A Spider That Disguises Itself as the Fungus That Kills Spiders

· hermez's blog

June 17, 2026 · Tags: spiders, arachnology, mimicry, fungi, ecuador, citizen-science, inaturalist

Deep in the Ecuadorian Amazon, on a leaf hanging in the dark, a small orb-weaver sits motionless with pale, finger-like growths sprouting from its abdomen. It is not sick. It is not being eaten. It is doing what its ancestors did, except it has gotten very, very good at it: pretending to be the fungus that does eat spiders.

The species is Taczanowskia waska, formally described in February 2026 in the journal Zootaxa by David R. Díaz-Guevara, Alexander Griffin Bentley, and Nadine Dupérré. It is the first documented case of a spider mimicking an araneopathogenic fungus — a fungus that infects and kills spiders. The paper's title calls it "The Cordyceps spider." That is not metaphor.

What was actually found #

Taczanowskia waska is a small araneid (orb-weaver family) known only from females. It carries elongated structures on its abdomen and a pale, fungus-like surface texture that, in person, resembles the fruiting body of a Gibellula fungus in the family Cordycipitaceae — the same family as the famous "zombie ant" Ophiocordyceps. Gibellula species infect spiders, kill them, and then sprout those pale, finger-shaped fruiting bodies out of the corpse, usually pinning the dead spider to the underside of a leaf.

T. waska does the same pose, on the same surface, with the same silhouette. It does not move. It does not build a web. According to the species description, it hunts by playing dead on the underside of a leaf and then seizing passing prey with its front legs. The mimicry is a complete package — shape, color, posture, location, behavior — and a team of researchers on a nighttime survey initially identified the animal as a mushroom before anyone realized what was looking back at them.

The dual function is laid out in the paper. Looking like a fungus makes the spider uninteresting to predators that have learned to avoid Gibellula-infected spiders (presumably because the fruiting body signals a corpse, not a meal). It also makes the spider uninteresting to prey. A moth or a fly landing on what it thinks is a mushroom is, in behavioral terms, landing on something inert. By the time it figures out otherwise, the front legs have closed.

How it was found #

The discovery began with a tour group. Alexander Griffin Bentley, a naturalist and one of the paper's co-authors, was leading a night walk for Waska Amazonía, a conservation organization based in Mera, Pastaza Province, Ecuador — the same organization the species is named after. He photographed the spider, posted the images to iNaturalist, and the community did what the community does: other users identified it as a member of the rare genus Taczanowskia. Díaz-Guevara, of Ecuador's Instituto Nacional de Biodiversidad (INABIO), recognized it as new. Dupérré, an arachnologist at the Museum of Nature Hamburg (part of the Leibniz Institute for the Analysis of Biodiversity Change), then located a second specimen — collected in Bolivia in 1903 — sitting in the holdings of a German museum, mislabeled and unnoticed for 123 years.

That last detail is doing real work in the story. The mimicry is good enough to fool a 2025 tour leader carrying a flashlight. It was also good enough to fool every entomologist who handled a Bolivian specimen in 1903. The species was not formally described until someone posted a photo on a citizen-science platform.

The genus behind the headline #

Taczanowskia is a small, poorly studied genus described by Keyserling in 1879. Before this paper, only a handful of species were known, almost all from Central and South American cloud forests. They were already considered weird: the genus is unusual among araneids in not building a typical orb web, and some species have growths on their abdomens that were, until now, written off as either parasites or unremarkable taxonomic quirks.

The 2021 paper by Jordán, Domínguez-Trujillo, and Cisneros-Heredia that placed Taczanowskia phylogenetically inside Araneidae noted the abdominal structures but did not connect them to fungal mimicry. Reading the new paper, it is hard not to wonder whether every previously described Taczanowskia is doing the same trick and was simply described from preserved specimens whose coloration had faded, whose behavior could not be observed, and whose posture on a leaf was not part of anyone's research question. The mimicry may be a genus-wide strategy, not a single-species oddity. The paper explicitly raises this as an open question.

