Starter "cycles"?
My starter often seems to work in cycles. It will be happy and active and bubbly and wonderful, making loaves that puff up like nothing you've ever seen out of whole wheat, then after a number of months it will falter. I'll get a month or more of half-hearted bubbles and mediocre rises paired with an increased tang. After a little while, it will pick up again. I don't think this is normal, but I'm not sure if anyone else has had this particular problem. I want my starter to be consistent, and I want to understand what is causing the fluctuation. I keep a fairly liquid starter (I believe it is 100% hydration) and keep it in the fridge. I typically feed it every 3 days, sometimes more when I bake more often during the week. I'm not sure if this is related, but I've noticed that often during this time, the hooch does not rise to the top, but settles in the bubbles in the starter or at the bottom of the jar. That makes it difficult or impossible to pour out. Any advice?
I think there are a number of potential factors that influence your situation, but we need more data to do even a speculative diagnosis.
How long is the starter out before you refresh it? And at what temperature would it be when you refresh?
What is your refresh ratio?
How long do you leave it out before you again refrigerate it?
What temperature is your refrigerator?
Ok, here are some specs on my starter:
I keep my starter in the door of the refrigerator, so it will not get too cold. I keep my starter in the refrigerator all the time, refreshing it twice a week. On wednesday and sunday I feed the starter I have not used a 1:1:heaping 1 ratio. If I have one cup starter left, I add one cup water (cool) and one heaping cup flour (whole wheat) to make a runny oatmeal consistency. I put it right back in the fridge after feeding. I like my starter very mild, so that is why I keep it strictly in the fridge. Any clues?
If you want a less sour bread, put a larger fraction of the flour into the preferment, use white not whole grain flour containing a small amount of malted barley flour (sometimes it is listed as "enzymes" in the ingredients list) then ferment and proof warm (25-27°C). The extreme end of this is when all of your liquid is in the starter and you just add white flour and salt to make your final dough, then ferment a the temperature a few degrees above the point that corresponds to the minimum ratio of LAB to yeast growth rates (20-25°C) which is close to the point of maximum CO2 production/acid production. Again Ganzle's model confirms the practice and provides quantitative justification.
First, the following is based on a starter growth model developed and validated by Michael Ganzle. Second, since Ganzle's model does not account for any lag phase growth characteristics, I use a rule of thumb that the delay before the beginning of the log phase is one generation time [ln(2)/growth rate]. And while your starter may be slightly different, the general observations should be close since at all temperatures the lactobacillus has a higher growth rate than the yeast.
At 4°C (39.2°F) which is a reasonable temperature for a refrigerator door, the doubling (generation) time for the LAB is about 39 hrs and for the yeast it is about 120 hours. Your 3 days between refresh cycles is only 72 hrs so the yeast never gets out of lag phase and the population declines by a factor of .3 (or so) every time you refresh. With initial population densities of 10^5 per gram, and a reduction to 30% of that each time you refresh, you lose a factor of 10 in population density every 2 weeks.
The LAB will grow more quickly, but will not quite fully triple in three days if you are at 4°C and we allow one doubling time for lag phase. Your description of a sour starter is indicative of excess acid so again we have indirect evidence that your refrigerator door is perhaps slightly warmer than 4°C. But allowing one doubling time (39 hrs@ 4°C) for lag phase is a relatively large fraction of the 72 hrs total refresh cycle so we are sensitive to the unvalidated approximation.
This leads to a result that has the LAB maintaining it's population reasonably well until the acid in the jar drives the post-refresh pH below about 4.3 at which point the LAB growth rate slows down.
So we have two data points that point to perhaps a warmer refrigerator door than 4°C.
The model predicts that under these conditions, the starter would lose yeast population monotonically from the beginning, but the LAB population would remain stable for while before beginning to decline.
Depending on how you elaborate the refrigerated starter up to the quantity you need to bake with, there may be opportunities for everything to move back toward a more normal population density and ratio of LAB to yeast (normally about 100:1). If you allow the starter to return to room temperature after feeding, then refresh/feed a few cycles at room temperature before using it to bake, you are probably doing enough to keep it from dying sooner.
So what should you be doing (given that you want a mild starter)?
Probably the best solution would be to feed at room temperature in a ratio of 1:10:10 for a few cycles, waiting for it to fully peak (but not fall back completely) before refreshing again. This high refresh ratio will dilute the LAB-produced acid and assure that the post-refresh pH is high enough to not inhibit LAB growth. Letting the starter grow until it just begins to fall back assures that you have maximum yeast population before you refresh (remember that the LAB grows faster than the yeast and has stopped replicating at a pH of 3.8, well before the yeast population peaks even though it continues to produce acid down to a pH of ~3.6 or lower).
When you want to refrigerate the starter, take some of the starter that has been refreshed at room temperature, feed it in a ratio of 1:1:1 at room temperature, let it sit for an hour, and then refrigerate it. If your refrigerator door is really at 5°C (41°F) [ which is consistent with what you report] a 1:2:2 refresh that has already transitioned to log phase growth will fully refresh the LAB (5x - at which point LAB metabolism will shut down because the pH drops below 3.6) in 3 days, but will not fully refresh the yeast which will grow to only about 1.7x the original population (and you need 3x to be fully refreshed). At this point you need to either let it run a whole week (168 hr) before bringing it up to room temperature and refreshing at least one cycle at 1:10:10 before repeating the feed-wait-refrigerate cycle. The LAB will be refreshed after 2 days (about 48 hrs) and will go dormant. The yeast will not be fully refreshed until day 6 and will not suffer substantial population loss in one day of dormancy. One or preferably two room temperature 1:10:10 refresh cycles before baking should renormalize just about everything and set you up for success.
So this has been a roundabout way of suggesting that you change your refresh cycle to get on a schedule that keeps both the LAB and the yeast at robust levels.
And again a caveat: Ganzle's model is just that, and your specific culture is likely to be different. But the processes at work are all the same, just the times and temperatures and relative growth rates will slightly different. Just take this discussion as guidance and not a prescription.
[edited to repair grammar, spelling, and clarity]
Wow, thank you so much! I admit I had to read it 3 times before everything really started to sink in ;) but I think I get it now. I have some purely scientific questions for you though, cause this starter-science stuff fascinates me.
Does the LAB population produce all of the acid that is also responsible for inhibiting its growth once the PH falls below 3.6?
Does a balanced yeast/LAB population keep the PH in check?
Does a strong yeast population inhibit or compensate for the production of this acid in a stable starter?
As a busy mother of young children, keeping my start on the countertop is not ideal, because I can't be sensitive to its feeding needs based on variable growth rate. What if I fed it once a week, and at that feeding used lukewarm water in a 1:2:2 ratio, left it out for an hour or so, and then refrigerated it. Would the warmth of the water and the hour at room temperature be enough to maintain the metabolism of the yeasts and the LAB population needed for proper growth during the refrigeration?
I use my starter in ALL my baking (pancakes, muffins, cakes, pizza, etc), not just one day a week. I really want to find a way to keep an active culture population without switching to countertop, because my countertop days were a total disaster.