Experimenting with Different Flours in Sourdough Starter (2024)

Last spring I read this Serious Eats article about different sourdough starters: their fermentation power and flavors. The article was posted in this Breadtopia forum thread by @djd418 and it documented an experiment of chef Tim Chin where he created five sourdough starters with different flours and baked breads with each starter. I was super impressed by the experiment and his breads; I struggle to keep variables constant in experiments with just two loaves. I did, however, wish his final dough formulas had been identical.

Each of his breads was 80:20 bread flour to sifted wheat flour plus the particular sourdough starter. So, for example, a rye starter test loaf had rye flour in it (from the starter), but an all purpose starter test loaf didn’t have any rye in the bread at all. And since the starter was 20% of the total flour weight that meant that 9% of the flour varied between some of the doughs (some of the starter types were 50:50 two flours). The differences in leavening and flavor could have come from the different flours only, not from the starters having a different character due to their flour types.

This past week, I did a scaled down version of this experiment: three starters and loaves, but with the modifications I felt were important. My experiment had different weaknesses, though:

1. I didn’t feed the new starters their new food source for multiple feedings, which likely means there wasn’t a tremendous shift in microbial population. They did smell and behave differently, but the differences might have been more pronounced if the starters were long-established with their different foods.
2. I used a no-knead approach to not introduce the variable of gluten development at different fermentation stages, but this left me with floppy doughs that I shaped inconsistently. Luckily, the loaf that suffered the most in this regard was the starter/dough that I’d decided to disqualify because it had a different shaped proofing basket compared to the other two doughs.

Despite these two methodological concerns — the rapid starter transition and the shaping issues — I think my results are interesting and worth reporting on.

Starter Process

I set up the following three starters.

25g all purpose flour starter
50g water
50g whole grain sprouted hard red spring wheat (sprouted)

25g all purpose flour starter
50g water
50g bread flour

25g all purpose flour starter
50g water
50g whole grain rye flour

With cool house temps, I had to let my fresh-milled whole grain flours cool to the same temperature as my bread flour before adding the starter. All three starters were at 74°F after feeding, and I placed them on the Raisenne dough warmer and their temps climbed to the mid-80s.

Starter Findings

The starters all contain 12.5g all purpose but for efficiency, I’m going to refer to them by the name of their feeding flour only. I’m also going to report on them in the order that they are depicted in all the photos, from left to right.

Sprouted flour starter smelled sweet and malty even when peaked. It also felt the thinnest and peaked at the lowest level. These findings may be explained by sprouted wheat reportedly having more enzymatic activity than other flours, which releases sugar (providing more available yeast food). Also, the gluten strength of whole grain flour is not as strong as a refined flour, so fewer bubbles are trapped in the expanding paste. (The rye starter, which also was whole grain, was thirstier and trapped CO2 with its thickness, not gluten.)

Bread flour starter smelled mildly of acetone throughout. The surface bubbles were the largest, likely due to the gluten strength of the flour. This starter peaked “second” between rye and sprouted.

Rye flour starter smelled earthy and it was much stiffer feeling at the same hydration as the other two, which forum member @singkevin had actually warned me about here. I had to lightly wet my fingertips to press it flat in the jar so I’d be able to mark the level on the jar, and when it peaked, this starter had a dome and was the tallest. For the microbial flow in the starters to be more similar, I should have used more water in the rye starter. The stiffness of this starter is likely why it peaked last.

Dough Formula

I mixed up the doughs one after another, one minute in my stand mixer, with sprouted starter first because it seems to have stopped expanding first, then bread flour starter, then rye starter. The ingredients of the three doughs are as follows:

Sprouted Flour Starter Dough

350g bread flour
50g rye flour
125g sprouted hard red spring wheat flour starter
300g water
9g salt

Bread Flour Starter Dough

300g bread flour
50g rye flour
50g sprouted hard red spring wheat flour
125g bread flour starter
300g water
9g salt

Rye Flour Starter Dough

350g bread flour
50g sprouted hard red spring wheat flour
125g rye flour starter
300g water
9g salt

Baker’s Percentages (Total Flour Weight 462.5g)

75.5% bread flour
11% rye flour
11% sprouted hard red spring wheat flour
2.5% all purpose flour
78% water
1.9% salt

Bulk Fermentation Process

The three doughs began fermenting around 7pm at kitchen temps of 70°F. A couple of hours later, I put them outside where temperatures overnight ranged from 58-64°F. I did this because I knew they would over-ferment if I left them inside, but my refrigerator didn’t have three spots with identical temperatures.

