William A Shear. American Scientist. Volume 87, Issue 3. May/Jun 1999.

Over the course of the past 20 years I have been studying a group of milliped families distributed around the Pacific Rim. These forms have been found in New Zealand, Australia, Japan, eastern Siberia, western North America and Chile. As I studied these millipeds it became apparent that despite the broad expanse of ocean separating them, the various families were quite closely related. Indeed, their evolutionary relationships and geographic distribution fit well with the notion of a lost Pacific continent, an idea first proposed by Amos Nur of Stanford University and Zvi Ben-Avraham of Tel Aviv University in 1977. This “missing” land mass-called Pacificasupposedly rifted into fragments, which subsequently collided with the continents around the Pacific margin. If pieces of Pacifica remained above sea level, they could have carried plants and animals to the various sites along the Pacific Rim. The milliped distributions are otherwise quite difficult to explain.

Clues to the possible existence of a lost continent are only part of the allure that millipeds hold for scientists. The millipeds are among the largest terrestrial invertebrates, with certain forms that may be as thick as one’s thumb and up to 20 centimeters long. In tropical environments, where earthworms are scarce, millipeds may be the major soilforming animals (converting vegetable debris into humus), thus providing a vital link in the cycling of energy and nutrients. Millipeds may also prove to be useful sources of compounds for medicine and industry, since they produce an enormous array of chemicals that they use for self-defense.

Nevertheless, all the interesting work that could be done with millipeds is impeded by how little we know of their diversity and evolution. Only a few specialists at any one time in the past two centuries have worked on them. Although they are one of the most important components of the globe’s biodiversity-perhaps numbering 80,000 species worldwide (nearly twice the number of vertebrate species)-only about a tenth of these species have been named and described! The astonishing things we know about milliped behavior and diversity must represent only a small fraction of the whole story. Here I provide a brief introduction to the diversity and life-styles of these curious animals.

What are they?

Millipeds are arthropods, as are insects, crabs and spiders. In arthropod fashion, they possess an exoskeleton (containing the remarkable biopolymer chitin) and legs made up of tubular sections. As a group, the arthropods are richly blessed with legs, and the millipeds especially so. Yet the term “milliped” (which translates as “thousand-legs”) is not quite accurate. The current record stands at 750 legs, held by Illacme plenipes of California. (Oddsmakers are betting that the tropics conceal a true “thousand-legger.”) The bodies of most arthropods are divided into repeating segments, and millipeds are among the most obviously segmented. But the apparent segments of millipeds are not the real fundamental units of the body. Each “segment” (except perhaps the first four) represents two real segments that are fused (forming a “diplosegment”). In token of this peculiar arrangement, millipeds have not one, but two pairs of legs on each segment posterior to the first four.

Although it is abundantly clear that millipeds are arthropods, their evolutionary position within that phylum is not settled. Millipeds, centipeds and two lesser known arthropod classes were once conveniently grouped as myriapods, but careful cladistic analyses have revealed that this group may not be monophyletic. Some “myriapods” appear to be more closely related to insects than to each other. Other recent studies using molecular evidence place millipeds at the base of the arthropod family tree, suggesting that they may not be very closely related to any other arthropod group.

Whatever their phyletic relationships prove to be, the evolutionary story of the millipeds is a very old one. Indeed, the oldest fossils of landdwelling animals are millipeds, dating to more than 425 million years ago. Incredibly, those archaic forms are nearly indistinguishable from certain groups living today. Ancient burrows in Pennsylvania suggest that the millipeds may even have come ashore millions of years earlier. Although the early terrestrial environment may have offered no more greenery than some algal carpets, by about 375 million years ago forests had covered much of the continents, allowing millipeds (and other animals) to diversify. By 300 million years ago an array of milliped types had appeared in the coal forests of the Carboniferous Period. These included foot-long crawlers with enormous defensive spikes and burrowing forms that could roll up into a seamless sphere. Today taxonomists recognize about 15 distinct orders of millipeds.

A Milliped’s Life

In North America and Europe, one comes to think of millipeds as the quintessential forest litter animals. And, in fact, millipeds are most abundant and diverse where damp leaves have collected over a humusy soil, preferably derived from limestone. The calcium content of the soil is important because the milliped exoskeleton (like that of the crustaceans) is strengthened with calcium carbonate. The large number of legs gives millipeds considerable pushing power, allowing them to force their way into rotted logs and soft soil. Here in their moist, cool haven, millipeds feed on the decaying leaves and twigs that have fallen from the trees above.

