No evolutionary convergence between Australia and southern Africa in snakes and legless lizards, part 1

@herping_with_berks @hamishrobertson @ockertvs @alexanderr @bushboy @jouberth @tyroneping @ptexis @snakesrcool @halvard_midtun @sanjoaquinserpents @wolfgang_wuster @michael_jacobi @wildlife1607 @vinaygogula @ludwig_muller @asimakis_patitsas @owenlishmund @tommyh44 @adammyates @graytreefrog @zdunek_herp @chrisvankalken @alexey_katz @wyn_russell @calebcam @cr_hundermark

Fitzgerald River National Park ( and and its environs in Western Australia, and Agulhas National Park and its environs in South Africa, have similar climates.

Indeed the climates are so similar that they provide a near-ideal opportunity for testing the notion of evolutionary convergence ( and

The substrates too are extremely similar: nutrient-poor and mainly sandy.

In both study areas there is a littoral strip including calcareous dunes, partly fixed by vegetation.

The vegetation inland, on plains and low hills, is broadly similar between continents. Typical vegetation is mallee-heath ( and in the Australian study area and proteoid fynbos ( and in the South African study area. However, there are also patches of thicket/tall shrubland, and a few small patches of woodland/forest in both study areas.

This constitutes perhaps the closest matching in overall environmental regimes that has ever been achieved in intercontinental comparisons ( and

However, it is clear that any prediction of evolutionary convergence is not borne out by the faunas of either mammals ( or snakes and legless lizards.

Please see part 2 for the lists of species.

The only well-matched pair of species is Pseudonaja affinis ( vs Naja nivea (

However, this is weak confirmation of evolutionary convergence because both belong to the same family (Elapidae). Furthermore, even in this case:

Anilios australis ( and in the Australian study area is a seeming counterpart for Rhinotyphlops lalandei ( and in the southern African study area. Unsurprisingly for members of the same family of snakes (, these species are similar in shape, colouration, and diet (brood of ants).

However, these species are too closely related to be considered convergent. Furthermore, the Australian species is the larger-bodied.

Furthermore, there is no typhlopid counterpart in the Australian study area for Leptotyphlops nigricans ( - which belongs to the same family as the smallest-bodied species of snake on Earth (

The elapids Acanthophis antarcticus ( and Echiopsis curta ( and show considerable evolutionary convergence with the viperids, Bitis spp. However, these elapids remain less specialised than the viperids in e.g. body girth, fang mechanism, and skin-texture. Furthermore, A. antarcticus, relatively small-bodied, is rare in the study area and has neurotoxic venom, whereas Bitis arietans, relatively large-bodied, is common in the study area and has cytotoxic venom.

Beyond the above cases, the differences I form and function outweigh the similarities in the faunas compared.

There are fewer species, in the composite category of snakes and legless lizards, in the Australian than the southern African study area.

This is partly because there is far less specialisation in the Australian than in the southern African study area for staple diets of:

  • termites (Leptotyphlops nigricans, Rhinotyphlops lalandei),
  • slugs (Duberria lutrix),
  • frogs (Amplorhinus multimaculatus, Crotaphopeltis hotamboeia),
  • blind snakes and fossorial lizards (Homoroselaps lacteus),
  • arboreal vertebrates, i.e. birds and chameleons (Dispholidus typus),
  • eggs of birds (Dasypeltis scabra) and reptiles (Prosymna sundevalli, if present), and
  • rodents (Pseudaspis cana and Bitis arietans).

This is balanced by greater specialisation in the case of Australia only for:

  • large spiders (Pygopus lepidopodus), and
  • ants (Aprasia spp. and Anilios australis).

Another way of expressing the above is that the community of snakes is less differentiated in the Australian than in the southern African study area.

One aspect of this differentiation is venom.

There are virtually no snakes specialising on a diet of snakes in either study area. Closest is Homoroselaps lacteus, which eats legless skinks as well as typhlopid snakes. Its closest counterpart in the Australian study area is Suta nigriceps, which is not fossorial to the same degree but has been recorded eating typhlopids.

