Madagascar's Lucky Break and Endemic Invaders

Evolutionary consequences of Madagascar's isolation on extinct and extant vertebrate fauna

© O'Reilly, X. (2010)
Madagascar lies off the heal of the African continent. The world’s fourth largest island, on the map it is dwarfed by Africa’s shadow – but only in size.  Famous for the immense variety of weird and wonderful organisms it is home to, most of which are found nowhere else, Madagascar’s unique biological catalogue is the result of millions of years of evolution in isolation, as well as its dramatically varied relief. Its most famous inhabitants, by far, are the lemurs – a whole group of endemic primates that come in a variety of forms, each uniquely adapted to specific niches throughout island’s amazing range of habitats.
  But lemurs are thought to have arrived on the island after its split from Africa – what was living on Madagascar at the time of its birth? What proportion of its current terrestrial vertebrate fauna actually split away with the island and what proportion was to invade it after isolation? Certainly Madagascar received terrestrial migrants despite its geographical isolation – but how did they get there?

  Madagascar was ripped from Africa’s side some 165 million years ago (Rabinowitz et al., 1983), together with what 88 million years ago split off to charge into the Asian mainland and become India (Storey et al., 1995).

  For millions of years Madagascar has been cut off from any other land mass, giving natural selection full reign to explore and exploit the island’s staggering variety of habitats born from it’s massive geological diversity. Today it is home to an incredibly diverse array of organisms, most of which are endemic, such as the aforementioned lemurs.

  Lemurs are descended from an ancient line of primates which once thrived across the ancient continents (Rose, 2006), the members of which were out-competed by the more advanced monkeys and apes with whom they share their roots. They were gradually wiped out everywhere except Madagascar, where – isolated from the continent, which was cradling the new forms of primate, and formidable new predators, for that matter – they were left to radiate and diversify into a myriad of different species, flourishing in a world lacking most of the competitors and threats faced by their mainland cousins. This all seems very straightforward and logical, but there’s a catch: 165 million years ago, when Madagascar started breaking away from Africa, mammals were still scuttling about between dinosaurs’ feet and primates had not yet hit the scene (Silcox, 2007).

  Research into the origin and evolution of these curious little primates and the rest of Madagascar’s weird and wonderful fauna (and flora!) is ongoing and in many cases controversial, but phylogenetic and fossil evidence converge on one now undisputed point: the founders of Madagascar’s major extant lineages of vertebrates appeared on the island after its isolation (Shubin and Sues., 1991; Flynn and Wyss, 2003; Tattersall, 2007), meaning some means of crossing the water was necessary to establish the rich biodiversity it boasts of today, whether this be across intermittent continental bridges (McCall, 1997; Hay et al., 1999) or rafting on meshes of vegetation (Simpson, 1940; Ali and Huber, 2010). Separated by more than 400 km of ocean from Africa’s closest edge, 4000 km away from India, and over 5000 km from both Antarctica and Indonesia (Krause 2003), this is no mean feat for any land-dwelling animal.

  As well as posing the problem of determining a means of dispersal across such distances – how terrestrial vertebrate animals managed to reach the island and from where –, the appearance of Madagascar’s endemic present-day lineages subsequent to its isolation incites some less frequently asked questions: what was living on the land that was to become the world’s fourth largest island, and what happened to it?

  Madagascar’s entire journey from its separation from the rest of the ancient Gondwanan supercontinent, through its southbound migration and settling as “Indiagascar” in its present day position relative to Africa, right up to its final split with the India-to-be, all happened within the boundaries of the Mesozoic (see Figure 1), more commonly known as the “Age of Dinosaurs”. It is logical to assume, therefore, that the island, like the rest of the world, was ruled by this mighty dynasty of animals; and indeed the fossil record, sparse though it may currently be, certainly indicates that this was the case and that the shifting landmass boasted a wide range of reptilian diversity. Proof of archaic turtle relatives, ancient crocodyliforms, and the oldest dinosaur fossils known worldwide, have all been found on Madagascar, dating back to a time prior to its tectonic migration. Along with them, several other examples of diverse but extinct reptilian lineages as well as various synapsid reptiles (the line from which mammals later arose) represent most of what is known of the island’s terrestrial Triassic fauna (Flynn and Wyss, 1999). Interestingly, the fossil record of this period already seems to suggest some degree of endemism of Malagasy fauna by the mid-Triassic; however, this is more likely to be the consequence of a presently poor collection rather than a unique biota during this period, since the area was still surrounded by the rest of Gondwana (Flynn and Wyss, 2003).
Figure 1  Rough representation of Madagascar’s known fossil record,
1999, and geological timescales from the Permian till the present day.
Rectangles represent the most fruitful fossil beds, all located on the
Western side of the island. 
© Krause et al. (1999)

  As already mentioned, the current state of Madagascar’s known fossil record is sparse, comprised of remains from a few sites rich in fossilised diversity widely scattered over the geological time scale between large gaps, of which there is simply no biological data available. One of these great gaps spans from the aforementioned finds dated to the Early and Mid Triassic until the Mid Jurassic, when the India-Madagascar complex finally became detached from Gondwana.

