eTawau > Insects of Borneo > Ants > When a weaver ant met a nettle caterpillar

When a weaver ant met a nettle caterpillar

Kampung Sayap, Kota Belud — and the ant walked away

A weaver ant came along a blade of grass and found a small green caterpillar sitting on it. Anyone who has watched kerengga at work knows how that ought to end. These are the fiercest small hunters in the Bornean forest, and the caterpillar was soft, slow and ten millimetres long.

The ant walked up to it, stopped, touched it briefly with its antennae — and left. No bite, no grip, no call for help. Fifty-six seconds later it was gone from the grass blade altogether.

It took me a long time to work out why, and the answer is in the photograph if you know where to look.

A reddish-brown weaver ant on a sunlit green grass blade facing a small yellow-green caterpillar covered in clusters of radiating black spines. The ant's left antenna is laid across the caterpillar's rear spines; its jaws are closed. Labels mark the weaver ant, its antenna resting on the spines, and the stinging spines of the caterpillar.
10:48:38 in the morning, 20 December 2022. The ant's antenna is lying right across the caterpillar's spines, and its jaws are shut. This is not an attack. It is an inspection — and it is the whole story of the page.

What everybody expects to happen

The Asian weaver ant, Oecophylla smaragdina, is the ant that binds living leaves into nests using silk squeezed from its own part-grown larvae. Colonies run into the hundreds of thousands and hold whole trees against all comers. They are aggressive, fast, extremely territorial, and they hunt everything they can overpower.

They are so good at it that people have been putting them to work for a very long time. Weaver ants are among the oldest biological control agents on record — Chinese citrus growers were carrying their nests from tree to tree centuries ago — and they are still used that way in tropical orchards today, including in this region. A caterpillar on a leaf is exactly the kind of thing a weaver ant colony is employed to remove.

Close view of the same weaver ant in profile on the green grass blade: reddish-brown, with very long legs, a large convex head with a prominent black eye, long slender antennae, a narrow waist and a smooth rounded abdomen.
How you know a weaver ant. Reddish-brown, very long legs, a big convex head with an obvious eye, and a long slender first antennal segment. About 10 mm — this is a major worker. If you see them running in file along a branch and binding leaves together overhead, you have found the nest.

Now look at the caterpillar

It belongs to the Limacodidae, the slug caterpillars — known across Malaysia as the nettle caterpillars, and the name is a warning rather than a description.

Close view of the yellow-green caterpillar on the grass blade, fifty-six seconds later and alone. Its flattened sluglike body carries rows of star-shaped tubercles bearing long radiating black spines, with blue and yellow marks along the back and a glassy translucent stripe down the middle. Labels mark the spined scoli, the hollow stinging spines, the blue marks, and the absence of legs.
The same caterpillar, fifty-six seconds later, alone. Look for what is missing: no legs. No clasping feet, no prolegs, nothing to grip a twig with. It slides along on a sucker-like sole, which is why the family is called the slug caterpillars.

What it has instead is armour. Those star-shaped clusters are called scoli, and every black spine radiating from them is hollow and drains a poison gland under the skin. They are built to break off on contact, so that the contents run out into whatever touched them. Brushing against one is often compared to a bee sting, and the pain lasts a day or two.

Why the spines beat the ant

Here is the part worth carrying away. Think about how an ant actually kills something.

It has to get its jaws onto the animal, hold on, and spray formic acid into the wound. Every single step of that requires contact with the body surface. And a dense field of long, brittle, venom-filled spines standing out in every direction does exactly one thing: it holds the jaws away from the skin. There is nowhere on this caterpillar for an ant to take hold. The defence is not really poison. The defence is distance — the poison is what makes trying it expensive.

And the ant never had to be stung to find that out. Ants assess almost everything by touch and smell, with the antennae, before they commit to anything. That first photograph is the assessment happening: one antenna laid across the spine field, jaws still shut. A second of touching, and the ant knew.

That is a defence which works by being recognised rather than by being used — the best kind there is, because nothing gets damaged. The caterpillar did not rear up, did not thrash, did not drop off the leaf. It did not need to.

Fifty-six seconds

I photographed the meeting at 10:48:38, and the caterpillar on its own at 10:49:34. I did not time anything — I only found the interval afterwards, in the camera's own record of when it opened the shutter.

It is worth saying to anyone who photographs animals: your camera is writing down the time whether you ask it to or not, and when a record is a sequence, the gap between the frames is itself an observation. It turned "the ant walked away" from something I remembered into something I can show.

What I cannot rule out

This site tries to keep what was seen apart from what was worked out afterwards. Three honest alternatives:

  1. The ant may not have been hunting at all. Weaver ants patrol territory and tend sap-sucking insects for honeydew as well as hunting. A single worker walking along a grass blade may simply have been on its way somewhere.
  2. One ant is not a hunting party. Weaver ants fetch nestmates for anything too big to carry alone. A scout that finds unmanageable prey often leaves in order to come back with help. What I can say is that 56 seconds later it had not come back, and it did not while I stayed.
  3. I did not see the antennae touch. I saw the ant stop; the photograph shows an antenna lying across the spines. Whether it met a spine tip or the air a millimetre above it, I cannot swear to.

