Two of These Four Columns Are Guesses.

There is a table on the internet. You've seen it. Four predatory mites down the top — cucumeris, swirskii, limonicus, andersoni — and rows for temperature range, humidity requirement, prey stages, feeding rate and best use case. Every supplier has a version. Ours included.

We went looking for the papers behind each cell.

What we found is that the table is not a synthesis of four comparable datasets. It is one reasonably well-measured species, one that has been thermally characterised and nothing else, one whose own reviewers state in print that no temperature demography has ever been published, and one with essentially no published thrips data at all.

Two of the four columns rest on supplier copy. One widely quoted feeding rate is roughly three times what the only peer-reviewed measurement found. And one row — the humidity row, for limonicus — is backwards against its own primary source.

None of which means the species are interchangeable. They aren't, and the real differences are clear and useful. They're just not the differences the table is about.

Temperature: two solid columns, two empty ones

Swirskii is the best-characterised of the four. A 2011 life-table study ran it at nine constant temperatures and produced the numbers a table row ought to be built from.

Swirskii, measured	Value
Lower developmental threshold	52.3 °F (11.3 °C)
Lower threshold for population growth	59.9 °F (15.5 °C)
Optimum for population growth	86.2 °F (30.1 °C)
Upper developmental threshold	99.3 °F (37.4 °C)
Development at 55.4 °F (13 °C)	None observed

The row that matters there is the second one, and it's the one no table carries. A population can develop at 52.3 °F (11.3 °C) and still not replace itself; the number below which swirskii is quietly shrinking is 59.9 °F (15.5 °C). A grower holding 55.4 °F (13 °C) has a population running down, not a slow one.

Cucumeris has two independent datasets. Across nine temperatures its lower threshold came out at 50.5–51.8 °F (10.3–11.0 °C) depending on diet, with development from 59 to 95 °F (15–35 °C) and the fastest rates at 89.4–93 °F (31.9–33.9 °C). Note the diet dependence — both the threshold and the degree-day requirement shifted with food, which is a caution against any single-number table entry for anything.

Limonicus: no published temperature demography at all

This is stated plainly in the most authoritative review of the species, whose four authors all worked for the company selling it: "Although no detailed studies on the influence of temperature on demographic parameters have been published, preliminary tests at Koppert Biological Systems have shown that A. limonicus … remains active at lower temperatures."

The supporting numbers in that review — 87% of eggs developing to adult at 55.4 °F (13 °C), taking 22.5 days, with females laying 0.8 eggs a day — are labelled as company data and have not been published. That is the entire basis for "limonicus works cooler than swirskii", and it is worth saying that it is a plausible claim from people with the equipment to test it. It is also not a citation.

The only published temperature series for the species is from 1965, on a citrus red mite diet: oviposition of 0.1 eggs a day at 50 °F (10 °C), 2.7 at 80.1 °F (26.7 °C), 1.0 at 90 °F (32.2 °C).

Andersoni: no threshold exists

A 2019 study measured andersoni at five temperatures on spider mite prey. It reports development times and reproduction, and it does not report a lower developmental threshold or a degree-day requirement — and we could not find those figures published anywhere else.

What the study does show is awkward for how the species is sold. At 62.6 °F (17 °C), andersoni's intrinsic rate of increase was 0.0003 — a population that is, for practical purposes, standing still. Its highest rate was at 95 °F (35 °C). Andersoni is marketed as the cold-tolerant broad-spectrum option, and the one temperature study available says it barely reproduces in the cold.

Humidity: the limonicus row is backwards

Here is the row we got wrong, along with most of the internet.

There is one good primary source. A 1995 study took eggs less than sixteen hours old, floated them over saturated salt solutions at 77 °F (25 °C) across a humidity range from 92% down to 32%, and measured the vapour pressure deficit at which half of them failed to hatch. Bigger number means tougher eggs.

Species	Critical deficit (kPa)	Equivalent humidity at 77 °F (25 °C)
A. hibisci	1.67	≈ 47%
I. degenerans	1.48	≈ 53%
N. cucumeris	1.11	≈ 65%
N. barkeri	0.94	≈ 70%
A. limonicus	0.88	≈ 72%

Of the seven species screened, limonicus eggs were the most humidity-demanding. The paper says so in as many words. Its authors rejected limonicus for Dutch winter greenhouses on exactly those grounds.

Three independent labs have since converged on the same place — critical deficits between 0.82 and 0.92 kPa, which is 71–74% humidity at 77 °F (25 °C). A 1965 study found only half of limonicus eggs hatched at 60% humidity and none at 50% or below. Its natural distribution in California is coastal, absent from the drier inland areas, and one field failure on nursery citrus was attributed partly to its humidity requirement.

So limonicus is not the tolerant one. It is the demanding one, and the difference between it and cucumeris is not small — roughly seven percentage points of humidity at the same temperature, in the direction nobody advertises.