  So despite more Mid Jurassic fossils being recovered, the question of whether Madagascar’s dinosaur fauna during this time was still similar to that on Africa, or whether it was already more like the island’s Cretaceous diversity for now must remain a mystery (Flynn and Wyss, 2003). The reason this information might be significant is as an aid in assessing at what point the island’s fauna started showing signs of provinciality, which could give clues to how accessible Madagascar was during its separation and at what time its continental connections were truly severed. For although the dating of the split from Africa, the settling in its current position, and the split from India, is widely accepted and today rarely questioned (Krause, 2003), what remains controversial is the suggested presence of land bridges or lingering connections with either of the aforementioned land masses, or others such as Antarctica or islands in the geographic vicinity, due to volcanic activity, rises and falls in sea-level, etc. The degree of dinosaur cosmopolitanism might prove a helpful indicator of such possibilities because they were generally large animals – and excavations in Madagascar have yielded very large specimens indeed –, and so if the same were to be found somewhere else, connections between lands would certainly have to be greater than those needed by smaller animals, which in some cases may not need any directly terrestrial connection at all, as discussed later.

  Therefore, at the time of its split, Madagascar beheld a rich variety of reptiles. Which other vertebrate groups were at home on the diverging land? The scattered Jurassic finds include more reptilian remains and fragments of at least one mammal (Flynn and Wyss, 2003). Late Cretaceous strata, however, yield terrestrial representatives of all the major vertebrate groups.

  Dinosaurs were still present and diverse at this time, although, interestingly, only members of one of the two orders is represented: Saurischia, with no trace of the other (Krause, 2003). A noteworthy discovery among these saurischian remains are various Avian dinosaurs, including some specimens that have played a crucial role in supporting the theory of birds’ dinosaur origins (Forster, 1998), although volant vertebrates do not concern us here. Other terrestrial reptiles include snakes and an astonishing compilation of crocodyliforms (Krause, 1999), one of the most diverse found worldwide, that spans tiny specimens to giants and even includes a form that appears to have been herbivorous (Buckley et al., 2000). Mammal-like reptiles haven’t been found in these assemblages, instead various true mammal fossils have been uncovered, such as a marsupial and a gondwanathere (a type of mammal known from the Southern lands of the Late Cretaceous and early Tertiary) who’s closest relatives are forms that have only been found to occur in India and South America. Amphibians, too, are represented, although – just as the Malagasy amphibian fauna is today – only by anurans, i.e. frogs and toads (Krause et al. 2003).

  It would seem then, that all major “types” of vertebrates found on the island today are accounted for from these Late Cretaceous sites – with the exception of lizards, who’s presence during this time has now been affirmed (Krause, 2003), but the record of which is certainly scant even in comparison to that of other Malagasy vertebrate groups of the Mesozoic. From then on, Madagascar’s contiguous isolation would allow this basal stock of animals to evolve independently for the next 85 million years; some taxa, such as the dinosaurs, were to become extinct, while others would flourish and diversify to produce the rich array of forms we see on the island today. As simple and evolutionarily plausible as such an account may be, it is not what happened.

  As mentioned before in reference to lemurs, the ancestors of Madagascar’s modern terrestrial vertebrates reached the island after its isolation; this would not rule out the smooth scenario just presented, whereby today’s fauna evolved from the Cretaceous taxa, because Madagascar by then had long since split from Africa and any remaining connections with other land masses could have been intermittent (Hay et al., 1999), thus it is indeed appropriate to refer to Madagascar as an island at this point. However, no part of the late Cretaceous Malagasy vertebrate diversity has left descendants amongst today’s wildlife, and so none of Madagascar’s present day terrestrial vertebrate fauna is represented at all during the Mesozoic era (Krause et al. 1999). In other words, despite the fact that the main vertebrate groups were present and successful on the land throughout the Mesozoic, none of the lineages are continuous with today’s, the founders of which therefore had to have arrived during the Cenozoic.

