As children, we learn about the various kinds of simple machines that underly most human constructions. Levers, pulleys and inclined planes all serve to help humans turn force in one direction into motion in another or otherwise make it easier to move objects. The six kinds of simple machines (levers, pulleys, wedges, inclined planes, screws and wheels) were identified and defined in the Renaissance and have formed the basis of more complex machines ever since. Indeed, simple machines are basic enough consequences of physical laws that animals can evolve some of them as parts of their bodies: a carnivore’s teeth are wedges and most of our joints are levers. However, one simple machine in particular stands out as something that, by and large, does not occur in nature—and it reveals some of the limitations of biology.

Why are there no animals with wheels?

Why Are Wheels Useful?

Anyone who has put something in a wagon or bicycle basket, or had to push a car in neutral, knows that moving something with wheels is much, much easier than moving that same object by dragging it on the ground. By reducing the contact area with the ground and introducing rotation, wheels replace ordinary friction with rolling friction. Rolling friction is a much weaker force than ordinary friction, by up to two orders of magnitude, so it’s little surprise that inventing the wheel was a seminal moment in humankind’s social evolution and that wheeled vehicles dominate transportation of both people and objects around the world. But why are humans the only animals that do this?

Challenges of Wheels

The key is that, unlike most other simple machines, a wheel is necessarily three parts: the wheel proper, the axle and the housing that attaches the previous two to the rest of the object. The axle can be joined to either of the other two, but not both, and the whole assembly’s function as a movement aid depends on that lack of direct connection. The wheel must be able to rotate relative to the object whose movement it facilitates in order to work as a wheel. As a result, the wheel-and-axle assembly is less a part of the object than something carried by it. Similarly, the rotating part of this structure exerts friction against its non-rotating neighbors, which necessitates that the axle, wheel, housing, or all three be periodically replaced as they physically wear down from that friction.

An additional wrinkle is that a wheel is not its own power source. Cars have wheels, but what actually makes a car move is its motor, and skateboards and bicycles rely on muscle power. A hypothetical wheeled animal would also need to invest in sails, blowers, legs or some other way to actually exert force against its environment in order for wheels to be helpful.

Animals can carry many things, such as shelter (think sea urchins wearing rocks as camouflage) or tools (birds carrying nesting materials), but they rarely create these things from their own bodies. One of the better examples is the shell of the paper nautilus, genus Argonauta, a kind of octopus that secretes an egg case it can also ride, hide within and control as an underwater vehicle. This submarine shell, striking as it is, does not interact with any other hard structures and so does not provide a window into how a wheel and axle could evolve.

Bioengineering Does Not Work Like That

Unfortunately for hypothetical bio-wheels, the growth mechanisms available to biology are, for the most part, wet and contiguous. Growing hard parts, such as shells or bones, usually involves glands that require those structures to stay reasonably still relative to the organ secreting them until their growth is complete. Maintaining those structures, in turn, requires a similar level of connection between the organs that maintain them and the structures themselves. Since physical separation is absolutely critical to the function of a wheel, this does not bode well for the ability of biology to create wheels.

Once created, that wheel and axle would then need to be maintained, periodically replaced, or both. Biology has evolved mechanisms for maintaining hard parts once they are made, but these mechanisms all have limitations incompatible with how wheels and axles work. Bones are maintained on their entire surface at once, but they are nearly all completely encased in flesh, and the ones that aren’t (antlers) are periodically shed and done without. Rodent incisors and other continuously growing teeth wear out in one part of their length and are grown in another, whereas a hypothetical secreted wheel or axle would be rotating about the exact site where it is grown, subjecting fragile glandular tissue to endless friction. Snails manhandled badly enough to separate their shells from their mantles tend not to survive the experience—a fatal flaw in the wheel as a biological construct. It is telling that the shell of the paper nautilus is not a shell in the conventional mollusk sense, secreted and maintained throughout its life from a single organ to which it is permanently glued, but a temporary structure that can be abandoned and made anew when it is damaged. An argonaut can function while it is replacing its shell, and male argonauts don’t even produce shells, but a wheeled organism without its wheels is rather less likely to survive.