An approximate match for Leptotyphlops nigricans can be sought in the pygopodid ( genus Aprasia, which contains up to three species in the Australian study area. However, these legless lizards:

  • are specialised on ants (eggs, larvae and pupae) instead of termites,
  • are not as slender or small-eyed as L. nigricans,
  • lay fewer eggs at a time,
  • have tails that are proportionately longer than in snakes and can be lost by autotomy, and
  • possibly spend more time at the surface, associated with ant-dominated habitats on extremely oligotrophic soils.

What this means is:
In both the Australian and the southern African study areas there are myrmecophagous (, small, legless reptiles. However, these differ in being specialised on ants in the former and termites in the latter - undermining even the similarity between typhlopids.

In both study areas there are forms specialising dietarily on lizards. However the Australian species, Lialis burtonis, is morphologically remarkably different from the southern African species in genus Psammophis, partly because the former is a lizard whereas the latter is a snake.

The elapid Rhinoplocephalus bicolor ( and the lamprophiid Lamprophis fuscus ( are comparable in appearance, diet and habits. However, R. bicolor is smaller than L. fuscus, and live-bearing instead of oviparous. The Australian species eats mainly skinks (, whereas the southern African species eats mainly rodents.

Elapognathus coronatus and Suta nigriceps are not as specialised, in either habitat or diet, for a diet of frogs as are Amplorhinus multimaculatus and Crotaphopeltis hotamboeia.

Evolutionary convergence may remain a valid principle, conforming as it does with those most basic concepts in biology, natural selection and adaptation.

However, even the most similar climates and soils on different landmasses remain just too different in their overall environmental regimes to allow a real test. This is a particularly intractable problem because such an important aspect of the regimes is the biotic aspect itself. Even in these extremely well-matched physical environments and even when the families are the same (Elapidae and Typhlopidae), the biotic opportunities/pressures are crucial.

Hence the cobra-like snake in the Australian study area still differs from the Cape cobra in having a minimally-expressed defensive hood, and in having smaller offspring. This is explained by the Cape cobra having to contend with an intense predatory regime compared with quasi-insular Australia.

In the case of blind snakes the lesser productivity of termites in Australia (see means that the dietary opportunities remain significantly different between continents.

Please see part 2 for the lists of species...

Posted on 02 March, 2022 21:12 by milewski milewski


The legless lizards Delma fraseri (in the Australian study area) and Chamaesaura anguina (in the southern African study area) both take grasshoppers as a significant part of their diets.

Delma fraseri is comparable with the snake Duberria lutrix inasmuch as both are small and legless with diets of invertebrates. However, the former is far less fecund than the latter. Delma fraseri produces only two eggs at time, whereas Duberria lutrix produces 3-12 and up to 22 active offspring at a time.

Along similar lines, Aprasia spp. are less fecund than Leptotyphlops nigricans.

Posted by milewski over 2 years ago

Typhlopids more similar to Leptotyphlops nigricans occur in tropical Australia. Herp elements associated with tropical grasses reach the southern tip of Africa but not Australia, and the same pattern can be seen in the frog Hyperolius.

Posted by milewski over 2 years ago

Termites may possibly be eaten by skinks (with legs) to a greater extent in the Australian than in the southern African study area.

Posted by milewski over 2 years ago

A specialist on large spiders (Pygopus lepidopodus) is common in the Australian study area. No reptile of any kind specialises on ants in the southern African study area. Nucras lalandii takes ants as 0% of the volume of its diet (

Posted by milewski over 2 years ago

There is a notable lack anywhere in Australia of legless lizards with tails shorter than snout-vent length (check Aprasia) and with very small eyes (compare Acontias).

Posted by milewski over 2 years ago

Well reasoned, well written, and an excellent use of the data and images available at iNat. It's a pleasure to see a serious use of the amazing resources offered by iNat. If convergence occurs because of similar selective pressures in disjunct environments, I think it is demonstrated that very different resources available to organisms in similar environments can result in some apparent convergences in similar appearances . However, underlying modifications more reflect the availability of food to create those organisms, and this results in many specializations and many significant differences that are not readily apparent. I think some of the animals do superficially look alike, but they are very different creatures. Interesting!