  Once again, the frustratingly incomplete state of the fossil record poses another obstacle for palaeontologists to overcome in the narration of Madagascar’s biological history – another huge gap, spanning from the late Cretaceous finds until the late Pleistocene, more than 80 million years later. This gap is possibly the biggest problem when trying to explain the origin of Madagascar’s vertebrate fauna, because it means that from studying the fossil record the only conclusion one can draw regarding the modern vertebrate is that their foundations were simply not present before the Cenozoic; but their is no hard evidence to show when the basal stock of today’s groups arrived, let alone how.

© O'Reilly, X. (2010)
  Since it has been established that Madagascar became tectonically isolated well before the commencement of the Cenozoic, it may seem obvious at this point – although perhaps hard to imagine – that the ancestors of its land animals would have had to disperse over an impressive water barrier in order to colonise the island. And that’s if we assume the ancestors of Malagasy vertebrates are of African origin; indeed some recent terrestrial groups appear to show greater affiliation with Asian taxa (Jansa et al., 1998), and this challenges our imaginations to envisage the dispersal of purely terrestrial vertebrate animals over a huge expanse of ocean.

  Other hypotheses have been proposed, however, in the attempt to explain the arrival of modern vertebrate groups. The existence of various land-bridges between Madagascar and different landmasses has been suggested for various stages since its separation from Africa and from India. Some would date to a time prior to when the basal taxa of extant vertebrates would have needed to cross and aim to explain the distribution of certain fossil forms (Hay et al., 1999). A prominent and highly controversial proposal is that of a bridge across the Mozambique Channel. This connection would in theory have been formed by the Davie Ridge, a geological structure currently submerged, but which some authors argue may have been exposed during a period of the Cenozoic when sea levels were lower (McCall, 1997).  The Davie Ridge runs right through the narrowest gap in the Mozambique channel, but in an only slightly inclined northwest-southeast direction, so even if its land were exposed some degree of transoceanic dispersal would be necessary, although the distance would evidently be greatly reduced. The real problem with this theory, however, becomes clear when considering the structure of the Davie Ridge: It is not a continuos elevation, rather a series of irregular peaks, the highest of which are not only far apart from one another, but are barely tall enough for their aerial exposure to be plausible even given prehistoric fluctuations in sea level. It has also been suggested that the structure itself could have once been taller and has since been submerged at a tectonic level, but at present there is no hard geological evidence to support either of these possibilities (Rabinowitz and Woods, 2006).

  The idea of land-bridges in general poses another problem: they are not selective. Such connections allow the migration of any animal without discrimination, and this would not help to explain why major vertebrate groups common throughout the African mainland and even Asia are completely absent on Madagascar, while others show an unrivalled degree of radiation or are entirely endemic. Perhaps only a fraction of the migrant fauna managed to secure a foothold on the island and the rest simply never established themselves? But the ecology of the island gives no reason to suggest why many African taxa would not be able to thrive on it. A Tertiary land-bridge connecting Africa and Madagascar is, therefore, very unlikely.

  All this makes transoceanic dispersal seem ever more likely, despite how far fetched it may appear that ancestors of an entire country’s terrestrial vertebrates arrived on floating bits of vegetation. However, this is no simple assumption made in the absence of an alternative explanation. Although the idea originally emerged back in the 1940s (Simpson, 1940), over the years increasingly reliable data has emerged which supports it. Most of the accumulating evidence arises from phylogenetic analyses, which have enabled the estimation of when the different groups diverged from their closest sister taxa – and, in many cases, which these taxa are. Phylogenetic studies can provide insight where the fossil record leaves no clues; for example, whereas none of the current vertebrate groups are present in Madagascar’s Mesozoic fossil record, from recent phylogenetic analyses, it has been concluded that some (although a minority) of terrestrial vertebrate forms ancestral to today’s were actually already present during the Cretaceous, such as iguanids and boine snakes (Yoder and Nowak, 2006). This might seem to turn the whole line of evidence for dispersal given here on its head and render the entire preceding argument vulnerable to “What if they merely haven’t been found yet?”; but then there would have been no point to this discussion. The various phylogenetic analyses that have been made on many different groups of organisms still place the divergence and radiation of almost every one of Madagascar’s current terrestrial vertebrate lines to sometime during the Cenozoic (e.g. Jansa et al., 1999; Rieppel, 2002; Krause, 2003; Vences et al., 2004; Harmon et al., 2008).

