Feeding Swirskii Predatory Mites Pollen: What the Studies Show

A predatory mite sitting in a pile of pollen kills 34 to 60 percent fewer thrips than a hungry one. Give Amblyseius swirskii, a predatory mite about half a millimeter long, a pinch of pollen every week on a cucumber plant, and the whitefly on that plant grew 1.8-fold over two months against 14-fold without the pollen. Both results are measured, and they do not contradict each other.

They fit together because how much each mite eats and how many mites there are are two different things, and pollen moves them in opposite directions. That difference decides when feeding helps and when it feeds your pest. This article follows the greenhouse experiment, what feeding costs each mite, the arithmetic that makes it worth doing anyway, thrips (the pest where it backfires), which foods have been tested, and what nobody has measured.

Forty mites, one cucumber plant, a weekly pinch

It sounds like sabotage. The point of releasing a predator is that your pest is the only thing on the menu, and here somebody was sprinkling an alternative onto the leaves.

The experiment is a 2010 study from the University of Amsterdam (Nomikou, Sabelis and Janssen, in BioControl), and the reason it carries weight is the unglamorous part of the design: whole plants.

Most of what is known about predatory mites was measured on leaf discs, a coin of leaf tissue in a dish of wet cotton wool with a mite on it and nothing else. That is how you get clean numbers, and it removes everything that makes a plant a plant: no stem to cross, nowhere for prey to be that the predator isn't.

Here the unit was a cucumber plant in a greenhouse at about 77 °F. Forty adult female Amblyseius swirskii per plant, twenty adult whiteflies to start and ten more pairs a week later. Half the plants got 25 to 30 milligrams of cattail (Typha) pollen weekly in a small vial, a pinch spread across a whole plant. Half got nothing. The experiment ran about two months, and then everything on every plant was counted.

The predator counts explain the pest counts. At the first count, 37 to 39 days in, there were 225 predatory mites per plant where pollen had been supplied and 45.7 where it had not, roughly a fivefold difference built out of a food that has nothing to do with whitefly. Whitefly on the pollen plants rose to 1.8 times their starting number, without pollen to 14 times, and on plants with no predators at all to about 350 times.

That result has company. Two later greenhouse trials pointed the same way: pollen improved thrips control on caged sweet pepper, and a live feeder mite raised predator numbers on chrysanthemum. The direction is well supported. How large the effect is in other crops is not settled.

This is what swirskii is for. It is a generalist, meaning it eats many kinds of food, and it feeds and breeds on pollen, so it can hold a standing colony on a plant with no pest on it. A specialist on a clean plant starves or leaves. That is why a release before the pest arrives works with this animal and not with most others, and the rest of what swirskii can and can't do runs on the same fact.

What feeding costs each mite

You may have read that a well-fed predator is a stronger, more vigorous hunter. In the studies that measured it, the opposite held for each mite. If you stopped at the cucumber trial you would conclude that feeding makes predators better at their job. Three research groups asked the same question in slightly different ways: with supplemental food on the leaf, how many pests does an individual mite still kill? The answer, every time, was fewer.

A mite sitting in a pile of pollen is not on strike. It is eating the nearer thing, which is what a generalist does.

Each mite loses 34 to 60 percent of its kills while food is on the leaf. If the colony had stayed the same size, pollen would have made control substantially worse.

More mites, each doing less

Ecologists name the two effects. How much each individual eats is its functional response. How many individuals there are is the numerical response. Pollen moves both, in opposite directions.

Five times as many mites each killing about half as much is still more killing, but only if the numbers land where they landed in that greenhouse. It is arithmetic, not a law. Change the ratio, the crop or the food and the sum can come out differently, and further down this page there is a field trial where the extra mites bought nothing.

There is a second, stranger line of evidence for the same mechanism. Messelink and colleagues ran greenhouse cucumber with thrips alone, whitefly alone, and both together. With two pests present, predator numbers ran up to fifteen times higher than with either alone, and whitefly control got better when thrips were also there. A follow-up with three pests found predator numbers up to fifty times higher and the least spider mite damage of any treatment in the three-pest plots.

