Elsevier

Earth-Science Reviews

Volume 232, September 2022, 104103
Earth-Science Reviews

A review of geological evidence bearing on proposed Cenozoic land connections between Madagascar and Africa and its relevance to biogeography

https://doi.org/10.1016/j.earscirev.2022.104103Get rights and content

Abstract

For a variety of reasons, Madagascar's rich and highly distinctive faunal assemblage has long attracted attention. Recurring questions with the associated studies have concerned when and how the ancestors of the various clades reached the island. For the land-bound animals, the common view is that they arrived after the Cretaceous on ‘rafts' that washed in from Africa. However, this centuries-old discussion has been periodically spiked with proposals for land-bridges, with a recent one arguing for three separate causeways or stepping-stone chains connecting Africa and Madagascar in the early, middle, and late Cenozoic (66–60 Ma, 36–30 Ma, and 12–5 Ma). Here, the general idea of former causeways spanning the Mozambique Channel is evaluated based on an extensive survey of the related geological and geophysical literature, a large portion of which dates from the last half-decade. The analysis, which makes use of a newly-developed topological schema, indicates that only a small number of the supposed dry-land sub-paths were actually subaerial during each of the postulated colonization windows. Notably, during the Early Oligocene, a 220–250 km2 pinnacle in the Sakalaves Seamount group formed a volcanic ‘high-island’. The other offshore sectors that were exposed would have been atoll crowns: Juan de Nova Island throughout the Cenozoic and the northern Sakalaves in the Late Miocene. However, it would have been challenging for most land animals to exist on those low-elevation carbonate platforms for any length of time. Therefore, the notion of Africa and Madagascar being linked in the Cenozoic by terrestrial walkways can be regarded as falsified. Over-water dispersal best explains how 26 of Madagascar's 27 land-vertebrate clades colonized the landmass. The exception, a group of small, soil-dwelling snakes, is likely a relict lineage whose ancestors were present on the island prior to Madagascar's geographical isolation that resulted from the tectonic block's breakup with India-Seychelles c. 85 Ma.

Introduction

Although Madagascar's iconic land-bound vertebrate assemblage comprises >900 species, this diversity has developed from no more than about thirty reptile, land-mammal, and amphibian clades (Crottini et al., 2012). For instance, the four extant land-mammal groups, (carnivorans, lemurs, rodents, and tenrecs), are today represented by between 10 and 100 species (Mittermeier et al., 2021). The marked difference between Madagascar's mammal suite and the one on Africa combined with its highly unbalanced/filtered nature led William D. Matthew (1915) to propose that the colonizations had resulted from over-water dispersal events (as opposed to range expansion via a former land-route) at various times in the early and middle Cenozoic, and possibly the Cretaceous. The issue was further considered by George G. Simpson (incidentally, Matthew's protégé), who developed the associated ‘Sweepstakes' colonization concept (1940). The land-mammal taxa that crossed from Africa were ‘drawn’ from a pool of small-bodied animals and transported passively to the island aboard vegetation rafts (see also O'Dea et al., 2016; Ali et al., 2021). Key to Simpson's argument was the idea that if a land-bridge had connected the two landmasses, then appreciably more colonizations should have taken place, as has been the case with the Panama Isthmus. In that example, the joining of South America with Central America (and North America) c. 3 Ma led to a large bi-directional transfer of land-vertebrate genera (e.g. Simpson, 1980; O'Dea et al., 2016). However, despite the Matthew-Simpson view having garnered wide acceptance (e.g. Krause, 2003; de Queiroz, 2005, de Queiroz, 2014; Vences, 2004; Rabinowitz and Woods, 2006; Yoder and Nowak, 2006; Van der Geer et al., 2010; Samonds et al., 2012), it has at various times been challenged (McCall, 1997; Stankiewicz et al., 2006; Masters et al., 2006; Mazza, 2014; Mazza et al., 2019). Resistance to the view is rooted in the belief that land vertebrates, especially the land-locked mammals, cannot survive such journeys, which would take a few to several weeks (Ali and Huber, 2010), due to a lack of food and an absence of freshwater. Furthermore, those waifs that endured the passages, as well as their near descendants, would have been susceptible to extirpation through selection pressures due to their restricted genetic diversity. Ali and Vences (2019) countered all three of these arguments and emphasized the fact that the sorts of land mammals that had established themselves on the remote islands and archipelagoes, for instance on the Canaries, Christmas Island, Galápagos Islands, and Madagascar, had one or more characteristics that facilitated the process: small body mass, low energy requirement, ability to go into torpor, and propensity to hibernate. When Mammalia are viewed as a whole, very few taxa emerge as victors in a ‘Simpson lottery.’

