Do you think I’m tasty?

As I hiked the upper stretch of the Shut-Ins Trail at Sam A. Baker State Park in southeastern Missouri, I encountered this 2-inch long millipede slowly making its way across the rocks.  Many millipedes, of course, produce hydrogen cyanide (HCN) as their primary method of defense against predation, and the bright yellow markings of this individual were an obvious sign that this particular species is no exception.  The wrinkled dorsal surface and black coloration with yellow wedge-shaped posterolateral markings identify it as a species of Pleuroloma (BugGuide), and of the four species known from North America (Shelley 1980) only the widespread Pleuroloma flavipes (literally meaning “yellow legs”) occurs as far west as Missouri (Shelley et al. 2004).  A similar pattern of coloration is seen in a number of related genera, e.g. Apheloria, Boraria, and Cherokia—all belonging to the order Polydesmida, presumably functioning across the group as an aposematic (warning) signal to predators that they should be left alone.  Another feature shared by the members of this group is the lateral expansion of the dorsal segments into “paranota,” giving the species a much more flattened appearance than other millipedes with the more typical cylindrical shape.  While all millipedes exhibit diplosegmentation (embryonic fusion of paired body somites and associated legs, spiracles, and ventral nerve cord ganglia), members of the Polydesmida have taken this condition to its culmination with no evidence of external sutures (

The bright coloration of this species was an interesting contrast to the cryptic invisibility of the copperhead snake I had seen just a few moments earlier during the hike—opposite strategies with identical goals.  Defense compounds are, of course, widely employed by many plants and animals; however, only millipedes and a few insects have developed the ability to utilize HCN, a highly toxic compound that halts cellular respiration in most animals through inhibition of the mitochondrial enzyme cytochrome c oxidase.  Evidence suggests that Pleuronota flavipes and other millipedes can tolerate HCN because they possess a resistant terminal oxidase that makes their mitochondria insensitive to the effects of HCN (Hall et al. 1971).

Perhaps some of you will be interested in this recent checklist of the millipedes of North and Central America (Hoffman 1999).

Update 6/13/11: My ID as Pleuroloma flavipes must be considered tentative, as Rowland Shelley has sent me an email with the following comment:

It could be Pleuroloma flavipes Rafinesque, 1820, or it could be Apheloria virginiensis reducta, I can’t really tell from the photos.



Hall, F. R., R. M. Hollingworth and D. L. Shankland. 1971. Cyanide tolerance in millipedes: The biochemical basis. Comparative Biochemistry 34:723–737.

Hoffman, R. L.  1999.  Checklist of the millipedes of North and Middle America. Virginia Museum of Natural History Special Publication No. 8, 584 pp.

Shelley, R. M. 1980. Revision of the milliped genus Pleuroloma (Polydesmida: Xystodesmidae). Canadian Journal of Zoology 58:129–168.

Shelley, R. M., C. T. McAllister, and S. B. Smith. 2004. Discovery of the milliped Pleuroloma flavipes in Texas, and other records from west of the Mississippi River (Polydesmida: Xystodesmidae). Entomological News 114 (2003):2–6.

Copyright © Ted C. MacRae 2011

Answers to ID Challenge #5 – Artrópodes em casca de árvore morta

Dead tree in Campinas, Brazil

After checking into my hotel in Campinas, Brazil I couldn’t wait to start exploring the grounds to see what insect life I might be able to find.  Almost immediately, I encountered this dead tree in back of the hotel.  To a beetle collector, a dead tree is an irresistible draw – especially one that is still standing and with loosely hanging bark, as in this one.  I approached the tree, gave it a look up and down the trunk to see if any beetles or other insects might be found on the outer surface of the bark, and when none were seen began carefully peeling sections of the bark away from the trunk.  Out from beneath the first section darted a small, black lizard – it reminded me in general form of our North American fence lizards (genus Sceloporus), but honestly it darted so fast up the trunk that I didn’t get a good look at it (much less even the chance to attempt a photograph).  Peeling the bark further away from the wood revealed a goodly number of what I took to be beetle larvae, although they were unlike anything I’d ever seen before.  They were fairly good-sized – about 25mm in length, and although there are a number of beetle families whose larvae may be encountered under the bark of dead trees, there aren’t many with larvae of this size.

Coleopteran larva (Tenebrionidae?) under bark of dead tree.

Despite their odd appearance, their basic gestalt suggested to me that they might be something in the family Tenebrionidae (darkling beetles).  Sadly, the state of beetle larval taxonomy is far from complete, especially in the tropics, and given the extraordinary diversity of the order as a whole I knew it could be difficult to impossible to identify them.  This task was further complicated by the fact that I did not collect any voucher specimens.¹

¹ Insect collecting permits are required in Brazil and are exceedingly difficult to obtain.  Although enforcement is lax, a few unlucky foreigners have been caught and suffered tremendous inconvenience at the hands of notoriously unsympathetic authorities.  This being a business trip, I had no desire to tempt fate for the sake of a few larvae in a group I don’t even study.

