“I’d been emailing my friend back and forth for several months, planning the trip. He was sure I wouldn’t be disappointed. And then one morning I left. Getting to where he was in Japan took a long time… when he and his students came to pick me up, I hadn’t slept for 30 hours since leaving NYC. The jetlag was tough—my mind was spinning and all I could think about was curling up in a ball. But then he drove out to a remote clearing in the forest. It was pouring with rain; I was umbrellaless and drenched. Then I turned over this rock, the ants cleared and it was just there, walking around—you could see it with your own eyes… and pick it up in your hands! Suddenly they were everywhere… I can’t explain exactly how it felt, but I should tell you, it’s possible that nothing more remarkable may ever happen to me.”

We live on a planet where subjugated minions turn the cogs. Naive to the cold efficiency of their existence, they’re subject to all manner of manipulation and exploitation. But this world could not be further from a depressing hell. It’s filled with profound beauty, bursting with curiosities, and once you know it exists, it pulls you in, enticing you to explore it further. For this isn’t the human world, but a parallel realm where microzoological enigmas parasitize the societies of insects.

Few people know about this world. Entomology luminaries like E. O. Wilson tell us routinely that if we don’t own and run the planet, then social insects do; global ant biomass approximates that of humans, and like us, the ecological pressures ants exert on habitats determine what and how much of which species live where. What’s less well advertised is that ants are victims of their own success. Pervasive habitat-sculptors and resource-hoarders, their nests are home to not just ants, but a diverse menagerie of other organisms that siphon off supplies, prey upon the colony, and live comfortably below its radar with biological stealth technology.

Meet the myrmecophiles: “ant-loving” creatures that rank amongst the bizarrest animals known. Their weird shapes and forms are matched only by their strange behaviours—all part of elaborate mechanisms of deception and mimicry that integrate these creatures seamlessly into colony life. Their biologies are cryptic and largely unexplored; most are scarce or hard to find, and seeing them in nature is a rare privilege. To find them, you could do what one of us did, and take a trip to a rain-soaked mountain in Japan.

Overturning rocks to expose the ant colonies, you’ll eventually find your quarry: a 2 mm-long, reddish-brown and very odd-looking beetle. This is Diartiger fossulatus, the consummate obligate-myrmecophile. Fully dependent on ants for its very survival, Diartiger walks calmly through the chaos of the disturbed nest. Its peculiar form and unhurried gait attest to a bewildering lifestyle, and seeing it in nature for the first time, you can’t help but wonder how it ended up doing what it does, looking the way it looks.

It seems inexplicable; fewer places could be more hostile for small insects than ant nests. They’re policed with brutal efficiency, using a pheromone code of nestmate recognition. Yet, while most intruders are readily dismembered and consumed, myrmecophiles like Diartiger have deciphered this code and made a living inside ant nests for millions of years. If you can pull it off evolutionarily, adopting this lifestyle brings obvious benefits: access to a climate-controlled home, one that’s well defended and devoid of predators. Then there’s the food—lots of it, be it stored booty, the ants or their brood, or, for the most highly integrated species like Diartiger, liquid food regurgitated directly to your mouth by the workers themselves.

So why don’t more arthropods join the myrmecophile club? Certainly, there are myrmecophilous butterflies, bugs, crickets, millipedes, silverfish, spiders; many different taxonomic groups include a myrmecophilous species or two. But given the species richness of each of these groups, such cases are statistical quirks. For most organisms, the path to becoming a myrmecophile must simply be too challenging, the ant fortress too impenetrable.

An inordinate fondness for myrmecophily

For two groups of organisms however, the same cannot be said. They have jumped the evolutionary hurdle dozens, perhaps hundreds of times, and in so doing, have produced some of the most extraordinary cases of ant exploitation and dependency known. Studying them, and their inherent bias towards myrmecophily, may tell us a lot about this mode of life and how it evolves.

They are two vast beetle subfamilies: Pselaphinae (SEH-LAH-FIN-EE, which includes Diartiger; >9000 species) and Aleocharinae (ALEE-OKA-RIN-EE; >13,000 species)—both members of the most species-rich beetle family of all, rove beetles (Staphylinidae).

There are 58,000 named species of rove beetle, but the true number out there is several fold this total. For the most part, they don’t look like typical beetles: the toughened wing cases (elytra) don’t cover the whole abdomen, leaving most of the abdominal segments exposed. In aleocharines, the exposed segments are flexible and telescoping, and the body form rather narrow; in pselaphines, the abdominal segments are rigid and tightly connected, the body shape more compact.

