On first acquaintance, the Apollo butterfly seems improbably ethereal for a harbinger of breakthrough technology. A scant two inches from wing tip to wing tip, it flaunts a bold delicacy of design more Miro than Microsoft, the wings translucent white flecked with glowing circles and confident brushstrokes of crimson, black, and brown. And striking though it is in close-up, Apollo (or, officially, Parnassius sminthius) is an easy creature to overlook outdoors. It lives in isolated upland meadows, surviving largely on one inconspicuous low-growing alpine plant, called stonecrop.
Yet for the past two summers, a few of these unpretentiously elegant creatures have been fluttering around the Kananaskis Range of Alberta’s Canadian Rockies outfitted like trend-crazed backpackers with the latest, the lightest, and the techiest of equipment; milligram for milligram, it’s also the priciest. The butterflies carry tiny radar probes that allow University of Alberta ecologist Jens Roland and his colleagues to track them across the meadows of Jumpingpound Ridge at altitudes of 7,000 feet and higher.
The Apollo butterfly lives a complex life--and, as an example of a species dependent on delicate environmental supports, a significant one. While it’s not endangered, it is an object lesson in ecological fragility, requiring an exact and rarefied environmental mix: Alpine meadow, yellow- flowered stonecrop, hot summer temperatures, and ample sunlight. Apollo is no traveler, and it can’t manage a long-distance search for optimal conditions. Though its ancestors found their way to North America from Asia, they must have done so in incremental stages, generation after generation. Individuals pupate, mate, and reproduce over a single summer, dying with the onset of cold weather. Only the eggs and the pupae winter over. Buried in the gravelly soil, they replace the water in their bodies with glycerol, a substance also used in antifreeze. The butterflies tend to stay put, content to live and die within the confines of a single meadow. Even there, Roland says, they like the hot little valleys, where there’s lots of stonecrop. So their well-being is highly contingent, fluctuating as meadow conditions change.
All in all, the Rocky Mountain site offers an ideal natural lab for studying the interaction between a vulnerable species with very particular needs and an environment that’s subject to change. Homebodies though they are, the 10,000 or so Apollo butterflies that live in the chain of high-lying meadows along Jumpingpound Ridge can and do sometimes migrate from meadow to meadow. That means a population isn’t necessarily doomed if, for example, the stonecrop in its home turf dies off; it also means that a freak summer snowstorm that kills off all the Apollos in one meadow can open up an attractive niche for pioneering migrants from a neighboring habitat. The Jumpingpound meadows form a beautifully linked miniature ecosystem, almost like a little Galápagos archipelago, with individual habitats separated by a few hundred yards of forest rather than miles of turbulent water. And that forest, while easy for humans to cross, is a real obstacle course for the butterflies. Their isolated populations and quick generational turnover mean that separate colonies can start to show genetic divergence with gratifying speed.
All these factors conspire to make the Apollo butterflies an inviting index species. They’re like a self-contained, easily studied test market for the investigation of ecological problems that may also turn out to bedevil longer-lived, wider-ranging, harder-to-handle animals. In a way, Roland explains, the Apollos are a little like grizzlies. Their habitats are shrinking and increasingly isolated. Will they be able to survive in the habitats they’ve got, and move if they have to?
The beauty of Roland’s research lies in its completeness: he can easily monitor large populations over a genetically significant time period while also tracking potentially fateful changes in their habitats. Unlike grizzlies, Apollo butterflies don’t get testy when you net them and don’t balk at having you write on them with a felt-tip pen. One of the reasons I started working on them, Roland admits, is that they’re easy to catch: relatively big, slow, and easy to spot. Much of Roland’s research depends on doing just that: ranging the meadows with his colleagues and students, catching the butterflies, marking them with identifying letters, and whenever they recapture a previously marked individual, recording the location and date. This mark-and-recapture technique allows them to monitor butterfly populations and track the broad outlines of their movements.
What it does not allow them to do, however, is monitor individual behavior. The technique can at most tell you where an individual butterfly starts and where it ends up, with perhaps the occasional good-luck interception in between. It can’t--as a grizzly’s radio collar can--track the nuances of individual wanderings, which can reveal significant patterns in an animal’s hourly and daily amble through its surroundings. And if you want to parse the secrets of a species’ behavior, you need to see it both in panorama and in miniature. What the species does, after all, ultimately rests on what the individual can do.