Posted by ptexis over 2 years ago

Interesting, how come Australia has a pythonid that can tolerate such mild climates, but in southern Africa they are restricted to more subtropical regions with summer rainfall? And why is this group so diverse in Australia compared with Africa?

Posted by alexanderr over 2 years ago

Very interesting, I enjoyed reading this. It is great to see data from the website go to good use.

Posted by wildlife1607 over 2 years ago

@alexanderr it is true that Pythonidae extend farther south in Australia than in Africa. However, even Python natalensis extends about to the latitude of Cape Town, i.e. about the same latitude as my study area in Australia (Fitzgerald River National Park). What is more different between the continents is that the python Morelia extends throughout the winter-rainfall area in Australia whereas its South African counterpart does not enter the winter-rainfall area at all.

One way to explain this difference in ecological distribution is to see Morelia as a suitably slow-metabolising replacement for warm-blooded predators missing from Australia, such as Genetta. The prey available in Australia is insufficiently productive to support a mammalian predator and so a reptile occupies the niche, as it were.

The closest counterparts for pythons in the Western Cape of South Africa are the lamprophiids Boaedon and Lamprophis, which belong to a family absent from Australia. There is broad convergence in the sense that in both cases we are dealing with non-venomous constrictors that prefer rodents, tend to be attracted to human habitation for the rats, and have sexual dimorphism in which females grow larger than males.

However, this convergence is outweighed by the fact that Morelia remains more arboreal than the lamprophiids, often eating birds such as the psittaciforms so common in Australia and so uncommon in South Africa. In Morelia the sexual dimorphism is extreme compared to the lamprophiids. For their part the lamprophiids are more terrestrial and more reliant on rodents, and thus more often commensal with humans.

And there is likely to be a considerable difference in fecundity, in favour of the lamprophiids - which is a common theme in comparisons of various animal lineages between these two continents. This is mainly because Morelia does not attempt to breed every year, whereas the lamprophiids do. The more intense predation in Africa goes with greater fecundity.

So this is yet another case of 'so near and yet so far', in terms of evolutionary convergence.

Moving northwards from the temperate zone: pythonids and varanids show a similar pattern in both being more diverse and differentiated in Australia than in southern Africa. Again, I suggest that an explanation lies in seeing these reptiles as metabolically inexpensive counterparts for mammalian and avian predators in Africa.

Compare, for example, the largest-bodied varanids on the two continents: Varanus giganteus vs Varanus albigularis. These are not merely counterparts, because the Australian form is a slender, fast-moving replacement for e.g. Vulpes chama, whereas the southern African form is a massive, sluggish semi-specialist on armoured prey (tortoises, large millipedes, etc.).

In summary, much of the intercontinental differences between Australia and southern Africa involves the more productive prey, particularly rodents, in southern Africa. Once this basic difference is recognised, various anomalies in the comparisons start to fall into place.

Posted by milewski over 2 years ago

@wildlife1607 Many thanks and I will continue to apply this use to other groups of organisms on iNaturalist.

Posted by milewski over 2 years ago

@ptexis Hi David, Many thanks for your supportive comments, with regards from Antoni.

Posted by milewski over 2 years ago

@milewski thanks. I was not aware that there was a lack of small mammalian predators in Australia. I was wondering if the archipelago of Indonesia was the engine for pythonid speciation in Australia and with the high diversity and competition, some species would inevitably adopt to unusual niches. Do you think that is part of the picture?

And on another topic, is the lack of mammals in Australia due to the nutrient poor soils, or because Placentals did not make it into Australia and overwhelm over groups?

Posted by alexanderr over 2 years ago


Maybe we can discuss your second question first.

Let's clarify what exactly is lacking in Australian mammals, relative to say southern Africa. This is important because there is actually a bewilderingly large number of mammalian spp. in the Australian fauna, including important eutherians.

The most glaring lack is in carnivores and large insectivores. There has never been a marsupial counterpart for felids, hyenids or herpestids, and surviving Dasyuridae are small and generalised even relative to that primitive family, Viverridae. This is one of the most 'island-like' features of the Australian fauna, and it is consistent with the remarkable tameness of e.g. relatively large birds in Australia.