And the whole row is temperature-dependent anyway, which no table admits. A critical deficit is a fixed physical quantity; the humidity it corresponds to moves. Cucumeris's 1.11 kPa is 52% at 68 °F (20 °C) and 74% at 86 °F (30 °C). Quoting a single percentage without a temperature beside it is the reason four institutions publish four different numbers for the same species.

Prey stages: everybody eats first instars

The row that sells the upgrade is the prey-stage row, and it does not survive the literature.

Every study we found agrees that phytoseiids take first-instar thrips larvae, take second instars poorly or not at all, barely touch adults, and do not reach pupae. The differences between species on first instars are real but modest:

Study	What was measured	Result
van Houten 1995	F. occidentalis first instars, 77 °F (25 °C), leaf disc	limonicus 6.9/day, cucumeris 6.0, degenerans 4.4, hibisci 3.5, barkeri 2.6
Arthurs 2009	Chilli thrips larvae, leaf disc	cucumeris and swirskii both ≈ 2.7/day — indistinguishable
Schoeller 2020	Chilli thrips, no-choice leaf disc	swirskii 4.6–6.3, limonicus 4.8–6.4 — no significant difference
Villamarin 2026	T. parvispinus first instars	degenerans 5.78, swirskii 4.14, cucumeris 3.23
Pijnakker 2026	T. parvispinus first instars	all four species 2–3.5/day

Those are not the differences a comparison table implies. In two of the five, the species are statistically indistinguishable.

As for second instars — the claim that swirskii and limonicus reach them where cucumeris doesn't — the canonical source is a 1995 pilot experiment that published no data, a 2024 behavioural study found swirskii unsuccessful against every second instar of three thrips species, and the one quantitative figure for limonicus is 0.60 second instars a day against 3.10 first instars, with first instars taken exclusively when both were offered. That argument has its own article, because it changed our mind.

The rows that are real

What happens when they're run against each other

This is the strongest evidence available and it is worth more than every table row above it.

Chilli thrips on pepper. Swirskii held thrips below one per terminal leaf. Cucumeris allowed up to 36. Untreated was 70. The detail that matters: in a leaf-disc assay in the same paper, both species ate about 2.7 larvae a day. The difference in the crop was entirely establishment and persistence, not appetite — which is the single most useful methodological warning in this whole subject. Leaf-disc predation rates do not predict field outcomes.

Ten species on cucumber. Limonicus was "clearly the best predator" of western flower thrips. Swirskii and E. ovalis also reached substantially higher populations than cucumeris. In related work, thrips stayed near zero for nine weeks under limonicus while rising to 130 per leaf under cucumeris.

Four species, cool short season, on pepper. This one is the awkward result. Under short photoperiod and cool temperatures — the exact conditions where the tables say cucumeris should be the pick — thrips suppression by cucumeris was the worst of the four species tested, in both trial years.

Seasonal comparison on chrysanthemum. Swirskii gave better control in summer, and in winter the two gave equivalent control with swirskii showing less damage. The authors then recommended cucumeris as "a more cost effective biological control agent for winter months."

The mechanism study. The most decision-useful of the lot measured functional and numerical responses for cucumeris and swirskii. Swirskii had the higher attack rate and less interference at high predator density; cucumeris had the shorter handling time and switched between prey stages as densities changed. Conclusion: swirskii is optimal at low thrips densities, cucumeris at relatively higher ones — which is close to the opposite of how they get sold. Swirskii's edge is preventive, not as a rescue.

Andersoni is in the wrong aisle

Of the four, this is the one where the gap between the marketing and the literature is not a matter of emphasis.

A 2025 study measured andersoni's prey consumption and preference across five temperatures. Against western flower thrips its maximum was 1.73 larvae per day, at the species' best temperature. Its preference for spider mites over thrips was strong at every temperature tested, and strongest when cold — the authors attributed the low thrips figures partly to the thrips' own antipredator behaviour.

The supplier product page states, with no citation, that under optimal conditions andersoni consumes four to five first-instar western flower thrips larvae per day.

That is roughly three times the only peer-reviewed measurement we could find.

We are not accusing anyone of bad faith — the number may come from an internal trial. But there is no published trial of any kind pitting andersoni against the other three for thrips control, no greenhouse or field efficacy trial against thrips at all, and in the one laboratory screen that included all four species, andersoni did not make the recommended list.

What it is genuinely good at

All of this is real, documented, and not what it is being sold for in the thrips aisle.

