Total Yeast and Mold Is Not an Organism

Why a Failed Harvest Needs an Investigation, Not a Remediation

By Tess Eidem, PhD · Cofounder, Intrepid Scientific · 2026-07-06

Your certificate of analysis (COA) comes back failing for micro contamination. It reads total yeast and mold, and it's over the limit. Or bile-tolerant gram-negative bacteria, over the limit. Or the plate is simply marked TNTC: too numerous to count. The harvest is on hold, the batch value is at risk, and the pressure is to do something. Bleach the room. Run an ozone generator. Send the flower out for remediation. Retest.

Here is the problem with all of that: you don't yet know what you're fighting. And you can't fix a source you haven't found.

The bucket is not the organism

"Total yeast and mold" (TYM) is not a single microbe. Neither is "bile-tolerant Gram-negative" bacteria (BTGN), or "total aerobic count" (TAC). These are regulatory screening categories. They are deliberately broad nets that a compliance lab casts to answer a single yes/no question: is the viable microbial load on this product over the action limit? They were never designed to tell you which organism(s) you're failing for, and they don't.

The categories span staggering diversity. "Yeast and mold" covers the entire fungal kingdom that can be cultured. That includes fungal species in the genera Aspergillus, Penicillium, Cladosporium, Fusarium, Botrytis, Cryptococcus, and Histoplasma, as well as yeasts and hundreds of thousands of other species. BTGN bacteria also represent a slice of the bacterial world, some of it environmental and harmless, some of it not. A single CFU (colony-forming unit) number rolls all of that into one figure. A TNTC result rolls it into no quantified number at all.

So a failing TYM count tells you the equivalent of "there are more cars on the highway than the limit allows." It does not tell you the make, the model, or the on-ramp they used. Yet almost every contamination response we're called into starts by treating the whole highway. Scrubbing every surface and spraying every room is done as if it were the same thing as finding the source of contamination. It isn't. Your failure is almost never the whole fungal kingdom or every BTGN species. It is usually a single or few specific culprits, and they left evidence.

A count is not a diagnosis

The industry uses simple culture-based tools as a broad standard compliance test. Culture-based plating grows whatever will grow on the medium and counts colonies, each representing a single microbial cell or spore. It is the right tool for a pass/fail gate. It is the wrong tool for a diagnosis, because it goes no further than the viable number.

The diagnostic step the response skips is identification. This means resolving the failure to the organism actually on the product:

qPCR speciation can confirm or rule out specific targets quickly. For cannabis, this most often means the four tested Aspergillus species (A. flavus, A. fumigatus, A. niger, A. terreus) that many states regulate.
DNA sequencing of the isolate goes further: sequencing the fungal ITS region or the bacterial 16S rRNA gene reads the organism's identity directly, so you learn the genus and species rather than the broad category. This is the difference between "TYM, TNTC" and "Penicillium citrinum is the dominant species on the final product."

That single change is a shift from how much to what. It is what makes everything downstream possible. Every organism has its own ecology: a preferred substrate and energy source, a moisture and temperature niche, a way it travels. Once you know the species, you get a good idea of where in a cultivation and processing environment it wants to live, and that tells you where to look. A CFU count alone cannot point anywhere, which is exactly why the common response is to inefficiently target everywhere.

Remediation is a treadmill, not a fix

When the source is unknown, remediation looks like the safe move. It rarely is, for four reasons.

1) It's expensive and it recurs. Remediation tolling runs on the order of $50-100 per pound, every time you fail. Because it only treats the flower in the moment, rather than treating the reason the flower was contaminated in the first place, the next harvest from the same room, the same moms, the same airflow, the same amendments, tends to fail the same way. Remediation devices are thousands of dollars to rent, hundreds of thousands to buy, and they still don't stop the source. You are renting a symptom suppressor on a monthly plan rather than fixing the underlying problem.

2) It can change the product. Aggressive remediation steps don't specifically target the microbe alone. Depending on the method, they can affect the color, aroma, and other aspects of the product you're selling. A passing retest on a remediated batch is not the same thing as a clean process and a passing product to begin with.

3) Disclosure is tightening. Patient and consumer pushback against remediation is growing, with more initiatives and markets moving towards requiring labeling of remediated product. Indeed, consumers increasingly treat "remediated" as a negative. Remediation is quietly shifting from an undisclosed cover-up to a labeled attribute. Relying only on remediation technologies to pass may not be a long-term viable solution as the industry evolves.

4) It never answers the question. Even when remediation "works" to get you a passing COA, you end the cycle knowing exactly as much as you started: that you failed, not why. The organism is still entering your process and wreaking havoc on your product. You've paid and you still haven't found out why you're failing.

Bleaching and ozonating everything in your facility won't fix the problem, and neither will remediation. An investigation identifies the reason it was contaminated.

Treat contamination like an outbreak

Think of a recurring microbial failure as a small outbreak, and outbreaks are solved the same way whether the host is a person or a plant: identify the agent, find the reservoir, map the transmission route, and interrupt it.

That means walking the entire process the way an investigator walks a crime scene or an epidemiologist maps disease transmission. Go room by room, from moms and veg through flower, dry, cure, trim, and package, and back up the chain through raw materials, growing media, water, biological inputs, and beneficials. Every one of those is a candidate point of entry. When you overlay the identified organism onto that map, find the overlapping high-risk areas. A post-harvest spoilage mold that shows up after a specific process step is not a mystery about the whole facility; it's a question about that step.

As an illustration, consider a grow that keeps failing for TYM on its flower from one part of the building while an identical strain elsewhere passes clean. Identifying the primary isolate(s) that are on the failing product, then tracing that species against the walkthrough, treats them as what they are: one room with one entry point. The organism's ecology narrows the map; the walkthrough closes it. You stop guessing and start pointing.

The output of that work isn't a longer, haphazard sanitation or IPM checklist. It's a case report: the top suspect organism named, the most probable entry point identified, and a prioritized action list. That list covers what to monitor, what to sample and how often, and what root cause to actually fix. Critically, it also says what not to spend money on. You don't have to run around throwing everything at the wall to see what sticks, because you now have an idea of the source of your micro issues.

From root cause to a reliable process that stays clean

Finding the source of your micro issues, even just once, is worth a great deal. Building a process that keeps finding it, and preventing it is worth much more. That's where root-cause work hands off to preventive controls.

A traced contamination source becomes the anchor for an environmental monitoring program: settle plates, surface swabs, and water sampling placed where the investigation showed the organism travels, not scattered at random. Where it's warranted, it also anchors a written preventive controls program in the spirit of the Food Safety Modernization Act's 21 CFR Part 117: hazard analysis across the process, defined monitoring and verification, and corrective actions that trigger before product is at risk rather than after a lab tells you it already failed. This is the same logic that underwrites GMP everywhere: compliant by design beats remediated band-aids later. It applies to the one hazard that most reliably costs cultivators their harvests.

The result is a genuinely safer product for the people who consume it and a safer workplace for the people who work around the plant and the potentially harmful microorganisms around it every day. Both follow from the same basic knowledge: knowing what you're dealing with.

The short version

A failing COA is a symptom. TYM, BTGN, and TAC are categories, not organisms; a CFU count and a TNTC are alarms, not diagnoses. Bleaching, ozonating, and spraying everything in your facility, or remediating the failed product don't specifically target the organism(s) entering your process and interfering with your product's compliance. The move that ends the cycle is to identify the organism to its species, trace it back to where it enters, and close that door. Then develop a monitoring program to ensure that door stays closed. That's an investigation and prevention, not remediation.

This is exactly the method we use behind our productized Contamination Case File. It takes six hours from facility walkthrough to a written case report, with the isolate sequenced to species. It's the fastest way to stop guessing.

Written by Tess Eidem, PhD, PCQI. Cannabis microbiology and bioaerosol depth. Senior Research Associate, University of Colorado Boulder, Department of Civil, Environmental and Architectural Engineering. This article is general technical guidance, not a compliance determination or legal advice. Microbial action limits, permitted remediation methods, and disclosure requirements vary by state and change over time. Verify against your current state regulations and your testing laboratory's methods. © 2026 Intrepid Scientific, LLC.