A new ant with eyespots for defense?

Great to see Alex back, and with such a beautiful shot. I had a little post ready, so I figured I would go ahead and put it up. Maybe it will give Alex a little bit more time to recuperate after what sounds like a tough journey back.

One of the things I have discovered about studying a diverse arboreal group in a system that allows such easy access to the canopy is that undescribed species are relatively plentiful. In my main cerrado site, I have 17 Cephalotes species and 3 of them are undescribed. In my fist study at a second site, just 30Km down the road, a fourth new species showed up. All of these species are exciting in their own way, but one is particularly striking. I thought I would share a few shots with you. The main point of interest that I wanted to talk about is the coloration, and particularly the eyespots in the gyne.


Why might these eyespots be there? Well, evidence suggests that Cephalotes are quite distasteful, so the best explanation is that this is some kind of aposematic coloration. While eyespots are remarkably rare in ants as a whole, they are quite common in Cephalotes. In my experience, though, they are rarely this pleasing to the human eye (or at least this myrecologist’s eyes). After mating, Cephalotes gynes roam the canopy in search of a suitable cavity to start their colony. Hanging out in trees for a couple of years, I have seen this searching behavior many times, but never managed to get a decent shot of it (unlike more talented photographers).

The soldier caste has a similar coloration to the gynes. While this could be nonadaptive developmental spillover from the gyne (it fades out in smaller members of the soldier caste), it may also have some adaptive value. Soldiers, the relatively rare and expensive sterile caste, shuttle between the colony’s various nests on a daily basis, so the eyespots may help ward off possible predators while they do it.


As for what I should call this gorgeous ant, I have a few ideas, but I would love to hear what you all think in the comments.

Waste not, want not

What do arboreal ants eat? This is not such an easy question to answer as one might think. Nitrogen, vital for building proteins, is typically in short supply in the tops of trees. Ants as a group are often viewed as scavengers, getting nitrogen from dead arthropods that they find in the environment. But dead arthropods in the canopy will tend to fall, or be blown, to the ground.

So where do arboreal ants get the nitrogen they need to build their colonies?  One important source is the trees themselves. Arboreal ants drink nitrogen-poor but extremely abundant plant sap by tending sap-sucking insects. In essence, the ants use the bugs as drinking straws, and some groups can be seen as herbivores in terms of the source of much of the nitrogen they acquire.

But are there other nitrogen-rich snacks in the canopy that ants rely on? The answer, at least for Cephalotes, would seem to be yes. Cephalotes just love to eat bird droppings. In fact, a nice fresh dropping will elicit stronger recruitment than any other food source. They also love leaves that have freshly deposited urine on them, from arboreal mammals. Although these sources of vertebrate waste most likely have less than ideal forms of nitrogen, Cephalotes have a gut packed full of microorganims that may well help is converting it to something useful.

Cephalotes eduarduli (large) and Cephalotes maculatus (small) feeding on a large bird dropping in the cerrado canopy.
Cephalotes persimilis feeding on bird droppings on a long blade of grass coming into contact with their home tree

The fun part of Cephalotes’ affinity for this unusual resource, however, is that it makes for a very convenient and highly effective “natural” bait. I have now switched exclusively to urine baits harvested from, er, a bipedal cerrado primate in my work with Cephalotes. Aside from being free and unlimited (as long as the primate(s) in question have a good water supply), it is more effective than traditional baits, like sardines. For those interested in behavior, like me, it is also a natural food resource for this group and therefore invaluable in studying natural foraging and recruitment behaviors (ants don’t tend to find many naturally occurring sources of fish protein on the canopy). Moreover, it makes for good photographs, by pulling a large number of ants from nearby nests and becoming invisible once the water has evaporated off. It turns out that this food is also highly attractive to many other groups of arboreal ants. This is intriguing from a  scientific standpoint, but also very useful for any myrmecologist interested in finding and photographing some nice arboreal ants. Something to keep in mind next time you are all in the field and nature calls.

Cephalotes depressus in a fun-for-all feeding frenzy
Cephalotes depressus in a fun-for-all feeding frenzy at a urine bait

One of these things is not like the others

Cephalotes pusillus is ever-present in cerrado. In fact, I have never encountered another ant that is so abundant in a natural system, tropical or temperate. They are generalist nesters and can be found in almost every piece of standing dead-wood and many live trees. The workers are particularly robust, even for Cephalotes, and will often bulldoze their way to foods already overwhelmed by other ants.

C. pusillus worker

But even for these tank-like ants, trouble lurks in a surprisingly familiar form. The aphantochilid spider Aphanlochilus rogersi is a very striking mimic of C. pusillus, but not for the purposes of protection. It is an infiltration tactic. A spider will sit on the edge of a foraging trail of its model, seemingly undetected by the ants. When one worker strays too far off the beaten track, the spider strikes and runs with its prey before the large and dangerous foraging force has time to react. These spiders are remarkably abundant…but very fast. It took me a good year to get this shot of one of these hunting mimics with its prey and model.

