Spider-eating Spiders

Robert R Jackson & R Stimson Wilcox. American Scientist. Volume 86, Issue 4. Jul/Aug 1998.

In the conventional view, “intelligent” function-complex and flexible behavior, the capacity to innovate and learn-requires the large, highly developed brains of animals such as chimpanzees and dolphins. More primitive creatures, whose behavior is commonly guided by instinct, have small, comparatively simple brains made up of relatively few neurons. A spider’s brain, for instance, fits comfortably on a pinhead, and it seems unlikely that such a small collection of neural components could organize complex and flexible behavior. Yet recent work on Portia, a genus of jumping spider, illustrates that although a spider may have a small brain, its behavior need not be simple.

The visual capability of jumping spiders sets them apart from other spiders. The eyes of most spiders lack the structural complexity required for acute vision, but jumping spiders have unique, complex eyes with resolution abilities unparalleled in animals of comparable size. A jumping spider has a pair of principal eyes, located front and center on its head, and they provide acute vision. Three pairs of secondary eyes surround the principal ones, and they detect motion. Jumping spiders use their vision when stalking prey. Although many spiders spin a web and wait for something to get captured, most jumping spiders hunt prey without using webs. In general, jumping spiders prey on insects, but Portia prefers to hunt other spiders.

Portia is a versatile predator, which means that an individual animal uses a repertoire of disparate predatory tactics, each specific to different circumstances or different types of prey. We shall show that these spiders invade the webs of other spiders, where they use mimicry, detours and deception to capture the resident spider. In addition, they are capable of changing their tactics through trial and error.

Aggressive Mimicry

Although the genus Portia includes about 15 species, our knowledge of their behavior comes primarily from studies of five species: P. africana and P. schultzi from Kenya and Uganda; P. albimana from Sri Lanka; P. labiata from Malaysia, the Philippines and Sri Lanka; and P. fimbriata from Australia, Malaysia and Sri Lanka. Appearance alone suggests that Portia is unique. It does not look like a spider at all, or even an animal. Instead, it resembles detritus in a web. Such an appearance may conceal Portia from prey that might see it coming, not to mention concealing it from its own visually hunting predators, especially birds.

We have concentrated on another form of mimicry, aggressive mimicry, which means that the predator deceives its victim by imitating something desirable. To study this behavior, we watched spiders in the field as well as in laboratory experiments. In addition, one of us (Wilcox) developed a computer-based system for recording and playing back signals on webs, which is rather like listening and talking to spiders in their own language. A transducer-rather like the stylus on an old-fashioned record player-is used to pick up web signals (vibration and movement patterns) from the silk and turn them into electrical signals that can be fed to a computer and analyzed. Signals can be played back by playing signals from the computer through an electric coil to make a magnet on the spider’s back force the spider to move in whatever way we choose. Using this system, we found that Portia uses aggressive mimicry when pursuing another spider on its web. After entering the other spider’s web, Portia sends web signals across the silk, and sometimes the resident spider responds as if a small insect were ensnared in its web. When the duped spider approaches, however, Portia lunges out and catches it.

Portia makes its web signals by manipulating, plucking and slapping the silk with any one or any combination of its eight legs and two palps, relatively small appendages near the mouth. Each appendage can be moved in a great variety of ways, and the movements of any one appendage can be combined with different movements of any number of the other appendages. In addition, Portia makes signals by flicking its abdomen up and down, and this can also be combined with virtually any of the appendage movements. Consequently, Portia can make an apparently unlimited number of different signals on the web of another spider.

Instead of relying on vision, typical web-building spiders detect and interpret web signals. Portia needs to vary the web signals that it makes because different species interpret them differently. How a spider interprets a web signal may also vary with its sex, age, previous experience and hunger level. Despite that variability, we have observed Portia using aggressive mimicry to capture just about any kind of webbuilding spider, from about one-tenth to twice its own size.

Tuning the Trembles

How does Portia generate an effective web signal for each victim? We think that the spider relies on two basic ploys: using preprogrammed tactics and deriving signals by trial and error.

Using preprogrammed tactics is consistent with the popular portrayal of spiders as animals governed by instinct. For instance, Portia makes specific web signals when hunting a specific species of spider. Trial and error, on the other hand, is an unexpectedly flexible behavior for a spider.

When Portia enters a web of a species for which it lacks a preprogrammed tactic, it makes a vast array of web signals. If one of these signals has an appropriate effect-such as enticing the resident spider to approach as if Portia were a small insect ensnared in the web-then Portia stops varying its signals and concentrates on producing the signal that worked. Nevertheless, aggressive mimicry can be dangerous for Portia, especially when the resident spider is large and powerful. By pretending to be prey and provoking a full-scale attack, Portia runs a risk of becoming what it pretends to be. Portia compensates for that danger by gaining fine control over the victim’s behavior. For example, Portia may make signals that slowly draw in a dangerous spider. Alternatively, Portia may move in slowly for the kill while using a monotonous web signal that keeps the victim calm. Sometimes, Portia entices its victim into a particular orientation before attacking. Interactions with pholcids, spiders with very long legs, illustrate this. If Portia hits a pholcid’s leg first, the pholcid defends itself and sometimes kills the would-be predator. So Portia uses trial and error to generate a web signal that coaxes the pholcid into a position that offers a clear shot at the body.

