This article was published in Scientific American’s former blog network and reflects the views of the author, not necessarily those of Scientific American
A few weeks back – during the Tet Zoo frog event – I wrote about the peculiar African brevicipitid frogs, variously termed short-headed frogs or rain frogs. The plan when compiling that article was to write about a second group of frogs, closely related to brevicipitids. But time was short, the article became too long, blah blah blah... anyway, here we are with the part II article on those ‘strange bedfellow frogs’. That second group is the hemisotids, the nine species of pig-nosed or shovel-nosed frogs, also known as snout-burrowers. All are included within the genus Hemisus.
Hemisus frogs are odd little ant- and termite-eating afrobatrachian frogs, endemic to Africa, that possess fused vertebrae, lack a sternum, and have robust skeletons and highly muscular forelimbs. A spade-like structure – very similar to that also present in spadefoot frogs – is present on the plantar surface of the foot. Hemisotids occur in savannahs, forests and cultivated areas and spend the dry season beneath ground, emerging when it rains. They burrow head-first (highly unusual for anurans) using a bullet-shaped skull. Some of their anatomical features seem eminently logical in view of what's seen in other burrowing tetrapods, like vertebral fusion. But others don’t: absence of a sternum seems odd given that big, bladed sterna that anchor various forelimb muscles are often considered typical for animals that burrow head-first. But then, these are frogs. Most things about them are weird. In the larger hemisotid species, females have an SVL of c 80 mm while the smaller species are c 40 mm or less in SVL (Channing 2001). [Adjacent photo by Ryanvanhuyssteen.]
The hemisotid tongue is remarkable – it’s a uniquely prehensile organ that’s ‘telescoped’ out of the mouth “extraordinarily slowly” relative to that of other long-tongued frogs (Anderson et al. 1998). “Extraordinarily slowly” means that tongue protrusion takes about 300 milliseconds. Frogs ordinarily have a hydraulic tongue but that of Hemisus seems to be a muscular hydrostat – that is, muscular control over fluid contained in a lingual sinus allows precise control over the tongue’s contraction, distension and direction (Nishikawa et al. 1999). So, it can be aimed three-dimensionally once it leaves the mouth: it isn’t just shot out in a straight line, but the frog can extrude it and then bent it sharply along its length to the right, the left, or up, or down. Sure, you can bend your extruded tongue in any direction too, but you don’t protrude your tongue at super-fast speed in order to grab objects with its tip. This superb control is linked to the presence of an especially high number of motor neurons that are placed in an unusual posterior position in the brainstem and spinal cord relative to the condition in other anurans (Anderson et al. 1998). And it gets even better, because the tongue has two muscular finger-like organs at its tip (Ritter & Nishikawa 1995).
On supporting science journalism
If you're enjoying this article, consider supporting our award-winning journalism by subscribing. By purchasing a subscription you are helping to ensure the future of impactful stories about the discoveries and ideas shaping our world today.
After emerging from the ground following heavy rain, male hemisotids call from the temporary pools that have now appeared. After pairing, males and females then retreat to a location where an excavation is made and the eggs are laid. These ‘excavations’ are variable in form: some seem to be proper subterranean chambers or burrows well below ground level while others are shallow scrapes beneath logs, stones or leaf litter. Clutches consist of between 30 and 200 eggs (depending on the species) and are protected on their upper surface by “a few top layers of empty jelly capsules” (Channing 2001). The female stays with the eggs throughout development, and with the tadpoles.
Sometimes, later rains flood the nest chamber and the female then constructs a tunnel that connects the chamber to a pool. The tadpoles then follow her to the water, sometimes slithering onto her back and moving with her for the journey. The tadpoles then complete development in the pool they’ve travelled to. Hemisus tadpoles can be large (up to 65 mm long); they have a deep tail fin and are clumsy and slow-swimming. However, it seems that they don’t always end up completing their development in a pool: on occasion, they complete their development while remaining in a moist mass inside the brood chamber.
