Imposters abound in the animal kingdom. Peruse any textbook description of mimicry—in which one species evolves to resemble another—and you will encounter various classic examples, such as the king snake, which copies the coral snake, or the hoverfly, which masquerades as a bee. Less familiar, but in many ways even more fascinating, are the mimics in a genus of jumping spider known as Myrmarachne, which look for all the world like ants.

Unlike other jumping spiders, with their furry, round bodies, Myrmarachne species have smooth, elongate bodies that give the appearance of having the three distinct parts—head, thorax and abdomen—of ants, despite having just two. To complete the charade, the spiders walk on their three rear pairs of legs and raise the fourth pair overhead, waving them around to simulate ant antennae. They even adopt ants' characteristically fast, erratic, nonstop mode of locomotion in place of the stop-and-go movements other jumping spiders make. It is an Oscar-worthy performance and the secret of this group's success: more than 200 species of Myrmarachne thrive in the tropical forests of Africa, Asia, Australia and the Americas. This rich diversity makes ant mimicry the most common form of mimicry. Yet it is the least known.

New research is exposing the mind-boggling complexity of the ant mimics' charade, however. Like the king snake and hoverfly, Myrmarachne species gain a survival advantage by looking like other species—in this case, lethal ant species, because predators of spiders steer clear of both the ants and their look-alikes. But, it turns out, the spiders pay for that advantage: to give a convincing performance, they must expose themselves to considerable risk. The evolutionary forces that led to their fakery have left the ant-mimicking spiders living on the knife's edge, walking a fine line between avoiding one enemy and falling prey to another. In revealing the unexpected perils of mimicry, studies of these remarkable arachnids show the phenomenon of mimicry in a new light.

Faking It
My fascination with mimicry began one day in 1995 in the office of my then supervisor, Robert R. Jackson, while discussing potential research topics for my master's degree. Jackson, a spider expert at the University of Canterbury in New Zealand, had cemented his reputation as a leading arachnologist through his work on Portia, a genus of jumping spiders renowned for their mammalianlike levels of clever behavior. Accordingly, he suggested that I work on a species of Portia. As an afterthought, he mentioned the antlike jumping spiders found in the tropics. I was instantly intrigued. Now, 20 years down the track, Jackson and I are colleagues who share a laboratory and have traveled throughout Africa, Australia and Asia to research these remarkable creatures. Throughout our journeys we have discovered many unusual consequences of mimicry that underscore just how much more complicated the business of deception is than conventional wisdom would suggest.

The standard view originated with English naturalist Henry Walter Bates, who in 1861 provided the first scientific theory to explain mimicry in nature, based on his observations of Amazonian butterflies. Bates supposed that an edible species that resembled an unpalatable or downright toxic one would gain a survival advantage by tricking potential predators into leaving it alone. In Bates's scenario, predators would learn from experience that eating the nasty species was a bad idea. After that unpleasant encounter, the predators would avoid the toxic species and would then avoid the mimics, too—even though the mimics themselves were harmless. This “parasitic” charade, in which one species exploits another's defenses, is now known as Batesian mimicry.

But it turns out that mimicry does not work exclusively in the simple, straightforward manner Bates described—far from it. For one thing, some mimics use their resemblance to another animal not to avoid getting eaten but to deceive their own prey and thus obtain a meal through dishonest signals—so-called aggressive mimicry. And animals exploit mimicry for various other reasons. No group of organisms illustrates the complexities of the strategy, and the evolutionary forces that shaped them, better than ant-mimicking spiders do.

Upsides and Downsides
To the uninitiated, ants might seem unworthy of imitation. But in the tropical rain forest, where their total biomass exceeds that of all vertebrate animals combined, ants are strong shapers of the environment and have great power over its inhabitants. As such, they are prime candidates for being imitated.

Myrmarachne spiders trade on the ants' fearsome reputations: ants avidly defend their nests by biting or stinging intruders, and an individual can recruit an entire colony to its cause—often with lethal consequences for the interloper. Predators are thus wise to avoid trying to eat any prey that look to be such ants. Yet for the spiders to trick predators into avoiding them, they must take some real risks. For instance, they need to live near the ants to avoid standing out to predators as being not antlike. Living in close quarters, which is unusual for spider species but common in ants, puts the spiders directly in harm's way; if they are found to be fraudsters, odds are they will become lunch.

