The Bite of the Bear-Dog

A restoration of Amphicyon by Charlene Letenneur. From Argot 2010.

Blogging on Peer-Reviewed ResearchA common wasp on a foraging mission catches an enticing scent on the breeze. It's a set of chemicals given off by plants that are besieged by hungry insects and it means that there is food nearby for the wasp's grubs - caterpillars. The wasp tracks the smell to its source - a flower - and while it finds nectar, there are no caterpillars and it leaves empty-mandibled. The smell was a trick, used to dupe the wasp into becoming a unwitting pollinator for the broad-leaved helleborine.

helleborine.jpgThe broad-leaved helleborine (Epipactis helleborine) is an orchid that grows throughout Europe and Asia. It is but one deceiver in a family that is rife with them. About 10,000 species of orchids trick pollinators into visiting their flowers. Some attract males by mimicking the sight and smells of females. Others resemble orchid species that provide rich nectar rewards, while providing none themselves. But while thousands of species offer the potential for sex or food, only the broad-leaved helleborine advertises itself by promising fresh meat.

Darwin himself noted that even though the helleborine packs a substantial reservoir of nectar, it is pollinated by only two species of insects - the common wasp and the European wasp. Until now, no one knew how the orchid was attracting its pollinators. Jennifer Brodmann from the University of Ulm in Germany solved the mystery by testing how wasps responded to the smells and sights of orchids.

She found that the smell of the helleborine alone attracted just as many wasps as the whole flowers. In contrast, the sight of a flower in a glass box that didn't let any scents through was far less attractive. Luring wasps with odours makes sense for the helleborine, for it grows in shady parts of dark coniferous forests, where they are difficult to see.

A smelly trick

Vespula_germanica.jpgBrodmann used gas chromatography to analyse the chemicals released by the flower, and recorded the electrical responses of wasp antennae as the scents wafted over them. She detected several organic molecules such as hexanal and hexyl acetate that are collectively known as "green-leaf volatiles". They are produced by other plants when they are set upon by insects and those of the cabbage, for example, responds to caterpillar infestations by producing a very similar spectrum of chemicals to the helleborine.

In these other plants, green-leaf volatiles are a call for reinforcements. They summon predatory insects that feast on the caterpillars, or parasitoids that use them as living larders for their own eggs. Common wasps are no exception. By placing them in a Y-shaped tube with different scents at the prongs, Brodmann found that wasps were consistently drawn to the smell of helleborines over empty chambers. They even preferred chambers containing synthetic mixtures of the green-leaf volatiles released by the orchids.

Any wasp that is duped into visiting a helleborine flower still receives a drink of nectar for its troubles. In the future, it may associate the smell of green-leaf volatiles with a sugary reward, and be more likely to visit flowers of the same species. That suits the helleborine, which receives a specific pollination service.

By releasing the right chemicals, the broad-leaved helleborine has effectively hijacked the lines of communication that other plants use to recruit wasps. It's the first species known to do this, but unlikely to be the only one.

Brodmann also found that the closely related purple helleborine (Epipactis purpurata), which is also pollinated by wasps, produces similar levels of green-leaf volatiles. On the other hand, a third species from the same genus, the royal helleborine (Epipactis atrorubens), which are pollinated by bees, releases few if any of these chemicals.. The strategy seems to only work on wasps.

Reference: BRODMANN, J., TWELE, R., FRANCKE, W., HOLZLER, G., ZHANG, Q., AYASSE, M. (2008). Orchids Mimic Green-Leaf Volatiles to Attract Prey-Hunting Wasps for Pollination. Current Biology DOI: 10.1016/j.cub.2008.04.040

Images: Wasp by Maciej Skorecki; orchid by BerndH

Between 23 and 16 million years ago, just outside of where the city of Lisbon, Portugal sits today, there lived a unique mix of mammals which would have seemed both strange and familiar. From bones and footprints left in fossilized feces, paleontologists have found that rhinoceros, deer, horses, antelope, and elephants browsed and grazed in the ancient ecosystem, and many were preyed upon by archaic carnivores such as the fearsome amphicyonids (popularly known as “bear-dogs“). That such confrontations occurred can readily be inferred by the presence of large predators and prey in the same place, but direct evidence of interaction is rare. It does not require a stretch of the imagination to envision a large bear-dog grappling a fleeing antelope or rhinoceros to the ground, but how do we really know that such events took place?