There are, however, a number of species that do not fit this stereotype. Ecologist Cliff Crawford of the University of New Mexico has found millipeds in the harsh deserts of the American Southwest, where they hide in ancient lava flows, emerging to feed when the rains dampen the dead leaves and twigs of sagebrush. Jerry Payne of the U.S. Deparatment of Agriculture and Richard Hoffman of the Virginia Museum of Natural History have described carnivorous and even predatory habits in several milliped species, including one North American parajulid that appears to specialize on insect pupae. Joachim Adis of the Max Planck Institute for Limnology and his colleagues discovered that many millipeds spend months in the trees during the periodic flooding of the Amazonian forests. Cave biologists often find that millipeds are dominant forms of life deep under the earth. Here they may assume ecological roles that are not possible for them on the surface. Henrik Enghoff of the Copenhagen Zoological Museum has described a species that spends much of its time submerged in cave streams, feeding by scraping or filtering bacteria with mandibles modified for the purpose. No aquatic or filter-feeding millipeds are known in surface habitats. However, certain Amazonian species can survive months of floods by breathing air bubbles that are trapped on their bodies. Despite the ability of some species to tolerate months of submersion, there are no millipeds in the sea, and only one European species has adapted to life in the intertidal zone.

There is little evidence that the milliped lifestyle poses a problem for human beings. A small European species, the “spotted snake milliped,” may feed on living roots, damaging certain crops such as potatoes and sugar beets. And a few species simply annoy by virtue of their large numbers. For example, North American populations of Oxidus gracilis (originally from Southeast Asia) frequently invade homes. If stepped on, they can release evil-smelling defensive chemicals. This species also takes part in what appear to be mass migrations consisting of millions of individuals. Their crushed bodies can impede traffic on roadways and railways, and in some instances the migrants may form heaps several inches high. Despite the description of these aggregations as “migrations,” little is known of their true function and the conditions under which they take place.

A Leg Up on the Competition

Lying at the base of the food chain, millipeds are a potential resource for predatory arthropods and vertebrates. Consider one predator, a small species of ant (Gnamptogenys ingeborgae) in Colombia, described by Bill Brown of Cornell University. Gnamptogenys appears to specialize on polydesmid millipeds, running them down and attacking them with a paralytic sting. The milliped is then slung over the ant’s back and hauled to the nest, where the head is removed and ant larvae clean out the diplosegments, one by one. Similar tactics are employed in California by the larvae of glowworms (Zarhipis integripennis) studied by California naturalist Darwin Tiemann. Finding a milliped, the glowworm gallops alongside its prey, finally seizing it, flipping the milliped over and biting through the ventral nerve cord. The successful predator eats its way through the paralyzed milliped from front to back, shrugging off diplosegments as it goes. No vertebrate predators are known to specialize on millipeds, but toads, lizards, birds, rodents and at least one African mongoose have been observed feeding on them.

Although their enemies may be formidable, millipeds are not entirely defenseless. For many millipeds the first line of defense is mechanical, relying on their hardened exoskeleton and an ability to roll into a tight spiral or sphere that conceals the vulnerable head and legs. Millipeds that roll up for defense are apt to be very smooth, making it difficult for predators to get a grip on them. This defense doesn’t always work, however. Tom Eisner of Cornell University watched a mongoose overcome the defense of an African giant pill milliped, the size of a golf ball when enrolled. The mongoose picked up the milliped in its forepaws and threw it between its hind legs, smashing it against a rock.

Other millipeds, obscure forms of the order Polyxenida, have adopted a porcupine-like defense, evolving a coat of stiff bristles. Eisner discovered that polyxenids turn tail to attacking ants and wave tufts of easily detached bristles that look like miniature grappling hooks. Efforts by the ant to get the bristles loose simply cause more entanglement, while the milliped ambles away. With enough bristles, the ant may become completely immobilized and die.

Many milliped species do not have a mechanical defense and so rely on chemical warfare against their predators. (However, even species with a good mechanical defense may use chemical defenses.) Eisner and his colleagues have found that millipeds produce a wide array of noxious compounds, including quinones, cresols, phenols and, most remarkably, hydrogen cyanide. Cyanide is produced by segmental glands in the flat-backed millipeds of the order Polydesmida (among the most common millipeds). The glands of these millipeds have two parts, including a portion that produces a cyanophore (an organic carrier that renders the cyanide temporarily harmless) and a reaction chamber into which the cyanophore is dumped. The walls of the reaction chamber produce an enzyme that separates the cyanide from the cyanophore. The hydrogen cyanide then escapes through a pore and can kill an attacking insect or sicken a predatory toad, which soon learns to leave millipeds alone. Millpeds appear to survive their own gas attacks by simply closing their breathing pores and reducing their metabolism until the danger passes. Curious people can sample this defensive strategy by lightly shaking a polydesmid in a closed hand. The amount of cyanide is too small to pose a danger, but a benzaldehyde component released with the odorless cycanide can be detected as a strong smell of bitter almonds. The secretions of some tropical millipeds are powerful enough to raise blisters and erode human skin. These chemicals have not been characterized, and these millipeds should not be picked up!

Perhaps the most sophisticated chemical defense is wielded by the European species Glomeris marginata. Jim Carrel of the University of Missouri found that their segmental glands produce quinazolinones, a class of substances that includes the synthetic drug Quaalude (a powerful sedative). Attacking wolf spiders exposed to this defense simply go to sleep, some for days. Additional components of the secretion have an antifeedant effect-the predator cannot bring itself to eat its preyprotecting the milliped until the sedative can work. Carrel also found that Quaalude itself did not sedate spiders, suggesting that the milliped secretion specifically evolved for the purpose.