Argonauts and other aquatic animals that grow elaborate motile armor for themselves are especially instructive, because they illustrate the other issue bedeviling biological wheels: the lack of biological drive trains. As mentioned earlier, wheels do not provide locomotor force on their own. Their role is to reduce friction and enable whatever force otherwise exists to operate more efficiently. A wheeled organism still needs some other way to move around, and all those ways—literally all of them—still work without wheels. An organism with legs for pushing its wheeled body along the ground could still push its body if it didn’t have wheels, just less well; one with a sail for catching the wind could still catch the wind without wheels. These facts mean that natural selection is far more likely to enable more efficient sliding (with smooth bellies, slime or lifting the body off the ground entirely) than it is to engineer a complex solution like wheels, which matches what we observe in nature. It is far less likely that axles somehow evolve to connect to some kind of biological drive train.

Every Step Matters

The ways humans construct wheels have little in common with how animals construct their locomotor organs. Humans build wheels, axles and the part of a vehicle meant to interface with wheels and axles separately, and then assemble them in ways that make perfect sense for an animal with advanced cognition and flexible grasping limbs. This means that whatever place or set of tools facilitates constructing the parts can be separate from the vehicle that will use them, enabling each to be designed for its purpose rather than an awkward compromise between the two very different priorities. It’s not completely impossible that an animal could, say, secrete its wheels or axle in one part of its body and have a separate wheel/axle housing for actually using them, especially since that housing could then take advantage of how good animals are at secreting lubrication (such as the synovial fluid in human joints or the slime that helps land snails move), but a complex biological system being imaginable all at once does not make it evolvable.

Natural selection does not abide failure and it does not build systems piecemeal. Every evolutionary step, through the ages, must be at least neutral, and none of them anticipate what’s to come. The mechanisms of biological growth do not lend themselves to complex interchangeable parts the way metalworking does, and almost everything must be made by the organism near where it will be used. If a path does not deliver incremental improvement toward a fitness peak, with no troughs along the way, it is abandoned. Calling a wheel a “simple” machine is a bit of a misnomer, as we have observed. A wedge is all one piece, and levers are a basic consequence of physics acting away from an object’s centre of mass, but wheels have multiple parts that all need to meet certain requirements to make the whole functional. When multiple pieces are needed before any of them are useful, evolution becomes the wrong way to make them happen. Wheels can be made, but, it seems, they cannot be grown.

Or Can They?

There is one way organisms can take advantage of the labour-saving power of rolling friction: they can roll. Many animals and even plants can turn their entire bodies into wheels. Armadillos, tumbleweeds, the flic flac spider and more can roll up or take on round shapes to enable themselves to roll down hills, with the wind or in whatever direction another animal might push them. These creatures do not solve the “biological drive train” problem and cannot move under their own power or exert much control over the direction of their travel while rolling, but they can move with little effort or friction while in this ball shape.

And against all these improbable odds, one animal nearly arrives at a biological wheel, and it does so by turning itself into a Flintstones car: the dung beetle.

Dung beetles shape animal feces into balls that they then grasp and roll toward their dens. Some of their legs latch into their dung balls, acting as axles. Dung beetles’ motion is clumsy, often resulting in them latching onto dung balls and rolling down hills instead; it requires that they already have well-developed legs capable of moving them around on their own; and it is temporary, executed as part of their breeding behavior. Axle wear is resolved by the animal not incurring much—animal dung is soft and exoskeletons are hard—and by replacing its exoskeleton periodically. The entire wheel is food for the dung beetle’s offspring and explicitly does not need to be maintained. And while the dung beetle pushes its prize across the savanna, it is the rare animal with what amounts to an actual, honest-to-goodness wheel.

So the answer to why animals don’t have wheels is that some of them do, but they’re almost universally…a bit crappy.