Having to cohabit with their enemies is not the only price these ant-mimicking spiders pay. The dissemblers are so convincing that predators that specialize in eating ants—including some other species of jumping spiders—attack them as prey. And competition between males for access to females has raised this predation risk. The choosy females have driven Myrmarachne males to evolve enlarged mouthparts that can increase their body length by up to 50 percent. Exactly why the females prefer a big mouth is not known, although it may be an indicator of health. At first glance, one would be forgiven for thinking that this enlargement would hurt the spiders' chances of surviving by detracting from their antlike appearance. It does hurt them but not in that way. The trait makes them look like ants that are carrying something in their mouth. Because an ant's mouthparts are very dangerous, ant-eating jumping spiders tend to preferentially target ants that are carrying objects in their jaw and that are thus unable to bite their predators. So although having a big mouth may help male Myrmarachne spiders score with the ladies, it also has the unwelcome effect of making them more attractive to predators.

The cunning mimics can actively defend themselves against some of these threats, exhibiting a surprising degree of behavioral flexibility. For example, when an ant-eating jumping spider initiates stalking, the mimic makes a display toward the potential predator, raising its front legs from their normal antennae posture to a position vertically above the head and staring fixedly at the other spider without moving. The display seems to communicate that it is a spider or at least that it is not an ant after all. Whatever the message, it effectively deters the predator. Similarly, when a pesky scientist (and presumably other potential predators) comes along and tries to catch a Myrmarachne spider clinging to a plant, the mimic will abandon its antlike behavior, drop off the vegetation and hang out of sight on a thread of silk—the best of both worlds.

One particularly Machiavellian species of ant mimic, Myrmarachne melanotarsa, gets the best of both worlds in yet another way and upends the notion that parasitic and aggressive forms are separate phenomena that arise from distinct selective pressures. The spider's resemblance to ants is so terrifying to other, ordinary jumping spiders that in addition to avoiding predation, M. melanotarsa uses its antlike appearance to capture prey. It drives hapless jumping spider mothers out of their nests; then it penetrates the nest to raid the eggs or the brood of spiderlings. Ants have trouble raiding spider nests because their legs get caught in the spider silk, but spiders have adaptations that enable them to negotiate the sticky strands—and M. melanotarsa takes full advantage of them.

Learned or Instinctive?
To fully tease out the forces that have caused mimicry to evolve and take the forms it does, researchers need to know the factors that cause predators to avoid imposters. Back in the 1800s, Bates thought that the predator must experience, in some way, the danger posed by a creature that another organism is mimicking before it grasps that it ought to steer clear of the real McCoy and anything that looks like it. But here again the ant-mimicking spiders flout that conventional wisdom. The ordinary jumping spiders that abstain from eating both ants and Myrmarachne do so from instinct, not as a result of learning through bad experiences. In other words, the forces that shape evolution have baked that avoidance into the predators' hard wiring.

In hindsight, this avoidance instinct is not surprising: after all, if you die in an encounter with an ant, there is no room for learning. In some ways, it is easier to envision how hardwired avoidance could have evolved: predators that happen to dislike approaching ants are more likely to survive and reproduce, and their genes get passed on; ultimately instinctive ant aversion dominates the population, and those that lack the trait are quickly weeded out by the ants themselves.

A Glorious Mess
The complexity my colleagues and I have discovered in the mimicry system of Myrmarachne serves as a cautionary tale: the tangled principles at work here almost certainly apply to other cases of mimicry. And we still have much to learn. Scientists have tended to view mimicry in terms of its being an adaptation to selective pressure from a single predator using a single sense: vision. (Because humans are so dependent on vision, this sense tends to be the one researchers focus on.) But we now know from Myrmarachne that multiple predators shape a mimic species: my own work has shown that ordinary jumping spiders and mantises are influential in this regard; birds, lizards and frogs probably are, too. And studies of other creatures hint that mimicry can involve smell and sound, among other senses. For example, a palatable species of tiger moth mimics the acoustic signals of a noxious one to avoid predation by echolocating bats. And some butterfly species copy the chemical signals emitted by ants to enter their well-defended nests, where the butterflies deposit their eggs for safekeeping.

Excitingly, scientists now have the technology to probe the sensory experiences of other species. High-frequency recording devices allow researchers to visualize noises above our own hearing threshold—including those emitted by tiger moths and bats; mass spectrometry lets them see the hydrocarbon profiles of ants and their mimics, providing a picture of their chemical interactions. Applying these techniques to the study of mimicry and other natural phenomena will no doubt expose more of the spectacular solutions and trade-offs that have evolved in the eternal arms race between predators and prey.