Although there are many and various folivorous/herbivorous marsupials and a few largely herbivorous rodents in Australia, the only species with a pace of life to rival the eutherian norm is Rattus vilosissimus ( All the rest lag in metabolism and/or growth and reproduction. And the rat solves nothing because it belongs to a genus with fast pace of life

The rodents are smallish and generalised, with no counterparts for porcupines, squirrels, mole-rats, etc.

There is no counterpart for monkeys.

These are probably enough to go on with.

All these major lacks can indeed be explained at least partly by nutrient-poverty, but this only goes so far because even the poorest soils in southern Africa retained mammals larger-bodied and with a faster pace of life than their closest counterparts on even the richest soils in Australia.

So is this residual difference owing to an accident of history in that eutherians, other than rodents, bats and pinnipeds, failed to reach Australia?

This is a question I've pondered for years, and I'm still unsure.

Suggesting that there has not been this sort of phylogenetic limitation are the following observations:

1) rodents and pinnipeds have been in Australia for so long that there would seem to have been ample time for them to evolve into the lacking forms. Why can a rat not evolve into a squirrel-like form in five million years? Or a sealion (see,Holocene%20from%20both%20the%20Northern%20and%20Southern%20hemispheres.) into a hyena-lookalike?

2) mole-rats and similar forms have evolved independently several times elsewhere on Earth, so why not in Australia?

3) reproductive strategies are actually remarkably varied among marsupials, and Didelphis in the Americas has managed to achieve fecundity to rival eutherians

4) some marsupials (particularly Peramelidae) do actually have placentas, but without the accompanying features seen in eutherians (see,delivers%20nutrients%20to%20the%20embryo.)

5) most significantly of all, the pattern in mammals is echoed by birds, herps and even invertebrates. Even among orthopterans and isopterans, the same theme of 'lacking' of metabolically/reproductively powerful forms in the Australian fauna crops up in family after family. And even in the case of eagles, crocodiles, varanids, pythonids, etc., the largest-bodied species in Africa remains more massive - and also more fecund if fecundity differs - than the largest-bodied species in Australia.

So there is great phylogenetic plasticity in marsupials, suggesting that the absence of most eutherian families has limited explanatory value. And the 'straitjacket' imposed on clade after clade of Australian fauna suggests that a failure to cross the sea barrier to the island continent hardly explains the problem. Why has there never been a vulture here, or a woodpecker, or a fungus-culturing termite, or a large-bodied dung beetle?

Does this make sense?

Posted by milewski over 2 years ago

@alexanderr Hi again Alex, to your first question:

Is there really a mechanism in evolutionary biology whereby adaptive radiation, however rapid, in one region propels, as it were, the competitiveness of emigrants in a nearby region? How would this work? I can see that the radiation of pythonids in Indonesia could mean that many taxa would have a chance of reaching Australia. However, surely what happens then depends only on whether Australia is suitable for survival and further evolution?

Let's take fynbos plants as a perverse example. There has been a great radiation of proteas in the Cape Floristic Region. And there are reltively few species in tropical/subtropical Africa. Does the radiation in the southwest inform the occurrence in the tropics? Or is it just the case that, once a clade is present, species arise and persist purely according to the ecological opportunities?

Imagine that the miombo ecosystem has a) no proteas, or b) 10 spp. in four genera, instead of the current reality of just a few spp. in one genus. Would the 'engine' in the far south explain the situation, either way?

Posted by milewski over 2 years ago

Wow! Amazing write up and incredible research! Well done @milewski !

Posted by ludwig_muller over 2 years ago

@alexanderr we have a feral python living in this area- very happy and successful. There have also been reports of Boa constrictors from Montagu pass. So Pythons can survive in the winter rainfall area, though I guess their inability to increase is due to the cold here.

Posted by ludwig_muller over 2 years ago

Fascinating information, thanks.

Second question: I would think that part of the reasons why some of those mammalian adaptations did not happen is because other groups filled the niche first, like varanids taking over the role of scavenging. But were Australia connected to other continents and accessible to the Placentals I'm sure they could outcompete most of these groups.

