Ask an Ant Expert: April 2012 Archives


I have been interested in certain insects & also arachnids for many years.I happened to encounter your website today while doing a google search for images of velvet ant varieties.They have absolutely beautiful coloration so I was doing a search on those insects.I realize velvet ants are not ants & are wasps but nevertheless I thought your site was interesting.Is it true that Bull Horn acacia ants in South America have a very painful sting? Its not often insects have a mutual-beneficial relationship with plants & other insects though it does occur in a number of species.


Hi Nick,

Bullhorn acacia-ants have a sting rated as a 1.8 on the Schmidt Sting Pain Index. While not excruciatingly painful by itself, when these ants attack, they attack with the full force of the colony and will continue stinging for as long as they can. They are also incredibly aggressive and will launch themselves from their trees onto invaders with no regard for their own safety. The wisest course of action is usually to avoid these trees altogether. Once you come into contact with the ants it is easy to understand how they can so effectively protect their plants because they are so aggressive and unpleasant.

The mutualism between the acacia-ants and their host plants is rather unusual but similar behaviors have evolved convergently both within the acacia-ant genus (Pseudomyrmex) and in unrelated genera such as Crematogaster and Azteca. Mutualistic Crematogaster ants in Eastern Africa are even able to protect their acacia hosts from megaherbivores like elephants and giraffes. You can check out the citation below for more of the story.

Goheen JR, Palmer TM. 2011. Defensive plant-ants stabilize megaherbivore-driven landscape change in an African savanna. Current Biology 20: 1768-1772.

Thanks for your question and keep up the interest in arthropods!
Ben Rubin & the AntAsk Team

Hello AntWeb Associates,

My question in short is, do cockroaches eat ants?

Let me expand on this a bit. Over the last several years, I have noticed a strip of sidewalk near my home where both cockroaches and Argentine ants seem to congregate. Often times at night when the temperatures are mild, I'll go for a stroll on this particular strip of sidewalk and I'll notice a great number of cockroaches (all sizes large and small) traveling and loitering among the ant trails. What do you suspect is going on here?


Robert in San Jose


Hi Robert,

Thanks for your question!

In all likelihood, the association you observed between the cockroaches & ants is merely coincidental rather than predatory. American cockroaches and Argentine ants alike are omnivorous scavengers that feed on a wide variety of organic matter (almost invariably dead if animal-based) and take eager advantage of the breadth of available foodstuffs in and around our homes. As with most domestic insect pests, they are more or less trophic equals, exploiting the same basic resources with similarly indiscriminate fervor. The concentration of these two opportunistic feeders in the same place says more about their overlapping diets than anything else, leading one to suspect that if you followed this particular nighttime procession, you might find some discarded fruit or an unattended waste bin or any number of other mutually appetizing treats.

While there are no known instances of cockroaches actively preying on ants, the scientific literature offers several examples of these two organisms engaging in other unexpected ways. Facultative myrmecophily (literally "ant-love", describing any meaningful association between ants and other organisms) has been documented in several species of cockroaches ranging from the miniscule Myrmecoblatta wheeleri, which has been found in the nests of both Solenopsis and Camponotus ants (Hebard, 1917; Fisk et al., 1976; Deyrup & Fisk, 1984), to the parthenogenetic Surinam cockroach, Pycnoscelus surinamensis, which exhibits a strong commensal association with the rover ant Brachymyrmex cordemoyi (Moretti et al., 2011). An earlier study describes additional, if somewhat less permanent, associations between P. surinamensis and seven other ant species, primarily the widespread tramp ant species Pheidole megacephala and Paratrechina longicornis (Deleport et al., 2002).

Perhaps the best-known cockroach myrmecophiles are those in the genus Attaphila, which maintain close host-specific associations with fungus-growing ants in the genera Atta and Acromyrmex. Most extensively studied of these is Attaphila fungicola (pictured), a minute cockroach that inhabits the fungus gardens of Atta texana and obtains nourishment by licking the exteriors of passing soldiers (Wheeler, 1910). These cockroaches will also instinctively mount the males and queen ants before a mating flight, thereby ensuring a wider dispersal than would otherwise be possible on their own.

Photo courtesy of LSU AgCenter

In many cases, myrmecophilous cockroaches display a marked ability to detect and orient to ant trail pheromones, especially in the more established Attaphila-Attine relationships. While it's certainly possible that the cockroaches you observed were somehow attuned to the scent trails of the scavenging Argentine ants, no functional association between these two species has ever been reported. Who knows? Maybe it's more than just a coincidence. Even without any permanent association, the Surinam cockroach was found to follow the scent trails of one of its hosts, Camponotus brutus, to a food source. Surely the enterprising American cockroach is capable of doing the same?

Thanks for your interest,

Alexandra Westrich & the AntAsk Team

Hello antweb team,

I am a huge fan of the ants and the antweb team. I have learned so much about ants from you guys, and have also taken away important life lessons from many of your eloquent responses.

