Origin of the Lemurs

The island of Madagascar once formed part of the great southern landmass of Gondwana. This enormous supercontinent began to break apart about 180 million years ago into the individual landmasses we know today as Africa, South America, Antarctica, Australia and the Indian subcontinent. Initially, Madagascar remained attached to Africa, but about 160 million years ago, together with the Indian subcontinent (in addition to Australia and Antarctica, which collectively formed “East Gondwana”), it began to break away from Africa (and South America, which collectively formed “West Gondwana”) (see Scotese, 2000; Yoder and Nowak, 2006). East and West Gondwana fragmented further, but the Indian subcontinent and Madagascar remained joined together as “Indo-Madagascar” for another 70 million years before finally separating about 90 million years ago. By that time, Madagascar had assumed its present position relative to the east coast of Africa, separated by the Mozambique Channel. The channel is just over 400 km wide at its narrowest point.

This has interesting implications for the origin of lemurs. In the early days of the science of plate tectonics—the subdiscipline of geology that deals with continental drift—it seemed that we had found a simple explanation for Madagascar’s biological uniqueness. The island, it was widely assumed, had merely parted from Africa with a sampling of that continent’s fauna aboard. And while Africa’s fauna saw a great deal of replacement of one group by another over the course of the epochs that followed, the fauna of isolated Madagascar did not.

The clarification of Madagascar’s geological history has resulted in much rethinking of this scenario. If Madagascar has been essentially where it is now in relation to Africa for the past 90 million years or so, we cannot explain the presence of lemurs in Madagascar by an African “founder effect” (although this may explain the presence of some of the more ancient members of the island’s biota). Primates comparable in evolutionary level to the lemurs (the so-called “primates of modern aspect” or euprimates) are not known in the fossil record until they appear abruptly in beginning of the Eocene epoch, some 55 million years ago (Martin et al., 2007). These are distinct from the plesiadapiforms (so-called archaic primates) which have a fossil record dating back to 65 million years ago. They are not necessarily primates, and they are not the ancestors of the euprimates—the radiations are separate. Tavaré et al. 2002; see also Soligo and Martin, 2006; Martin et al., 2007; Soligo et al. 2007) argued that the origin of the primates dates back to the Cretaceous. Most of the Eocene euprimates are either Adapiformes or Omomyiformes and, while quite similar, some authors link the former to the strepsirrhine primates (prosimians—all but the tarsiers) and the latter to the haplorhines (tarsiers, monkeys and apes) (see Kay et al., 1997; Fleagle 1999). The euprimates have, however been documented by fossils very largely from North America and Europe and the only fossil fragments from Africa are from Egypt and the Arabian peninsula (Yoder et al. 1996). There is virtually no fossil record of terrestrial mammals in Africa for the first half of the Age of Mammals, and there is no known last common ancestor of the euprimates. The adapiform primates went extinct in the Early Oligocene (34 million years ago) and were a radiation of euprimates very largely restricted to the northern continents. Martin (1993) and Martin et al. (2007) proposed that the adapiform and omomyiform euprimates together represent a parallel northern continental radiation quite separate from the strepsirrhines and haplorhines, but already sharing derived, primate-defining, features in common with them. They argue that the last common ancestor of the euprimates would, therefore, be much than is widely thought and would go back to the Cretaceous 85.9 million years ago (with the 95% confidence limits of 73.3–95.7 million years ago (Tavaré et al., 2002; Soligo et al., 2007).

Divergence time estimates based on a number of studies have indicated the lemuroid dispersal to have occurred between 80 to more than 50 million years ago. A mean age was calculated by Masters et al. (2006) from 10 studies as 60 million years (that is when lemuroids diverged from the African-Asian lorisoids). A more recent analysis by Horvath et al. (2008) indicated an earlier date for the split between the African lorises (and galagos) and the lemurs at 75 million years ago (with a 95% credibility interval of 66.9–84.4 million years ago).

Nevertheless, even if the ancestors of the primates were much older than was previously thought, and even that certain key events in mammalian evolution took place in Africa a little earlier than in the northern continents (for which relatively complete fossil records are available), it would still seem that the ancestors of today’s lemurs most likely reached Madagascar by a sea crossing (from Africa: Yoder et al., 1996; Roos et al., 2004). Large, matted clumps of floating vegetation are routinely washed down major rivers and out to sea, sometimes with unwilling mammals and other passengers aboard. There are numerous cases in which this is the only plausible mechanism for the dispersal of terrestrial mammals (and other terrestrial vertebrates, see below) to oceanic islands (Simpson, 1940). Despite the width of the Mozambique Channel, Madagascar’s lemurs most likely represent one more such case.

There are a number of aspects, however, which need to be considered before accepting this as an explanation, not least among them whether sea currents and winds would favor, or permit even, a passage from the continent where the ancestors were presumed to have lived, and also whether the passage would be short enough for the survival of the unwilling passengers. This last aspect was discussed by Kappeler (2001), who indicated that torpor or month-long hibernation, a trait of Microcebusd and Cheirogaleus, would have contributed to the likelihood of survival of a sea passage. This was based on the incorrect supposition by Purvis (1995), that the mouse and dwarf lemurs are the most ancestral of the living lemurs. They are in fact one of the more recent lineages (see below). The galagos are the mainland relatives of the lemurs and they dos not show torpor or hibernation as an adaptation for stressful enviornent (Mzilikazi et al., 2004) and Stankiewicz et al. (2005) also discounted the evidence for this. Stankiewicz et al. (2005) reviewed and investigated the possibilities concerning the likelihood of a sea crossing, taking into account even such fortuitous events as freak winds and tornados or trees on the floating vegetation which could act as a sail. Their unequivocal conclusion was that “current and wind trajectories [as they are today] show that the most likely fate for a raft emerging from an estuary on the east coast of Africa is to follow the Mozambique current and become beached back on the African coast.”(p.1). That is, transport from Madagascar to Africa is very much more likely than the reverse. Even if the passage was possible, the time it would take would make survival of any mammals unikely. They concluded that an alternative explanation was needed for the ancestral lemur’s colonization of Madagascar.

Another possibility that has been raised is that islands or even a bridge was formed from peaks along the north-south running Davie Ridge in the Mozambique Channel at times when sea level was lower, which could have facilitated a crossing (e.g., Tattersall, 1982; McCall, 1997; Masters et al., 2006). This would have occurred during the second half of the Eocene and inot the Oligocene and the idea that it might have allowed for the lemur colonization of Madagascar by the ancestral lemur was suggested by McCall (1997) based on a divergence time for the lemuroids and lorisoids at about 40 million years ago. This, as discussed above, would be too recent (Martin, 2000). A land bridge has also been discounted for geological reasons (Rabinowitz and Woods, 2006) and even if there were small islands at some time during the Cenozoic, they still would have been separated by hundreds of kilometers of ocean (Ali and Huber 2010). Krause et al. (1997) suggested that the currents in the Mozambique Channel may have been favorable for a crossing from Africa to Madgascar in the Late Cretaceous, about 65 million years ago, when Madagascar and Africa lay about 15º further south from where they are now. This possibility was strongly supported by Ali and Huber (2010) who, using paleo-oceanographic modeling, concluded that in the Eocene epoch strong currents in the Mozambique Channel certainly would have flowed eastward and that there were occasions when so-called “trajectories” starting in the region of northeast Mozambique and Tanzania would have fast, jet-like eddies with velocities indicating that crossings to Madagascar could have been possible in 25–30 days or even less. The region was also then subject to tropical cyclones. Transient storms and ocean current activity were consequently favorable to sea crossings; the cyclones providing not only the rafts (large, floating islands of vegetation) but also the drinking water. Ali and Huber (2010) concluded that favorable currents (from Mozambique and Tanzania to the north coast of Madagascar) would have persisted through the Oligocene, but that Madagascar, moving north, would have breached the margin of the subtropical and equatorial gyres by the early Miocene, and currents thereafter would have been perennially directed westwards, as they are today.

With the possible exception of a podocnemine turtle, the diverse record of Cretaceous terrestrial fossils on Madagascar, which includes representatives of several mammalian clades, has not (yet) revealed links to the extant Malagasy fauna. What is more, the few molecular-clock data that exist imply Cenozoic origination of the major vertebrate groups, from fishes to mammals, at a time when Madagascar was already isolated. For example, the closest relatives of Madagascar’s indigenous freshwater fish (known as cichlids) are found in the East African great lakes, and present molecular estimates of divergence between major cichlid lineages considerably younger than the separation of Madagascar from East Africa. Similar patterns are seen for all of the terrestrial groups of Malagasy mammals, including lemurs. There is even evidence of Cenozoic origins of certain groups (e.g., chameleons) on Madagascar that then purportedly emigrated to other landmasses. The global chameleon fossil record goes back 20 million years, yet molecular phylogenies suggest a Malagasy origin with multiple “out of Madagascar” radiations to Africa and other Indian Ocean islands. If this interpretation of the chameleon radiation is correct, there was “two-way traffic” across the Mozambique Channel, and Madagascar provided a source for at least some biodiversity elsewhere, probably from sometime during the Oligocene.

