Flies disguised as wasps can’t fool birds

0 7 minutes read

Despite their bee-like appearance, hoverflies are all buzz, no bite. The harmless insects, more closely related to midges than wasps, imitate their distant stinging cousins with stripes, high contrast colors, and narrow waists. In theory, the “flies in wasps’ clothing” use this strategy to ward off would-be predators, without having to pay the cost of evolving venom and an appendage to inject it. But in practice, it’s an open question how well the tactic works.

The quality of hoverfly mimicry can vary– from detailed disguises to the insect equivalent of slapping on a pair of cat ears for a Halloween party. Evolutionary biologists have long questioned why it is that some hoverflies bother with such inaccurate impersonations, and why poor mimicry persists in populations.

New research offers a possible explanation. Even questionable copycats can trip up certain predators like spiders and praying mantises, according to a study published July 2 in the journal Nature. In contrast, other predators are tougher to trick. Using unconventional methods, the researchers showed just how good birds are at distinguishing wasp mimics from the real sting, and what particular traits make the difference.

Some insects imitate the appearance of a harmful insect, such as a stinging wasp, to help protect themselves from predators. Using 3D printing, Taylor et al. examined how accurate mimics need to be for predators to avoid them. a) The common wasp (Vespula vulgaris). b) The hoverfly Chrysotoxum arcuatum is generally considered to be an accurate mimic of the common wasp. c) The hoverfly Syrphus ribesii is generally considered to be a less-accurate mimic of the common wasp. CREDIT: Taylor et al. 2025

Printing insect chimeras

The team created intricate, 3D printed models of hypothetical insect chimeras to explore the whole range from 100 percent wasp to fully fly. Across four experiments with wild birds, domestic chicks, and invertebrates, they observed how different potential predators responded to a variety of artificial mimics.

It’s “a novel way to ask questions that have been discussed in the literature a lot, but not very rigorously tested,” Tom Sherratt, an evolutionary ecologist at Carleton University in Ottawa, tells Popular Science. Sherratt wasn’t involved in the study, but co-wrote an accompanying News & Views piece about the research. “I thought it was a really nice paper,” he says, “they had a great idea for these 3D printed insects.”

Most prior research of mimicry uses real insect specimens or crude stimuli, like colored squares, to examine predator response. But those approaches come with limitations. Either the cue is rudimentary and it’s difficult to evaluate outcomes, or “you’re restricted to only what’s in nature,” Christopher Taylor, lead study author and an evolutionary biologist at the University of Nottingham in England, tells Popular Science. In contrast, 3D printing opened up an entire world of possibilities. Taylor and his colleagues were able to finely tune each of the models used across their experiments with the desired mix of features, and better assess the importance of some changes over others.

*3D printed models of a common wasp. The model in front is life-sized. CREDIT: Tom Reader*

They found that great tits (Parus major), a common species of European songbird, have a hyper discerning eye. After a short period of training, the wild birds were able to pick out most of the fake wasps. They could also readily discern between the “jack of all trade” mimics that mix traits from different targets and models of actual, stinging insects. Taylor and his many co-authors further found that color and size were the most critical aspects of successful mimics, over shape and pattern, in trials with recently hatched domestic chickens. Finally, they discovered that crab spiders and praying mantises are much more easily put off by inaccurate wasp mimics than birds, offering an explanation for why evolution has guided some flies to invest in half measures.

“In some situations, but not all, [poor mimics] persist because they can get away with it. They don’t need to be super accurate in all respects in order to fool a predator–with the caveat that it depends on what their main predators are,” explains Taylor.

In other words: If your main threats are other insects or spiders, appearing unappetizing and potentially dangerous is a lower lift. If you’re particularly tasty to birds, you’ll probably have to try harder.

