The Mites Are Alive. The Thrips Are Winning.

Most biocontrol failures announce themselves. The ladybugs leave. The lacewings hatch and you find the shells. The sachet dries to a crisp and you can see that it dried out.

Cucumeris mostly doesn't do that. In the great majority of failed thrips programs the mites are still there — alive, walking around, findable on a leaf under a lens — and the thrips count is going up anyway. Which is genuinely confusing, because the one piece of evidence everyone knows how to collect says everything is fine.

It's also the way in. The mites being alive is not an absence of information; it's a diagnostic. What they're doing — whether there are eggs, whether there are juveniles, whether the white feeder mites in the sachet are still moving, whether the mites are in the flowers or only on the leaves — narrows eleven documented causes down to one or two, usually inside five minutes.

This is that narrowing, and then the eleven causes with what is actually measured behind each one.

Five minutes with a hand lens

Before any of the causes, four observations. They cost nothing and they eliminate most of the list.

The whole thing as one table

Eleven causes, and the column that separates them. Read across the row you match.

Cause	Alive on arrival?	On the crop at week 2–3?	Eggs / juveniles?	Feeder mites alive?	The tell
Dead on arrival	No	No	No	Often no	Ammonia smell; warm box
Humidity	Yes	Adults only, thinning	No	Yes	Warm and dry; any season
Diapause	Yes	Adults only, thinning	No	Yes	Short days and cool nights
Heat	Yes	Few	Few	No — dead	Sachet quiet by week 2–3
Cold	Yes	Few, sluggish	Few	Yes	Sachet quiet, mites intact
Pesticide	Yes	No — dead fast	No	Often yes	A spray inside the residual window
Prey stage	Yes	Yes, thriving	Yes	Yes	Larvae present are the large ones
Released too late	Yes	Yes	Yes	Yes	Counts were rising before release
Wrong pest	Yes	Yes, thriving	Yes	Yes	It isn't WFT under the scope
Eaten by an ally	Yes	Low despite good emergence	Variable	Yes	Open bran; or absent from flowers
Crop in flower	Yes	High and rising	Yes	Yes	More mites, less control
Never got there	Yes	Patchy	Yes near release	Yes	Spaced pots; plants moved

1 · They were dead on arrival

First because it settles the question fastest and because it is the only cause where the crop never had mites at all.

There is no published study measuring how often commercial predatory mite shipments arrive short, and that absence is worth saying out loud — there is no pharmacopoeia for live invertebrates. What is measured is the shipping itself. A USDA engineering study puts the appropriate transit temperature for beneficial insects and mites at about 50 °F (10 °C), with an acceptable range of 41–59 °F (5–15 °C), and then measured a real fifteen-hour air shipment averaging around 69.8 °F (21 °C) — with insulated containers doing their job.

An uninsulated box in a summer van is outside that window for the entire journey.

• Ammonia on opening is the classic tell and it means the material has gone anaerobic.
• No tan mites moving at 10–15× magnification. Extension guidance is specific about the magnification because the animal is half a millimetre.
• No white feeder mites either. That means the whole colony died rather than just the predators — a transit problem, not a storage one.

Storage after arrival is its own small failure mode. Above 53.6 °F (12 °C) for cucumeris, above 60% humidity, and not in a refrigerator — a domestic fridge is both too cold and far too dry, and office air at 40–50% humidity kills the moulds the feeder mites depend on. If the box is going to sit for more than a day or two, open it, because carbon dioxide builds up in a sealed carton.

2 · Adults but no juveniles — humidity

This one has a genuinely elegant answer buried in it, and the answer resolves a contradiction that has confused this subject for years.

Ask four institutions what humidity cucumeris needs and you get four numbers. Ontario says above 60%. UConn says above 65%, ideally 75%. Cornell says above 70%. The UK's levy body says 65–75% in the plant microclimate. None of them cites a primary measurement, and they can't all be right.

They are all right. They are the same number read at different temperatures.