These groups, with their inherent propensity to jump the ant-shaped hurdle, are particularly illuminating if we want to get out how this lifestyle—with all its contrivances and idiosyncrasies—comes into existence. We can treat each instance of myrmecophily these beetles show as a kind of independent “experimental replicate”. With enough replicates, a picture emerges of what usually happens as these beetles crawl along the evolutionary path, from free-living to an advanced state of myrmecophily.

Evolution of the (predictably) bizarre

What both pselaphines and aleocharines show is that species that exhibit an advanced state of myrmecophily tend to demonstrate analogous elaborate morphological and behavioural modifications. Although moving to life inside ant colonies massively distorts how these beetles appear, the evolutionary endpoints are somewhat predictable. The bizarre, it turns out, can be the norm.

Here’s an example. In pselaphines, the tribe Clavigerini is large (over 350 species, in ~100 genera) and comprised entirely of highly integrated ant colony inhabitants. They rank amongst the most evolutionarily derived myrmecophiles known, with strange fusions of segments within the abdomen and antennae. Glands cover their bodies, and they glisten with oily secretions. They’re fed directly by the ants, via reduced mouthparts that are recessed inside the front of the head; they’re even carried around the nest by the workers, as if they were brood. Diartiger is a member of this tribe.

With a couple of exceptions, all Clavigerini have structures called “trichomes”: bundles of hairs at the at the base of the abdomen and tips of the elytra, that exude chemicals which ants seem to find attractive and tasty. Trichomes likely act as “wicks”, conducting secretions from nearby glands. The chemicals are mysterious, but might be appeasement substances that encourage ants to “adopt” rather than attack the beetles. According to our own unpublished, DNA-based work on pselaphine evolutionary relationships, trichomes have arisen independently an additional four times (at minimum) across the Pselaphinae tree of life. And in two cases, the trichomes have appeared at exactly the same positions on the body as in Clavigerini.

Attapsenius, a guest of the fungus gardens of South American leaf cutter ants. It has trichomes like members of Clavigerini.

Attapsenius, a distantly related genus found solely within the fungus gardens of leaf cutter ants in South America, serves to demonstrate. The genus is a few paces behind clavigerines on the myrmecophile road, but you can see how it’s converging on the same body plan as Diartiger. It has small trichomes in the same positions; what’s more, its antennal segments are becoming consolidated (normally, pselaphine antennal segments are very distinct, but here the segments are tightly pressed together). You can’t tell, but its mouthparts are also tiny, approaching the state found in Clavigerini.

In Songius, a Chinese genus recently described by our friend Zi-Wei Yin, trichomes are again present, but this time sprouting from the sides of the head! Note the shiny, oily surface of this beetle—again similar to Clavigerini.

To maintain permanent acceptance in the nest, these different advanced myrmecophiles seem to have converged on the same solution to the problem. That the same regions of the body have been co-opted suggests that some shared ancestral precursor structure may exist there—a gland, some bristles—which can be readily modified into a device for myrmecophily. So, even when creating the bizarre, evolution seems to follow the path of least resistance.

While these pselaphines show a sophisticated level of ant exploitation, the evolutionary path taken revolves around developing trichomes on their robust and compact bodies. In contrast, aleocharines are a wholly different lineage of rove beetles, and with their flexible bodies, have run a totally different evolutionary track to advanced myrmecophily.

One theme aleocharines repeatedly exploit is the evolution of a myrmecoid or “ant-formed” body plan. This is a phenomenon unique to myrmecophiles of that famous poster child of tropical biology: army ants. Famous for forming endless trails that spread through the forest, devouring everything in their path, army ant colonies contain astronomical numbers of workers and an endless, cyclical production of eggs. Although menacing, they’re caloric treasure troves ripe for exploitation. And among the plethora of myrmecophiles that turn up to the party, myrmecoid rove beetles are among the most numerous, and by far the most greatly modified for this lifestyle.

They walk with the ants, and the resemblance is striking; the abdomen is constricted to resemble the narrow waist of an ant, and their segments and limbs are sculpted into forms intricately similar to their hosts. In fact so similar are ant and beetle that an evolutionary tale of mimicry to deceive the vicious host ants is tempting.

Digging into the evidence deeper though, we’re left wondering how looking like an ant should help these beetles at all. For army ants are, for all intents and purposes, blind. They live in a world dictated by odours, pheromonal signals and trails that’s foreign to our visually-biased view of life. It may well be that myrmecoid aleocharines are not trying to deceive the ants with their appearance, but birds and other larger animals.