Despite a millipede-like appearance, six legs and loose cluster of ocelli indicate its true identity.

After consulting all of the print and online resources at my disposal and failing to find a convincing match at even the family level, I began to second guess not only whether these were tenebrionids, but larvae or even beetles.  I’m not aware of any tenebrionids with larviform adult females, but such are common in the Lampyroidea.  That didn’t seem to fit, however, as the latter tend to be much more flattened and armored in appearance, and the round head is really unlike the elongate and narrow head so often seen in that group.  The actually began to wonder if it was even a beetle – most xylophagous beetle larvae are light-colored and rarely so heavily sclerotized, and the antennae are unlike the typical 3-segmented antennae seen with most xylophagous beetle larvae.  In fact, the antennae and the shape of the head actually reminded me of a millipede, but the obvious presence of six legs (and no more) made this untenable (even though 1st instar millipedes are hexapod, the large size of these individuals precludes them from being 1st instar anything).  Eventually, I could only conclude that they were coleopteran – possibly a larviform adult, but more likely larval.  As a last resort I sent photos to Antonio Santos-Silva, a coleopterist at the University of São Paulo.  Although he specializes in Cerambycidae, I reasoned this might be a fairly common species since I had found good numbers on a single tree in an urban area near São Paulo, and as such it might be something he would recognize.  Antonio quickly replied saying that he agreed it was the larva of a species of Tenebrionidae, with an appearance similar to the larvae of Goniodera ampliata (a member of the Lagriinae, formerly considered a separate family).  I’ve not been able to find photos of the larva of Goniadera or related genera, but these do bear a striking (if more glabrous) resemblance to these presumed tenebrionid larvae from Australia.  Until a more convincing opinion is forthcoming, Tenebrionidae seems to be the consensus.

Polyxenid millipedes and two types of Collembola (several Poduromorpha and one Entomobryomorpha)

Three tiny adult coleopterans (family?) surround a large larval coleopteran

Although nobody zeroed in on Tenebrionidae for this challenge (#5 in the ID Challenge series), I must say that I enjoyed the diversity of opinion about what it might represent.  Moreover, congratulations to those who ‘took nothing for granted’ and noted the presence of several other organisms in the photo – this is where the big points were to be earned, and several participants successfully ID’d what I take to be a number of poduromorph collembolans, a single entomobryomorph collembolan, a central cluster of polyxenid millipedes, and several indistinct but clearly coleopteran adults (see super crops above).  David Hubble got the most correct answers to earn 15 points and the win in this inaugural post of BitB Challenge Season #2, while Dave and Troy Bartlett earned 13 and 10 points, respectively, to complete the podium.  Seven other participants got in on the fun and earned some points – I hope you’ll join the fun next time, too!

Copyright © Ted C. MacRae 2011

Millipede assassin bug

Ectrichodia crux

I continue the hemipteran theme begun in the last post with this photograph I took in South Africa below the Waterberg Range in Northern (now Limpopo) Province. I recognized them as members of the family Reduviidae (assassin bugs), and since to my knowledge species in this family are exclusively predaceous (except for the so-called “kissing bugs” of the mostly Neotropical subfamily Triatominae, large distinctive bugs that feed exclusively on vertebrate blood), I found what I took to be a case of scavenging on a dead millipede to be rather curious.  It had rained the previous evening, resulting in a burst of millipede (and insect) activity that night, and this scene was rather commonly encountered the following morning. Of course, appearances can be deceiving, and it turns out that I actually was witnessing predation – and a most unusual case at that.  The individuals in this photo represent Ectrichodia crux (millipede assassin bug), a common species in many parts of southern Africa.  Although nearly 500 species of assassin bugs are known from the region (Reavell 2000), E. crux is easily recognizable due to its large size (adults measure up to 22 mm in length), stout form, and coloration – shiny black, with a distinctive black cross incised on its dull yellow thorax and with yellow abdominal margins (Picker et al. 2002). The nymphs as well are distinctive – bright red with black wing pads. Clearly, these insects are advertising something.

Ectrichodia crux belongs to the subfamily Ectrichodiinae, noted for their aposematic coloration – often red or yellow and black or metallic blue, and as specialist predators of Diplopoda (Heteropteran Systematics Lab @ UCR).  Species in this subfamily are most commonly found in leaf litter, hiding during the day under stones or amongst debris and leaving their shelters at night in search of millipedes (Scholtz and Holm 1985). They are ambush predators that slowly approach their prey before quickly grabbing the millipede and piercing the body with their proboscis, or “beak.”  Saliva containing paralytic toxins and cytolytic enzymes is injected into the body of the millipede to subdue the prey and initiate digestion of the body contents, which are then imbibed by the gregariously feeding assassin bugs.