The actual measurement isn't a humidity at all — it's a vapour pressure deficit, which is how much drier the air is than saturated air at that temperature. A 1995 study measured the point at which half of a species' eggs fail to hatch, and for cucumeris it came out at 1.11 kPa. That single constant, converted to relative humidity, is where all four extension numbers come from:

Air temperature	The humidity at which half the eggs fail
59 °F (15 °C)	35%
68 °F (20 °C)	52%
77 °F (25 °C)	65%
82.4 °F (28 °C)	71%
86 °F (30 °C)	74%
95 °F (35 °C)	80%

65% is the 77 °F (25 °C) figure. 70% is the 82.4 °F (28 °C) figure. 75% is the 86 °F (30 °C) figure. 60% is roughly the 71.6 °F (22 °C) figure. Four institutions picked the temperature that suited their crop and published the humidity that fell out of it. Nobody is wrong and nobody said which temperature they meant.

An independent lab confirmed the permissive end: at 68 °F (20 °C), cucumeris egg mortality across 60–82% humidity was 3%, and egg stage duration was unaffected across the whole range. Which is precisely what the vapour pressure figure predicts.

The other half of the diagnosis is that adults are far tougher than eggs. Adult survival stays above 90% until the deficit exceeds about 2.5 kPa — more than twice the level at which the eggs are already failing. Larvae are the weakest stage of all.

So dry air doesn't hand you dead mites. It hands you living adults that leave no next generation, and a population that thins out over three or four weeks while every mite you find looks perfectly healthy. That is a completely different picture from a kill, and it is why this cause gets misdiagnosed as a bad batch.

One honest caveat. In a well-run high-gutter greenhouse, leaf boundary layer deficits were measured at 0.1–0.7 kPa — never close to the 1.11 threshold — and had no significant effect on predation or oviposition. In that setting, humidity is probably not your problem and blaming it is a misdiagnosis of its own. In a dry room with an open canopy, it plainly can be.

3 · Adults but no juveniles — daylength

Identical symptom, completely different cause, and it only happens between roughly October and February.

Short days during juvenile development switch adult females into reproductive diapause. She survives; she just stops laying. The population then declines by pure attrition over three to five weeks while you watch healthy adults walk around.

The measured induction conditions are specific and useful. The critical day length is 12.45 hours. But temperature is a second switch, not a footnote: under a short-day regime with a 71.6 °F (22 °C) day, diapause incidence fell from 100% at a 59 °F (15 °C) night to zero at a 69.8 °F (21 °C) night. Cool nights are doing as much work as the calendar.

It is also stage-specific. Adults moved from permissive conditions into inductive ones kept reproducing. Once she is laying, she keeps laying — which means the risk lives in the rearing and shipping window, not on your windowsill.

The practical discriminator against humidity: supplemental lighting makes this one go away and does nothing for the other.

4 · The sachet cooked before the mites did

The mite and its packaging have different thermal limits, and the packaging fails first. Almost nobody makes this distinction and it explains a specific, common disappointment.

Cucumeris itself is more heat-tolerant than the trade tables suggest. Measured across nine constant temperatures, development runs from 59 to 95 °F (15–35 °C) with a lower threshold around 50.5–51.8 °F (10.3–11.0 °C) and the fastest development at 89.4–93 °F (31.9–33.9 °C) depending on diet.

The sachet gives up around 89.6 °F (32 °C). In a study measuring release across six temperatures, output was good at 17, 22 and 80.6 °F (27 °C) and collapsed at 12, 32 and 37 — and inside the pouch, numbers of both the predator and its feeder mites started falling within a week at 32 and 37. The colony starves before the predator overheats.

That work used a close relative rather than cucumeris, so read it as the mechanism rather than a precise threshold. But the mechanism is enough, because sachets on an exposed bench have been measured peaking at 101.7 °F (38.7 °C) in a real greenhouse.

The tell is the feeder mites. Cut a sachet open. If the white mites are gone, the colony cooked, and no amount of environment-fixing on the plant will bring that sachet back. This is the single cleanest discriminator on the entire list.

5 · Too cold to matter

Less interesting but worth ruling out. Below about 59 °F (15 °C) development effectively stops — no development was observed at all at 50 °F (10 °C). The mites are alive and slow, the sachet is quiet, and the thrips, whose own lower threshold sits below the mite's, carry on developing.

That asymmetry is the whole problem with cold. It doesn't pause the fight; it pauses one side of it.

6 · Something was sprayed

Fast, total, and indiscriminate — mites dead within days, including on the outside of the sachet, while the feeder mites sealed inside are often still fine because they were never exposed. If a spray went on inside the residual window, stop reading; this is your answer.

The part worth spelling out is that organic does not mean compatible.

