Why most flies fly alike: a new study pins the uniformity on physics, not ecology

On 10 July 2026, a team of biomechanists published the most comprehensive cross-species comparison of insect flight to date: 133 species of true flies, mosquitoes, and their close relatives, filmed in tethered flight and reduced to a handful of clean numbers. The result, in PLOS Biology, is a near-uniformity no field guide has ever quite captured — wingbeat frequency, stroke amplitude, and the ratio of one to the other converge across most of the order, even between species separated by tens of millions of years of evolution and ecological niches as different as a stream-bed larva and a carrion-feeding adult.

The finding recasts a textbook puzzle. For decades, entomologists treated wingbeat as a species-specific trait, a fingerprint useful for taxonomy. The new data argue the opposite: physics does most of the sculpting, and evolutionary history fills in the rest. Mosquitoes, the paper shows, are the outliers — slower, larger-amplitude, kinematically distinct — and that distinction maps onto a body plan optimised for very different aerodynamic demands.

A convergent kinematic envelope

The team, based at the University of Cambridge and the Royal Veterinary College, used high-speed video to track wing motion across each species and reduced each recording to two variables: wingbeat frequency and stroke amplitude. Plot the 133 species against those axes and the cloud tightens. Mean wingbeat frequency clusters between roughly 100 and 200 hertz across most flies regardless of size, and stroke amplitude sits in a narrow band of around 90 to 150 degrees. The geometry of the envelope is dictated by the physics of flapping at small scales: air viscosity, the inertial cost of reversing a wing at the end of each stroke, and the muscle mechanics needed to drive both.

That convergence is what the paper's authors call the "fly flight envelope." It is not absolute — bumblebees and other heavy fliers sit outside it — but it is tight enough that evolutionary novelty operates within a corridor. A species can shift its frequency or its amplitude, but pushing either too far quickly runs into an aerodynamic or energetic wall. The mathematics of unsteady aerodynamics, not ecological opportunity, sets the perimeter.

Where mosquitoes break the rule

Mosquitoes fly differently, and on the new data the difference is not subtle. Their wingbeat frequencies run slower, often below the lower edge of the fly envelope, and their stroke amplitudes are larger — closer to 150 degrees or more. The reason is structural. Mosquitoes carry unusually long, narrow wings on slender thoraces, optimised for hovering at low speeds and for the slow, controlled flight that blood-finding and mating require. The paper argues that mosquito flight is not the primitive state from which other fly flight evolved, but a secondarily derived solution to a different aerodynamic problem.

This matters beyond taxonomy. Mosquitoes are the vectors of malaria, dengue, and a long list of arboviruses; understanding how they fly, and how their flight envelope differs from that of other Diptera, is a precondition for any genetic or acoustic control strategy. Acoustic traps tuned to mosquito wingbeat frequencies, for instance, assume the frequencies stay within a known range. The new data sharpen the boundary of that range and explain why it is where it is.

What evolution gets to vary

Within the envelope, evolution still has room to act. The paper documents a correlation between wingbeat frequency and body mass — heavier flies beat slower, as mass and muscle physics demand — and a weaker but consistent correlation between stroke amplitude and ecological niche. Flies that hover over flowers or patrol streams tend toward higher amplitudes; flies that pursue fast, straight-line escape trajectories tune for frequency instead. None of these shifts break the envelope; all of them use the slack that physics leaves behind.

That is the structural frame the paper offers, and it is one biologists more familiar with vertebrates may find counterintuitive. Bird and bat wings vary enormously in shape and motion, because the inertial and viscous regimes they operate in are different. Insects live in a regime where air is, in engineering terms, comparatively thick relative to their wing inertia, and the consequence is that flapping kinematics collapse onto a narrow set of solutions. Convergence is the default. Divergence is the exception, and mosquitoes are the exception's loudest example.

Stakes: control, modelling, and what comes next

The practical stakes are concrete. Vector-control programmes, drone and micro-air-vehicle engineers, and anyone building acoustic or visual lures for Diptera all need accurate wingbeat data to design against. The new dataset — 133 species with standardised kinematics — is the largest such resource yet published, and its immediate use is as a calibration set. Models of insect flight that have relied on a handful of well-studied species (Drosophila, blowflies, honeybees) can now be tested against the wider order, and the envelope provides a falsifiable prediction: new species should fall inside it, and any that does not warrants a closer look.

The deeper stake is conceptual. Evolution is often taught as a process of unbounded tinkering, producing endless variety wherever niches allow. The fly data say something more constrained: where physics draws a tight perimeter, evolution works the interior with great subtlety and leaves the perimeter largely alone. The mosquitoes that slip past the perimeter are the proof that the constraint is real — without it, there would be nothing to break.

What the paper does not yet settle

The dataset is tethered-flight, and tethered flight is not free flight. Aerodynamicists have long argued that kinematics measured on a tether can shift systematically — the animal pulls against the restraint rather than accelerating its own body. The authors acknowledge this and report that their key findings hold across multiple tethering protocols, but the caveat is real: the "fly flight envelope" is best read as an envelope of what flies are physically capable of doing under standardised conditions, not a direct map of what they do in the wild. Field validation, with free-flying individuals tracked in three dimensions, is the obvious next step, and several of the same labs are understood to be building the hardware for it.

A second open question is developmental. The paper shows that mosquitoes escape the envelope, but it does not resolve when, in evolutionary time, the mosquito lineage made its escape — whether it was an early branch of Diptera, or a later refinement tied to the blood-feeding lifestyle. The phylogeny in the supplementary material is suggestive rather than conclusive. For now, the cleanest statement the data support is also the simplest: most flies fly alike because the air won't let them fly any other way, and mosquitoes fly differently because they found a different air to fly in.

Desk note: this article frames the PLOS Biology result as a physics-constrained convergence story with mosquitoes as the documented outlier, anchored to the paper's kinematics dataset and the authors' own aerodynamic interpretation.