What would be the explanation for the absence of pythons in Madagascar?
This is a fantastic question, and one that is rather contentious, and takes quite a lot of answering. So I’m sorry this has taken weeks to put together, but here is the extremely long answer to the question:
Until recently, it was believed that mainland Africa had only pythons and a single group of boas (sandboas, genus Eryx), while the rest of the boas boas were restricted to the Americas… and Madagascar.
Madagascar’s large constricting snakes were assigned in their early descriptions to the Boidae. Indeed, so alike are Madagascar’s boas to those of the Americas that it misled phylogenetic reconstructions of them:
In 1991, based solely on morphological characters (molecular phylogenetics was in its infancy), Kluge found Acrantophis and Sanzinia to be sister to the South American genus Boa. And so he synonymised them, transferring the Malagasy boas to the South American genus, thus rendering a genus Boa with the following members: Boa constrictor (and subspecies thereof), B. madagascariensis, B. dumerili, and B. manditra - Sanzinia madagascariensis obviously couldn’t become Boa madagascariensis, because Acrantophis madagascariensis had already taken that name, so he erected a new name for this species.
As it turns out, moving Madagascar’s boas to the genus Boa was a huge mistake. But Madagascar’s boas are boas (Family Boidae: Subfamily Sanziniinae), they just don’t belong in the genus Boa. A phylogeny produced by Noonan & Chippindale (2006a) showed that the genus Boa belongs to a neotropical clade, together with Epicrates, Eunectes, and Corallus. These are closest related to a group containing Afro-Indian Eryx plus Candoia from Papua New Guinea. These together are then related to the North and Central American boas (Exiliboa+Lichanura+Charina). Finally, sister to this whole group of New World+Australiasian boas is a clade containing Calabaria, Acrantophis, and Sanzinia. This whole group is formalised as Boidae, and is sister to the Pythonidae in the Noonan & Chippindale phylogeny.
Until this point, Calabaria were thought to be allied to pythons and not to boas, but the molecular work of this and subsequent studies showed that to be quite incorrect. So boom, another group of boas is present in Africa. Woo.
All subsequent phylogenies (and several earlier ones that I can’t be bothered to go find) have supported this view that Madagascar’s boas belong in their own genera (Sanzinia and Acrantophis), and yet still Wikipedia would have you believe that they are in Boa. I will fix this in the next few weeks, but in the mean time, ignore it. Focus on my argument here.
In light of more complete taxonomic and genetic sampling by Pyron et al. (2013), it was shown that boas and pythons aren’t even sister groups (which we knew before, but never with quite such nice resolution and such high taxon sampling). Rather Boidae+Calabaridae are sister to a group containing Anomochilidae+Cylindrophiidae+Uropeltidae+Xenopeltidae+Loxocemidae+Pythonidae. Together, these two clades comprise the superfamily Henophidia (ancient snakes), which is paraphyletic with regards to all other snakes except the blind snakes, but shh we can ignore that for now.
Anyway, the important thing is that both of these reports (Noonan & Chippindale 2006a and Pyron et al. 2013) agree on one key point: Madagascar’s boas are unequivocally the oldest radiation of boas.
How, then, did Madagascar’s boas get to Madagascar? Well, Noonan & Chippindale (2006a, 2006b) suggested a vicariant origin; that they were in Madagascar when it broke off from Gondwana as it split. This vicariant origin is further supported by the rapid diversification in deep branches in the boids (Noonan & Chippindale 2006a).
However, the age of the African clade (Calabaria+Acrantophis+Sanzinia) suggested by Noonan & Chippindale (2006a, 2006b) is 30 million years after Madagascar lost contact with Antarctica, and through Antartica, South America; and 11 million years after Madagascar lost contact with India and became isolated (Samonds et al. 2013). It would be prudent therefore to interpret this as either incorrect inferrence by Noonan & Chippindale in both papers, or indication of oceanic dispersal. I lean towards the results of Noonan & Chippindale being wrong because of the way they calculated their clade ages (poor outgroup choice with an incorrect date).
After the split from Africa, Madagascar maintained connection with India and, through a land bridge connecting to Antarctica, was connected to South America, and could continue to exchange species with it for at least a few million years (Noonan & Chippindale 2006a; Samonds et al. 2013). This left Madagascar with its own boas, one of which apparently dispersed to Africa to give rise to Calabaria. Not until much later did Madagascar’s modern boa genera arise on the island.
Today, most authors (e.g. Pyron et al. 2013; Reynolds et al. 2014) tend to assume that Calabaria, which is a troublesome group, genetically, is actually basal to all boas, and do not include it in the true boas. They do not, however ,make any comment on the implications of this, in terms of origins.
Bizarrely, the Round Island Boa (Casarea dussumieri) is not related to those of Madagascar at all (Round Island is very close to Madagascar - just north of Mauritius), but is instead most closely related to a Malaysian lineage, and only distantly related to other constrictors at all (Reynolds et al. 2014). How it fits into this picture is very unclear.
The implication then is that pythons and boas split long before Madagascar broke off from Africa. Very recent (read: not yet physically published but available online) research by Reynolds et al. (2014) has suggested an origin for the modern pythons in what is today northern Africa, whence they then spread south, north, and east. The boas, on the other hand, are suggested to have arisen in southern Gondwana by some authors (Reynolds et al. 2014) and Asia by others (Noonan & Sites 2010).
But the problem is that Reynolds et al. (2014), and indeed all authors working on the relationships of the Henophidian snakes, have not attempted to figure out where and when the other families arose. Thus, we have no idea where the ancestors were at the time of the split of Gondwana. All we can postulate is that Madagascar had boas when it split off (perhaps getting them after Africa was isolated from the rest of Gondwana via land bridges to Antarctica+South America; Samonds et al. 2013), while pythons were further up in Africa somewhere. Then, when pythons radiated down into Africa, they didn’t make it across to Madagascar - or if they did, they went extinct soon afterwards.
tl;dr: Nobody knows for sure. Nobody has really even tried to find out. Probably because it’s not a very interesting question (relatively speaking).
Kluge, A. G. (1991) ‘Boine snake phylogeny and research cycles' Miscellaneous Publications, Museum of Zoology, University of Michigan 178:1-58
Noonan, B. P. & P T. Chippindale (2006a) ‘Dispersal and vicariance: The complex evolutionary history of boid snakes' Molecular Phylogenetics and Evolution 40:347-358
Noonan, B. P. & P T. Chippindale (2006b) ‘Vicariant Origin of Malagasy Reptiles Supports Late Cretaceous Antarctic Land Bridge' The American Naturalist 168(6):730-741
Noonan, B. P. & J. W. Sites Jr. (2010) ‘Tracing the origins of iguanid lizards and boine snakes of the Pacific' The American Naturalist 175(1):61-72
Pyron, R. A., F. T. Burbrink & J. J. Wiens (2013) ‘A phylogeny and revised classification of Squamata, including 4161 species of lizards and snakes' BMC Evolutionary Biology 13:93
Reynolds, R. G., M. L. Niemiller & L. J. Revell (2014) ‘Toward a Tree-of-Life for the boas and pythons: Multilocus species-level phylogeny with unprecedented taxon sampling' Molecular Phylogenetics and Evolution 71:201-213
Samonds, K. E., L. R. Godfrey, J. R. Ali, S. M. Goodman, M. Vences, M. R. Sutherland, M. T. Irwin & D. W. Krause (2013) ‘Imperfect isolation: factors and filters shaping Madagascar’s extant vertebrate fauna' PLoS One 8(4):e62086