• Spider mites and eriophyid mites in perennial crops. This is its actual specialism — vineyards and orchards, where it is one of the standard generalists controlling tetranychid and eriophyoid mites.
• Conifer pests. Development from egg to female in 5.12 days on a spruce spider mite diet.
• Pesticide robustness, which is underrated. Field populations in Italian apple orchards were unaffected by pyrethroid applications, and the resistance has been mapped to specific mutations. The catch is that it is strain-dependent — one commercial strain was highly susceptible to deltamethrin while others were unaffected, and no supplier discloses which strain you are buying.
• Persistence on pollen. Airborne pollen availability drives its abundance in vineyards, and spraying pollen raises its numbers.

If you have spider mites or rust mites in a perennial planting and you spray, andersoni is a good answer. If you have thrips on a houseplant, the published evidence for it is one lab screen where it wasn't recommended.

What happens if you use two

The obvious response to all this is to buy two species and cover more ground. The evidence here divides unusually cleanly into what has been measured and what has only been assumed, so it is worth taking those separately.

Swirskii is an intraguild predator of cucumeris juveniles, with both a high predation rate and a preference for cucumeris juveniles over thrips. Both halves matter — this is not incidental predation, it's a choice. Cucumeris eats swirskii juveniles back at a lower rate, so the interaction is asymmetric and not in cucumeris's favour. Contrary to what theory would predict, eating each other was as good a diet as eating thrips for both species: there is no nutritional penalty discouraging them.

Andersoni is worse. In a four-species comparison it was the superior intraguild predator — the one most likely to eat the others — while also being the least efficient at converting what it ate into eggs. The same generalism that makes it a broad-spectrum product makes it the species most likely to eat whatever else you released.

Choosing, honestly

Stripping out everything that isn't measured, this is what's left.

If	Then	How solid
Warm and humid, thrips pressure	Swirskii or limonicus. Both establish better than cucumeris and reach higher populations, which is the difference that shows up in crops	Strong — multiple head-to-heads
Whitefly as well as thrips	Limonicus first, swirskii second. Cucumeris does not control whitefly	Strong
Cool greenhouse, winter production	Cucumeris, on cost. Not because it performs better cold — one cool-season trial ranked it worst of four	The cost claim is sourced. The performance claim isn't
Dry environment	Fix the humidity. Of the two species with published egg data, limonicus is the more demanding, not less — and there is no data at all for the other two	Strong for the ranking, absent for half the table
Broad or cyclamen mite with your thrips	Cucumeris or swirskii. Not limonicus — survival and oviposition on broad mite were both very low	Moderate
Spider mites or rust mites in a perennial planting	Andersoni is genuinely good at this	Strong — and it's the only thing it's strong at
Thrips, and you were sold andersoni	Ask what it's for. One lab screen, no greenhouse trial, and a published rate a third of the advertised one	The absence is the finding

Common questions

Is limonicus more humidity-tolerant than cucumeris?

No — the opposite, and this is the entry we most often see backwards. In the one study that measured egg hatch across a humidity gradient for both species, limonicus was the most humidity-demanding of seven species tested and cucumeris was more tolerant. Three independent labs put limonicus's requirement at around 71–74% humidity at 77 °F (25 °C), and a 1965 study found no limonicus eggs hatched at 50% humidity or below.

Is swirskii worth the extra money over cucumeris?

In warm conditions, on the head-to-head evidence, usually yes — but for the right reason. It is not that swirskii eats more or eats bigger thrips; in leaf-disc assays the two are often indistinguishable. It is that swirskii establishes better and reaches higher populations in the crop, which is what produced a 36-fold difference in one greenhouse trial. In cool winter conditions the published recommendation for cucumeris is explicitly about cost, not performance.

Which one is best in the cold?

Nobody can answer this properly, which is itself worth knowing. Swirskii has a measured population-growth threshold of 59.9 °F (15.5 °C). Cucumeris has a measured developmental threshold around 50.5–51.8 °F (10.3–11.0 °C). Limonicus has no published temperature demography at all — the cold-tolerance claim rests on unpublished company data. Andersoni's only temperature study shows it barely reproducing at 62.6 °F (17 °C), which is the opposite of how it is marketed.

Can I use two species together for better coverage?

You can, and blends that do are sold widely, ours included. What you should not expect is two independent lines of defence. Swirskii preys on cucumeris juveniles and prefers them over thrips, and andersoni is the strongest intraguild predator of the four, so a mixture tends to end up dominated by one species. Being fair to the evidence in both directions: the mechanism comes from confined lab arenas, nobody has ever tested a mixture against the better single species in a crop, and suppliers openly disagree about it. Cucumeris with californicus is a simpler case — different prey.

Why is andersoni sold for thrips at all?

We genuinely don't know, and we sell biocontrol. It is a well-documented predator of spider mites and eriophyid mites in perennial crops, it tolerates pesticides better than most, and it persists on pollen. Against thrips, the only peer-reviewed measurement we could find puts it at 1.73 larvae per day against a supplier claim of four to five, there is no greenhouse efficacy trial, and the one laboratory screen including all four species did not recommend it.