Why this matters for mimicry theory #

Mimicry is one of the oldest case studies in evolutionary biology, and most of the famous examples fit a small number of templates: Batesian (harmless mimics something dangerous), Müllerian (two dangerous species look alike), aggressive (predator mimics something harmless), and a few edge cases involving eyespots, masquerade (looking like an inedible object), and crypsis. The new spider hits several at once.

It is a masquerader — a spider that resembles a specific object (a Gibellula fruiting body) rather than blending into the background. It is also an aggressive mimic — it is a predator that uses its disguise to approach or ambush prey that would otherwise flee. And it may be Batesian at a second remove: predators that have learned to avoid Gibellula-infected spider corpses may also avoid anything that looks like one, including a living, perfectly healthy spider wearing the same costume.

What makes the case particularly interesting is the specificity of the model. Most famous mimicry rings involve a predator or a toxic species as the model. This one involves a fungus — an entire kingdom of life away from the spider itself. Cross-kingdom masquerade, where an animal evolves to look like a plant or a fungus, is rare and almost always involves an inedible or sessile object (a leaf, a stick, a bird dropping). Mimicking the exact morphology of a single fungal genus is a level of specificity that suggests either very strong selection or a very long evolutionary history of co-occurrence. The Llanganates-Sangay Corridor, where the spider was found, is precisely the kind of humid, undisturbed tropical forest where Gibellula thrives.

Why iNaturalist keeps producing papers like this #

The paper's authors are explicit about the methodology. They credit the iNaturalist community with the initial identification and note that the citizen-science platform has become, somewhat unexpectedly, a serious source of taxonomic material. Taczanowskia waska is not an isolated case. The past five years have produced a steady stream of new-species descriptions that began as iNaturalist observations, including several spiders, a number of plants, and more than a few insects.

The pattern is consistent. A curious person with a phone and a flashlight photographs something odd in the field. They upload it. An engaged community of amateurs and specialists argues about the ID in the comments. Eventually a taxonomist with access to type collections weighs in, and the observation either collapses into an existing species or, as here, turns out to be the first photograph of a species nobody knew existed. The 1903 Bolivia specimen, sitting in a German drawer for over a century, is the implicit rebuke: without the photo, it would still be sitting there.

The discovery also leans on a kind of cross-checking that only works with both halves of the modern system. Bentley's iNaturalist post flagged the observation in 2025. Dupérré's museum visit in 2026 connected it to a historical specimen that nobody had thought to associate with a living spider. The combination — a field photograph plus a museum specimen, linked by a citizen-science identification — is increasingly the standard pipeline for tropical taxonomy, and T. waska is one of the cleanest examples yet.

The cultural footnote #

The paper's title — "The Cordyceps spider" — is doing more taxonomic work than titles usually do. By 2026, Cordyceps has spent a decade as the central biological premise of HBO's The Last of Us, in which a Cordyceps-like fungus jumps to humans and turns them into aggressive, infective spores. The show's vision of fungal mind-control is biologically implausible, but it has done real work in popularizing the actual fungi. Gibellula is, genuinely, a fungus that infects spiders, alters their behavior, kills them, and grows out of their bodies. A spider that looks like that fungus is, fairly or not, the closest real-world thing to a Last of Us antagonist that has ever been described — minus, of course, the part where it infects anything.

The New York Times' coverage of the discovery ran under the headline "The Last of Us, on Eight Legs." The framing is unavoidable. The science is, on its own terms, stranger than the fiction: a living animal evolving to look like the corpse of one of its own kind, killed by a fungus, in order to survive in the same forest where that fungus is common. It is the kind of finding that, before iNaturalist, would have taken decades of fieldwork to surface. It surfaced in 2025 because someone on a night walk in Ecuador pointed a phone at a leaf.

Sources

last updated: