How to track contamination rates and actually reduce them

Most growers describe contamination as a feeling. "I had a bad month." "The grain has been rough lately." That is enough to worry about and not enough to fix, because the useful question is never whether you have contamination — everyone does — but which specific step is leaking, and by how much. You get to that with counting, and counting takes less discipline than people expect.

Define the number before you collect it

A contamination rate is contaminated units divided by units started, over a period. The two words that cause all the trouble are unit and started.

Pick the unit that matches the decision you want to make. If you are trying to judge grain sterilization, the unit is the jar or spawn bag, not the run. If you are judging bulk substrate and your process fills five blocks from one mix, the unit is the block, but you also want to know which mix each block came from — five blocks failing out of one bad tub is one process failure, not five.

"Started" means everything you inoculated, including the ones you threw out on day three and would rather forget. A rate that quietly excludes the obvious failures is worse than no rate, because it will look fine while your throughput does not.

Record the stage, not just the fact

The single most valuable field is the stage at which the failure appeared. Roughly:

• Agar — plates showing bacteria, Trichoderma or cobweb before transfer.
• Grain, early — visible within about five days of inoculation, usually bacterial (sour smell, wet grey patches) or a fast green mould.
• Grain, late — appears once the jar is half colonized, often at the lid or injection port.
• Substrate, before pinning — the block never fully takes, green or black shows through the bag.
• Fruiting — the block produced but went over: cobweb on pins, bacterial blotch, or Trichoderma in the casing layer after flush one.

Each of those points at a different cause, and mixing them together produces one meaningless average. A 20% rate that is entirely early grain failure is a pressure cooker problem. A 20% rate that is entirely fruiting-stage cobweb is a humidity and air exchange problem. Same number, opposite fix.

The four breakdowns that pay for themselves

Once you have a stage on every failure, slice the rate four ways. You are looking for a group that is clearly worse than the others, not for statistical significance.

1. By stage. Where is the leak? Sort your last fifty units by stage of failure and the biggest bucket is your project for the month.
2. By recipe. Same organism, two substrate recipes, very different outcomes is common. Check hydration first: the majority of "mystery" substrate contamination is a mix that ran two or three points too wet.
3. By genetic line and culture. A line that has been transferred a dozen times colonizes slower, and slow colonization loses races it used to win. A line that suddenly gets worse is usually one specific culture that picked up a light bacterial load you cannot see on agar.
4. By batch and date. Group by sterilization batch. If failures cluster in one load, suspect the load: overfilled cooker, jars packed too tight, cold spot, a run where pressure dropped while you were doing something else.

Fixes that follow from each pattern

Early grain failure

Almost always sterilization or grain prep, not your inoculation technique. Check hydration of the grain first — grain that is too wet holds pockets that do not reach temperature, and endospores survive. Then check load size and time. If you moved from a small cooker to a bigger one and kept the same 90 minutes, your centre jars are undercooked. Add time for larger loads and let pressure come up before you start the clock.

Late grain failure at the port or lid

This is handling. Injection ports that have been used four times leak. Lids that have been autoclaved twenty times warp. Shaking a jar with a hairline crack in the filter patch introduces air you did not filter. Replace consumables on a schedule and log it, because the failure shows up two weeks after the cause.

Substrate failure before pinning

Usually spawn rate, hydration or the environment where you mix. Raising spawn rate from 10% to 15% buys speed, and speed is the whole game in bulk substrate: the block that colonizes in ten days beats the mould, the one that takes eighteen does not. If you mix in the same room where you break open finished blocks, you are seeding your own substrate with spores. Separate those two activities in space or in time.

Fruiting-stage failure

Cobweb and bacterial blotch are usually stagnant, saturated air rather than dirty technique. More air exchange, less standing water on the surface, and do not let a contaminated block sit in the fruiting chamber "just to see". Pull it out the day you see it, outside, and record it.

Make the recording take five seconds

All of this dies if logging a failure is a chore. Three rules keep it alive:

• Log at the bin, not at the desk. Whatever tool you use has to work with wet hands, standing up, on a phone.
• Label everything at inoculation. If a jar cannot tell you its line, generation, recipe and date, it cannot contribute to any breakdown. A QR label on the lid solves this permanently.
• Never fix and forget. Write down the change you made and the date. Your rate before and after that date is the only evidence that the change worked.

What good looks like

Numbers vary with species, scale and setup, so treat these as orientation rather than targets. A hobby grower doing still-air grain work often sits between 10% and 25% and can usually get under 10% with better sterilization discipline. Growers running a flow hood and clean grain routinely hold grain failures at a few percent. Bulk substrate on oyster mushrooms is more forgiving than it looks; on shiitake or lion's mane, small hydration errors show up fast.

The important comparison is not against other growers, it is against yourself last quarter. A rate you measure every week, broken down by stage and recipe, will drift downwards almost on its own, because you stop guessing which change helped.

A simple monthly review

1. Pull every unit started in the last month, with its stage of failure if it failed.
2. Compute the overall rate, then the rate by stage.
3. Take the largest bucket and slice it by recipe, line and sterilization batch.
4. Pick one change. Write it down with the date.
5. Review the same numbers next month.

One change a month is twelve process improvements a year, each with evidence behind it. That is how a 20% rate becomes a 5% rate without any single dramatic intervention.

This is exactly the bookkeeping our software is built around: mark a run or jar contaminated with the stage it turned, and the contamination rate report breaks down by line, recipe and stage for you. If you would rather do it in a spreadsheet, do it in a spreadsheet — just do it consistently enough that the number means something.