Tardigrades are cosmopolitan micrometazoans that inhabit everything from marine and freshwater sediments to moss cushions, lichens and leaf litter. Although they are fundamentally aquatic animals, they thrive on land through a survival strategy called anhydrobiosis. When their microhabitat dries, a tardigrade folds its body into a compact barrel-shaped form known as a tun, losing nearly all of its body water and suspending its metabolism to an undetectable level. In this cryptobiotic state the animal tolerates desiccation, freezing and extreme temperatures, and it can revive within minutes of rehydration. What tardigrades cannot do well is disperse under their own power; their distribution across continents has therefore been attributed to passive transport, chiefly by water currents and wind. Wind dispersal has been demonstrated experimentally in Antarctica and Central Europe, and specimens of the genus Echiniscus have turned up in Greenland after hot, dry Foehn storms. Vertebrates contribute as well: tardigrades have been recovered from bird feathers and feces, from the carapaces of freshwater turtles, from mud rubbed onto trees by wild boars, and even from the footwear of Antarctic tourists.

Until now, however, no highly mobile invertebrate had ever been shown to serve as a tardigrade carrier. Enter the ants, an extraordinarily successful group of more than 14,000 described species that dominate terrestrial ecosystems and account for an estimated 30 to 45 percent of the planet’s entire insect biomass. Many ant species are arboreal, and in temperate Europe several of them nest inside oak galls, the woody, marble-like structures induced by cynipid gall wasps such as Andricus quercustozae and Andricus kollari. Once the gall wasp leaves through its escape hole, ants of the genera Crematogaster, Colobopsis, Temnothorax and Camponotus move in, converting the tree’s parasite-induced growth into a fortified domicile with a stable foothold on the plant. To find out whether these ants also relocate the microscopic fauna living around them, a team co-led by Daniele Giannetti, Ilaria Giovannini and Edoardo Massa of the University of Parma and the University of Modena and Reggio Emilia turned to a mixed oak forest near the village of Fornoli in Lunigiana, northern Tuscany. Galls, they reasoned, are ideal experimental units: self-contained ant nests that can be sealed off from the surroundings and later dissected to reveal exactly which passengers have arrived.

The study began with a field survey between March and May 2024. The researchers collected 129 galls bearing the telltale exit hole of the gall wasp from 40 oak trees spread over roughly three hectares. Every gall was sectioned, its outer and inner surfaces were scraped and rinsed in distilled water, and the suspensions were filtered through sieves with mesh sizes of 500 and 38 micrometers to capture tardigrades and their eggs. The ants were frozen, air-dried, identified under a stereomicroscope and individually inspected for hitchhikers. The census revealed that 83 galls were occupied by ants, 10 by spiders and 36 were empty, and that four ant species were present: Crematogaster scutellaris, by far the most abundant, occupying 40 percent of the sampled galls; Colobopsis truncata with 17 galls; Temnothorax italicus with 9; and Camponotus fallax with 3. Tardigrades surfaced in the internal material of five galls, all of them C. truncata nests, and never on the external surfaces of any gall. The team also found moss and lichen fragments inside several nests, including a lichen fragment in a T. italicus nest that contained a Macrobiotus specimen—nesting material the ants evidently use to plug and narrow their gall entrances.

Correlation alone could not prove delivery, so the team ran a multi-year exclusion experiment on two oak trees. In March 2022, 43 galls in which the gall wasp was still enclosed were protected with safety nets that blocked colonization by ants and other arthropods. A year later, after 22 galls showed the wasp’s exit hole, the nets were removed from twelve of them, which were left open to ant colonization, while ten were kept as ant-free controls, isolated with adhesive strips at the base of their branches. When the galls were dissected in March 2024, eleven of the twelve experimental galls held complete ant colonies: six of Colobopsis truncata, two of Camponotus fallax, two of Dolichoderus quadripunctatus and one of Crematogaster scutellaris. The statistical contrast was unambiguous. Tardigrade presence differed significantly between ant-colonized and ant-excluded galls, with a chi-square test with Yates correction yielding p below 0.01, and tardigrades were significantly associated specifically with C. truncata nests by a Mann–Whitney U test at p below 0.05. Five of the six C. truncata galls yielded six tardigrade taxa, and the nests contained eight live tardigrades alongside moss and lichen fragments. Most strikingly, scanning electron microscopy revealed desiccated Hypsibiidae tardigrades clinging to the abdomens of two C. truncata workers, one on the ventral and one on the dorsal surface, while in one C. scutellaris nest a Macrobiotus specimen was found on the underside of a worker’s abdomen.