But difficult questions remain unanswered: chameleons and small terrestrial mammals (e.g., tenrecs, rodents, carnivora, and lemurs) might have crossed significant sectors of the Indian Ocean on vegetation floats; but what about larger terrestrial mammals (e.g., the recently extinct pygmy hippos), frogs (which are relatively intolerant of marine conditions), and freshwater fish in the apparent absence of land bridges? Could the larger mammals have swum significant distances? Could the rafts of vegetation been large enough to harbour freshwater and its inhabitants? These questions, going back as far as the time of Alfred Russel Wallace and the birth of the science of biogeography, still loom large and will not be answered until a better Cenozoic fossil record on Madagascar is brought to light.

The phylogenetic relationships among lemurs above the species level have been obscure for a long time. Several arrangements among lemur families—the main branches of this adaptive radiation—have been proposed based on analyses of various morphological traits (Tattersall and Schwartz, 1974; Tattersall, 1982). All subsequent genetic analyses have confirmed lemur monophyly, i.e., all lemurs are the descendants of a single successful colonization event (Yoder et al., 1996; Roos et al., 2004; Yoder and Yang, 2004, Karanth et al., 2005; Horvath et al., 2008). Supporting the findings of Roos et al. (2004) and Yoder and Yang (2004), Hovarth et al. (2008) concluded that the diversification of lemurs on Madagascar began with the separation of the lineage leading to the aye-aye (Daubentonia), about 66 million years ago (54.9–74.7 million years ago), extending as such across the Cretaceous/Tertiary boundary of 65 million years ago. Current evidence indicates that further splits of lemurs into families did not begin to occur until about 42 million years ago, more than 20 million years later in the middle Eocene, but the possibility remains that there were unrecorded lineage extinctions during this time (Yoder and Yang, 2004). The first major divergence in evidence today, following that of the separation of Daubentonia, was between the “true” lemurs (Lemuridae) and the ancestors of the Cheirogaleidae, Lepilemuridae and Indriidae. The indriids subsequently separated, and then the lepilemurids, all within a period of about 10 million years (Yoder and Yang, 2004).

Madagascar is the last place on Earth to have experienced the disappearance of diverse groups of large indigenous mammals (megafauna), and one of the last habitable continental regions to be colonized by humans. Most of the documented extinctions occurred after the first arrival of humans on Madagascar, about 2,000 years ago (Burney et al., 2004). The earliest date for man’s presence came from cut marks suggesting the removal of flesh from a radius bone of a sloth lemur (Palaeopropithecus ingens) dated at 2,325±43 years before present (Perez et al., 2003).

Elephant birds, relatives of the ostrich, the largest of which stood 3 m tall and at 450 kg probably the heaviest bird that ever lived, were still present on the island near the end of the of the first millennium A.D., as were at least 16 species of large, now extinct lemurs. The stratigraphic resolution of this megafaunal “extinction window” is estimated to have been between 500 and 1,500 years ago. At least 53 species of large Malagasy mammals, birds, and reptiles have become extinct, leaving the region with no indigenous terrestrial vertebrates of a body weight greater than 12 kg.

The list of extinct land vertebrate species (see Goodman and Jungers, 2014) includes: a huge crocodile (Voay robustus), two species of giant tortoise (Gerlach and Canning, 1998); two aardvark-like mammals Plesiorycteropus (Order Bibymalagasia), three dwarf hippopotami (Stuenes, 1989; Faure and Guérin, 1990), a tenrec (Microgale) (Goodman et al., 2007), a giant fossa (Cryptoprocta) (Goodman et al., 1997, 2003), a giant jumping rat (Hypogeomys) (Goodman and Rakotondravony, 1996) and two other nesomyid rodents (Mein et al., 2010), two hipposiderid bats (Samonds, 2007), 21 birds, including three cuckoos (of an endemic genus Coua), a ground roller (of the endemic genus Brachypteracias), two shelducks (Alopochen and Centrornis), a mesite (Monias), a large gallinule (Hovacrex), a plover (Vanellus), a cormorant (Phalacrocorax), eight kinds of elephant birds (the endemic Aepyornis and Mullerornis), and three eagles, one of which was very large (Stephanoaetus and Aquila), besides 17 giant lemurs, some of which grew to be as large as a female gorilla (Goodman and Rakotondravony, 1996; Hawkins and Goodman, 2003; Goodman and Hawkins, 2008; Turvey, 2009).

In the next chapter we discuss in more detail the nature, natural history of the 17 large and remarkable lemurs that tragically are now extinct. Their loss emphasizes that the lemurs surviving today could face a similar fate if major conservation efforts currently underway are not effective.