Millipedes employ powerful chemical defenses – primarily benzoquinones and sometimes hydrogen cyanide gas as well, which are discharged from specialized glands along the millipede’s body – to protect themselves from predation.  Thus, specialized predation of millipedes is a niche that has been exploited by relatively few predators, and little is known about the mechanisms used for circumventing these defenses. The recently reported millipede specialist, Deltochilum valgum (order Coleoptera, family Scarabaeidae), has been observed killing its prey by violently decapitating and disarticulating it before feeding on the body contents (Larsen et al. 2009, summary here); however, the exact manner by which the beetle avoids or withstands the millipede’s chemical discharges remains unknown.  For ambush predators such as Ectrichodia crux and other ectrichodiines, a strategy similar to that described for another millipede specialist predator, larvae of the phengodid beetle, Phengodes laticollis (order Coleoptera, family Phengodidae), might be employed. This species subdues its millipede prey by piercing thinner regions of the millipede’s integument (e.g., intersegmental membranes on the ventral surface) with its hollow sickle-shaped mandibles and apparently injecting gastric fluids that abruptly paralyze the millipede, thereby preventing it from discharging its gland contents (Eisner et al. 1998).  These undischarged benzoquinones remain confined to the glands and are prevented from diffusing into the body cavity by the glands’ impervious cuticular lining, thus allowing the phengodid larva to safely imbibe the liquified systemic contents of the immobilized millipede.


Eisner, T., M. Eisner, A. B. Attygalle, M. Deyrup and J. Meinwald. 1998. Rendering the inedible edible: Circumvention of a millipede’s chemical defense by a predaceous beetle larva (Phengodidae).  Proceedings of the National Academy of Sciences USA 95(3):1108–1113.

Larsen, T. H., A. Lopera, A. Forsyth and F. Génier. 2009. From coprophagy to predation: a dung beetle that kills millipedes. Biology Letters DOI:10.1098/rsbl.2008.0654.

Picker, M., C. Griffiths and A. Weaving. 2002. Field Guide to Insects of South Africa. Struik Publishers, Cape Town, 444 pp.

Reavell, P. E. 2000. The assassinbugs (Hemiptera: Reduviidae) of South Africa.

Scholtz, C. H. and E. Holm (eds.). 1985. Insects of Southern Africa. Butterworths, Durbin, South Africa, 502 pp.

Copyright © Ted C. MacRae 2009

Add to FacebookAdd to NewsvineAdd to DiggAdd to Del.icio.usAdd to StumbleuponAdd to RedditAdd to BlinklistAdd to Ma.gnoliaAdd to TechnoratiAdd to Furl

Done with dung, meat please!

ResearchBlogging.orgNo feces for this species.” “Carnivorous dung beetle shuns dung and decapitates millipede.” “Little dung beetle is big chopper.” “Dung beetle mistakes millipede for dung.” These were some of the clever headlines that I had to compete with in coming up with my own opener for a remarkable beetle that titillated the science blogosphere last week. At the risk of being redundant, I’d like to revisit that beetle and offer a few (hopefully novel) thoughts of my own. I can say that I have a unique and special treat for those willing to read further.

First the background. Deltochilum valgum is a so-called “dung beetle” in the family Scarabaeidae that lives in the lowland rain forests of Peru. As suggested by its common name, it belongs to a group of beetles that are well known for their dung feeding habits. Over 5,000 species of dung beetles are known throughout the world, all of which carve out balls of dung and bury them as provisions for larval development – or so it was thought.  As reported by Trond Larsen of Princeton University and colleagues in Biology Letters, D. valgum has apparently abandoned its ancestral dung ball-rolling behavior in favor of a predatory lifestyle. Its prey – millipedes! Moreover, the species exhibits several distinct morphological traits that appear to have evolved as a direct result of their predatory behavior. Adult beetles were repeatedly observed killing and eating millipedes, and their disdain for dung was rather conclusively demonstrated by an exhaustive, year-long trapping program in which pit-traps were baited with a variety of bait types known to attract dung beetles (e.g., various kinds of dung, carrion, fungus and fruit) – and millipedes.  In all, over 100,000 dung beetles representing 132 species were trapped (what a nice collection!), 35 of which were found to scavenge on dead millipedes, but only five of these dared tackle live millipedes.  Of these, only D. valgum ignored all other foods – it only came to traps baited with live millipedes.