First question:
I think what I am trying to say is that in Africa pythonids (at least P. sebae and natalensis) are widespread generalists and there are few opportunities for vicariance and subsequent specialisation. In Australia you have a mechanism for vicariance, and you also have a higher pythonid diversity.

Whether 'adaptive radiations' could propel the competitiveness of emigrants in neighbouring regions is a slightly different topic I think.

@ludwig_muller you do? can you give more information on species/population size please? Yeah one individual surviving short-term is very different from the long-term persistence of a viable population. A population has to make it through the highs and lows.

Posted by alexanderr over 2 years ago

The one near our place is probably natalensis. he has been here for around 25 years, and lives on a north facing cliff with plenty of Dassies around.
The other records are not so detailed. I know that a large boa (over 2m in length) has been seen on Montagu pass and photographed once by cyclists who found him crossing the road at night. There is a small population of feral rabbits in that area, but apart from them I can't imagine what he lives on, as fynbos habitat is usually poor in nutrients and large mammals.

Posted by ludwig_muller over 2 years ago

Sho crazy, does anyone know the origin of the python?

Posted by alexanderr over 2 years ago

yes- released by the reserve manager when it grew too large.

Posted by ludwig_muller over 2 years ago

@alexanderr Hi again Alex,

Many thanks for the last point you made about varanids precluding the niches of mammals in Australia.

But are you aware that varanids themselves invaded Australia from the north relatively late (, as opposed to being original in the 'gondwanan' sense? This means that marsupials or pre-marsupial ancestors have been on what became Australia right from the start about 60 million years ago (see, whereas varanids invaded only about 15 million years ago when Australia approached Asia closely enough.

The varanids would have had to establish themselves and radiate evolutionarily in the face of not only the already-established marsupials but also an already-established fauna of lizards/lizard-like reptiles.

This to me suggests a contest of competitive superiority rather than a contest based on precedence. Do you see this differently? with regards from Antoni

Posted by milewski over 2 years ago

"But are you aware that varanids themselves invaded Australia from the north relatively late". I did not, thanks for that.

Egg-eating snakes are interesting, and probably worth thinking about outside of Australia too since Dasypeltis is restricted to the African continent (and although it marginally enters North Africa it has not expanded into Europe) and the only other obligate egg-eater (that I found with a quick search is Elachistodon westermanni in India. Can we link these snakes to gregariously nesting birds? Many snakes eat eggs, but to limit themselves to only eat eggs would require such an exploitable resource. Interestingly, at least in South Africa, Daypeltis seem to be able to survive in regions without gregariously nesting birds... Interested to hear your take on this.

Posted by alexanderr over 2 years ago

@alexanderr I agree that we should find out more about the habitats of Dasypeltis spp., particularly in semi-arid areas with largely treeless vegetation. There are enough naturalists familiar with Dasypeltis on iNaturalist that we should be able to get some idea by asking around. As a working hypothesis, I suggest that Dasypeltis depends mainly on Ploceidae, and that many/most of the observations in semi-arid areas and on the Highveld are associated with homesteads where non-indigenous trees have been planted, attracting Ploceus. I note that there are several spp. of Ploceus in India, which might also help to explain Elachistodon. Ploceidae are notably absent from Australia. Your further thoughts?

Posted by milewski over 2 years ago

I think there are enough ground nesting birds in those areas for Dasypeltis to survive. I do not think that Dasypeltis has expanded due to homesteads (but perhaps has increased in abundance), across Africa, Dasypeltis are present in forest, desert and almost everything in-between.. Perhaps the gregariously nesting birds facilitated the specialisation of Dasypeltis, which could then move on to other nesting birds. But the question is why did they not expand into Europe or Asia, and why have similar adaptations not occurred in snakes from other parts of the world? Perhaps they have, but are just not as specialised as in Dasypeltis.

Here are some publications that discuss the subject a bit:

Posted by alexanderr over 2 years ago

@alexanderr Many thanks for these interesting papers. Let's keep working on this puzzle, which goes beyond Dasypeltis. There are about 16 spp. of Prosymna, and I wonder if all of them are truly specialised for reptile eggs ( Do Asia or the Americas have similarly specialised snakes? Does any African snake eat only amphibian eggs, and are African snakes more specialised in this way than snakes on other continents?