[ Related: Scientists discover how some flowers mimic the smell of death. ]

Great tit puzzles and spider arenas

To reach that conclusion, the scientists began by conducting 3D scans of real wasps and flies, and then using those digitized insects to construct fake ones. They printed plastic versions of the real insects, fly/wasp hybrid variants ranging from 25 to 75 percent similarity, hybrids between two different species of real wasps, and versions of the wasp/fly chimeras where color, size, shape, or pattern were the focus of the hypothetical mimicry.

In their first experiment, the scientists presented wild great tits with puzzle grids made up of small lidded dishes. On top of each was one of their 3D insect models. Beneath the fly or fly hybrids, the researchers placed a tasty mealworm. Nothing was placed in the wasp containers.. Over weeks of observation, they recorded what order the birds approached containers in. After a short period of teaching the tits that flies advertised a treat while wasps advertised wasted time, the birds readily separated even convincing fly mimics from wasps. The birds significantly favored the fly and fly hybrid boxes over the wasp boxes, according to the findings.

The team also verified that what the birds learned from the figurines would apply to the real thing by running a trial where some of the plastic critters were swapped with pinned, dead insects. They found that the birds associated the real flies with food and the real wasps with less likelihood of reward.

The video shows two great tits at one of our feeding stations. The birds examine various 3D models before selecting which dish to open. If they have successfully avoided the wasps, they find a tasty treat inside (a mealworm). (The wire mesh through which the video is recorded was there to exclude larger birds and squirrels from knocking over dishes and stealing the mealworms.) CREDIT: Christopher Taylor.

In another experiment, they conducted similar trials, but with 3D models of two different species of wasp and hybrids spanning the gap between both. They found that the birds were easily able to tell the difference between the insect models, and readily learned that the plastic replicas of the true wasps had no food to offer, while the hybrid “mimics” were worth foraging for.

Then, Taylor and co presented baby chickens with different versions of hypothetical wasp mimics, to tease out which traits were the most important for keeping predators at bay. After food reward training, they timed how long it took for the chicks’ to attack or examine the insect models. When the insect models were yellow and black or accurately sized to match a common wasp, the birds delayed pecking for longer. Wasp-like patterns and shapes, on the other hand, were less of a deterrent.

In their invertebrate trials, the team placed jumping spiders, crab spiders, and praying mantises in an arena with a series of 3D printed insects, one at a time, spanning from full fly to true wasp. In the presence of the plastic wasps, the researchers administered a “punishment” to most of the test subjects by prodding them in the abdomen. With and without training, the mantises and crab spiders showed a significantly different response to the flies than to the wasps, and were much more cautious of lesser intermediate mimics than the birds had been. Jumping spiders, which have better vision than many other invertebrates, were better able to tell hybrids and true wasps apart.

On their own, each of these experiments “would be good papers,” says Sherratt. “But the fact that we’ve got all of them under one roof is what I think really helps make this paper quite interesting and imaginative.”

[ Related: This spider pretends to be an ant, but not well enough to avoid being eaten. ]

A new biological frontier

Though imaginative, 3D printing does come with limitations. For one, the plastic insect figurines are purely visual cues. They can’t replicate the behavior, sound, or chemical aspects of an actual, living hoverfly or wasp, notes Taylor. It’s quite possible these less printable features might be part of how predators pick their meals and learn to avoid unpleasant experiences. Then, there’s the fact that the birds in the study were trained and tested on a reward/no reward basis. In nature, the risk of picking on a wasp isn’t just that you might miss a meal, it’s also that you could be hurt, says Sherratt. Under those higher stakes conditions, birds in reality might be less willing to approach wasp mimics than they were in the experiments.

Still, the study is an “ambitious” foray into a new frontier of biology research, Sherratt says. He imagines scientists using similar methods to explore sexual selection and other forms of mimicry.

Taylor hopes their approach for exploring imagined adaptations allows for clearer answers about why and when species settle into a long-term form.

“It feeds into this larger question about evolution,” he says. “You see a broad range of adaptations in nature, but how well-adapted does a particular species need to be before it kind of reaches the peak?”.