Larsen et al. determined that adults of D. valgum are opportunistic hunters and were much more likely to attack injured millipedes than healthy ones, even those weighing 14 times as much as the beetle.  Ball rolling behavior was never observed by D. valgum.  Most dung beetles have wide, shovel-shaped heads used to scoop and mold dung balls, but D. valgum has a much narrower head with sharp “teeth” on its clypeus (Fig. 1A vs. 1B).  The teeth apparently aid in killing the millipede by piercing the ventral surface behind the head and prying upwards (decapitating it), and the narrow, elongate head facilitates insertion into the millipede body for feeding.  Further, the hind tibia are elongate and curved, which are used to “grip” millipedes by holding them up against the dorsally reflexed pygidium (Fig. 1C vs. 1D).  This allows the beetle to drag its coiled up victim with one hind leg while walking forward on the other five (Fig. 1E).  Once killed, the beetles proceeded to break their prey into pieces and consume their meaty innards, leaving the disarticulated millipede exoskeletons licked clean (Fig. 1F).  One of these “attack” episodes was filmed (using infrared lighting so as not to affect their nocturnal behavior) and can be seen in this BBC News video.

Deltochilum valgum

Figure 1. (a) Dorsal view of D. valgum head. Sharp clypeal teeth and angled clypeus act as a lever to disarticulate millipede. Narrow, elongate head permits feeding inside millipede; (b) dorsal view of Deltochilum peruanum head, lacking characters described in (a), head used to mould dung balls; (c) lateral view of D. valgum pygidium and hind tibia. Dorsally reflexed pygidial lip is used to support millipede during transport. Elongate, strongly curved hind tibia is used to grip millipede. (d ) Lateral view of D. peruanum pygidium and hind tibia, lacking characters described in (c), hind tibia used for rolling dung balls. (e, f ). Predation strategy by D. valgum. (e) Dragging live, coiled millipede with one hind leg, walking forwards; ( f ) feeding on killed millipede with head inside
segments; disarticulated empty millipede pieces nearby.
Credit: Larsen et al. (2009).

Much has been made about this remarkable shift from coprophagy to predation, which Larsen et al. speculate was driven by competition for limited resources with the many other dung beetle species that occur in the Peruvian rainforests. In fact, adult dung beetles are known to feed on a variety of resources besides dung, as exemplified by the range of baits used in their survey. Thus, my first thought after reading the coverage was actually a question: “Has this species abandoned dung provisioning completely as a reproductive strategy?” Everything I had read focused exclusively (quite understandably) on the bizarre feeding habits of the adults, but there was no mention of what the species’ larval provisioning strategies were. Wanting more information about this, I contacted Trond Larsen, who graciously sent me a PDF of the paper. Unfortunately (though not a criticism of the paper), no further insight about this was found in the paper either. Indeed, in all of the observations recorded by Larsen et al., millipedes killed by D. valgum were consumed entirely by the adults, and no mention was made of how or whether millipedes were utilized for larval provisioning. I wondered if D. valgum had truly abandoned dung provisioning for larval development (a remarkable adaptive switch), or if in fact the species might still utilize the strategy for reproduction (perhaps having specialized on a dung type not included in their survey), while also exploiting millipede predation as adults for a nutritional advantage. I asked Trond about this, to which he replied with this juicy tidbit (I told you I had a special treat!):

Yes, I would very much like to know what the reproductive/nesting behavior of D. valgum is. My best guess is that they also use millipedes as a larval food source, but as you say, we haven’t observed that behavior yet. I have observed other generalist dung beetle species rolling balls out of dead millipedes, presumably to bury for the larvae, so I certainly think it would be an adequate food source. Many dung beetle species use carrion for their larvae.

I am quite confident that D. valgum does not use any kind of dung. I have sampled these dung beetle communities very thoroughly, with many dung types and other bait types, and also with passive flight intercept traps that catch all beetles. Every dung beetle species that feeds on dung is at least sometimes attracted to human dung (this is not the case in African savannahs though, but is in neotropical forests – that is a whole different story). There are still a small handful of species we catch in flight intercept traps that we don’t know what they eat, although some of these mysteries have recently been solved – many of them live in leaf-cutter ant nests for example.

While predation of millipedes by a dung beetle is itself a fascinating observation, demonstrating the abandonment of dung provisioning in favor of captured prey for larval development would be a truly remarkable example of an ecological transition to exploit a dramatically atypical niche. I hope Trond (or anybody for that matter) actually succeeds in observing millipede/prey utilization for larval provisioning by this species.

Many thanks to Trond Larsen for his delightful correspondence.

Larsen, T. H., A. Lopera, A. Forsyth and F. Génier. 2009. From coprophagy to predation: a dung beetle that kills millipedes. Biology Letters DOI: 10.1098/rsbl.2008.0654.

Copyright © Ted C. MacRae 2009

Add to FacebookAdd to NewsvineAdd to DiggAdd to Del.icio.usAdd to StumbleuponAdd to RedditAdd to BlinklistAdd to Ma.gnoliaAdd to TechnoratiAdd to Furl