Your Soil Mites Didn't Fail. Nine Things Did.

Soil predatory mites have an unusual failure profile. Most beneficials fail visibly — the ladybugs leave, the lacewings hatch and you find the shells, the sachet dries out and you can see that it dried out.

Stratiolaelaps scimitus fails invisibly, into an inch of dark peat, and the only symptom is that the gnats keep coming.

The useful thing is that the list of causes is short. Nine of them account for very nearly everything, each one leaves a different fingerprint, and working through them in order will almost always tell you which one you've got.

1. You released into an established population

The most common cause by a wide margin — and it's a mechanical limitation rather than a quality problem, not a bad bottle.

S. scimitus eats fungus gnat larvae, every instar, the youngest by preference. What it doesn't do is eat pupae — those aren't attacked at all — and it barely touches eggs, where predation was described as negligible in the foundational laboratory work.

So in a pot that's been producing gnats for a month, the population is spread across four larval instars, a pupal stage and a layer of eggs — and the predator can only reach part of that.

The fingerprint. Adult catch stays flat or rises for two to three weeks after release, then falls. That isn't failure — that's the pupal reservoir emptying.

The fix. Pair the mite with something that reaches the stages it can't. UConn's guidance is explicit that when fungus gnat populations are already established, S. scimitus should be used with Bti or with entomopathogenic nematodes rather than alone.

Kansas State goes further, and it's worth reading twice. "In most cases, using S. scimitus alone will not be effective in managing fungus gnat and western flower thrips populations."

2. You judged it at week one

Related, but worth separating — plenty of releases get abandoned before they ever had a chance.

Adult fungus gnats live roughly seven to ten days. The ones flying around your room on release day aren't affected by anything happening in the substrate — they'll finish their lives on their own schedule, in front of you, looking like proof that nothing is working.

Meanwhile the mites are eating larvae you can't see, in a stage that produces no visible signal at all — which is a hard thing to feel encouraged by.

The fix. Yellow sticky cards at substrate level, counted on a fixed schedule, and a note of the count before you release. Without a baseline you're comparing a memory to an impression.

3. You dosed by pot volume instead of surface area

Every source that's looked agrees on where this mite lives — the top half-inch to one inch of substrate. Kansas State says the top 2.5cm. UConn says the top half-inch. Cornell's scouting guidance sends you to the top inch at the base of the plant.

It isn't a deep-soil animal — and it doesn't become one just because you bought a bigger pot.

Which means the correct dose scales with the surface area of the substrate rather than its volume. A 20-inch container has roughly eleven times the surface of a 6-inch pot, so if you spread one bottle rated for ten plants across ten containers of wildly different sizes, the big ones got a fraction of the density they needed.

The fingerprint. The small pots cleared and the large ones didn't.

The fix. Dose the big containers as though they were several plants. And accept the real limitation — in a deep, coarse profile a meaningful part of the pest population is sitting below the working layer entirely — where nematodes reach and the mite never will.

4. The substrate dried out

Dry substrate doesn't kill this mite quickly — it makes it leave, which gets you the same result and a much more confusing diagnosis.

A 2022 study measured how far the mites moved and how many walked out of a test arena across a range of conditions. At 20–29% relative humidity, more than 85% left. At 80–89%, only 19% to 33% did. Dispersal in search of moisture is normal behavior for this species, and a pot with a dry crust on top is a pot they'll walk off.

The fingerprint. It worked for a few weeks and then stopped, and you can remember the week things got busy and the watering slipped.

Or you bottom-water exclusively, so the root ball is soaked and the top inch — the entire habitat — is bone dry, which is an excellent way to grow nothing at all.

The fix. Keep the surface inch damp rather than the whole pot wet. If you bottom-water for gnat control, mist the surface anyway.

It feels contradictory, and it is. You're maintaining a habitat for the predator in the same layer you were trying to make hostile to the pest — and the predator is the better long-term investment.

5. The substrate never dried out

The opposite error, and a worse one — it helps the pest at the same time as it hurts the predator.

S. scimitus doesn't survive standing water. It's a soil-surface animal working in air-filled pore space, so saturate that space and there's nowhere left for it to be.

Meanwhile fungus gnat larvae do best in wet media — optimal survival is reported around 52% moisture content — and in a controlled greenhouse comparison, constantly saturated medium produced significantly higher gnat populations than medium that was watered once and allowed to dry down.

So an overwatered pot is a habitat improvement scheme — for the thing you're trying to kill.

The fingerprint. A saucer that's never empty, a pot with no drainage, a dense peat-heavy mix that stays wet for a fortnight, or algae growing on the substrate surface.

The fix. A normal moist-then-dry-back cycle that never reaches bone dry at the surface. Empty the saucers, and fix the drainage before you buy more predators.

6. The medium was too cold

Substrate temperature, not room temperature — a pot on a cold windowsill, a tiled floor, or against an exterior wall in winter runs several degrees below the air around it.

A release into a cold room isn't a dead product. It's a paused one — and because the mites survive three to four weeks without prey and don't diapause, a paused population will often pick up again in spring if the substrate stayed moist through the winter.

The fix. Check the medium with a probe instead of guessing from the thermostat. Below 60°F, either move the plants or accept that the program is idling until it warms up.