The ant-mimicking Aphanlochilus rogersi with a paralyzed worker of its model, Cephalotes pusillus. This shot is on the underside of a branch. The foraging column from which the C. pusillus worker was plucked was on the upper surface.

For those wanting to read more, there is a nice older paper on this interaction (Oliveira, P.S. & Sazima, I. 1984. The adaptive bases of ant-mimicry in a neotropical aphantochilid spider (Araneae: Aphantochilidae). Biol. J. Linnean Soc. 22: 145-155.)


In cerrado, one of the most striking features of the vegetation is the dense covering of lichens on the trunks and larger branches of the diminutive trees within the system. This patchwork of pale greens makes for a great background for photographing ants. Below are two workers of one of my favorite cerrado species, Cephalotes borgmeieri, taking a moment to share some food.

Cephalotes borgmeieri
Cephalotes borgmeieri workers engaging in trophallaxis

More significant than the benefits to the camera-toting myrmecologist, though, is that the lichen cover has had strong evolutionary implications for the native fauna.

Weevil mimic

Lichen mimics are both abundant and diverse in the cerrado, and I tried to snap a quick shot whenever I encountered a new one. Unfortunately, I have not put names to these animals, but I think you’ll agree that the taxonomic diversity is quite amazing. Whereas some no doubt gain protective benefits from being green and crusty, others, like the spider, may be better suited to surprise potential prey. Scroll down to see the complete bunch that I managed to get decent shots of (others were too good at avoiding my camera lens). Some clearly did a better job than others in finding their model!

Mantid mimic
Mantid nymph
Psocopteran (maybe)
Tree frog
Katydid nymph
Hemipteran nymph

An army ant that likes a bit of fruit

Army ants are big meat-eaters. In fact, it would be fair to consider them among the most important of the big predators in tropical regions.

Most of the 150+ species of New World army ants (Ecitoninae) seem to be specialist predators of other ants, using strength in numbers to overpower the defenses of prey colonies. A few species also prey on a range of large arthropods, including spiders and scorpions. It might come as a bit of surprise, then, that members of the army ant genus Labidus also like a bit of fruit in their diet, especially in the cerrado.

I had read a few accounts of Labidus species eating nuts and even used peanut butter to trap them in a collaborative project that I worked on while in Panama. But it wasn’t until I started working in cerrado that I saw a natrual example of this behavior.

The Pequi tree (Caryocar brasiliense) is a common cerrado species with seeds that are covered in bright orange, richly flavored, and oily flesh. During the fruiting season (usually October-January), the fruits drop and the fleshy seeds are collected by a variety of animals, including Labidus coecus.

Labidus coecus workers simultaneously burying Pequi seeds and removing the flesh.

The ants simultaneously bury whole fruits (large green pods about the size of an orange) or individual seeds and harvest the flesh. The burying is achieved by removing small balls of soil from below the seed and piling it up the sides, ultimately forming a dome over the seed. It happens so quickly that you can actually see a seed sinking in real time. Burying behavior is quite common among Labidus species and it is presumably to protect large items while they are being harvested. These trees seem to be so attractive to L. coecus that they frequently move their nest (all army ants frequently shift nest sites) to the root structure below fruiting trees, taking up residence for a couple of months. Once the ants are done harvesting the  fruits and have moved on, the signs of their activity remain for some time, in the form seeds that are fully or partially buried with thousands of tiny balls of soil. The ants no doubt gain greatly from the oil-rich flesh, but what benefits or costs this has for the tree remains to be discovered.

The telltale signs of Labidus activity. A Pequi fruit has been partially buried by small balls of soil. The seeds inside were buried deeper still, with the pulpy flesh stripped by the ants.

A cerrado themed Myrmecos

As Alex mentioned, I will be standing in here at the Myrmecos blog for a few weeks. I thought I would try and stay true to Alex’s main theme of photo-based posts, but with my own little twist. So, the theme for the next couple of weeks will be ants (and other beasts) of the cerrado.

Even hardened tropical biologists are often unfamiliar with cerrado, which is a unique savanna-like habitat that covers much of central Brazil and small areas of neighboring countries. It is the poor relation in terms of research effort in the Neotropics, as most are drawn to The Rain Forest, but it is a myrmecologists dream. Much like wet tropical forests, an area of nice cerrado will be home to hundreds of ant species. The wonderful thing about the cerrado, though, is that the ‘canopy’ usually tops out at around 5m. This brings all those great arboreal ant species within arms reach, or at least from the end of a small step ladder. That’s me on the ladder, with a motley crew of biologists helping out with the harvesting of an experiment.

Working with arboreal ants in the cerrado
Working with arboreal ants in the cerrado

My plan is to share with you all a few of my own snaps from within and below the cerrado canopy. They will, I am sure, be biased towards my two favorite groups of ants, the Ecitoninae (a.k.a. the New World army ants) and the genus Cephalotes (a.k.a. the turtle ants), but I will try to squeeze in a few other interesting animals along the way. My hope is that this will be an enjoyable diversion until I hand the reins back to Alex.