The paper is titled “two pests are better than one,” and it means it literally. The authors' explanation is dietary: juveniles survived better and developed faster on a mixed diet than on any single food. Not merely more food, but different food. Pollen buys a similar boost to numbers without introducing a second pest to get it.

If the pest is thrips, pollen is the wrong food

Western flower thrips is an omnivore, and one of the things it eats is pollen. A dusting of cattail pollen is not a subsidy to your predator. It is a subsidy to whoever gets there.

Pirayeshfar and colleagues tested six supplemental foods against cattail pollen on greenhouse chrysanthemum and measured thrips egg-laying alongside predator numbers. Cattail pollen produced the highest thrips egg-laying of every treatment in the experiment, including the unfed control. Leman and Messelink had already found the same direction on the same crop: cattail pollen, corn pollen and Ephestia (moth) eggs each roughly tripled thrips egg-laying.

Stack that against the per-mite cost and the shape of the risk is clear. On a thrips problem you are feeding the pest, cutting the rate at which each predator removes it, and hoping the extra mites outrun both. Before there is a thrips population to feed, that bet is reasonable and the trials support it. On a collection that already has thrips, it is a plausible way to make the problem worse, and a commercial grower can flush a crop where someone with twelve pots cannot.

The best-performing food, and why it stays in the greenhouse

The same chrysanthemum experiment found something better than pollen. Living Thyreophagus entomophagus, an astigmatid mite (the family that includes the ones in flour and stored grain), gave the highest swirskii numbers of anything tested, significantly above cattail pollen. Neither the living nor the frozen form supported thrips reproduction: egg-laying on those plants was statistically indistinguishable from plants fed nothing at all. It is the only food in this literature that wins on both counts.

It also meant releasing about 500 living mites per plant, and this is where the recommendation stops. Astigmatid mites can cause dermatitis, allergies and anaphylaxis in the people who handle them, and related species damage plants. Releasing them by the thousand into a sealed greenhouse compartment is a defensible research decision. Doing it in a bedroom is not. The finding is real and it is the best result in that trial, and it is not an indoor recommendation.

The documented middle option is Artemia, brine shrimp cysts sold decapsulated as a predator food. In Vangansbeke's arena work it raised the thrips' own rate of increase by about 5% where cattail pollen raised it 38%, while still supporting swirskii nearly as well. Those are bean leaf-disc figures rather than whole-plant ones, so take the direction rather than the decimals. The ingredient does not settle it either: a commercial decapsulated Artemia product failed to establish swirskii on two crops where a particular Artemia strain succeeded.

More predators is not more control

Feeding trials are cheap, and “we fed them and nothing happened” is a hard paper to place, so the negatives deserve more attention than the positives.

The cleanest one is a field study. Warburg and colleagues grew a Rhodes grass cover crop between citrus trees in Israel, which supplies windborne pollen to whatever phytoseiid (predatory) mites are in the canopy. It worked exactly as designed on the predators: A. swirskii and Euseius stipulatus numbers both rose. Rust mite control improved only in the orchards where E. stipulatus dominated, and the laboratory half of the same study found that swirskii does not control citrus rust mite at all.

There is a second null inside a paper cited above as evidence for pollen. In Ghasemzadeh's caged sweet pepper trial, swirskii plus pollen significantly reduced thrips, and Euseius ovalis, in the same cages on the same food, reached four times higher density and still did not significantly improve its thrips control. The numbers arrived and did not cash out.

We found no paper whose main result is that supplemental food made control worse. That is consistent with the effect being real, and equally consistent with a file drawer of unpublished nulls. The direction is well supported; the size of the effect may be flattered.

Which foods have been tested

“Pollen” on its own is not advice. Goleva and Zebitz screened twenty-one pollen species as the sole food for swirskii and found two, Lilium martagon and Hippeastrum, that killed 100% of the mites. What has been measured, food by food:

Two rows are worth carrying around. Maize pollen ranks below apple, cattail and almond in two independent studies, and the second found the deficit compounds across generations until it shows up as reduced predation. Pine pollen has no evidence behind it for this species. The work that exists is on Amblyseius andersoni, a different mite, and there it was judged unsuitable for rearing. We are not saying it doesn't work. We are saying nobody has looked.