Extinct Holocene vertebrates of Madagascar (not included are 17 extinct lemurs, see chapter 3).


Order Bibymalagasia

  Family Plesiorycteropidae Patterson, 1975

    Plesiorycteropus madagascariensis Filhol, 1895

    Plesiorycteropus germainepetterae MacPhee, 1994

Order Afrosoricida

  Family Tenrecidae Gray, 1821

    Microgale macpheei Goodman, Vasey and Burney, 2007

Order Chiroptera

  Family Hipposideridae Lydekker, 1891

    Hipposideros besaoka Samonds, 2007

    Triaenops goodmani Samonds, 2007

Order Artiodactyla

  Family Hippopotamidae Gray, 1821

    Hippopotamus lemerlei A.Grandidier in Milne-Edwards, 1868

    Hippopotamus laloumena Faure and Guérin, 1990

    Hippopatamus guldbergi Fovet, Faure and Guérin, 2011

Order Carnivora

  Family Eupleridae Chenu, 1850

    Cryptoprocta spelea G. Grandidier, 1902

Order Rodentia

  Family Nesomyidae Forsyth-Major, 1897

    Brachytarsomys mahajambaensis Mein, Sénégas, Gommery, Ramanivosoa, Randrianantenaina and Kerloc’h, 2010

    Hypogeomys australis G. Grandidier, 1903

    Nesomys narindaensis Mein, Sénégas, Gommery, Ramanivosoa, Randrianantenaina and Kerloc’h, 2010


Order Aepyornithiformes

  Family Aepyornithidae Bonaparte, 1853

    Aepyornis gracilis Monnier, 1913

    Aepyornis hildebrandti Burckhardt, 1893

    Aepyornis maximus I. Geoffroy St. Hilaire, 1851

    Aepyornis medius Milne-Edwards and A. Grandidier, 1866

    Mullerornis agilis Milne-Edwards and A. Grandidier, 1894

    Mullerornis betsilei Milne-Edwards and A. Grandidier, 1894

    Mullerornis grandis Lamberton, 1934

    Mullerornis rudis Milne-Edwards & A. Grandidier, 1894

Order Falconiformes

  Family Accipitridae Vieillot, 1816

    Stephanoaetus mahery Goodman, 1994

    Aquila sp. A

    Aquila sp. B

Order Anseriformes

  Family Anatidae Vigors, 1825

    Alopochen sirabensis (Andrews, 1897)

    Centrornis majori Andrews, 1897

Order Pelecaniformes

  Family Phalacrocoracidae Reichenbach, 1850

    Phalacrocorax sp.

Order Mesitornithiformes

  Family Mesitornithidae Wetmore, 1960

    Monias sp. (undescribed)

Order Gruiformes

  Family Rallidae Vigors, 1825

    Hovacrex roberti (Andrews, 1897)

Order Charadriiformes

  Family Charadriidae Vigors, 1825

    Vanellus madagascariensis Goodman, 1996

Order Cuculiformes

  Family Cuculidae Vigors, 1825

    Coua berthae Goodman and Ravoavy, 1993

    Coua delalandei (Temminck, 1827)

    Coua primaeva Milne-Edwards and A. Grandidier, 1895

Order Coraciiformes

  Family Brachypteraciidae Bonaparte, 1854

    Brachypteracias langrandi Goodman, 2000


Order Crocodylia

  Family Crocodylidae Cuvier, 1807

    Voay robustus A.Grandidier and Vaillant and 1872

Order Testudines

  Family Testudinidae Batsch, 1788

    Aldabrachelys abrupta (A. Grandidier, 1868)

    Aldabrachelys grandidieri (Vaillant, 1885)


Goodman, S. M., Vasey, N. and Burney, D. A. 2007. Description of a new species of subfossil shrew tenrec (Afrosoricida: Tenrecidae: Microgale) from cave deposits in southeastern Madagascar. Proceedings of the Biological Society of Washington 120: 367–376.

Mein, P., Sénégas, F., Gommery, D., Ramanivosoa, B., Randrianantenaina, H. and Kerloc'h, P. 2010. Nouvelles especes subfossiles de rongeurs du Nord-Ouest de Madagascar. Comptes Rendus Palevol., 9(3): 101−112,

Samonds, K. E. 2007. Late Pleistocene bat fossils from Anjohibe Cave, northwestern Madagascar. Acta Chiropterologica 9(1): 39–65.

Goodman, S. M. and Jungers, W. L. 2014. Extinct Madagascar: Picturing the Islands Past. The University of Chicago Press, Chicago, IL.