  The collected phylogenetic data on Malagasy vertebrates to date hints towards a great number of dispersal events taking place during the Tertiary period (Rabinowitz and Woods, 2006), not only bringing unsuspected migrants to Madagascar, but there is also evidence to suggest numerous outward dispersals taking place, as is suspected in the case of the charismatic chameleons (Rieppel, 1987; Raxworthy et al., 2002) and the island’s only native rodents, the nesomyines (Jansa et al., 1999). Recent palaeo-oceanographic simulations and analyses have only lent further support; due to different global tectonic arrangements, it is no surprise that prehistorically, ocean currents were very different to today’s. A very recent palaeogeographic reconstruction and study of the Mozambique Channel has revealed that there was a time during the Tertiary when forceful ocean currents between Africa and Madagascar would have made it more likely than it is presently for animals to be swept away on floating vegetation or “floating islands” resulting from storms, or even to have swam, and seem to have pinpointed seasons of the year when these currents were strongest and the odyssey would have been considerably shorter. The data obtained from this study fits the dates of divergence for Malagasy mammalian taxa predicted from the phylogenetic analyses, thus it may explain why there have been no further terrestrial mammal arrivals on the island since the early Miocene, when the simulations suggest the present day currents were finally becoming established (Ali and Huber, 2010).

  What is more, the multiple cases of herpetological dispersal – including many out of Madagascar – need not all fit into this oceanographic model, as reptiles and amphibians have been shown to be better oceanic dispersers than mammals, and recent instances of such movements have even been recorded elsewhere, such as in Caribbean iguanas (Censky, 1998).

  Colonisation of Madagascar through multiple dispersal events thus seems to have provided the basis for a great part of the magnificent biodiversity it boasts of today, the primary material to produce such a large and varied array of endemic forms. One further question remains, that of the island’s original fauna, which had diverged from its natal stock of Gondwanan vertebrates – What of the Late Cretaceous diversity the fossil record implies? Could the late Mesozoic forms have been pushed to extinction in light of the arrival of other vertebrates on drifting vegetation? Due to the absent Tertiary fossil record, it’s impossible to tell exactly when the Cretaceous forms went extinct, but it seems very unlikely that the arrival of involuntary migrants could cause the disappearance of an entire fauna that had evolved to thrive in its habitat. Certainly, the mass extinction that marks the boundary between the “Age of the Dinosaurs” and the “Age of Mammals” rid the planet of just about every animal over 10 kg., but even relatively few types of small animals survived. For example, at that time the largest mammal is likely to have been no bigger than a beaver, yet most mammalian taxa worldwide went extinct (Rose, 2006).

  Thus, the most likely scenario is that most terrestrial vertebrate forms disappeared from Madagascar during the K-T extinctions as they did on a global scale, so throughout the millennia that followed, the unfortunate vertebrates from distant lands that were swept away and marooned on Madagascar, actually found themselves in a place not only climatologically and geologically diverse in itself, but with plenty of empty niches to fill, providing enormous scope for radiation and specialisation, the result of which we only see a fraction of today – and this fraction still boggles the mind of even the most experienced field biologist.

Sara-Xaali O’Reilly Berkeley, March 2011


Ali, J. R. and Huber, M. (2010). Mammalian biodiversity on Madagascar controlled by ocean currents. Nature, Vol. 462, pp 653–656.

Buckley, G. A., Brochu C. A., Krause, D. W., and Pol, D. (2000). A pug-nosed crocodyliform from the Late Cretaceous of Madagascar. Nature, Vol. 405, pp 941-944.

Censky, E. J., Hodge. K., and Dudley, J. (1998). Over-water dispersal of lizards due to hurricanes. Nature Vol. 395, p 556.

Flynn, J.  J., Parrish, J. M., Rakotosamimanana, B., Simpson, W.  F.,Whatley, R. L.,  and Wyss, A. R. (1999). A Triassic Fauna from Madagascar, Including Early Dinosaurs. Science, Vol. 286, pp 763-765.

Flynn, J.  J. and Wyss, A. R. (2003). Mesozoic Terrestrial Vertebrate Faunas: The Early History of Madagascar’s Vertebrate Diversity. In The Natural History of Madagascar. Goodman, S. M., and Benstead, J. P. (eds.). University of Chicago Press, pp 34-40.

Forster, C. A., Sampson, S. D., Chiappe, L. M., and Krause, D.W. (1998). The Theropod Ancestry of Birds: New Evidence from the Late Cretaceous of Madagascar. Science, Vol. 279, pp 1915-1919.