What nobody has measured

The holes here are large, and two of them sit directly underneath the sentence “feed them and they will hold.”

• A persistence curve with the pest absent. We found no study that released swirskii with supplemental food, withheld all pest, and reported what the colony did over time. Every persistence figure here comes from a trial where a pest was present, arrived partway through, or where the plant made pollen continuously. “A standing colony” is a strategy statement, not a measured curve.
• What happens when you stop feeding. Whether the elevated colony converts to predation, disperses or crashes is unmeasured in either direction. A crash is a plausible failure mode, and no finding is not a finding of safety.
• Any indoor study. We found none, not in a home and not on a mixed houseplant collection. Every trial on this page is a greenhouse, a field orchard or a screened cage inside a heated greenhouse. Potted chrysanthemum and potted pepper are encouraging about plant scale and say nothing about a living room.
• Commercially reared mites against prey-reared mites on a real pest. Commercial swirskii is raised on factitious feeder mites (a substitute prey that is easy to farm), and diet is known to change the predator's own reproduction and predation across generations. Whether that matters for the mites that arrive in the mail is untested.
• Food does not substitute for temperature. Swirskii's colony stops growing at about 60 °F, and we work to a range of 68 to 95 °F. The trials on this page ran at about 77 °F. A pollen-fed colony still cannot grow on a cold windowsill in February.

What to do about it

Four rules, each of which is just the arithmetic restated.

Format matters too. Sachets are for prevention: they keep a colony on plants that have no pest. The treatment bottles are for treatment: loose mites released onto plants that already have the pest.

There is also a cheaper option than any of this, and it suits a shelf of pots better than an applicator does.

An ornamental pepper plant makes its own pollen continuously and carries tuft domatia, dense little hair clusters in the leaf-vein junctions where mites shelter and lay eggs. Avery and colleagues compared four cultivars for exactly this purpose and Red Missile came out best on both counts. It needs no dosing schedule, no applicator and no storage. Like everything else here it has never been tested indoors, but it is the finding on this page that transfers furthest toward a home, and the ongoing cost is watering a pepper.

The amounts and timing for dusting pollen, if you go that way, are in the application guide with the trials they came from.

And the sentence to keep: feeding a predator does not make it hunt better. It makes there be more of it.

Common questions

Does feeding predatory mites make them lazy?

Individually, yes. With supplemental food on the leaf, a swirskii kills 34 to 60% fewer thrips than a hungry one, measured by three research groups on two species of thrips. Feeding still improves control because the colony gets around five times larger. You are not buying a better hunter, you are buying more hunters, each of them slightly worse.

Should I dust pollen on plants that already have thrips?

No. Western flower thrips eats pollen too. Cattail pollen produced the highest thrips egg-laying of every treatment in one greenhouse chrysanthemum experiment, and a second study found cattail pollen, corn pollen and Ephestia eggs each roughly tripled it. Before thrips arrive, the trade is a reasonable one.

Is pine pollen good food for predatory mites?

There is no study of pine pollen on Amblyseius swirskii. The evidence that exists is for a different species, Amblyseius andersoni, and it was not recommended for rearing there. Cattail pollen is the food with the whole-plant greenhouse results behind it.

Should I buy live feeder mites to feed my predators indoors?

No. Living Thyreophagus entomophagus is the best-performing food in the trial that tested it and does not feed thrips, but the trial released about 500 living mites per plant inside a sealed greenhouse compartment, and astigmatid mites can cause dermatitis, allergies and anaphylaxis in people. That is a research decision, not a houseplant one.

If I stop feeding, will the colony crash?

We found no published answer either way. No study has released swirskii with food, withheld the pest and reported the colony over time, and none has followed a pollen-built colony after feeding stopped. Anyone who tells you a fed colony holds for a specific number of weeks is quoting something that was never measured.

If pollen builds the colony, why did it fail in citrus?

Because the pest was one swirskii cannot eat. A pollen-supplying cover crop in Israeli citrus raised swirskii numbers as intended and rust mite control did not improve, and the laboratory half of the same study found swirskii cannot develop on citrus rust mite at all. Food changes how many mites you have, not what they can eat.