Posted by milewski over 2 years ago

I'm afraid I do not know enough about that subject to provide much insight.
Do Asia or the Americas have similarly specialised snakes? - The other snakes listed in the abstract are not African, so yes I think so.

Does any African snake eat only amphibian eggs - I don't know of any, but I know it is well document in the Central America (I think). Africa's Dipsadoboa may be potential candidates, but I doubt any exclusively feed on them. America is way ahead on Africa on these kinds of Natural History observations.

Bryan Maritz is the person to talk to if you seriously want to bounce around snake dietary ideas.

Posted by alexanderr over 2 years ago

@ludwig_muller Hi Ludwig, I just spent a whole day writing a Post about vertebrate-eating birds, only to have it mysteriously vanish as if deleted. However, I did not consciously delete it and so I have no idea what went wrong. Do you know how to restore items such as journal Posts on iNaturalist? With many thanks from Antoni

Posted by milewski over 2 years ago

Hi Uncle Antoni,
Did you save it before closing the tab? I had the same experience with my Habenaria article- inat does not save drafts unless you click the Save draft button beneath the article.
Sorry to hear about your loss- if you did save it it should show up in Drafts.

Posted by ludwig_muller over 2 years ago

Hi Ludwig,

I saved the Post many times today as I worked on it. I can only imagine that, on my most recent save, I accidentally pressed 'unpublish', which is right next to 'save'. If it is saved somewhere in the system despite having been (presumably) unpublished, how can I locate the saved copy?

Posted by milewski over 2 years ago

@ludwig_muller I think I understand what you mean by 'save 'draft'. I assumed it simply to mean an incremental saving but it is more like an initial setting-instruction, not so? If so, I have learned a hard lesson today.

Posted by milewski over 2 years ago

Hi Uncle Antoni,
If you saved your draft it should show up in your journals- on the right side of the screen- under Drafts.
Maybe you should ask @tonyrebelo about this?

Posted by ludwig_muller over 2 years ago

@tonyrebelo I spent all day on a Post only to have it vanish. I don't know what I did wrong, and I certainly did not confirm any deletion. I saved innumerable times as I went along. Do you think the work can somehow be retrieved? With many thanks from Antoni

Posted by milewski over 2 years ago

Ask the help desk on iNaturalist ( - it is possible that you might be able to revert to an earlier edition. But I am not certain.
Personally, complicated pieces I save in Word, and only paste it here at the end. And I use Typio Quick Restore in case it does happen.

Posted by tonyrebelo over 2 years ago

G Swan and S Wilson (2008), on page 60 of 'What snake is that? Introducing Australian snakes', state: "Most Australian elapids eat lizards...Nearly 80 per cent of the 97 terrestrial Australian elapids rely heavily on lizards as a major part of their diet, and almost half eat nothing but lizards from birth to adulthood. The large snakes with generalist diets start life by eating lizards before graduating to include more diverse prey."

Posted by milewski over 2 years ago

The elapid Rhinoplocephalus bicolor ( is particularly associated with the stick-nest' of the ant Iridomyrmex conifer ( and This small-eyed species of snake, about 450 cm long, eats mainly small skinks - which also shelter in these 'nests'. Ants play a negligible role in the diets of either the this species of snake or the skinks it predates. No species of ant in the southern African study area builds a similar 'nest'. The closest counterpart to R. bicolor in the southern African study area is the lamprophiid Lycophidion capense (40 cm and up to 64 cm). However, the dentitions of the two species of snakes differ too much for this to be interpreted as a case of evolutionary convergence.

Posted by milewski over 2 years ago

Australia seems to be the only continent lacking any species of snake specialising on molluscs.

Africa: Duberria lutrix ( and several congeners

North America: Contia tenuis ( and Storeria occipitomaculata (

Central and South America: Sibon nebulatus (, many species of Dipsas (, and several other genera

Southeast Asia: Pareas carinatus ( and various congeners

The above species belong to three different families, namely Lamprophiidae, Colubridae, and Pareidae.

Posted by milewski over 2 years ago

Add a Comment

Sign In or Sign Up to add comments