Unencumbered by these considerations, however, Masters et al., 2021, Masters et al., 2022 have recently proposed that Madagascar's land-vertebrates colonized the island via three temporally-separate causeways at 66–60 Ma, 36–30 Ma, and 12–5 Ma. It should be noted, though, that the notion of former terrestrial connections is not new. McCall (1997) surmised that a land-bridge existed 45–26 Ma based on a combination of biological and geological data. Much earlier, and well before the theories of continental drift and plate tectonics had been formulated, Wallace (1880) argued for a Cretaceous-age routeway. This was preceded by Sclater's (1864) proposal for Madagascar having once being connected to Africa, the Indian subcontinent and Southeast Asia via a now-sunken landmass that he labelled ‘Lemuria’.

As Madagascar's biota is the focus of an enormous amount of research activity, and where the interpretation of a clade's history is dependent upon knowing the time since it ancestral colonization event, that in turn reflects the mode of arrival, this review evaluates the geological evidence for former land-bridges. The proposals of Masters et al., 2021, Masters et al., 2022 are the primary focus, but it is straightforward to extend the analysis to the model of McCall (1997). Also, it is possible to consider other issues and to touch upon the palaeogeographical scenarios of Sclater (1864) and Wallace (1880).

Section snippets

Publications underpinning the Masters et al. ‘three-causeways' model

The palaeogeographical aspect of Masters et al., 2021, Masters et al., 2022 was constructed around three physiographical maps of the Mozambique Channel at the aforementioned time windows. To facilitate discussions, we label the scenarios/land-bridges ‘Early Paleocene’ for 66–60 Ma, ‘Eo-Oligocene’ for 36–30 Ma and ‘Late Miocene’ for 12–5 Ma. However, with the first and the third there are slight discrepancies with regards to the formally defined geological intervals (e.g. Walker et al., 2018);

Basis for McCall's 45–26 Ma land-bridge

McCall (1997) proposed that the Davie Ridge (Fig. 1) played a key role in the lemur, tenrec, carnivoran, and rodent colonizations. First, he presented biological arguments suggesting it was impossible for land-mammals to over-water disperse from Africa to Madagascar. He then made the case for a land-bridge, summarizing key geological data from areas along and adjacent to the Davie Ridge. Notably, the sedimentary sequence recovered at DSDP Site 242 (676-m-long) on the Macua Seamount's northeast

Davie Ridge

The Davie Ridge (Fig. 1) underpins all the land-bridge/stepping-stone routes between Africa and Madagascar. Although no parts of the high are presently exposed, the feature is now reasonably well known due to it having been studied for over half a century through geophysical surveys, dredging and drilling (e.g. Heirtzler and Burroughs, 1971; Simpson et al., 1974; Mougenot et al., 1986; Mascle et al., 1987; Coffin and Rabinowitz, 1987; Leclaire et al., 1989; Hall et al., 2017). In the last

Masters et al. land-bridge models

To support discussions of the proposals of Masters et al., renditions are presented of their three palaeogeographical maps (Fig. 11, Fig. 12, Fig. 13). Each shows land, seabed 0–100 m deep, seabed >100 m deep, and the inferred locations of volcanic centres. Notably, although they are less complex than the originals, they retain all of the information for making an assessment. Alongside each plot is a related figure from the PhD thesis of Antoine Delaunay (2018). Although Masters et al.

Evaluating the land-bridge models of Masters et al. and McCall

Due to the large amount of information that is contained within the various land-bridge scenarios, particularly those of Masters et al., it is a challenge to comprehend the various elements. It was thus decided to simplify matters by converting the four proposed pathways into topological models, and to then compare each with data-validated ones (see below). Here, the obvious physiographic elements have their own ‘sectors' for instance ‘Macua Seamount’, ‘Juan de Nova Island’, but so do the

Early Paleocene land-bridge proposal of Masters et al

The Masters et al. model for 66–60 Ma and the data-evaluated configuration are shown in Fig. 15. With the former, all sectors are assumed to been exposed. However, the analysis carried out here indicates that only the coast strip at Nacala was truly dry, whilst Juan de Nova Island was an atoll. All other elements were covered by water.