Why Colobopsis truncata? Behavioral observations offered the clues. When these small workers walk, they lower their abdomens until the gaster scrapes across the substrate, whether moss cushions or bare bark, and the researchers repeatedly recorded droplets of water on both the upper and lower surfaces of their abdomens. Compared with C. scutellaris and T. italicus, C. truncata also grooms itself less frequently, and its brisk walking speed of roughly 2.4 centimeters per second means that a moist fragment of substrate grasped in the mandibles stays hydrated long enough to reach the nest. Laser distance measurements of marked workers showed foraging trips averaging 10.3 meters, with a maximum of 21 meters, for C. truncata, and an almost identical mean of 10.3 meters, up to 19 meters, for C. scutellaris. Moss samples taken from the ants’ trails on the trunks of both experimental trees harbored six tardigrade taxa, matching the diversity recovered from the galls and confirming that the ants patrol substrates densely populated with microscopic life.

To close the causal loop, the researchers brought three entire C. truncata colonies—queens, workers and brood—into the laboratory and installed them in standardized artificial nests held at 25 degrees Celsius and 50 percent humidity. Each colony’s arena contained a sterilized, autoclaved bamboo stick that served as an experimental nest, and at the opposite end sat what the team called a patrolling facility: a small petri dish packed with half a gram of dry-weight moss cushion collected near Modena and deliberately overpopulated with more than 150 additional tardigrades. The moss naturally hosted rotifers, nematodes and five tardigrade species—Macrobiotus cf. sapiens, Mesobiotus sp., Minibiotus sp., Ramazzottius sp. and Echiniscus sp.—at densities well within the 400 to 4,600 animals per gram routinely measured in natural moss. Ants were then allowed to explore for 24 hours before the entire apparatus was frozen at minus 20 degrees Celsius. The subsequent audit found tardigrades on the body surfaces of workers in two of the three replicate setups, with three individuals recovered from ant abdomens. Living and dead tardigrades and rotifers were retrieved from the bamboo sticks, the nest areas and the arenas, and dead nematodes were also recovered from the experimental nests. The ants, meanwhile, had visibly redistributed moss fragments to every corner of their new territories.

The researchers distinguish two modes of what they have named myrmecophoresy, the ant-mediated form of phoresy. In direct phoresy, a tardigrade clings to the ant’s cuticle, as documented by the electron micrographs; if drying proceeds slowly under humid conditions, the animal can desiccate gently into its survival tun during transport. In indirect phoresy, the ants carry substrate—water droplets, moss flakes and lichen chips—whose embedded meiofauna travels invisibly and can be deposited far from its origin. Either way, the measured rates rival established dispersal vectors. Wind studies recorded a maximum of 64 tardigrades per square meter over four weeks, about 16 animals weekly, while Antarctic fieldwork logged roughly 2.8 tardigrades per site per week. A land snail, the only invertebrate previously shown to carry tardigrades, covers ten meters in 48 hours, though tardigrades can survive even passage through its gut; an ant crosses the same distance in under seven minutes and then shuttles back and forth continuously, patrolling both ground and vegetation. Because humidity inside galls fluctuates with the weather, hitchhiking tardigrades can dry slowly enough to enter anhydrobiosis en route, and the galls themselves become new colonization sites—micro-refuges that could connect otherwise isolated moss patches, lichens and leaf litter.

Scale is what turns a curious observation into a potentially transformative one. With about 14,000 species, billions of individuals and an unrivaled presence in soil, bark, canopy and leaf litter, ants weave the terrestrial landscape into a three-dimensional transport network that no wind-current model can replicate. The authors caution that individual rides are rare—only a handful of tardigrades were ever seen on an ant in the wild—but infrequent transport combined with the colossal volume of daily ant traffic can accumulate into ecologically meaningful dispersal. The work, conducted within Italy’s National Biodiversity Future Center, raises questions well beyond Tuscan oak forests: soil-nesting ants may ferry tardigrades and even smaller organisms through the litter layer, and such pathways could matter for species persistence when habitat fragmentation and climate change make long-distance dispersal increasingly difficult. As the authors conclude, the myrmecophoretic dispersal of micrometazoans may be just the tip of the iceberg—an unexplored highway system that water bears have apparently been using all along, one patient ant ride at a time.