But how do you follow the travels of one two-inch butterfly? Manageable though the butterflies are when you net them, try shadowing an individual in the wild for more than ten seconds, let alone a whole day or week, and you’ll appreciate the difficulty. Apollo is no speed freak, but when active it constantly crosses paths with other identical-looking individuals, plummets into spiky labyrinths of dense alpine vegetation, and occasionally heads away in a beeline, vanishing into the light before it’s a hundred feet away. The slightly darker females are even more elusive than the males, more likely to hug the ground beneath the plant cover.
Clearly a radio collar would do the trick. An individual Apollo, however, weighs only about a hundredth of an ounce; you might as well outfit it with the pyramid of Cheops as with a conventional collar. But Roland, working with Graham McKinnon and Chris Backhouse, engineers at the Alberta Microelectronic Center in Edmonton, has come up with an ingenious solution: a super-lightweight, nearly invisible radar transmitter that’s about a thousandth the weight of the butterfly--about the equivalent of a wristwatch to a human. The tag doesn’t seem to interfere at all with the insect’s behavior; in tests it has stayed in place and functional for at least two weeks.
And it supplies a giant missing research link. For the first time, field-workers can scope out the potentially revealing secrets of an Apollo butterfly’s day. Ecologists now have an easy-to-manage model system that will let them test the mettle of a multipronged top-to-bottom ecological study--beginning with individual behavior, correlating that with the movements of the group, then seeing what happens to a whole population over multiple generations as its habitats shrink, expand, or change. If, as Roland thinks it will, such a study reveals significant secrets of how the species meets environmental challenges, that would justify the far more expensive, longer-term projects it would take to investigate bigger, longer-lived, wider-ranging, and endangered animals--like grizzlies.
The tag looks like an unusually frail strand of human hair, with a pinpoint-size diode attached to the middle of the three-inch-long superfine aluminum wire strand that serves as an antenna. Drop one on the floor and it’s gone forever. McKinnon and Backhouse assemble the tags from off-the-shelf components--the diode, for example, can be ordered from Hewlett-Packard, and the cost of the parts is only about $8. But the labor of wiring them together is highly specialized and finical, so they end up costing $30 each.
The technology the tag implements is also ingenious--a relatively recent invention called harmonic radar, commonly used in those bulky antishoplifting tags riveted to every sweater and T-shirt in the mall. The system’s been available for some years, and Roland credits the idea of tracking insects to colleague Henrik Wallin. Henrik showed me some tracking equipment he was using on beetles, Roland recalls. James Riley in England had the same idea and developed a relatively powerful stationary tracking system to follow the movements of bees. But being relatively burly, bees can handle heavier probes. I started wondering if we might get something really small for butterflies, Roland says. I approached Graham and Chris, and we started working on it in 1993.
The Alberta team’s innovation thus lay in paring the tags down to a butterfly-friendly weight and tracking them with a battery-powered, hand- held device that weighs about five pounds and fits easily in a day pack. It’s a standard unit, manufactured in Sweden and originally designed for search-and-rescue missions trying to find people buried in avalanches.
The tracking unit sends out a 1.7-watt microwave pulse at a 917- megahertz frequency, which bounces off everything it hits over the unit’s 150-foot range. According to Chris Backhouse, the physical principle is similar to what happens when you crank up the volume on your stereo. The distortion you hear comes from harmonics--grating overtones that vibrate at two to four times the frequency of the original note. In harmonic radar the same principle is at work, but with electromagnetic waves. The energy from the transmitter sets up a resonating frequency in the antenna wire on the probe, Roland explains, and the diode acts as a kind of one-way gate, absorbing the original frequency and converting it into the higher harmonic frequency, which then bounces back to the receiver. That means the tag-- powered only by the radiation striking it--returns a signal at 1,834 mhz, which the tracker is set to pick up. This neatly eliminates the barrage of 917-mhz noise bouncing back from every rock and tree trunk within range.
