Caterpillar Fungus Identification Guide

The Setting You Would Be Standing In

This is a fungus of extreme altitude. To find it you would be on the Tibetan plateau or the Himalayan arc between roughly 3000 and 5000 metres, in short alpine turf above the treeline, on slopes of sedge and cushion herbs where the snow has recently pulled back. The ground is thin, stony and cold, the grass is only a few centimetres high, and there is nothing that looks like a mushroom in the usual sense anywhere in view.

The search is done at very close range, scanning the turf for a single dark, wiry, upright stalk that reads at first glance as a dead grass stem. It is found by silhouette against pale ground rather than by colour, and by the fact that the stalk stands straight where the surrounding vegetation lies flattened by snow.

First Impression and the Two-Part Body

The visible portion is a single dark club standing 4-11 cm tall, sometimes less. What makes the identification is what is underneath. Excavate around it with a knife, working from ten centimetres out and undercutting so that the soil block lifts intact, and the stalk will prove to be attached to a caterpillar lying in the soil in a vertical position, head uppermost. That two-part structure, one club emerging from the head end of one intact insect larva, is the identification. Everything else is confirmation.

Reading the Stroma

The fungal club, properly the stroma, is cylindrical to very narrowly clavate and 2-6 mm thick. It is dark brown, olive-brown or nearly black, sometimes with a slightly paler base, and it is usually unbranched; occasional specimens carry two stromata from a single host, and a few show a shallow longitudinal groove or a flattened section.

Divide it into two zones. The lower portion, from soil level upwards, is the sterile stalk: smooth, tough, fibrous, and slightly paler. The upper 1-4 cm is the fertile head, which is subtly thicker than the stalk, duller in surface sheen, and under a hand lens covered in fine regular pimples. Each of those bumps is the protruding mouth of a flask-shaped perithecium buried in the tissue below. Above the fertile zone the club tapers to a short sterile tip, blunt or pointed, which is often paler and shows no pimpling.

Running a fingernail up the club is a useful test: the fertile zone feels like fine shagreen, the stalk feels smooth.

Reading the Host

The insect is a larva of a ghost moth, the group of genera around Thitarodes and Hepialus whose caterpillars feed underground on the roots of alpine plants. Cleaned of soil, the mummified larva is 3-6 cm long, cylindrical, ochre to yellow-brown or dull reddish-brown, and remarkably well preserved.

Work through it as you would any specimen. The body segmentation is still crisp, with fine transverse wrinkling. The head capsule is darker, reddish-brown to chestnut, and distinctly harder. Rows of short prolegs are visible along the underside, and the tiny hooks on them can often still be seen with a lens. The whole thing is rigid rather than soft.

Break or section the larva across the middle. The interior should be completely filled with dry, solid, white to pale cream fungal tissue with no trace of insect gut contents, no cavity and no liquid. A hollow larva, or one filled with dark decayed matter, is not a mature specimen of this fungus.

The join matters too: the stroma emerges from between the head capsule and the first thoracic segments, not from the middle of the body and not from the rear.

Sectioning the Fertile Head

With a razor blade take a thin longitudinal slice through the fertile zone and mount it in water. The perithecia are flask-shaped and 300-500 micrometres long, arranged in a dense palisade and set either at right angles to the surface or slightly obliquely, immersed to about half their depth in the stroma tissue with only the ostiole breaking the surface.

Inside each perithecium sit long cylindrical asci with a thickened apical cap. The ascospores are filiform, running nearly the full length of the ascus, and divided by many cross-walls; at maturity they fragment into short cylindrical part-spores, which is the normal dispersal unit for the genus.

The Equivalent of a Spore Print

There are no gills, so an ordinary spore print is impossible, but a deposit can still be obtained and it is a good confirmation that your material is mature. Suspend the fertile head a few millimetres above a glass slide or a piece of dark card inside a closed humid container overnight. Discharged part-spores appear as a faint whitish bloom on the surface below, which you can then mount and examine.

If nothing is discharged, take a squash mount from the perithecial zone and look directly for asci and part-spores; immature stromata will show developing perithecia without ripe asci.

Substrate, Season and Range

The substrate is not soil or wood but a living insect. The fungus infects the larva underground, where the caterpillar spends years feeding on the roots of alpine herbs and sedges, then kills and mummifies it and orients it vertically with the head close to the surface before the stroma is produced. This makes it an entomopathogen and ties its distribution exactly to the distribution of its ghost-moth hosts.

Stromata appear for a short window after snowmelt, broadly May to July depending on elevation and aspect, with higher slopes fruiting later. Meadows on north-facing slopes with deep, long-lying snow cover are typical. The natural range is the Tibetan plateau and the Himalaya, taking in Tibet, Qinghai, Sichuan, Yunnan and Gansu, plus Nepal, Bhutan and northern India. Records of the species from Europe, Japan or North America almost always refer to a different fungus, so geography is a legitimate part of the identification here.

Separating the Look-alikes on Morphology

Cordyceps militaris is the commonest global confusion and is easy to separate visually: it is bright orange to orange-red, with a distinctly clavate swollen head, and it arises from a buried lepidopteran pupa or larva in temperate woodland and grassland, not from an alpine meadow.

Ophiocordyceps gracilis also grows from a buried caterpillar but has a slender pale stalk topped by a clearly differentiated orange-brown clavate head, giving it a matchstick outline instead of a uniform dark rod.

Ophiocordyceps nutans has a thin, black, wiry, often nodding stalk with a bright orange fertile head, and its host is a true bug rather than a caterpillar; digging it out reveals a shield bug, which settles it at once.

Tolypocladium ophioglossoides and T. capitatum are the classic traps because both produce dark clubs from the ground. Follow the base and you will find yellow to golden rhizomorphs running through the soil to a false truffle of the genus Elaphomyces, which is a fungal host, not an insect. T. ophioglossoides has a flattened dark olive-black head; T. capitatum has a rounded head on a yellowish stalk.

Ophiocordyceps robertsii of New Zealand and Australia also grows from a caterpillar, but its stroma is much longer and more thread-like relative to the host, and geography separates the two absolutely.

Ophiocordyceps sobolifera and relatives on cicada nymphs are separated by the host: excavate and you find a squat nymph with enlarged front legs rather than a segmented caterpillar with prolegs.

Finally, plain plant material fools people constantly. A dark sedge stem, a wiry rush, or a fine root can look identical above ground. The only reliable method is to dig, and the identification is not complete until the host is exposed intact.

Working Checklist

Alpine meadow above 3000 metres in the Himalaya or on the Tibetan plateau, shortly after snowmelt; a single unbranched dark brown to black club 4-11 cm tall with a finely pimpled fertile zone below a sterile tip; the club rising from the head of a rigid, well-segmented, ochre-brown ghost-moth caterpillar filled solidly with white fungal tissue; immersed flask-shaped perithecia in a palisade; filiform multiseptate ascospores that fragment into part-spores.