IPM principles for cannabis: the system, not the spray

Most cannabis "IPM" content is a shopping list: buy these predatory mites, spray this when you see that. That is pest control, not integrated pest management (IPM). IPM is a six-step decision loop: sample, identify, learn the pest's biology, set a threshold, choose a tactic, evaluate ^[1]. Only one of those six steps is about killing anything, and it is usually the last one you reach for, not the first.

IPM is a six-step decision loop, not a spray schedule

The standard framework, used across horticulture and adapted for cannabis facilities, runs in a fixed order ^[1][2]:

1. Sample. Scout on a schedule, not when something already looks wrong.
2. Identify. Confirm the actual species. "It's a mite" is not identification; Tetranychus urticae or Phytoseiulus persimilis is.
3. Learn the biology. How fast does it reproduce at your room's temperature, and does it have a life stage that hides from a spray.
4. Set a threshold. Decide, in advance, the count that means act.
5. Choose a tactic. Cultural first, then biological, then mechanical, then chemical, in that order (more on why, below).
6. Evaluate. Did the count fall, and did anything else break in the process.

Notice what is missing from that numbered list: exclusion. Keeping pests out in the first place does not appear as a step, because the framework files it under "choose a tactic," one option among many. In practice it deserves top billing. Exclusion and sanitation run continuously, all cycle, and they are the one part of the system that is invisible right up until the week it fails. The rest of this guide treats it that way: not as step five's first option, but as the work that decides how often you need the other five steps at all. The dedicated article on exclusion and sanitation covers the physical detail; this one covers why the whole system is shaped the way it is.

Why cannabis breaks the vegetable-crop IPM playbook

Generic IPM advice assumes a few things that are not true of cannabis, and each one changes what "good IPM" looks like in practice.

There is almost no pesticide residue tolerance on the harvested product. Flower is smoked, vaporised or otherwise consumed close to how it was grown, with no cooking or washing step to strip residues the way a vegetable crop gets one. As a direct consequence, very few products are registered for cannabis anywhere, and the ones that exist are tightly restricted the closer the plant gets to harvest: "concerns over possible fungicide residue carry-over in the inflorescence tissues used for medical or recreational purposes has limited the registration of synthetic fungicides" for cannabis specifically ^[3]. A lettuce grower can spray, wait out a pre-harvest interval measured in days, and wash the leaf. A cannabis grower is usually choosing between a much shorter list of options and a hard stop once flowers set.

The canopy is dense and stays that way. Cannabis is typically grown as an unbroken, closely spaced mass of flowering material, and denser inflorescence structure is directly linked to more disease: cultivars that produce large, tightly packed inflorescences develop more Botrytis bud rot than ones with smaller, loosely arranged flowers ^[3]. The same dense, still-air microclimate that favours Botrytis also favours spider mites and powdery mildew: warm, humid pockets deep in the canopy that a fan and a glance at the room thermometer never reach.

Cropping is continuous, with overlapping generations. A field vegetable crop gets a fallow break between plantings that resets the pest population. A tent or a commercial room rarely does: clones go in as the previous batch comes out, so a pest that survives one cycle walks straight into the next one's seedlings, and successive, overlapping generations build resistance to a repeated tactic faster than a seasonal crop would ^[1].

Sprays are effectively banned deep into flower. Between the residue problem above and most jurisdictions' testing regimes, almost nothing is legal, labelled and sensible to apply once flowers have set. That pushes the whole programme's weight earlier: by the time you are three weeks into flower, prevention and monitoring are close to the only tools left.

None of this means cannabis needs a different six steps. It means steps one and five (sample, and choose cultural and biological tactics) have to do more work, because step five's chemical option is off the table for most of the crop's life.

Most of the budget belongs in prevention and monitoring

As a rule of thumb, roughly 80% of the effort and cost in a working IPM programme sits in the bottom two bands of that pyramid, prevention and monitoring, rather than in reacting once something is found. Both sources describe prevention and monitoring as the foundation the rest of the system is built on ^[1][2], but neither gives a percentage split, on cannabis or any other crop, so treat 80% as a grower- and extension-practice heuristic, not a measured number.

A worked comparison makes the ratio concrete.

The point of the pyramid is not that prevention and monitoring are free. It is that they are cheap relative to the alternative, and cheapest of all compared with what a single missed week can cost once a pest population is past its first few generations.

Setting the action threshold: when a count becomes a decision

A threshold is not a feeling that something looks bad. It is a specific number, at a specific location, decided before you find anything, so the decision to act is made in advance rather than in the moment.

The underlying idea comes from the economic-threshold concept in agricultural entomology: the economic injury level is the pest density at which damage costs more than treating it would, and the economic threshold (or action threshold) is set below that, at the density where you need to act now to avoid ever reaching the injury level ^[1]. On a food crop, the injury level is set by yield loss and the cost of that particular product at that particular market price. Cannabis changes the calculation in two ways: there is very little tolerance for damaged, contaminated or unsellable flower once it is close to harvest, and, as above, your intervention options shrink the further into flower you go. Practically, that means cannabis thresholds usually sit tighter (you act sooner) than the equivalent threshold on a crop you can still treat right up to harvest.

A commonly used starting point for a threshold, on spider mites at vegetative or early-flower stage, looks like this: roughly 2 mites per leaflet on 10% of scouted leaves. Treat that as a grower- and extension-practice starting figure carried over from other greenhouse crops, not a cannabis-specific trial result, and verify it against a current source before you build a programme around it; the dedicated thresholds article works through pest-by-pest starting numbers and how they should tighten as flower progresses. What matters here is the shape of the decision: below the number, you watch and recount in 48 to 72 hours; at or above it, you act, and you already know what "act" means because you decided the tactic before you needed it.

