Stripes do a poor job of hiding a zebra. Lions and hyenas hunt at close range in low light, and a large black-and-white target in tawny grass reads as exactly what it is. The strongest current explanation for the pattern is aimed at much smaller eyes: blood-feeding flies. In field experiments, tabanid flies circled zebras and plain-coated horses at similar rates, but they landed on the zebras far less often — and when horses wore striped coats, the flies failed to land on the coats too. Stripes work at the last second of an insect's approach, a completely different visual problem from the one involved in why camouflage fails the moment you move.
The experiment that moved the debate
Researchers working at a horse farm in Britain watched horse flies interact with captive plains zebras and domestic horses sharing the same fields. Counts showed no real difference in how many flies circled or briefly touched the two species, but far fewer flies managed to settle on zebra pelage per unit of time. The decisive part came next: horses were dressed sequentially in black, white, and striped cloth coats. The same animals, the same days, the same flies — yet striped coats received drastically fewer touches and landings than black or white ones, while the flies continued biting the uncovered heads. Whatever the stripes were doing, they were doing it at very short range.
Video analysis explained the misses. Flies approaching a zebra came in fast and failed to decelerate on final approach, then either bumped the surface and bounced off or aborted entirely. On horses, the same flies executed a normal, controlled descent. The pattern was not repelling flies from a distance; it was breaking the visual autopilot that guides landing.
What a stripe does to a fly on approach
A horse fly on final approach needs to see the surface coming and judge its speed against it. Zebra stripes present alternating dark and bright bands whose apparent width, at the distance where a fly commits to landing, sits right in the range where the insect's motion-detection circuits produce contradictory information. One laboratory analysis showed that evenly spaced stripes generate spurious motion signals of the kind vision scientists call the aperture effect: edges seen through a small receptive window can be interpreted as sliding in almost any direction. A fly flying across that field receives nonsense about its own velocity relative to the surface.
There is a second, independent channel. Many biting flies locate water and hosts by the polarization of reflected light, and tests with polarotactic tabanids found that striped patterns with strong brightness modulation were among the least attractive surfaces to land on. Both results point the same way: stripes sabotage the sensory task that immediately precedes a bite. Judging how quickly one odd element pops out against a patterned field gives a small sense of how much a busy background changes detection.
Why the lion story never fit comfortably
Predator camouflage has been the popular answer since the days of early naturalists, backed by the appealing image of stripes blurring together into a running herd. But field data undercut it: predators typically attack from tens of meters, where stripes resolve perfectly, and zebra lions catch zebras regularly anyway. Other proposals have fared no better in rigorous tests — social signaling, or stripes as a cooling system that channels breeze across the hide — and the geographic overlap between heavy horse-fly seasons and the most finely striped zebra species remains one of the pattern's most consistent correlations. Notably, the animals an ordinary person worries about, such as deer, face the opposite problem from hunters, which is why blaze orange is loud to human eyes and quiet to deer; prey and parasite vision both run on rules that differ from ours.
The entomological answer also explains a feature the camouflage story never could: the pattern is not uniform across a zebra. Striping is finest on the torso and legs where biting flies prefer to land, and comparatively coarse on the neck and face that flies approach differently. A coat designed to fool a lion would profit from blurring the whole silhouette evenly; a coat designed to crash an insect’s landing circuit profits precisely where the insects commit. The geographic argument runs the same way — stripe density across the striped equids tracks the historical range of tsetse and tabanid flies far better than it tracks predator density, and the non-striped horses and asses that live outside heavy fly country manage perfectly well in front of the same big cats.
An answer whose edges are still unfinished
A 2025 review of the whole field describes biting-fly deterrence as the leading explanation rather than a settled verdict. The precise mechanism is still contested, several hypotheses have overlapping predictions, and one imaging study even argues that stripe-scale patterns interfere with the way flies visually isolate warm targets such as blood vessels near the skin surface. That is how a good scientific answer looks: a strong effect demonstrated in the field, a set of candidate mechanisms narrowing, and honest disagreement about the final step. The next time horse flies swarm around grazing animals, watch the last ten centimeters of their flight. The bump, the overshoot, the second circuit — that is the zebra's stripes doing their job, one failed landing at a time.
Three surfaces with identical average brightness; only stripe width changes.