Without a time machine, it can be extremely difficult to tease out the relationships between extinct organisms, but every now and then paleontologists find a rare specimen which records the interaction of two species. One such specimen, a bit of the left lower jaw from the rhinoceros Iberotherium rexmanueli, was described by scientists Miguel Antunes, Ausenda Balbino, and Léonard Ginsburg in 2006. Although rather plain-looking at first sight, the specimen is remarkable for exhibiting a series of pits and scratches that were most probably made by the bear-dog Amphicyon giganteus.

A line drawing of the anterior portion of the rhinoceros jaw showing furrows made by the incisor teeth of a large carnivoe. From Antunes et al, 2006.

Blogging on Peer-Reviewed ResearchSpecific language impairment (SLI) is a language disorder that affects growing children, who find it inexplicably difficult to pick up the spoken language skills that their peers acquire so effortlessly. Autism is another (perhaps more familiar) developmental disorder and many autistic children also have problems in picking up normal speech and communication. These two conditions have a common theme of language difficulties running through them, but a new study reveals a deeper connection - both are linked to a gene called CNTNAP2.

FOXP2.jpgThe story of CNTNAP2 actually begins with another gene, whose name will be familiar to anyone with a passing interest in the genetics of language - FOXP2. Earlier this year, I wrote a long feature on the history of FOXP2 for New Scientist, but here's a potted version.

FOXP2 was catapulted into the limelight earlier this decade when it became the first gene to be linked to an inherited language disorder. Initially heralded as "a language gene", the hype surrounding FOXP2 was soon pierced by a number of studies which showed that the gene is an ancient one - it is present in a variety of different animals and has changed very little over the course of evolutionary time. In other species, it is hardly involved in language and in some, it isn't even involved in communication.

The latest evidence suggests that FOXP2 affects the learning and production of complex sequences of movements. Such sequences are, of course, crucial for speech so it's understandable that faults in FOXP2 leads to linguistic difficulties. So much for the hype, but the FOXP2 story isn't over yet.

One of the most interesting things about it is that it's a 'transcription factor', an executive gene that controls the activity of several subordinates. It was the quest to identify the genes that FOXP2 lords over that led to CNTNAP2. And lo and behold, it too plays a role in language.

The discovery was made by a group from the University of Oxford led by Simon Fisher, the scientist who first demonstrated FOXP2's involvement in language disorders. Sonja Vernes from Fisher's lab began by looking for areas of human DNA that the FOXP2 protein sticks to.

One of these regions piqued their interest for it lay within CNTNAP2, an extremely compelling candidate for a language-related gene. Previous studies had shown that CNTNAP2 is switched on in nerve cells and affects the connections between these cells. It is involved in the development of the brain and its levels are particularly high in neural circuits that are involved in language. And faults in the gene had been linked to other conditions including Tourette's syndrome and autism.

Lips.jpgVernes found that FOXP2 switches off CNTNAP2, for parts of the developing brain with the highest FOXP2 activity also have the lowest CNTNAP2 activity. The team went on to sequence the gene in members of 184 British families with a history of specific language impairment. They assessed the language abilities of each child with a battery of tests, including one that analysed whether they were able to repeat made-up words. It seems like a strange test, but children with SLI do characteristically badly at it. 

Vernes's group identified nine different versions of CNTNAP2 that affected a child's language abilities, each differing from the normal one by a single change within a relatively narrow section of the gene. Some of these variants cropped up in groups and one such combination, known as ht1, had a particularly strong influence on language skills. Children who carry one or two copies of the ht1 version of CNTNAP2 score much lower on the nonsense-word repetition test than their peers who lack any such copies.