Harmon, L.J., Melville, J., Larson, A., and Losos, J. B. (2008). The Role of Geography and Ecological Opportunity in the Diversification of Day Geckos (Phelsuma). Systematic Biology, Vol. 57, No. 4, 562-573.

Hay, W. W., DeConto, R., Wold, C. N., Wilson, K. M., Voigt, S., Schulz, M., Wold, A. R., Dullo, W-C., Ronov, A. B., Balukhovsky, A. N. and Söding, E. (1999). Alternative global Cretaceous paleogeography. In: Evolution of the Cretaceous Ocean-Climate System. Barrera, E. and Johnson, C.C. (eds.). Geological Society of America, Special Paper, pp 1-47.

Jansa, S. A., Goodman, S. M., and Tucker, P. K. (1999). Molecular Phylogeny and Biogeography of the Native Rodents of Madagascar (Muridae: Nesomyinae): A Test of the Single-Origin Hypothesis. Cladistics, Vol. 15, Issue 3, pp 253-270.

Krause, D. W., R. R. Rogers, C. A. Forster, J. H. Hartman, G. A. Buckley, and S. D. Sampson (1999). The Late Cretaceous vertebrate fauna of Madagascar: Implications for Gondwanan paleobiogeography. GSA Today, Geological Society of America, Vol. 9, no. 8, pp 1-7.

Krause, D. W. (2003). Late Cretaceous vertebrates of Madagascar: A window into Gondwanan biogeography at the end of the Age of Dinosaurs. In The Natural History of Madagascar. Goodman, S. M., and Benstead, J. P. (eds.). University of Chicago Press, pp 40–47.

McCall, R. A. (1997). Implications of recent geological investigations of the Mozambique Channel for the mammalian colonization of Madagascar. Royal Society of London Proceedings, Vol. 264, pp 663–665.

Rabinowitz, P.D., M.F. Coffin, and D.A. Falvey (1983). “The separation of Madagascar and Africa”. Science, Vol. 220, pp 67-69.

Rabinowitz, P. D. and Woods, S. (2006). The Africa-Madagascar connection and mammalian migrations. Journal of African Earth Sciences, Vol. 44, No. 3, pp 270-276.

Raxworthy, C.J., Forstner, M. R. J., and Nussbaum R. A. (2002). Chameleon radiation by oceanic dispersal. Nature, Vol. 415, pp 784-787.

Rieppel, O. (1987). The phylogenetic relationships within the Chamaeleonidae, with comments on some aspects of cladistic analysis. Zoological Journal of the Linnean Society, Vol. 89, pp. 41-62

Rose, K. D. (2006). Primates and plasiadapiformes. In The Beginning of the Age of Mammals. The John Hopkins University Press, pp 166-192.

Shubin, N. H. and Sues, H-D. (1991). Biogeography of early Mesozoic continental tetrapods: patterns and implications. Paleobiology, Vol. 17, No. 3, pp 214-230.

Silcox, M.T. (2007). Primate Taxonomy, Plesiadapiforms, and Approaches to Primate Origins. In Primate Origins: Adaptations and Evolution, Part 1, pp 143-178.

Simpson, G. G. (1940). Mammals and land bridges. Journal of the Washington Academy of Sciences, Vol. 30, pp. 137-163.

Storey, M., Mahoney, J. J., Saunders, A. D., Duncan, R. A., Kelley, S. P., and Coffin, M. F. (1995). Timing of hot spot-related volcanism and the breakup of Madagascar and India. Science, Vol. 267, pp 852–855.

Tattersall, I. (2007). Origin of the Malagasy Strepsirhine Primates. In Lemurs: Ecology and Adaptation. Gould, L. and Sauther, M. L. (eds.). Developments in Primatology: Progress and Prospects, Springer, pp 3-17.

Vences, M, Kosuch, J., Rödel, M-O., Lötters, S., Channing, A., Glaw,  F., and Böhme, W. (2004). Phylogeography of Ptychadena mascareniensis suggests transoceanic dispersal in a widespread African-Malagasy frog lineage. Journal of Biogeography, Vol. 31, pp 593–601.

Yoder, A. D. and Nowak, M. D. (2006). Has vicariance or dispersal been the predominant biogeographic in Madagascar? Only time will tell. Annual Review of Ecology, Evolution, and Systematics, Vol. 37, pp 405-431.


People read...

Happy World Tapir Day!

Twelve and a half months in Ecuador

Hips Don't Lie #1 - Hip hip, hurray!