Even when Portia has a preprogrammed tactic for a spider, trial and error remains relevant. Often the preprogrammed tactic simply gets a predatory sequence off to a good start, after which Portia finishes the job by trial and error. For instance, the victim spider might start approaching slowly but then lose interest, become distracted or begin approaching too fast. At that point, Portia might switch from a preprogrammed tactic to trial and error.

Deception and Detours

In most cases, Portia very slowly stalks a spider in a web. In some circumstances, especially when the wind blows, Portia might walk rapidly toward an intended victim. We used fans to generate artificial wind in laboratory experiments that showed that Portia opportunistically chooses to approach its victim when wind provides a vibratory “smokescreen.” This tactic works because windgenerated movement of the silk masks most web signals, including the faint signals from Portia stalking across it.

In still conditions, Portia can make its own vibratory smokescreen. That is, while stalking across the web, Portia can mask the faint vibrations it makes by producing smokescreen signals that mask forward motion as well as simulate the vibrations made when twigs and leaves hit a web. Perhaps surprisingly, Portia only uses opportunistic and selfgenerated smokescreens against spiders, not when it stalks insects ensnared in webs or the egg sacs of another spidertargets for which masking is irrelevant.

If Portia cannot get close to intended prey through deception, it might use a detour, as it often does in pursuit of Argiope appensa, a spider that builds orb webs on tree trunks. Although it might appear that Portia could simply walk straight from the tree trunk to the web, A. appensa is exceedingly sensitive to anything foreign touching the web and rarely lets Portia enter unchallenged. If an insect prey is detected, A. appensa attacks. If a potential predator is detected, it uses a specialized defense called “pumping,” or rapidly flexing its legs over and over again. That sets the web in motion and either drives or throws Portia out of the web.

Portia often walks up the tree trunk toward A. appensa and then stops, looks around, goes off in a different direction and reappears above the web. If there is a vine over the web, for example, Portia seems to look at the web, the vine and the neighboring vegetation before moving away, perhaps going to where the web is completely out of view, crossing the vegetation and coming out on the vine above the web. From above the web, Portia drops on a silk line alongside but not touching the intended victim’s web. Then, when parallel with the spider in the web, Portia swings in to make a kill.

In laboratory experiments with Michael Tarsitano (then Jackson’s graduate student), we found that Portia consistently takes detours to reach prey, and these detours often appear to be planned ahead of time. For example, if presented with a choice of two routes on artificial vegetation, only one of which leads to prey, Portia consistently takes the appropriate path even when this means initially going away from the prey and losing visual contact.

Getting to Know the Locals

Portia can also adapt to local prey. For instance, a population of Portia fimbriata in Queensland, Australia, evolved predatory tactics against local cursorial jumping spiders, which are exceptionally abundant there.

Although jumping spiders that are strictly cursorial do not spin prey-capture webs, they do spin silk shelters, or nests, which are usually densely woven, tubular in shape and not much larger than the resident spider. When a spider of the same species but opposite sex finds a jumping spider inside a nest, the intruder might court the resident by making vibratory signals on the silk. If P. fimbriata finds an occupied nest, it uses vibratory signals that often cause the resident to poke its front end out, only to be grabbed and eaten.

For catching jumping spiders out in the open, P. fimbriata uses cryptic stalking, a special type of trickery that capitalizes on its unusual, detritus-mimicking appearance. Portia normally walks in a slow, choppy gait, but during cryptic stalking the Queensland P. fimbriata moves even more slowly, often going undetected until it is too late for the victim to escape. If the stalked spider suddenly turns around, the Queensland P. fimbriata freezes in its tracks, and the hunted spider apparently perceives nothing more than a piece of detritus. In addition, the Queensland Portia pulls in its palps during cryptic stalking, which makes it look even less like a spider. Other species of Portia also pull in their palps when they are quiescent in a web. We know that palp outlines are cues by which jumping spiders recognize Portia as a predator, making it all the more interesting that the Queensland Portia hides this cue when stalking other jumping spiders.

Another example of local adaptation exists between the Queensland P. fimbriata and another local jumping spider Euryattus, which is not known to coexist with any other population of Portia. The female Euryattus is unusual because, instead of making a tubular silk nest, she uses heavy silk guylines to suspend a rolled-up dead leaf from a rock ledge, tree trunk or vegetation. Euryattus males go down guylines onto rolled-up leaves and court by suddenly flexing their legs and making the leaf rock back and forth. Then the Euryattus female comes out on the leaf surface to mate with or drive away the male. Unlike other Portia studied, the Queensland Portia mimics the courtship of a male Euryattus. When the “courted” female comes out, the Queensland Portia attacks and eats her.