These results are a big step forward in understanding the genetics of language. For a start, it's the first time that faults in a specific gene have been linked to a common type of language impairment. The defect caused by faulty FOXP2 is very rare and SLI on the other hand affects 7-8% of preschoolers. That the gene is also implicated in autism just makes it that much more interesting, especially since the group took great pains to ensure that autistic children were excluded from their study, so as to avoid biasing their results.

Some variants of CNTNAP2 affect the length of time it takes for autistic children to develop language skills. And amazingly, both these variants and those implicated in SLI are the result of genetic changes within the same section of the gene. So faults in this single gene contributes to language problems that are shared by both SLI and autism.

Finally, the new study is a sign that the true promise of FOXP2's discovery is being fulfilled - the gene itself has been overly hyped, but its true worth lies in opening a door for more research into genes involved in language. It was the valuable clue that threw the case wide open. CNTNAP2 may be the first language disorder culprit revealed through FOXP2 but if Fisher's team have anything to say about it, it certainly won't be the last.

Reference: New England Journal of Medicine, citation tbc.

The Iberotherium specimen in question is the middle portion of the animal’s lower jaw; the front of the jaw is missing and the rear portion was accidentally broken off during collection. There are a number of small pits and perforations along the jaw which indicate that a carnivorous mammal stripped the flesh off it, with some of the most conspicuous damage being seen towards the front of the bone. Here, near the area of the left jaw which would have met with the front of the right jaw (called the symphysis), there are three indentations on both sides which were probably made when the carnivore gripped the jaw with its incisors.

The hypothesis that Amphicyon giganteus was probably the offending carnivore came out of a process of elimination. Although the large carnivore Dinocyon also lived around the same time, its remains have not been found from the same locality, and while the bear-dog species Amphicyon major also prowled nearby, its presence in the Lisbon beds has not been confirmed. Likewise, the size of the bite marks indicated the activity of an animal the approximate size of a brown bear, and other contemporary candidates were not large enough to fit the profile. Hence the scientists were left with only one valid candidate – Amphicyon giganteus.

But, even if we can be confident in the identification of Amphicyon giganteus as the predator in question, does the tooth-marked jaw represent a predation event or scavenging? It is impossible to know for sure. Antunes and colleagues propose that the pattern of toothmarks on the jaw suggest that the Amphicyon was holding onto and biting the rhinoceros mandible in a manner similar to how we eat corn on the cob. Hence they hypothesize that this means that the carnivore was eating a largely-intact carcass “on the spot.” If true, this means that the Amphicyon either killed the rhinoceros – perhaps preying upon an individual weakened by droughts which frequently occurred during the time – or was lucky enough to happen upon an intact Iberotherium which had been killed in a flood.

Yet, as a large carnivore, Amphicyon would have also been capable of tearing off portions of a carcass – such as a head or jaws – and carrying them away to consume in relative peace. This hypothesis could also apply to either a hunting or scavenging scenario, but without the rest of the Iberotherium skeleton, it is impossible to tell exactly what happened. The toothmarks indicate that an Amphicyon fed on the rhinoceros jaw, but they can’t tell us how the rhinoceros died in the first place.


ANTUNES, M., BALBINO, A., & GINSBURG, L. (2006). Ichnological evidence of a Miocene rhinoceros bitten by a bear-dog (Amphicyon giganteus) Annales de Paléontologie, 92 (1), 31-39 DOI: 10.1016/j.annpal.2005.10.002

ANTUNES, M., BALBINO, A., & GINSBURG, L. (2006). Miocene Mammalian footprints in coprolites from Lisbon, Portugal Annales de Paléontologie, 92 (1), 13-30 DOI: 10.1016/j.annpal.2005.09.002

Top image from: Christine Argot (2010). Morphofunctional analysis of the postcranium of Amphicyon major (Mammalia, Carnivora, Amphicyonidae) from the Miocene of Sansan(Gers, France) compared to three extant
carnivores: Ursus arctos, Panthera leo, and Canis lupus Geodivertistas, 32 (1), 65-106

Leave a Reply

Your email address will not be published. Required fields are marked *