7. Something in the pot was eating them

The one that catches out careful people — because it's caused by doing more rather than less.

Rove beetles (Dalotia coriaria) and S. scimitus are both sold for fungus gnats, so using both feels like thoroughness. The published work found intraguild predation running in both directions.

The commercial laelapid mites, S. scimitus included, attacked every larval stage of the beetle, taking more first instars than later ones. Beetle adults returned the favor by consuming all twenty eggs of the sister species offered to them, plus "almost all of the 2- to 5-day-old nymphs."

Kansas State's guidance on S. scimitus says plainly that the rove beetle preys on it and that simultaneous use should be avoided.

Springtails are the other case. Laelapid soil mites prey on springtails — the clearest direct evidence is for the sister commercial species Gaeolaelaps aculeifer, which kills and eats Folsomia candida. If you're running springtails as a cleanup crew in a bioactive setup, the two products are working against each other.

The fix. One soil predator per container. Nematodes are the standard partner instead — different depth, different mechanism, and extension guidance recommends the pairing once populations are established.

8. It isn't fungus gnats

Three impostors — each producing the same complaint, none responding to a standard soil mite release.

The thrips case deserves its own line, because it's the most oversold combination in biological control.

Soil-stage interception is real. Between 86% and 99% of a western flower thrips population pupates in the substrate — and soil agents reliably cut adult emergence by something like 45% to 58%.

Whether that translates into fewer thrips on the leaves is crop-dependent, and the published results point both ways. A cyclamen trial pairing canopy predators with a soil mite and Steinernema feltiae found the combination worked faster than either alone. The ornamentals trial above found it added nothing at all.

Anyone telling you it's settled hasn't read both.

9. It was never going to work on that pest

Some of the prey list on soil mite packaging isn't supported, and in two cases it's actively contradicted.

• Varroa. A field trial applying S. scimitus to honey bee colonies at the suppliers' own recommended rates, by the suppliers' own method, concluded it "did not succeed in controlling varroa populations." It doesn't attack varroa attached to bees, and in the laboratory it fed on every unprotected honey bee brood stage, with a preference for the eggs.
• Root mealybugs. Not one peer-reviewed paper or extension publication tests this mite against Rhizoecus. Every trail leads back to supplier copy.
• Snake mites. The entire peer-reviewed basis is a 2015 paper titled "Preliminary Notes," covering two bearded dragons, with no controls and no quantification. Keepers report success and the mechanism is plausible — but that's not the same as evidence, and it's worth knowing which one you're leaning on.

The strange part is that the two best-evidenced targets are barely marketed at all. This mite completes its whole life cycle on root-knot nematode juveniles and cut root knots by about half in a pot trial, and it has real published data against poultry red mite.

Neither of those is on the front of anybody's packaging.

When none of the above applies

Sometimes the moisture was right, the timing was right, the temperature was right, nothing was eating them, and it still didn't do much.

That happens — and the specialists say so.

A 2022 review of predatory soil mites notes that although S. scimitus did reduce thrips densities on greenhouse cucumber, the researchers who ran those experiments "found it difficult to predict the efficacy of S. scimitus." That's an unusually candid line to survive into a review, and it's worth taking at face value rather than treating every disappointing release as user error.

What follows from it isn't despair, it's program design. This species is a good preventive layer with a real establishment advantage — it survives weeks without prey, it doesn't diapause, and one application can hold a population for a season.

It's a poor single answer to an active problem. Build it into the substrate early, put something fast alongside it when there's pressure, and measure on a sticky card rather than by feel.

Common questions

How long should I wait before deciding it didn’t work?

Weeks three to four is the first honest read. Week one tells you nothing at all — the adults flying around on release day live seven to ten days and will finish their lives on their own schedule, in front of you, whatever is happening in the substrate. Week two is still not a verdict. A clear downward trend at weeks three to four means it worked.

My sticky card count went up after I released. Did I get a bad bottle?

Almost certainly not, and this is the single most common reason a release gets abandoned. S. scimitus does not attack pupae at all, and barely touches eggs. So every pupa in the pot on release day still emerges on schedule, and every egg already laid still hatches. A count that stays flat or rises for two to three weeks and then falls is the pupal reservoir emptying, which is what success looks like from the outside.

I used one bottle across ten pots and only the small ones cleared. Why?

Because the dose scales with the surface area of the substrate, not its volume. This mite works in the top half-inch to one inch and does not become a deep-soil animal because you bought a bigger pot. A 20-inch container has roughly eleven times the surface of a 6-inch one, so an even split across mixed sizes leaves the large pots at a fraction of the density they needed.

Can I run predatory mites and rove beetles together for better coverage?

No — they eat each other. The published work found predation running in both directions: commercial laelapid mites attacked every larval stage of the beetle, and beetle adults consumed whole clutches of mite eggs and nearly all the young nymphs offered to them. Kansas State says plainly that simultaneous use should be avoided. One soil predator per container; nematodes are the standard partner instead.

Will they work on my springtails, or alongside them?

Alongside is the problem. Laelapid soil mites prey on springtails — the clearest direct evidence is for the sister commercial species Gaeolaelaps aculeifer feeding on Folsomia candida. If you are running springtails as a cleanup crew in a bioactive setup, the two products are working against each other.

References