‘Bedfellows’ can be strange
Some of these aspects of reproductive behaviour are very similar to the behaviour also present in the brevicipitids we looked at last time. Coincidence? It seems not. Van der Meijden et al. (2004) were among the first to note that brevicipitids and hemisotids might belong together on the basis of molecular and anatomical data. Frost et al. (2006) then reported good molecular support for this hypothesis, and regarded it as robust enough for the naming of a newly discovered clade. Alas, the name they chose is fairly hideous: Xenosyneunitanura. The etymology is quite nice, the name meaning something like ‘strange bedfellow frogs’. The monophyly of Xenosyneunitanura has since been recovered by other authors (Pyron & Wiens 2011, Pyron 2014). Both ‘strange bedfellow’ groups share hydrostatic tongues of the sort discussed above as well as behavioural and molecular characters which seem to demonstrate close affinity, but they’re otherwise pretty different in appearance.
In fact, van Dijk (2001) compared the osteology of both groups and discerned “more differences than similarities”, noting how brevicipitids and hemisotids differ conspicuously with regard to their vomer bones, in the morphology of the vertebral column, in the shape of the vertebral centra, form of the articular surfaces on the atlas, shape of the clavicles, sternum, scapula, hindlimb bones, and so on and on. Van Dijk (2001) concluded that the two groups must be but distant relatives.
As we've just seen, however, this is very clearly contradicted by non-skeletal data. The molecular, behavioural and soft-tissue characters linking these two groups make a very strong case. I conclude that strange bedfellow frogs are an excellent example of the fact that very close relatives might end up looking very, very different – we shouldn’t just expect close kin to be obviously similar in all respects. Of course, there are examples of this sort of thing all over the tree of life. Don’t get me started on teleost fish.
Where do the ‘strange bedfellow frogs’ fit within the anuran family tree as a whole? It seems that they’re part of the same major assemblage as the reed and lily frogs (Hyperoliidae) and the narrow-mouthed frogs (Microhylidae). This large clade – termed Allodapanura by Frost et al. (2006) – is close to Natatanura, the giant clade that includes ranids and all of their relatives. And that, as they say, is awhole ‘nother story...
One more thing. If you can recall what the recently discovered purple Indian frog Nasikabatrachus sahyadrensis looks like, you might understand why some anuran workers have suggested that it might be allied with hemisotids rather than with Seychelles frogs [adjacent photo by Karthickbala].
However, molecular data consistently finds Nasikabatrachus to be part of the Seychelles frog lineage (Biju & Bossuyt 2003, Frost et al. 2006, Pyron & Wiens 2011, Pyron 2014) – any similarity with shovel-nosed frogs has to be considered the product of convergence.
Strange bedfellows, convergence between highly disparate lineages, and frogs frogs frogs frogs frogs. Such is the nature of tetrapod evolution.
For previous Tet Zoo articles on frogs and toads, see...
In pursuit of Romanian frogs (part II: WESTERN PALAEARCTIC WATER FROGS!!)
The toads series comes to SciAm: because Africa has toads too
North American spadefoot toads and their incredible fast-metamorphosing, polymorphic tadpoles
Frogs you may not have heard of: Brazil’s Cycloramphus ‘button frogs’
‘Strange bedfellow frogs’ (part I): rotund, adorable brevicipitids
Refs - -
Anderson, C. W., Nishikawa, K. C. & Keifer, J. 1998. Distribution of hypoglossal motor neurons innervating the prehensile tongue of the African pig-nosed frog, Hemisus marmoratum. Neuroscience Letters 244, 5-8.
Biju, S. D. & Bossuyt, F. 2003. New frog family from India reveals an ancient biogeographical link with the Seychelles. Nature 425, 711–714.
Channing, A. 2001. Amphibians of Central and Southern Africa. Cornell University Press, Ithaca and London.
Pyron, R. A. 2014. Biogeographic analysis reveals ancient continental vicariance and recent oceanic dispersal in amphibians. Systematic Biology 63, 779-797.