I ended the bulk fermentations as follows:

Sprouted starter dough after 13.5 hours.

Rye starter dough after 14 hours.

Bread flour starter dough after 16 hours.

Bulk Fermentation Observations

The sprouted and rye starters fermented their respective doughs faster than the bread flour starter. This is especially interesting given how little the sprouted starter had expanded in its jar compared with the other two starters. This tells us that the microbial population size can’t be equated with the degree of starter expansion when comparing starters of different flour types. In other words, your whole wheat starter may be more powerful when doubled than your bread flour starter is when doubled.

The state of the sprouted starter, almost flattened and no longer growing, could have played a role in its fermentation speed rather than its flour composition. However, the rye starter dough was in close second for fermentation speed, and its starter was domed and possibly still expanding when I mixed the dough. So, I believe it’s more likely that the whole grain nature of the two flour types, not slight differences in the stage when they were used, is the more likely cause of the speed with which they fermented dough.

The sprouted starter dough container was different than the other two containers, and it is possible that this impacted fermentation speed by influencing the speed of temperature shifts (i.e. the dough in it could have cooled down slower when I put it outside). But the rye dough container was the same as the bread flour dough container, so glass thickness is unlikely the cause of fermentation speed differences.

I believe I introduced a new, unintended variable at this stage of bulk fermentation by slightly under-fermenting the bread flour dough.

Pre-Shape, Bench Rest, and Final Shaping

Methods

With no active gluten development, these 78% hydration doughs were very difficult to handle. I used a bench knife, extra flour, and I pre-shaped aggressively. The doughs rested uncovered for 20 minutes. I then shaped and reshaped the doughs until they stopped splaying outward the moment I took my hands away (two rounds of shaping for the sprouted dough, and three for the rye and bread flour doughs).

In addition to shaping it less aggressively, I put the sprouted dough into a much wider banneton (6 3/4″ vs 5 1/2″ wide), so I disqualified it from shape and oven spring evaluations. (I also proofed it differently, which you’ll read below).

Observations

The bread flour starter dough, which was the final one to be shaped, was slightly easier to handle than the other two, mostly likely because it was a smidge less fermented than the other two doughs. You can see in the photo below how lamentably differently they were shaped.

Final Proof and Baking Methods

The two doughs that continued to be part of this experiment (bread flour and rye starters) proofed for one hour at room temperature, followed by 20 minutes in the freezer to stiffen up. (The sprouted starter bread came out quite flat and inspired me to add the 20 minute freezer stage to the subsequent two doughs.)

The loaves were baked for 35 minutes in a cast iron challenger bread pan. The oven and pan were heated to 500°F. After I loaded each dough, I dropped the oven temp to 475°F and baked for 20 minutes. Then I lowered the temp to 450°F, removed the lid and placed the pan base on it (to prevent the bread base from burning), and baked the bread an additional 15 minutes.

Baked Bread Observations

I found it difficult to make judgments on any aspect of these breads except flavor because of the unintended shaping variables that came into play. Nonetheless, the loveliness of the rye starter loaf does stand out in terms of height and wild crumb.

From a flavor standpoint, the bread flour starter bread was more sour than the other two breads, according to three of four taste testers. The bread flour starter bread process, from mixing to oven, lasted longer than the other two breads: 17.5 hours compared with 15 hours for sprouted and 16 hours for rye. This extra time in an overall long process doesn’t seem very big, but it could be relevant in that perceived sourness. I think it’s possible that both the speed of fermentation and the starter composition impacted the sourness of the bread.

Two of four taste testers thought the sprouted starter bread was the sweetest, while the other two taste testers picked the rye and the bread flour breads as sweetest. That’s probably not statistically significant, but since I am one of the taste testers who thought it was sweetest, I’ll hazard a guess that the increased enzymatic activity of the sprouted starter extends into the dough, releasing more sugars.

If I were to repeat this experiment, I would maintain the starters with their different flours for about four feedings and maybe two weeks time. I would go for a similar hydration feel to the starters, meaning more water with rye flour. I would use a similar dough formula except for hydration, dropping it down to may be 73% so that shaping is not challenging. Finally, I’d only compare two starters/doughs, or I’d buy a third identical dough container and banneton.

For my general sourdough baking, this experiment has made me consider the speedy fermentation benefits of mixing up a jar of flour for starter feeding that is 50:50 bread flour and whole rye flour.