Graduate student Sherri Fownes is Roland’s acknowledged master butterfly wrangler and an adept at the art of attaching the probes to insects in the field. In the August afternoon sunlight, the operation looks like pantomime, so nearly invisible is the tag. A typical operation begins with Roland snaring a butterfly, a male. He pinions it gently in his two hands, wings flat, while Fownes scrapes a few hairs off its abdomen, using fine forceps. Then she plucks the tag from an envelope. The butterfly doesn’t seem to mind--at least, it stays calm and doesn’t struggle. With a blade of grass as applicator, Fownes dabs a minuscule drop of rubber cement on both the tag and the bug’s abdomen. She’s careful not to get glue on the thorax, which might cause the butterfly to entangle itself in the antenna wire. Finally she taps the probe into place, again with a blade of grass. She straightens a kink in the filament and blows on the butterfly gently a few times to dry the cement. Then Roland releases it.
As it flutters off toward the northwest, Roland grabs the charcoal-gray tracker. By this time the butterfly has disappeared into a crowd of other male Apollos fluttering in the sunlight in search of nectar and females. Roland begins arcing the unit around in various directions until a sudden high-pitched chirp announces the tag--just as the sun moves behind a cloud and all the butterflies, deprived of their energy source, sink to the ground.
Roland loses the signal at that point. Apollo butterflies are like rubber-band-driven toy airplanes: they draw their energy from the sun, so when it disappears they sputter to the ground in a torpor, extending their wings as they wait for it to reappear. Their dark dorsal hair and the dusky tint on their wings near the body are designed to absorb light energy efficiently.
As soon as the sky brightens, the butterflies rev up and head aloft once more. Roland, brandishing the tracker, eventually picks up the wired male and soon spies him heading uphill toward the east. The butterfly disappears into the forest at the northeastern edge of the meadow, remains incognito for a few minutes, then reemerges, now heading south. Random motion, a hunt for food, a quest for a female, or a purposeful route the key to whose itinerary has yet to be discovered? These are some questions radar tracking may ultimately answer, though Roland emphasizes that the technique is still in its infancy. Before it becomes routinely possible to document an individual Apollo’s habits and talents, Roland will have to address certain technical problems. For example, the portable tracker can keep you homed in on your butterfly, but you still have to follow it across rough terrain before it flies out of the unit’s 150-foot range. Every so often the unit picks up a false signal--maybe from a resonating crystal in a rock. And it would be useful (though not so healthy) to sit still and track the insect’s movements on a stationary radar screen instead of having to risk losing a speeding bug while chasing it over rough ground.
Roland’s group is also refining the tags. We’re working on a loop antenna that would be only half as long as the current one, he says. And we’d ultimately like to use a printed circuit on Mylar in place of a wire. That would make the tags cheaper, still less of a nuisance for the butterflies, simpler to mount, and easier to see when you drop one on the ground.
A couple of days after chasing the male, Roland tags another Apollo--this time a female--in another meadow about a mile to the east. Her movements prove a bit less erratic than the male’s. As soon as Roland releases her, the unit shows her heading southeast for about 50 feet, whereupon she executes a quick dogleg to the north and lands. Then, after a few minutes on the ground, she takes off, completes a graceful curve, and heads southwest toward the forest at the meadow’s edge, where she speeds beyond the tracker’s range.
We’ve been seeing more meadow-to-meadow movement in our butterflies than natural historians would have predicted, Roland says. They’re dispersing more easily than we thought they did. But how? The mark-and-recapture studies, he says, seem to suggest that the butterflies achieve their mobility by zinging along the high ridges that snake in and out among the meadows. But that may be an artifact, he adds, because we do all of our capturing on the ridge; we don’t search in the forest. Radio-tagging individuals will help determine whether, when they pack up for a new meadow, they prefer the ridge route or the forest route. If it’s the latter, the method may help explain how the insects manage to get through foreign territory--whether they just bumble along or have some undiscovered navigating technique.