Given their chemical defenses, it is not surprising that many millipeds have aposematic coloration-exhibiting bright reds, yellows and orangeswhich advertises their distastefulness or downright lethality. Vertebrate predators, such as toads, quickly learn to associate the bright pattern with a bad experience and actively avoid similar millipeds in the future. Certain North American polydesmids are strikingly beautiful, with brilliant shades standing out against a glossy black background. In California, the cyanide-producing Motyxia species are brilliantly luminescent at night, shining greenish-white like “stars in the sky” as one observer put it. Rowland Shelley of the North Carolina Museum of Natural History speculates that this is the nighttime equivalent of aposematic coloration.

It seems that certain other soildwelling animals may have evolved color patterns similar to those of distasteful millipeds (a phenomenon known as Batesian mimicry). In such a situation, a harmless animal is protected by its similarity to a dangerous one that advertises its noxiousness by shape or color. Herpetologist Laurie Vitt of the University of Oklahoma has published images of lizards and worm snakes that closely resemble large millipeds that share the same habitat.

Bizzarre Sex

Courtship among millipeds has been little studied, but complexities are suggested by the fact that males often exhibit a wide array of secondary sexual characteristics, including strong modifications of the legs and head. Gland openings of unknown function are also common and may produce pheromones. In European Chordeuma, the males have special glands on their backs that produce a secretion on which females feed prior to mating. Male sphaeriotherds, tropical versions of glomerids, produce sound during courtship by rubbing the back edge of the body against an enlarged rear leg.

Sex in millipeds seems bizarre from a vertebrate perspective. In most species, mating occurs by indirect copulation, in which the male and female genital openings never make contact. Males arch their bodies so that the penes, located near the bases of the second pairs of legs, come in contact with strongly modified legs (gonopods) on the seventh or eighth diplosegment. These gonopods are then used to transfer either seminal fluid or packets of sperm called spermatophores to the female, who stores them in seminal receptacles near the front end of her body. At egg-laying time, the sperm is released and fertilizes the eggs.

Sexual selection of several types has caused the form of the male gonopods to evolve rapidly, and the details of their shapes are specific to each species. In most millipeds, the gonopods no longer resemble legs, and may be truly baroque in their complexity, with flagella, brushes, knobs and hooks. We are only beginning to understand the function of these ornaments. Some are evidently used to clear out sperm from the receptacles of females that have previously mated; the old sperm are then replaced by those of the current mate. Other parts of the gonopods may serve to stimulate the female and make her receptive to mating.

One of the most unusual mating mechanisms is found in glomerid millipeds. In this group, gonopods are not present, but the last pair of male legs is modified into a set of strong claspers, used to hold the female in place. The late Ulrich Haacker found that the males mold soil into a small cup into which they ejaculate sperm. The cup is then passed down the array of legs to the female, who is firmly held by the enormous claspers. This is the only animal known that uses a manufactured tool to mate!

Eggs are usually laid in the soil. Female Narceus, the largest North American milliped, pass the eggs down the length of the body from the gonopore near the head. Then they are taken into the rectum and coated with fecal material, forming a pellet that disguises the egg and protects the developing embryo. At least some chordeumatid millipeds can produce silk, making toughly woven bags for their eggs. Having thus provided for their young, most millipeds simply walk away. Although brood care by females has never been observed, males of the Appalachian species Brachycybe petasata have often been seen embracing egg masses or broods of young in their coils, a rare case of male brood care.


I began this article with the notion that millipeds may have something to tell us about the geologic history of our planet. Such geologic import for an invertebrate is not without precedent. It is an interesting episode in the history of science that the geographical distribution of earthworms played a key role in the recognition of continental drift. Closely related earthworm species were found in widely separated regions (such as South Africa and Chile), and it was realized that since earthworms could not cross the sea, their regions of distribution must once have been contiguous. The breakup of the supercontinent Pangea neatly explained the puzzle. Like earthworms, millipeds are an ancient group of animals with very limited powers of distribution. Most require moist, mild habitats, so seas and unforested regions act as barriers. These qualities suggest that millipeds might be ideal subjects for biogeographical studies, since vicariance events (such as the rifting of continents and island colonizations) would be the most important factors in their distribution. Regrettably, such studies are hindered because we know so little about the diversity and distribution of millipeds.

To help remedy this situation, the National Science Foundation, through its program to Preserve and Extend Expertise in Taxonomy (PEET), has made a five-year grant to Petra Sierwald of Chicago’s Field Museum and me to study the classification and phylogenetics of millipeds, describe new species, catalogue museum collections and train students in systematics and evolution. Our colleagues, Richard Hoffman and Rowland Shelley, will publish updated lists of all known milliped genera and families, and a checklist of all the milliped species already named from North America, Mexico and West Indies. We expect not only to uncover new aspects of milliped biology, but also to place taxonomic tools such as checklists and identification keys in the hands of ecologists, behavioral biologists and biogeographers, making it possible for them to use millipeds as model organisms in their studies.