The evolutionary origins of zebra stripes have been investigated—and debated—for centuries. The trait is rare, conspicuous, and intensely expressed, and thus appears to beg an adaptationist explanation. However, the utility of a complete coat of densely packed, starkly contrasting black-and-white stripes is not immediately apparent. Unlike many conspicuous visual traits, striped pelage is expressed with comparable intensity in both sexes and is thus unlikely to have arisen through sexual selection alone (although in plains zebras, Equus quagga, males have stripes closer to true black than females). Stripes are clearly not aposematic warning signals, nor do they provide camouflage in either the woodland or savannah habitats common across zebra ranges1,2. So, striping presents an ideal evolutionary puzzle: a trait so refined it seems it must be “for” something, but one that confers no clear advantage upon its bearers and imposes apparent costs (conspicuousness) that cannot be explained in Zahavian terms.
Scientists have proposed and investigated several possible explanations for the evolution of zebra stripes (reviewed in3). The hypotheses suggest various ways in which stripes may provide a social function (species or individual recognition or social cohesion1,4), a temperature-regulation benefit5,6, an anti-predator effect7,8, or an anti-parasite effect9,10. There is continued debate over both the merits of individual hypotheses and the likelihood of stripes having arisen via a single driver vs. a confluence or alternation of multiple selective pressures6,11.
The present study addresses the hypothesis that has thus far received the most empirical support: the anti-parasite hypothesis (also known as the ectoparasite hypothesis12). Zebras, like most ungulates, are harassed by tabanid, glossinid and Stomoxys species of biting flies, which can inflict significant blood loss, transmit disease, and weaken hosts when fly-avoidance behaviors reduce the host’s feeding rate9,13,14. Yet zebras are attacked far less than sympatric ungulates across their African range15,16, and also less than other equids9,17. Zebras also produce odors that may augment their anti-fly defenses18, but so do other sympatric ungulate species18,19, and a host of observations and experiments have demonstrated that black-and-white stripes alone are unattractive, or actively repellent to tabanid, glossinid, and Stomoxys flies17,20,21,22,23.
Though the effect of stripes on flies is well-established, the source of the effect remains unexplained. Since Waage’s foundational studies in the 1970s and 1980s9,24 most hypotheses have suggested ways that stripes might interfere with the visual and navigational systems of flies, making it harder for them to locate, identify, or successfully land on striped targets. These hypothetical mechanisms can be roughly grouped by the distance (and the attendant phase of a fly’s orientation and landing behavior) at which they would likely operate:
From afar: stripes might make it harder for flies to locate and distinguish zebras from background vegetation, perhaps by breaking up their outline9 or varying the way they polarize or reflect light17,31 especially from distances at which composite eyes support only low-resolution vision and cannot resolve zebra stripes as clear bands of alternating color on a single host (estimated at > 2.0 m22, > 4.4 m24, and even > 20 m25).
At close range (estimates range from 0.5 to 4.0 m26): stripes might interfere with orientation or landing behavior via any of several disruptive or ‘dazzle’-related visual effects27. For example, stripes might affect ‘optic flow’, or the fly’s perceived relative motion to its target as it approaches, by creating an illusion of false direction or speed of motion (e.g., via variants of the ‘barber pole’ or ‘wagon wheel’ effects28). Alternatively, relative motion to a striped pattern within the visual field may create the perception of self-rotation, inducing the fly’s involuntary ‘optomotor response’ and resulting in an avoidance turn in an effort to stay on a straight course29.
Finally, stripes might cause confusion in the transition between long- and short-distance orientation. If zebras appear as blurred gray from a distance and then, at closer range, suddenly resolve into a sequence of floating black and white bars, this abrupt ‘visual transformation’26 might disrupt the behavioral sequence that facilitates landing.