Each of our techniques for studying them works well on a different scale, Roland explains. The challenge will be to link those scales. Mark-and-recapture studies give an overview: what’s happening to the population in each meadow; how often do marked butterflies make a bold foray through the intervening woods and reappear in a nearby meadow; what’s the longest foray an individual is likely to make? (So far in Roland’s study the distance champion is a male that eventually made its way from the meadow where it was first tagged to another well over a mile--and three intervening meadows--away.)
Our radar tracking has shown us that males follow two types of movement, says Roland. There’s a very local patrolling behavior--that’s presumably to encounter recently emerged females. But sometimes they also make much longer movements, anywhere from 800 to 1,200 meters. Maybe that happens when they decide their habitat isn’t good. Females behave differently, staying closer to the ground and fluttering less. When they do fly, their movements appear more directed, even deliberate, possibly because they have to be methodical in looking for stonecrop plants to lay their eggs on. Females may call the shots in effective migration when they decide to head into the woods for parts unknown, since the males must follow them to reproduce. But is that really how migration takes place? And if it is, what makes the female decide to bolt? Temperature, too high a density of butterflies, some hitherto unnoticed feature of the landscape, or some change in the stonecrop of her home meadow that makes her restless? Radar tracking, Roland thinks, may answer some of these questions as well.
In recent years the tree line in the Canadian Rockies has been rising; alpine meadows have been shrinking, the woods that divide them expanding. The change may be the result of global warming, or of human forestry policy: the suppression of natural fires may prevent the natural resetting of the tree line. Or it may be a periodic effect that ecologists don’t entirely understand. Whatever its cause, though, a change of that sort is sure to have implications for the long-term survival of the butterflies--and, by implication, for all species. What’s the biggest forest an Apollo can successfully migrate through? If the meadows continue to shrink, growing ever farther apart, at what point will the butterflies become unable to migrate and therefore vulnerable to extinction if, say, stonecrop vanishes from the meadow where they’re stranded?
To get definitive answers to such questions, you need the research equivalent of a zoom lens, homing in on individual behavior (diode tracking), pulling back to monitor the behavior of populations in a system of neighboring meadows (mark and recapture studies), then backing off yet further to see how the species is doing in a whole region. Nusha Keyghobadi, a graduate student in Roland’s group, is interested in that larger picture: she’s embarked on a DNA study of Apollo butterflies, not only on Jumpingpound Ridge but in a 30- to 40-mile radius around it. She clips off a harmlessly tiny piece of the insect’s wing, then analyzes it for microsatellites--short, repeated sequences of apparently nonfunctional DNA of the same type used for genetic fingerprinting in humans. She’s built a library of genetic markers that will ultimately enable her to uncover kinship even in widely separated populations, determining how closely related they are, how long ago they separated, maybe even correlating population movements with historical records of environmental crises--a prolonged spell of adverse weather, for example, or a cataclysmic forest fire.
It’s a well-established research technique, but one that’s never been used in conjunction with the tiny radar tag, and the possibilities for synergy are rich. If the diodes can give me information on the behavior and movement of individuals, Keyghobadi says, that would help me interpret my DNA data. If the radar tracking data reveal a large number of individuals coming and going between two meadows, that would confirm microsatellite data suggesting that the two populations were interbreeding.
Apollo butterflies are simple creatures in the sense that most of their behavior seems hard-wired into their brains: they’re not given to displays of temper or creative eccentricity. But severely constrained by biology though it is, their behavior is also apparently quite complex, modulated by the butterfly’s need to interact in different ways with its habitat. It doesn’t just bumble around at boring random in purely Brownian motion. For example, Roland says, we’re about to publish data that show Apollo butterflies tend to make slow, torturous movements across meadows but directional movements through forests. Something, in other words, enables them to speed efficiently through territory where they can’t survive and head efficiently toward the sunlit, stonecrop-rich meadows where they thrive.
What enables them to modulate their behavior remains a mystery. But the survival of their species ultimately depends on their ability to behave in the right way at a critical juncture. And our own fate in a sense depends on only a vastly more intricate elaboration of the same ability. They aren’t aware of it, but in their own way the wired butterflies of Jumpingpound Ridge may be contributing to a deeper understanding of the universal, enigmatic, and fateful dialogue between an individual’s behavior and its habitat.