Eo-Oligocene land-bridge proposal of Masters et al

The Masters et al. model for 36–30 Ma and the data-evaluated configuration are shown in Fig. 16. With the former, the Juan de Nova-Melaky Province route was

Discussion

None of three land-bridge proposals of Masters et al., 2021, Masters et al., 2022, nor the one of McCall (1997), stands up to scrutiny. Africa and Madagascar were not connected by continuous causeways or stepping-stone chains at any time in the last 66 million years. As is the case today, the extent of dry land in the Mozambique Channel in the early, middle and late Cenozoic was negligible. A better explanation for Madagascar's land-fauna colonizations is centred around many over-water

Conclusions

Since the mid-19th century, biologists have proposed ancient land-bridges and stepping-stone paths linking Africa and Madagascar to explain Madagascar's unique biota. The hypothesized routeways are significant because they were thought to have enabled the mini-continent to be colonized by the ancestors of its extant and recently-extinct land-bound vertebrate clades (c. 30). As the Malagasy assemblage is scientifically important at a global level, the age and origin of its constituent lineages

Data availability

No primary data were generated for this contribution. At the time of writing, e-versions of the PhD theses of Delaunay (2018) and Ponte (2018) were available on-line, respectively at https://tel.archives-ouvertes.fr/tel-01865476 and https://tel.archives-ouvertes.fr/tel-01865479. If required, pdfs can be obtained from JRA.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

We thank Denis Mougenot for sharing information, and Wilfried Jokat and an anonymous reviewer for their constructive critiques of the manuscript. Sun-Yi Ali checked a late-stage draft.

References (98)

  • V.T. Leinweber et al.

    The crustal structure of the Central Mozambique continental margin – wide angle seismic, gravity and magnetic study in the Mozambique Channel, eastern Africa

    Tectonophys.

    (2013)
  • E. Leroux et al.

    Evolution of the Glorieuses seamount in the SW Indian Ocean and surrounding deep Somali basin since the Cretaceous

    Mar. Geol.

    (2020)
  • D. Macgregor

    History of the development of the East African Rift System: a series of interpreted maps through time

    J. Afr. Earth Sci.

    (2015)
  • C.O. Mueller et al.

    Geophysical evidence for the crustal variation and distribution of magmatism along the central coast of Mozambique

    Tectonophys.

    (2017)
  • J.P. Ponte et al.

    The Zambezi delta (Mozambique Channel, East Africa): high resolution dating combining bio-orbital and seismic stratigraphies to determine climate (palaeoprecipitation) and tectonic controls on a passive margin

    Mar. Petrol. Geol.

    (2019)
  • M.J. Pratt et al.

    Shear velocity structure of the crust and upper mantle of Madagascar derived from surface wave tomography

    Earth Planet. Sci. Lett.

    (2017)
  • P.D. Rabinowitz et al.

    The Africa–Madagascar connection and mammalian migrations

    J. Afr. Earth Sci.

    (2006)
  • G. Salman et al.

    Development of the Mozambique and Ruvuma sedimentary basins, offshore Mozambique

    Sed. Geol.

    (1995)
  • S.N. Stephenson et al.

    Disentangling interglacial sea level and global dynamic topography: Analysis of Madagascar

    Earth Planet. Sci. Lett.

    (2019)
  • T.H. Torsvik et al.

    Late Cretaceous magmatism in Madagascar: palaeomagnetic evidence for a stationary Marion hotspot

    Earth Planet. Sci. Lett.

    (1998)
  • M. Vormann et al.

    The crustal structure of the southern Davie Ridge offshore northern Mozambique – a wideangle seismic and potential field study

    Tectonophys.

    (2020)
  • C. Agnini et al.

    Calcareous nannofossil biostratigraphy: historical background and application in Cenozoic chronostratigraphy

    Lethaia

    (2017)
  • J.R. Ali et al.

    Mammalian biodiversity on Madagascar controlled by ocean currents

    Nature

    (2010)
  • J.R. Ali et al.

    Mammals and long-distance over-water colonization: the case for rafting dispersal; the case against phantom causeways

    J. Biogeogr.

    (2019)
  • J.R. Ali et al.

    Monkeys on a free-floating island in a Colombian river: further support for over-water colonization

    Biogeographia

    (2021)
  • Y. Bassias et al.

    The Davie Ridge in the Mozambique Channel: crystalline basement and intraplate magmatism

    N. Jb. Geol. Paläont. Mh. H.

    (1990)
  • V. Boullet et al.

    An updated account of the vascular flora of the Iles Eparses (Southwest Indian Ocean)

    Atoll Res. Bull.

    (2018)
  • P. Bouysse et al.