Within these categories, hypotheses have proliferated faster than experimental tests of many of the proposed mechanisms. The very active literature on this question has grown in somewhat haphazard fashion, as curious researchers test new possibilities without eliminating old ones6. Importantly, few experiments have controlled the distance from which flies are first able to view potential landing sites (but see23). While growing evidence supports a mechanism operating at close range22,26, failing to restrict the starting distance of the fly means that the full set of possible mechanisms outlined above all remain plausible contributors to most previous results.
Additionally, while many studies have, appropriately, used artificial stimuli to isolate basic effects of color, pattern, brightness, and light polarization of (usually flat) test surfaces, possible contributions of several aspects of natural zebra pelage remain untested. Controlled experiments have used various landing substrates, including striped and solid oil tray traps, sticky plastic, smooth plastic17, cloth (Experiment 2 in22), horse blankets or sheets26, and paint on live animals30. These have all clearly demonstrated a broadly replicable visual effect: stripes, and some other juxtapositions of black and white (e.g., checkerboard patterns26), repel flies. However, insofar as specific features of zebra pelage factor into proposed mechanisms of fly repellence—the reflective properties of “smooth, shiny” coats31; the orientation of the stripes17,32; the light-polarizing effects of black and white hair vs. background vegetation25; and the complex structure of hair25—there is a need for more experiments that present natural targets to wild flies (but see22,33). Similarly, most experiments have compared landing preferences between black-and-white striped, solid black, solid white, and occasionally solid grey substrates, which have served as important controls for determining that light polarization, rather than a combination of polarization and brightness, is sufficient to induce the effect of stripe avoidance17. However, it is now time to refocus on the original question by presenting flies with more realistic choices. Since biting flies seeking a bloodmeal on the African savannah seldom encounter solid black hosts, and even more rarely solid white hosts, landing choices should be compared between zebra stripes and common coat colors of sympatric mammals, namely various shades of brown. Further, tabanid, glossinid, and Stomoxys flies all avoid landing on stripes that are the same width or narrower than the widest zebra stripes 17,23, and there is some evidence that narrower stripes are even more repellent to tabanids17. This pattern is potentially significant in the application of the anti-parasite hypothesis to an adaptive explanation for the striking variation in stripe width across zebra species and between the different areas of the body on individual zebras22, but must first be confirmed with experiments using real zebra pelage.
Here, we present a simple experiment designed to address each of these gaps in the literature on the anti-fly benefits of zebra stripes. In this field experiment, the landing choices of flies were tested entirely within the range at which all estimates agree flies should be able to perceive the presented stripes (< 1.0 m). This restriction enabled us to determine whether hypothesized mechanisms that act at close range are sufficient to produce the fly-repulsion effect, potentially revealing other mechanisms to be unnecessary (all mechanisms that operate at or beyond the distance at which zebra stripes can be resolved). Our paired-choice experiments used real animal pelts, both striped (zebra spp.) and solid tan (impala; Aepyceros melampus), mimicking the actual choices made by flies cruising the African savannah. Using pelts over live animals allowed us to remove potentially interacting effects of fly-deterrent behavior (tail-switching, stomping, skin rippling, and other movement) that are less relevant to specific questions about the evolution of zebra stripes. We further isolated visual properties of the pelts by salt-curing the hides, a well-established gentle preservation method that (a) minimally impacts the specific luminance, reflectance, and pigment saturation, as well as the depth, density, and structure of animal hair, while (b) minimizing the persistence of volatile olfactory compounds that might be independently aversive to flies34. To further minimize the potential confound of odor, we chose impala skin as the alternative landing substrate to zebra pelts; like zebras, impala are consistently unattractive hosts for biting flies19,35,36. Finally, we tested pelts from the zebra species with the widest stripes overall (the plains zebra, the only species with wide stripes across most of its body17) and the narrowest stripes (the Grevy’s zebra, Equus grevyi) under the same conditions, to investigate whether cross-species variation in stripe width could represent differences in their abilities to repel flies.
Source: Ecology - nature.com