Many vertebrate predators have learned to scavenge forest floor critters that the roaming army ants scare out of hiding. But these opportunists seldom target the ants themselves, and so the beetles slip by, unnoticed (Kistner and Jacobson 1990). If you’re lucky enough to spend some time with a myrmecoid army ant myrmecophile, you’ll see it grooming the workers, procuring colony-specific odours that they rub over themselves. It’s maintaining a disguise of chemical mimicry, diverting unwanted attention from the ants. Myrmecoid rove beetles are essentially playing two dangerous games of deception: one with the outside freeloaders, the other with their marauding hosts.

Many different evolutionary lines of Aleocharinae have converged on this same ant-mimicking body plan. What is more, although the “classical” army ants found throughout the tropics—Eciton, Dorylus, Aenictus—are closely related, a similar behavioural syndrome has evolved in some distantly related ant genera, such as Letopgenys (Witte and Maschwitz 2000). Remarkably, each such occurrence comes with its own dedicated set of myrmecophilous aleocharines, some of which also have a myrmecoid body form (Hlavac and Janda 2009 provide a beautiful example). Aleocharines have also integrated into termite societies many times, repeatedly evolving a termite-looking “physogastric” body. Yet again, a demonstration of the predictably bizarre.

Preadaptations: breaking the code

With so many cases of myrmecophily, you have to wonder what it is that biases Pselaphinae and Aleocharinae to this ant-centric way of life. Free-living, non-myrmecophilous members of both subfamilies mostly inhabit tropical and temperate forest floors. They abound in leaf litter and soil where ants roam; some are also found up in the tree canopy—again, a place with lots of ants.

Sharing the same habitat probably makes the shift to myrmecophily much easier, but can’t itself explain the propensity to do so. After all, a great many other groups of organisms frequent the same parts of the forest as ants, and these don’t show the same bias to myrmecophily as aleocharines and pselaphines.

That leaves features of the beetles themselves: morphology, behaviour, perhaps even biochemistry or physiology—any characteristic that helps them avoid attacks, and ultimately, empowers them to break the colony code of intruder detection. What these traits—these preadaptations—are, we cannot be sure of at this time, but let us speculate: we suggest that both aleocharines and pselaphines possess preadaptations to myrmecophily engendered by their respective defence mechanisms.

Most aleocharine tribes with myrmecophilous members possess a defensive gland close to the tip of the abdomen. The gland puffs out a chemical to deter would-be predators, giving the beetle a window to escape. In some myrmecophilous aleocharines, this gland has been modified to produce different kinds of chemicals that appease or attract the ants, allowing the beetle to gain acceptance within the ant society (Stoeffler et al. 2011).

One problem with this explanation is that other groups of rove beetles also have defensive glands, and don’t show the same myrmecophily bias as aleocharines. It might be that, for genetic reasons, the glandular chemistry of aleocharines is easier to reprogram from defence to appeasement than it is in other rove beetles. In some myrmecophilous aleocharines, including many advanced myrmecoid species, the number and types of abdominal glands have multiplied, presumably as part of elaborate colony integration mechanisms. Perhaps compared to other rove beetles, aleocharines innovate this way with relative ease; they may be free from some kind of developmental constraint, and so evolve new glands in novel positions on their bodies quite readily.

In pselaphines, the preadaptation is even less clear. Defensive glands exist in some groups (Newton and Thayer 1995), but many myrmecophilous species don’t have them. Instead, the unique body plan of pselaphines might be the preadaptation. Pselaphines aren’t just more compact than other rove beetles; the cuticle is also especially thick, and internally buttressed by a kind of endoskeleton made of strong, chitinous scaffolding (Ohishi 1986; Chandler 2001). These features probably enable pselaphines to withstand physical compression in soil as they force their way through dense or heavy substrates.

Their distinctive physical form may bring an added advantage: it means pselaphines are essentially armoured, and conceivably pretty good at surviving encounters with ants. Indeed, amongst pselaphines, tribes where the body is relatively less “armoured” and compact tend not to evolve myrmecophily. In North American Batrisodes, the beetles have been seen to survive attacks from host ants (Park 1935). Moreover, it’s possible that these brawls serve a purpose, allowing Batrisodes to pick up the colony odour and go undetected inside the nest. One small scuffle for a beetle, one giant leap along the path to myrmecophily.

Myrmecophilology: a subject ripe for exploration

Myrmecophiles are undeniably fascinating, and we want to emphasize that precious little is known about them: the lives they lead, their evolutionary history, indeed the entire biological phenomenon is prime scientific real estate. We touched on behaviours and morphological features that might integrate these insects into colony life, but for the most part, information is scarce. At even the most superficial levels, countless questions are unanswered: what’s the purpose of trichomes? What do they secrete? Are the substances they produce the same in Diartiger, Adranes, Attapsenius and Songius? From what theorised ancestral precursor structure did they evolve? The more you look at these creatures with their curious structures and behaviours, the more they lure you in, begging you to make sense of them.