Euryattus is not defenseless. Either before or after Portia reaches the leaf, Euryattus may come out and make sudden, violent leaps or charges toward the invader, sometimes making head-on collisions with the predator and knocking it away. Once attacked, Portia always flees. Euryattus is exceptionally effective at recognizing and defending itself against a stalking Portia. This appears to be a local adaptation. At Davies Creek, only about 15 kilometers away from the habitat where Euryattus and P. fimbriata live together, there is another population of Euryattus living in the absence of Portia. When tested in the laboratory, Davies Creek Euryattus only rarely evaded or attacked stalking Portia, and the Queensland Portia readily caught the Davies Creek Euryattus.

Mate or Meal?

Some of the most unpredictable behavior in Portia involves mating. An apparent mate may become a meal, and what appears to be a mating tactic may also be a predatory or defensive tactic. In general, male jumping spiders approach females in rapid stop-and-go spurts of activity, punctuated with displays, such as stiffly extending the first two pairs of legs and waving or shaking them in front of the female. Most female jumping spiders scrutinize the male’s displays and occasionally display back. Conversely, Portia females often initiate intersexual interactions by displaying first. Once the interaction begins, the female is exceptionally active in displaying back to the displaying male. Portia’s intersexual interactions may take place in or outside of webs, with the female tending to display, move away a short distance and then turn back and look at the male. In this way, the female moves steadily to higher ground at the top of a web or up in the vegetation, with the courting male close behind.

The distinction between mating and predation blurs in the Sri Lankan P labiata and the Kenyan P. schultzi. Females of these two species drum-pounding on the silk with their two palps-and tug-making sharp pulls on the silk with their forelegs-solely in interaction with males of the same species. Once a female has been mounted by a male, she drops on a dragline with the male on board. A risky mating takes place with the pair suspended on thread. During or just after mating, while still suspended, females almost always swing around violently with fangs extended, in sometimes successful attempts to impale and eat their unfortunate suitors. Other times, the male does not mate at all because, as he approaches, the female suddenly and violently lunges forward, spears him with her fangs and eats him.

The twisting and forward lunges of P. labiata and P. schultzi females appear to be prey-specific predatory tactics designed to catch courting males-prey that deliberately approaches. By killing the male after he starts copulating, the female chooses him as both a sperm donor and a meal. If she kills and eats him before mating, she rejects him as a mate but accepts him as a meal. In some cases, females might encourage males under false pretenses. If eating is a female’s only interest in a male, then she is using aggressive mimicry against her own species.

The implication of feminine deceit is less ambiguous when P. labiata and P. schultzi females are immature and cannot mate. Males appear unable to distinguish a large juvenile from an adult female, and juvenile females drum and tug, apparently as a ploy to lure a male in for the kill. Often a male mounts and performs futile and dangerous pseudocopulations while suspended from the juvenile female’s dragline. By twisting and lunging, the juvenile often makes a meal of the frustrated male. Chemical deceit may also be part of the juvenile’s game. Pheromones from adult Portia females attract males and prime them to court, and juvenile females release similar, if not identical, chemical signals.

Portia females also react violently toward other females. Female-female interactions can become fights to the death in which the spiders fight over webs and eggs. Webs spun by one female can be used by another, and the eggs in a web-the resident female’s progeny-are potential food for a rival. Moreover, another Portia female is also a potential predator and prey because cannibalism is a frequent outcome of female-female interactions.

Vision and Beyond

Many other investigators have shown that typical jumping spiders distinguish between prey and nonprey through various optical cues, including shape, symmetry, presence of legs and wings, size and style of motion. Beyond using its visual skills, Portia seems to be making decisions to control the many components of its predatory strategy. Most of what we know so far concerns the cues that govern Portia’s decisions on whether to enter and stay in a web, whether to make signals once in a web and whether to persist at signaling.

Many cues influence Portia’s behavior. Seeing a web drives Portia to enter it. Seeing a spider in a web increases Portia’s inclination to enter. As far as we know, volatile chemical cues from webs or prey spiders appear to be unimportant. After entering a web, cues from the web largely control signaling behavior, but seeing a prey spider or detecting vibrations on the web make Portia more prone to signal by vibrating the web. Working with our colleagues, we have also shown that, with optical cues alone, Portia can distinguish between quiescent spiders, insects and egg sacs, as well as between different kinds of spiders, but we do not know precisely what cues they are using to make such distinctions.

Although Portia’s eyes are large and complex for a spider, this is no primate. The principal eye lens is only a few millimeters in diameter, and there are only about 100 photoreceptors in the fovea of a jumping spider’s principal eye. How does Portia do so much with so little?

If there is a lesson to learn from Portia’s predatory strategy, it might be that we should not underestimate the potential for complex behavior in a small and seemingly simple animal. There were good reasons to expect less from Portia. How an eye with so few components can achieve Portia’s feats of visual discrimination is only part of the mystery Portia’s tiny central nervous system somehow supports a predatory strategy more similar to what we would expect in a cat or a dog than in a spider. Much more work must be done to find out how a brain so small orchestrates such complex and flexible behavior.