The friendship this is not

When I showed these pictures around, the first thing people said was that the two knew each other — that the ant had greeted an old friend and moved on. It is a lovely idea, and it is not foolish, because ants and caterpillars really do have friendships.

The caterpillars of many blue butterflies (family Lycaenidae) carry glands that feed ants a sweet secretion, and in exchange the ants guard them from wasps and spiders. Some are even carried down into the ants' nest and raised there. If you ever see a smooth, flat, legless-looking caterpillar with ants fussing attentively over it, that is what you are watching, and it is worth stopping for.

But this is the opposite animal. A lycaenid caterpillar is smooth because it wants the ants to touch it. This one is covered in venom because it wants nothing to touch it at all. The ant came, found out, and went. Not friendship — a warning, read correctly.

The record

20 December 2022 · 10:48 am · Kampung Sayap, Kota Belud

On a grass blade at Eco-Edu Tourism Tinuman

Eco-Edu Tourism Tinuman, about 3 km before the centre of Kampung Sayap, in the Kota Belud district of Sabah, on the northern flank of Mount Kinabalu. I had car-camped there alone the night before.
One weaver ant, Oecophylla smaragdina, a major worker, and one nettle caterpillar, family Limacodidae — genus not determined. Both about 10 mm, each alone, on a blade of the long grass the people here call larang. Two frames, 10:48:38 and 10:49:34. Neither animal was handled or moved; both were left where they were.
— iNaturalist observation 399882665

Why the caterpillar has no name here. The Moths of Borneo records 95 limacodid species in 54 genera from this island, most of them generalist feeders and a great many of them green and spiny as larvae. What separates the major groups is the arrangement of the scoli — and reading that needs a proper view straight down on the back, or a caterpillar reared through to the adult moth. I have one photograph from the side. A green nettle caterpillar photographed from one angle is not nameable below family, and I would rather leave the record honest and unfinished than dress a guess up as a fact. The observation is on iNaturalist, where anyone who knows these larvae can take it further.
One thing I would like answered. It was on a grass. The Bornean records of this family are dominated by broad-leaved plants — palms above all, then Ficus relatives, legumes and their kin — and the pest species attack coconut, oil palm, cocoa, coffee and tea. A nettle caterpillar on a grass blade is not where the books put one. Was it feeding, crossing, or looking for somewhere to pupate? I did not see it feed, and there is no feeding damage on the blade beneath it.
Do not pick one of these up. I photograph insects alive and put them back, and I handle a great many of them. This family is the exception. The spines break in the skin and the poison comes out of them, and a small green caterpillar you barely noticed can spoil the rest of your day.

If one does get you: do not rub or brush the area — that snaps more spines and spreads the poison over more skin. Lift the broken spines off with sticky tape, pressed on and pulled away, then wash. It is painful rather than dangerous, but it is properly painful, and children should be told before they go walking in long grass.
Where this was. Kampung Sayap is in the Kota Belud district, on the quiet northern side of the Kinabalu massif — the back door into Kinabalu Park, reached from Kota Belud town rather than from Kundasang. Eco-Edu Tourism Tinuman is a village-run camping and walking place about 3 km short of the village centre.

I was not looking for anything. I was carrying a camera to breakfast.

Words you may meet

Kerengga
The Malay name for the weaver ant, Oecophylla smaragdina — the red ant that builds leaf nests in trees and bites hard.
Limacodidae
The slug caterpillars or nettle caterpillars: moths whose larvae have no prolegs and slide on a sucker-like sole, most of them armed with stinging spines.
Scolus (plural scoli)
A raised, often star-shaped tubercle on a caterpillar's body, carrying a cluster of spines.
Urticating
Stinging on contact, in the way a nettle does. The spines are hollow and drain a venom gland, and they break off in whatever touches them.
Antennating
What an ant does when it examines something by tapping it with its antennae — which is how ants smell, taste and feel all at once.
Formic acid
The chemical that ants of this group spray into a bite wound. It is what makes an ant bite sting.
Myrmecophily
Literally "ant-loving": the habit of living with ants for protection, as the caterpillars of many blue butterflies do.

Related pages

About the photographs. Taken by Wong Fook Yee with a Canon IXUS 285 HS at Eco-Edu Tourism Tinuman, Kampung Sayap, Kota Belud, on 20 December 2022 at 10:48:38 and 10:49:34 am. For this page they have been reduced and labelled, because a visitor wants to be shown what to look at. The unmarked originals are attached to the iNaturalist record above, which is where anyone working on the identification should start.

The Bornean limacodid figures and the larval characters follow The Moths of Borneo, Part 1; the structure of the stinging spines follows the Virginia Cooperative Extension account of stinging slug caterpillars. All photographs © Wong Fook Yee. Page written 13 September 2026.