Product	What was measured	Where it comes from
Spinosad	Over 90% mortality on direct exposure. Residues aged 2–24 hours repelled the mites. The authors recommend releasing six days after application	Peer-reviewed, 2011
Insecticidal soap	"Very harmful" to nymphs and adults on direct contact — but no residual activity once the spray has dried	UConn Extension
Neem / azadirachtin	No significant acute mortality at 60 ppm. But treated surfaces were avoided in choice tests and females laid fewer eggs after 48 hours	Peer-reviewed, 1996
Sulphur, oils, pyrethrin	No published data for this species. A real gap	—

Spinosad and insecticidal soap are both OMRI-listed. Neem is the outlier that is genuinely gentle on contact and still suppresses egg-laying.

The residual periods everyone quotes — Avid three to six weeks, Kontos and Pylon fourteen days, landscape oil two weeks — come from two suppliers' side-effect databases rather than from independent publication. They are the best list available and they are not peer-reviewed, which is worth knowing when they disagree with each other. They do: extension guidance gives spinosad two weeks where the published study says six days.

7 · The thrips are the wrong size

This is the cause where everything looks right. Mites present, breeding, feeder colony healthy, humidity fine — and the thrips count climbing anyway.

Cucumeris kills first-instar thrips larvae. Second instars are not merely less preferred; they physically win. When a predatory mite attacks one, the larva swings its abdomen and produces a droplet of rectal fluid, and the attack fails. In a 2024 study observing individual attacks, second-instar larvae of three different thrips species were never successfully attacked at all during the observation windows.

So the picture is not a predator declining to eat something. It is a predator attacking something and losing.

It gets worse in a way that matters for the population. Second instars and adults of two of those species were observed destroying predatory mite eggs. The larger thrips stages are not just out of reach — they impose losses on the predator.

There is a whole article's worth in this, and it is the next one in this series. For diagnosis, the fingerprint is enough: healthy mites, rising thrips, and the larvae you find on the leaves are mostly the larger, yellower second instars rather than the small translucent ones.

8 · It was released too late

Mechanically the same problem, arrived at a different way. Cucumeris takes about six first-instar larvae per day at best. A single western flower thrips female lays 150–300 eggs across a life of thirty to forty-five days, and a generation completes in one to two weeks in warm conditions.

Release into a population that is already established and most of what is present is in stages the predator cannot touch. It has no numerical response fast enough to overtake that, and the trials agree: Cornell says it should not be expected to give a quick knock-down of high-density pests; UConn says start preventively, at planting, before thrips are detected.

The fingerprint that separates this from cause 7 is the timing rather than the biology — sticky card counts were already climbing before you released, or you released because you saw damage. The same product at the same rate four to six weeks earlier would have worked.

9 · It isn't western flower thrips

Cucumeris is sold for thrips. The published evidence is overwhelmingly for one thrips, and for several of the others it is documented not to work.

Ontario's thrips identification guidebook also reports that on-farm Canadian trials indicate phytoseiids are not effective for Thrips parvispinus in ornamentals, and the UK levy body lists cucumeris as ineffective against Japanese flower thrips. Cucumeris also failed to suppress common blossom thrips in flowers while controlling melon thrips on the foliage of the same plants.

And it may not be thrips at all. Broad mite damage appears on new growth and growing points, looks like herbicide injury or a nutrient problem, and comes with no frass. Thrips damage is silvering with black specks on expanded tissue. The frass is the tiebreaker — broad mites don't leave any. The confusion runs both ways: Thrips parvispinus damage on tropicals is reported to resemble heavy broad mite damage.

10 · Something else you released is eating them

The failure caused by doing more rather than less.

Rove beetles. If you run Dalotia coriaria for fungus gnats and cucumeris in open bran or breeder piles in the same house, the beetles walk into the bran. Measured across two trials, sachets held six to eight times the proportion of cucumeris that open piles did, sachet populations grew 67–148% while open piles fell 92–100%.

Soil predatory mites. The same published work implicates Stratiolaelaps — which most growers release for fungus gnats and never think of as a competitor.

Orius. This is the one with the distinctive fingerprint. A predatory bug will eat cucumeris about as readily as it eats thrips, and in two years of sampling on greenhouse pepper, cucumeris was found mainly in flowers — except that in the presence of Orius, virtually no mites occurred in the flowers at all. Temperature and humidity gradients did not explain the distribution. Orius did.