The fact we can see evolution apparently repeating itself in instances of myrmecophily suggests that there may not be an infinite number of ways to integrate into an ant colony. This kind of approach, comparing diverse species, is, we think, a particularly good way to study myrmecophily. It may help identify preadaptations that make certain groups so prone to sacrifice a free-living existence for a myrmecophilous one. And it may yield generalizable “rules” that govern how myrmecophiles evolve once they’ve broken the code.

Broader scale ecological studies are also needed. Ant societies show different levels of size and complexity: those with derived or “complex” social structures are able to operate larger, “more successful” colonies. But like economic systems in humans, as colony size and complexity increase, so too do the number of loopholes that outsiders can exploit. Primitive ants with their small colonies house few, if any myrmecophiles, whereas elaborate and extensive colonies like those of army and leaf cutter ants can accommodate hundreds (Navarrete-Heredia 2002; Rettenmeyer et al. 2010). The evolutionary pressures these myrmecophiles exert on their hosts are unknown. Ants may be victims of their own success, but to what extent?

Finally, we think that the astonishing numerical diversity of pselaphines and aleocharines, and their predisposition to myrmecophily are no coincidence. When ants became ecologically dominant, they radically changed terrestrial ecosystems and the niches insects could occupy. As Holldobler and Wilson (2005) put it, ants “dominate the central, more stable areas of habitats, whereas solitary insects are best able to flourish in the peripheral, more ephemeral areas”. Not so for aleocharines and pselaphines. They radiated extensively in the very habitats dominated by ants, we think by exploiting the same defensive preadaptations that make these beetles so prone to myrmecophily. And today, they are two of the largest and most abundant animal groups on the surface of the planet.

It’s time we updated the mantra: if we humans don’t own and run the planet, then ants, aleocharines and pselaphines most certainly do.



Chandler D. S., 2001 Biology, Morphology and Systematics of the Ant-Like Litter Beetles of Australia. Memoirs on Entomology, 15. Associated Publishers.

Hlavac P., Janda M., 2009 A new genus and species of Lomechusini (Coleoptera: Staphylinidae, Aleocharinae) from Papua New Guinea associated with ants of the genus Leptogenys. Zootaxa 2062: 57–64.

Kistner D. H., Jacobson H. R., 1990 Cladistic analysis and taxonomic revision of the ecitophilous tribe Ecitocharini with studies of their behaviour and evolution (Coleoptera, Staphylinidae, Aleocharinae). Sociobiology 17: 333–480.

Navarrete-Heredia J. L., 2002 Beetles associated with Atta and Acromyrmex ants (Hymenoptera: Formicidae: Attini). T Am Entomol Soc 127: 381–429.

Newton A. F., Thayer M. K., 1995 Protopselaphinae new subfamily for Protopselaphus new genus from Malaysia, with a phylogenetic analysis and review of the Omaliine Group of Staphylinidae including Pselaphidae (Coleoptera). In: Pakaluk J, Slipinski SA (Eds.), Biology, Phylogeny, and Classification of Coleoptera: Papers Celebrating the 80th Birthday of Roy A. Crowson, Muzeum i Instytut Zoologii PAN, Warszawa, pp. 221–320.

Ohishi H., 1986 Consideration of internal morphology for the taxonomy of Pselaphidae. Papers on Entomology Presented to Prof. Takehko Nakane in Commemoration of his Retirement: 1–21.

Park O., 1935 Beetles associated with the mound-building ant, Formica ulkei emery. Psyche 42: 216–231.

Rettenmeyer C. W., Rettenmeyer M. E., Joseph J., Berghoff S. M., 2010 The largest animal association centered on one species: the army ant Eciton burchellii and its more than 300 associates. Insectes Sociaux 58: 281–292.

Stoeffler M., Tolasch T., Steidle J. L. M., 2011 Three beetles—three concepts. Different defensive strategies of congeneric myrmecophilous beetles. Behav Ecol Sociobiol 65: 1605–1613.

Wilson E. O., Hölldobler B., 2005 Eusociality: Origin and consequences. Proceedings of the National Academy of Sciences 102: 13367–13371.

Witte V., Maschwitz U., 2000 Raiding and emigration dynamics in the ponerine army ant Leptogenys distinguenda (Hymenoptera, Formicidae). Insectes Sociaux 47: 76–83.


The authors thank Taku Shimada (Tokyo; and Zi-Wei Yin (Shanghai Normal University; for allowing use of their images.