid reacts with the α-dicarbonyl via Schiff base formation, producing an α-aminocarbonyl compound and an α-imino acid, which hydrolyzes to an α-keto acid; the α-keto acid spontaneously decarboxylates to produce an aldehyde with one less carbon than the original amino acid, releasing CO&sub2;. Each amino acid produces its characteristic Strecker aldehyde:

Alanine → acetaldehyde (fruity, sharp; detection threshold ~25 ppb). Leucine → 3-methylbutanal (malty, roasty, “bread-like”; threshold ~0.2 ppb). Isoleucine → 2-methylbutanal (malty; threshold ~0.5 ppb). Valine → 2-methylpropanal (malty, green; threshold ~0.5 ppb). Phenylalanine → phenylacetaldehyde (honey, rose, floral; threshold ~4 ppb). Methionine → methional (boiled potato, cabbage-like in excess; threshold ~0.2 ppb). Proline reacts slightly differently—the cyclic secondary amine does not undergo classical Strecker degradation but reacts with methylglyoxal and other dicarbonyls to produce 2-acetyl-1-pyrroline (2-AP), the compound responsible for the iconic roasted bread aroma, the same compound found in jasmine rice, pandan leaves, and basmati. Its odor threshold is approximately 0.1 ppb; it is detectable at trace quantities and is the dominant volatility character of freshly baked sourdough crust. 2-AP degrades rapidly after baking, which is why the aroma of fresh bread cannot be sustained—and why a reheated day-old loaf (briefly restoring crust temperature above 120°C) regenerates a partial 2-AP burst.

The arabinoxylans in wheat bran (pentosans, 2–3% of flour dry weight) provide pentose sugars (arabinose, xylose) that produce furfural via the Amadori/Strecker routes. Furfural has an almond-like, caramel aroma (threshold ~3 ppb) and contributes to the aromatic complexity of whole wheat and high-extraction sourdough more than white-flour sourdough. Sourdough educators documenting the specific flour extraction rate and its predicted effect on Maillard-volatile profiles—as a scientifically grounded explanation for why higher-extraction loaves smell more complex—produce content that translates fermentation science into sensory prediction.

Starch gelatinization and crumb structure formation

The crumb structure of a sourdough loaf sets during a specific 15–20 minute window in the oven. Understanding the molecular events during this window explains why oven temperature, steam timing, and scoring angle produce their specific effects.

Native wheat starch granules are semi-crystalline structures, 5–50 μm in diameter, composed of amylose (25–28% of starch, molecular weight 100–500 kDa, essentially linear α-1,4 glucan chains) and amylopectin (72–75%, MW 10–100 million Da, highly branched α-1,4 glucan with α-1,6 branch points every 24–30 glucose residues on average). Amylopectin organizes into crystalline lamellae 15–25 nm thick through double-helical packing of adjacent side chains. These crystalline domains alternate with amorphous lamellae in a repeating 9 nm periodicity visible by small-angle X-ray scattering (SAXS). Amylose forms inclusion complexes with lipids (V-type single helices), which are visible as the “Maltese cross” birefringence pattern under polarized light microscopy—the same pattern that disappears during gelatinization.

Wheat starch gelatinization is a thermally irreversible process. By differential scanning calorimetry (DSC), wheat starch gelatinization onset is approximately 58°C, peak approximately 65°C, completion approximately 72°C (values shift with water availability; in excess water these are reliable; in a dough at 60–70% hydration, the transition shifts slightly higher and broadens). During oven spring (the first 10–15 minutes of baking, when the crumb interior is heating from initial dough temperature ~4°C to the gelatinization range), the following sequence occurs:

Below 55°C: K. humilis yeast remains active, producing CO&sub2; at an accelerated rate as temperature rises (enzyme activity increases with temperature up to the thermal inactivation point). Maximum CO&sub2; pressure within the crumb gas cells. Gluten proteins are still elastic and the gas cells expand without setting.

55–58°C: K. humilis approaches thermal death (typical yeast thermal death range 55–60°C). CO&sub2; production ceases. Gas cells at their maximum volume begin to depend entirely on the gluten network for containment. Gluten proteins (denaturation range 60–85°C for gliadins; glutenins are more thermostable) begin to lose native structure and form new heat-induced intermolecular bonds.

58–72°C: Starch gelatinization. Water penetrates amorphous regions first, then swells crystalline lamellae. Birefringence (the Maltese cross pattern) disappears as crystalline order is disrupted. Starch granules swell to 20–60 times their dry volume, absorbing the free water in the dough. Amylose leaches from swollen granules into the surrounding gel phase. Alpha-amylase is still active throughout this range, cleaving gelatinized starch—if alpha-amylase activity is too high (low FN flour), it liquefies the gelatinizing starch before the crumb can set, producing the sticky, gummy crumb defect.

75–80°C: Crumb structure sets irreversibly. Gelatinized amylopectin forms a continuous gel network around the gas cells; denatured gluten proteins coagulate within and at the cell walls, reinforcing the structure. The crumb is no longer elastic—the gas cells can no longer expand, and the loaf volume is fixed. The simultaneous gelatinization and gluten denaturation is why bread crumb structure is firm but not rubbery: starch provides the rigid gel matrix; gluten provides the flexible protein membrane within cell walls.

Steam injection during the first 10–15 minutes of baking serves a specific function in this sequence: it keeps the crust surface moist and extensible during oven spring, preventing premature crust formation that would restrict the expansion of the still-elastic interior. Without steam, a dry crust sets before the interior has fully expanded, limiting loaf volume and crumb openness. With steam, the surface remains pliable through the oven spring period; the Maillard browning and crust hardening begin only after steam is vented and the surface temperature rises above 120°C. This is why sourdough educators documenting the timing and source of steam injection (covered dutch oven lid for the first 20 minutes vs. dedicated steam injectors vs. ice cubes on a tray below the baking surface) are providing technically justified variable documentation, not preference.

iOS rates and Apple Tax

Sourdough and artisan bread creator iOS rates are high across all primary platforms. YouTube sourdough tutorial channels—bulk fermentation timing, scoring technique, open crumb troubleshooting, starter feeding schedules, grain-to-loaf milling workflows—track at 62–74% iOS, reflecting a mix of home bakers watching on iPad and iPhone in the kitchen during bake sessions and desktop researchers planning upcoming bakes. Instagram sourdough photography—scored loaves, crumb pull shots, starter bubble documentation, baking timeline posts—tracks at 74–84% iOS; the food aesthetics and artisan bread community is among the most iOS-concentrated on the platform. TikTok sourdough scoring videos, crumb reveal pulls, “my first sourdough loaf” transformation content, and active fermentation bubble videos track at 76–84% iOS, reflecting TikTok’s mobile-first audience.

Beginning November 1, 2026, Apple charges Patreon 30% on every subscription payment processed through the iOS app.

At $200/month with 70% iOS: approximately $42/month ($504/year) in Apple fees. At $350/month with 74% iOS: approximately $77.70/month ($932.40/year). At $500/month with 78% iOS: approximately $117/month ($1,404/year). Enable Patreon’s web-only billing toggle before October 31, 2026 and update all subscription CTAs—Instagram bio link, TikTok link-in-bio, YouTube About page URL—to the direct Patreon web URL. Verify with a test subscription from Safari on an iPhone before November 1.

KeepTier is a self-hosted membership page for creators who want 100% of their tier revenue and zero Apple Tax. Plans from $9/month.

Patreon for sourdough creators · Patreon for pottery creators · Patreon for natural dyeing creators · Eight Patreon alternatives compared