Releasing Orius on top of cucumeris doesn't just add a predator. It evicts cucumeris from the flowers, which is where the thrips are.

Other phytoseiids. Swirskii is a documented intraguild predator of cucumeris juveniles, with both a high predation rate and a preference for cucumeris juveniles over thrips. Cucumeris eats swirskii juveniles back, at a lower rate — the interaction is asymmetric and not in cucumeris's favour. Worth noting for balance: a four-week greenhouse trial that included the mixture did not see it underperform, and no trial has ever compared a mixture against the better single species. The mechanism is demonstrated; the crop-level consequence is not. So this is a reason to expect a blend to end up dominated by the more aggressive species — not a reason to treat running two as a mistake.

11 · The crop came into flower

The cause where the standard diagnostic actively lies to you.

Cucumeris is an omnivore, and pollen is a better food for it than thrips are — measured development is faster and survival higher on pollen than on thrips larvae. So when a crop comes into bloom, or you add a food supplement to boost establishment, mite numbers go up.

And predation goes down. Pollen presence produced a 55% reduction in thrips predation by cucumeris in one study, and 43–60% reductions for phytoseiids generally in another. In a multiple-predator experiment, cucumeris showed no association with thrips at all — the majority of the mites were on the leaves where the pollen was.

More mites, fewer kills. If your monitoring question is "are there enough predators", this is the one cause it will answer wrongly.

Whether the net effect is bad is genuinely unsettled — supplemental food raises predator numbers even as it lowers per-mite predation, and for swirskii the net came out positive. It has not been shown either way for cucumeris.

12 · They never got there

Released from a central plant in a tray, most cucumeris were recovered at the release point. The related work is blunt: they do not disperse far, only about a quarter will walk down to ground level to leave a plant at all, and inter-plant contact greatly improves movement between plants.

The fingerprint is unmistakable once you look for it. Control is good within a plant of a release point and absent two plants away; mite counts are high on sampled leaves near the sachet and near zero elsewhere; on spaced pots, control stops at the gap.

What makes this one dangerous is that a sample taken near the sachet looks like complete success. The mites are alive, reproducing and doing their job — in one place.

Also on this list: moving plants around. A commercial trial specifically noted that movement of plants during production made effects hard to detect. Whatever distribution you achieved, rearranging the bench undoes it.

Common questions

My mites look fine but the thrips are getting worse. Is that possible?

It's the single most common shape of cucumeris failure. Most of the causes on this list leave the mites alive — only a pesticide kill and a dead-on-arrival shipment don't. The useful next question isn't whether the mites are alive; it's whether there are eggs and juveniles, whether the feeder mites in the sachet are alive, and whether the larvae on your leaves are the small translucent first instars or the larger second ones.

I see adult mites but no eggs. What does that mean?

A reproduction failure rather than a kill, and there are two candidates. Dry air stops the eggs hatching while leaving the adults healthy — adults tolerate roughly twice the dryness eggs do. Short days put the females into reproductive diapause. Split them on season and night temperature: diapause needs day length under about 12.5 hours and cool nights, and it goes away under supplemental lighting.

Is 65% humidity enough?

It depends entirely on the temperature, and this is where most diagnoses go wrong. The real measurement is a vapour pressure deficit, and converted to humidity it moves: about 52% at 68 °F (20 °C), 65% at 77 °F (25 °C), 74% at 86 °F (30 °C). So 65% is comfortable in a cool room and below the failure line in a warm one. A warm dry space is harder on the eggs than a cool dry one, even though it reads higher on the hygrometer.

I sprayed neem two weeks ago. Am I clear?

For acute mortality, almost certainly — neem showed no significant kill of cucumeris at 60 ppm in the published work. But the same study found treated surfaces were avoided and egg-laying dropped for 48 hours after contact, so the effect is on reproduction rather than survival. Spinosad is the one to be careful about: over 90% mortality on direct exposure, and the study that measured it recommends waiting six days.

Can I run cucumeris and swirskii together to cover more ground?

No, and the reason is specific: swirskii preys on cucumeris juveniles and prefers them over thrips when given the choice. You are not doubling your thrips control, you are feeding one predator to the other. Cucumeris and californicus are a different matter — they pursue different prey and are routinely run together.