Four tools, reached for in order

An IPM programme reaches for tactics in a fixed order, cheapest and least disruptive first ^[1][2]:

1. Cultural. Change the growing conditions so the pest or pathogen struggles, not the plant. On cannabis: spacing and defoliating the canopy for airflow, so the still, humid microclimate that favours mites and Botrytis never forms in the first place, and keeping the media surface from staying wet enough for fungus gnats to breed.
2. Biological. Introduce or conserve something that eats or out-competes the pest. On cannabis: releasing a predatory mite such as Phytoseiulus persimilis onto a confirmed spider mite hot spot, either as a one-off curative release or on a preventive schedule before mites are found at all.
3. Mechanical and physical. Trap, exclude or physically remove. On cannabis: yellow sticky cards used for mass-trapping fungus gnats and whitefly as well as for monitoring, and simply removing a single heavily infested leaf before it becomes a source for the rest of the plant.
4. Chemical. A product, applied by the label, for the jurisdiction and the crop stage. On cannabis: an insecticidal soap or horticultural oil registered or permitted for cannabis where you grow, used early enough in the cycle that it clears before flowers set, and only for a pest a cultural or biological tactic has already failed to hold below threshold.

The order matters because each tactic changes the cost of the next one. Sanitation and airflow (cultural) reduce how often you need a predatory mite (biological). A healthy predator population reduces how often you need a spray (chemical). Reach for chemical control first and you skip straight past the tools that would have made it unnecessary.

Growers sometimes frame tactics three and four as a rivalry: "beneficials or sprays." It is a scheduling problem, not a choice. A broad-spectrum insecticide or miticide that kills a target pest will usually also kill the predatory mites, parasitoid wasps or other beneficials you have released or are relying on, because most of them are more sensitive to a given product than their prey is. Applied on the wrong day, a spray that was meant to help can wipe out weeks of established biological control and hand the room back to the pest it was meant to protect.

The practical fix is a calendar, not an either/or rule. Check a current compatibility source before combining any spray with an active release, and build in a gap, product-dependent, between a broad-spectrum application and the next beneficial release. Rotating which mode of action you reach for, when a spray is genuinely needed, also matters beyond compatibility: repeated use of one mode of action is exactly the selection pressure that breeds resistance, in the pest you are treating and, over enough cycles, in a beneficial population's ability to establish afterwards ^[1]. The rotation and resistance guide covers mode-of-action rotation for commercial programmes in more depth.

A home tent and a licensed room run the same loop on different numbers

The six steps do not change between a spare-room tent and a licensed facility. What changes is risk tolerance, the threshold numbers, the record-keeping obligation, and which tactics are even on the table.

Risk tolerance. With one to nine plants, a single infested plant can be a large share of total canopy, so a home threshold should generally sit tighter than a commercial room averaging counts across dozens or hundreds of plants. Losing one of four plants is a 25% loss; losing one of four hundred is not.

Record-keeping. A home grower can keep a single running notebook line a week. A licensed facility usually cannot choose to skip this. Malta's regulator, ARUC, has required licensed cultivation associations to log temperature and humidity daily since its Directive 1 standard took effect in July 2024 ^[4]; a January 2026 amendment went further, adding a dedicated log of pesticide types, quantities and pest-monitoring activity as its own compliance requirement, separate from the environmental log ^[5].

Available tactics. Regulator rules can remove options outright rather than just discouraging them. Under ARUC's cultivation standard, licensed cannabis associations in Malta are barred from using inorganic chemical pesticides and herbicides at all ^[4], which pushes a licensed Maltese programme further towards cultural and biological control by regulation, not by choice. A home grower elsewhere might have a legal pesticide available that a licensed facility down the same street cannot touch.

This guide stops at the system. It never diagnoses a specific pest or disease: for identification and species-specific control, the pests and diseases sections carry that detail, and each profile in this section links back to the pest or disease it targets.

What reactive-only spraying costs

A pattern IPM advisors see repeatedly, in cannabis and in other protected crops, runs like this. A grower skips scheduled scouting because nothing has looked wrong for weeks, then finds visible stippling and fine webbing on the upper leaves of several plants at once: a spider mite population already several generations deep, well past the point a threshold decision would have caught it early. Under pressure, they reach straight for a broad-spectrum contact insecticide, the fastest thing they have, and knock the visible population down within days.

Two weeks later the mites are back, faster and worse than before. The broad-spectrum spray killed the pest mites, but it also killed the small resident population of predatory mites that had been holding the canopy in check without anyone noticing they were there, and it did nothing to the spider mite eggs tucked into leaf folds that were never exposed to the spray. With no predators left and a fresh generation hatching into an empty canopy, the population rebounds unchecked, and by the time it is caught again the room is close to flower, when spraying is barely an option at all. The room is lost, not because nobody sprayed, but because spraying was the first move instead of the last one, and it took out the one thing that had been doing the job for free.

The fix was never a better spray. It was scouting on a schedule that would have caught the mites at 2 per leaflet instead of visible webbing, a threshold decision made before the outbreak instead of during it, and a first tactic that did not remove the room's own biological control on its way to solving one problem.

Building your own plan

The six-step loop is the same wherever you grow; the numbers, the record-keeping and the tools available are not. Start with whichever plan matches your scale: the home tent IPM plan sizes the whole system down to ten to fifteen minutes a week, and the licensed facility IPM plan sets out the SOP skeleton a regulator will expect to see.