My question is.. can you please tell me some things about Spechomyrma and other ant ancestors? What makes Sphecomyrma an ant and not a wasp? Also, what extant lineages of wasps are most closely related to ants? Lastly, do either ants or their wasp ancestors have free will?

Thanks so much!

Dear Jered,

Thank you for contacting AntBlog and sending in a question about ant fossils. We have asked a colleague, Dr. John LaPolla, who is currently working on some fossil ants to address your question and here is what he had to say:

"Great question about what makes Sphecomyrma an ant or not an ant. This is something ant experts have variously debated. There are a few morphological features of ants that are generally considered to define them and separate them from other hymenopterans (bees, wasps, and ants). There are two characters in particular that are important: an elongated scape (the first segment of the antenna) with an elbowed antennae and the metapleural gland (a gland found on the body of ants). While Sphecomyrma does have an elbowed antennae, it's scape is short. That being said, Sphecomyrma clearly possesses a metapleural gland (the exact function of this structure is unclear, but recent work strongly suggest it is involved in antimicrobrial functions and the secretion of various communication pheromones). The metapleural gland is particularly diagnostic for ants because no other hymenopterans possess anything remotely similar to it. Therefore, Sphecomyrma is considered an ant. It is not uncommon at all for ancient lineages of modern groups to possess some, but not all of the features we use to define them today (in this case the presence of short scape). Just so you know, there are now several known sphecomyrmines belonging to different genera (from two rather morphologically distinct tribes) that are even older (but not by much being around 95-100 million years old) than Sphecomyrma that have been discovered in French and Burmese ambers.

The next closest relatives to the ants is probably the Armaniidae which are an extinct group of large ant-like hymenopterans. We don't know much about them. Their exact placement has been hotly debated some have called them ants, some have not, and several researchers have variously flipped back and forth in their opinion. The reason is because armaniids, while being large, are only known from impressions in ancient rock (about 110 million years old) and they are very poorly preserved. Nothing resembling workers have ever been found for them (suggesting they were not eusocial), they have a wasp-like, very short scape, and they apparantly do not possess a metapleural gland.

Among living groups of wasps it remains unclear who is most closely related to the ants. Some have suggested vespids, others have suggested scoliids, and still other some of the more obscure wasp families, but I would say the jury is still out on this question. As for free will, I don't think ants have this - the concept of free will is very much a human construct. Ant colonies are more or less units that respond to their environment through chemical communication among nestmates."

Thanks again,
John LaPolla (guest expert), Corrie Moreau & the AntAsk Team

Dear Ajay,

As it turns out, your question was a very interesting one. After a bit of digging, we discovered that most citations on the vestigial traits of Eciton eyes trace back to Schneirla's seminal book "Army Ants: A Study in Social Organization" (1971). The most pertinent parts of the book that deal with eyes as vestigial structures are quoted below:

"Although often spoken of as blind, all army ants are sensitive to light, and many of them have eyes. New World species nearly all have two tiny degenerate compound eyes called "lateral ocelli," equipped with single lenses (Werringloer, 1932). Even the workers and queens of Old World species, which lack eyes, have a subdermal sensitivity to light." [pp. 28-29]

"By contrast, foraging army ants utilize light only in minor ways as by withdrawing in groups from bright light in a fallen tree area or at the forest edge. Light, although forcing changes in their local movements, affects the main direction of the raid very little." [pg.75]

As you can probably see, the reference you were looking for is the only real citation in Schneirla's book on the material. The rest is assumed to be observational by Schneirla himself. Although we could not locate an electronic copy, we did manage to find a hard copy of the journal in the Field Museum library, and it looks to be heavily morphological.

Why is this so interesting? It appears that there have been no direct studies targeting the neurobiology of Eciton, indicating that the evaluation of worker and queen eyes as vestigial is based on morphology alone. Males, which need functional eyes to navigate during their mating flights, have eyes of distinctly different morphology in addition to the triad of ocelli on the top of their head, which you have probably already seen on AntWeb.

Hope this helps!


Max Winston & the AntAsk Team



I was looking for information on eye structures of Eciton and discovered a comment made on the ant blog:

"Other vestigial traits are the absence of functional eyes in army ant species (see photo of Eciton burchellii below). These ants are blind, but show remains of the eyes."

Would you happen to have a reference for this? I have only got an old german reference which I have not found a pdf for!
Werringloer, A. 1932. Die Sehorgane und Sehzentren der Dorylinen nebst Untersuchungen ├╝ber die Facettenaugen der Formiciden. Z. Wiss. Zool. 141,432-524.

Any help would be greatly appreciated.

You guys are doing a fantastic job on the ant blog - keep it going!!

best wishes,

Ajay Narendra
Bldg 46
ARC Centre of Excellence in Vision Science
Research School of Biology
The Australian National University
Canberra, Australia

Recent Assets

  • strangeAnts3.jpg
  • strangeAnts1.jpg
  • Attaphila.jpg