    A new CGMW map: the structural map of the Indian Ocean: an attempt at ocean cartography

    Episodes

    (2004)
  • K. Burke

    The African plate

    South Afr. J. Geol.

    (1996)
  • K. Burke et al.

    The African Erosion Surface: a continental-scale synthesis of geomorphology, tectonics, and environmental change over the past 180 million years

    Mem. Geol. Soc. Am.

    (2008)
  • E.J. Censky et al.

    Over-water dispersal of lizards due to hurricanes

    Nature

    (1998)
  • M.F. Coffin et al.

    Reconstruction of Madagascar and Africa: evidence from the Davie Fracture Zone and western Somali Basin

    J. Geophys. Res.

    (1987)
  • J. Collot et al.

    Geodynamics of the SW Pacific: A brief review and relations with New Caledonian geology

  • S. Courgeon et al.

    The offshore East African rift system: new insights from the Sakalaves seamounts (Davie Ridge, SW Indian Ocean)

    Terra Nova

    (2018)
  • A. Crottini et al.

    Vertebrate time-tree elucidates the biogeographic pattern of a major biotic change around the K-T boundary in Madagascar

    Proc. Natl. Acad. Sci. U. S. A.

    (2012)
  • A. de Queiroz

    The resurrection of oceanic dispersal in historical biogeography

    Trends Ecol. Evol.

    (2005)
  • A. de Queiroz

    The monkey’s Voyage: How Improbable Journeys Shaped the History of Life

    (2014)
  • M. de Wit

    Madagascar: heads it’s a continent, tails it’s an island

    Annu. Rev. Earth Planet. Sci.

    (2003)
  • A. Delaunay

    Les mouvements verticaux de Madagascar (90–0 Ma): Une analyse couplée des formes du relief et de l'enregistrement sédimentaire des marges ouest Malgaches

    (2018)
  • F. Dixey

    The geology and geomorphology of Madagascar, and a comparison with eastern Africa

    Q. J. Geol. Soc. Lond.

    (1960)
  • C. Ebinger

    Tectonic development of the western branch of the East African rift system

    Geol. Soc. Am. Bull.

    (1989)
  • F.R. Fosberg

    Atoll vegetation and salinity

    Pacific Sci.

    (1949)
  • D. Franke et al.

    The offshore East African Rift system: structural framework at the toe of a juvenile rift

    Tectonics

    (2015)
  • C. Gaina et al.

    Microcontinent formation around Australia

  • C.W.D. Gibson et al.

    The vegetation of Aldabra Atoll: preliminary analysis and explanation of the vegetation map

    Phil. Trans. Roy. Soc. Lond. Ser. B

    (1983)
  • I.R. Hall et al.

    Site U1476

    Proc. Int. Ocean Disc. Prog.

    (2017)
  • J.R. Heirtzler et al.

    Madagascar’s paleoposition: new data from the Mozambique Channel

    Science

    (1971)
  • R.M. Key et al.

    Revised lithostratigraphy of the Mesozoic-Cenozoic succession of the onshore Rovuma Basin, northern coastal Mozambique

    South Afr. J. Geol.

    (2008)
  • D.W. Krause

    Late Cretaceous vertebrates of Madagascar: A window into Gondwanan biogeography at the end of the age of dinosaurs

  • Cited by (10)

    • Slow and steady wins the race: Diversification rate is independent from body size and lifestyle in Malagasy skinks (Squamata: Scincidae: Scincinae)

      2023, Molecular Phylogenetics and Evolution
      Citation Excerpt :

      On the other hand, the presence of rich and structured biotas on continental islands may in turn limit the opportunities for the diversification of new colonisers (McPeek and Brown, 2007; Toussaint et al., 2015; Ali, 2017). Madagascar is the fourth largest island in the world (ca. 587,000 km2), initially part of the Gondwana supercontinent until it separated from Africa ca. 160–130 million years ago (Ma), it remained isolated from all other landmasses since ca. 80–65 Ma (Ali and Aitchison, 2008; Ali and Hedges, 2022). Madagascar’s Cretaceous vertebrate biota was likely almost entirely depleted during the mass extinction at the boundary between the Cretaceous and Paleogene (K-T boundary) (Samonds et al., 2013) and time-calibrated phylogenies and biogeographic reconstructions suggest that most present-day biodiversity results from subsequent colonisation events rather than vicariance (Vences et al., 2003; Yoder and Nowak, 2006; Vences et al., 2009; Crottini et al., 2012; Samonds et al., 2012; Welt and Raxworthy, 2022).

    View all citing articles on Scopus
    View full text