ArticlePDF Available

Mammalian biodiversity on Madagascar controlled by ocean currents

February 2010
Nature 463(7281):653-6

DOI:10.1038/nature08706

Source
PubMed

Authors:

Jason R Ali

Senckenberg Society for Nature Research

Matthew Huber

Purdue University West Lafayette

Madagascar hosts one of the world's most unusual, endemic, diverse and threatened concentrations of fauna. To explain its unique, imbalanced biological diversity, G. G. Simpson proposed the 'sweepstakes hypothesis', according to which the ancestors of Madagascar's present-day mammal stock rafted there from Africa. This is an important hypothesis in biogeography and evolutionary theory for how animals colonize new frontiers, but its validity is questioned. Studies suggest that currents were inconsistent with rafting to Madagascar and that land bridges provided the migrants' passage. Here we show that currents could have transported the animals to the island and highlight evidence inconsistent with the land-bridge hypothesis. Using palaeogeographic reconstructions and palaeo-oceanographic modelling, we find that strong surface currents flowed from northeast Mozambique and Tanzania eastward towards Madagascar during the Palaeogene period, exactly as required by the 'sweepstakes process'. Subsequently, Madagascar advanced north towards the equatorial gyre and the regional current system evolved into its modern configuration with flows westward from Madagascar to Africa. This may explain why no fully non-aquatic land mammals have colonized Madagascar since the arrival of the rodents and carnivorans during the early-Miocene epoch. One implication is that rafting may be the dominant means of overseas dispersal in the Cenozoic era when palaeocurrent directions are properly considered.

Eocene ocean currents.a, Simulated annual mean vertically averaged (barotropic) currents (volume flux) and surface ocean currents (streamline vectors). The barotropic currents delineate the average positions of the major Eocene ocean gyres, with Madagascar at the heart of the strongest gyre on Earth, as described in the text. The modern location of Madagascar is shown in red outline, showing how the interplay of continental position and the gyres controls dispersal pathways. b, Ensemble-averaged monthly mean ocean surface currents for January, the month in which climatological ocean current directions were optimal for transport towards Madagascar. c, During sporadic events, as typified by this ensemble average of the four optimal ocean current events evaluated from model output saved at a temporal resolution of three days, rapid transport directly to Madagascar from Africa was possible at rates of >20 cm s-1. Currents are shown using vectors (with the scale shown), and the magnitude of the east–west current strength is shown in colour. The simulations are described in Methods. Mad, Madagascar.

…

Figures - uploaded by Jason R Ali

Content may be subject to copyright.

Content uploaded by Jason R Ali

Content may be subject to copyright.

LETTERS

Mammalian biodiversity on Madagascar controlled

by ocean currents

Jason R. Ali

& Matthew Huber

Madagascarhosts one of the world’s most unusual, endemic,diverse

and threatened concentrations of fauna

. To explain its unique,

imbalanced biological diversity, G. G. Simpson proposed the

‘sweepstakes hypothesis’, according to which the ancestors of

Madagascar’s present-day mammal stock rafted there from

Africa

. This is an important hypothesis in biogeography and evolu-

tionary theory for how animals colonize new frontiers

1,3–5

, but its

validity is questioned

5–9

. Studies suggest that currents were in-

consistent with rafting to Madagascar

and that land bridges provided

the migrants’ passage

5–8

. Here we show that currents could have

transported the animals to the island and highlight evidence in-

consistent with the land-bridge hypothesis. Using palaeogeographic

reconstructions and palaeo-oceanographic modelling, we find that

strong surface currents flowed from northeast Mozambique and

Tanzania eastward towards Madagascar during the Palaeogene

period, exactly as required by the ‘sweepstakes process’. Sub-

sequently, Madagascar advanced north towards the equatorial gyre

and the regional current system evolved into its modern configura-

tion with flows westward

from Madagascar to Africa. This may

explain why no fully non-aquatic land mammals have colonized

Madagascar since the arrival of the rodents and carnivorans during

the early-Miocene epoch. One implication is that rafting may be the

dominant means of overseas dispersal in the Cenozoic era when

palaeocurrent directions are properly considered.

Madagascar is home to one of the most intriguing inventories of

flora and fauna anywhere on Earth

. It is characterized by unusually

high levels of mammalian endemism combined with a uniquely broad

diversity from a limited number of orders: lemurs, tenrecs, carnivorans

and rodents

1,11–13

. Although our focus here is primarily on mammals,

similar patterns are observed for other terrestrial animals on the island,

includingamphibians and reptiles

14,15

, implying a broadly similar colo-

nization mechanism. It is widely acknowledged that Madagascar’s

mammals arederived from Cenozoic migrants becausethey share none

of the detailed characteristics of the island’s late-Cretaceous forms

16,17

and thus could not have evolved from them. The current stock’s

ancestors journeyed from Africa at various times during the

Cenozoic, since 65 Myr ago, and each colonization appears to be the

result of a single arrival event

. The timing of apparent arrival events of

new taxa is a fundamental constraint on hypotheses about how they

accomplished the journey

. From molecular-clock dating estimates, it

is possible to discern four distinct early-Cenozoic to mid-Cenozoic

events: the arrival of lemurs between 60 and 50 Myr ago, that of tenrecs

between 42 and 25 Myr ago,that of carnivorans between 26 and 19 Myr

ago, and that of rodents between 24 and 20 Myr ago

12,13

.Thus,by

,20 Myr ago the major non-volant and non-swimming faunal groups

were established, with no further evidence of transfer, except for a few

late-Quaternary arrivals such as the pygmy hippopotamus, which is

semi-aquatic and known to swim significant ocean distances

.As

Madagascar and Africa have been separated for ,120 Myr by the

430-km-wide Mozambique Channel

19,20

, the question is whether the

non-volant, non-aquatic migrants made their way to the island by

walking or by rafting.

One hypothesis is that land bridges might have enabled fauna to

walk. It has been argued that substantial portions of the Davie

Ridge—a prominent bathymetric feature running north–south

down the middle of the Mozambique Channel (Fig. 1)—may have

once been sub-aerially exposed and thus could have acted as a quasi-

continuous causeway linking Africa and Madagascar

. Although this

idea has received some support

5–7,21

, important weaknesses are

acknowledged because it would require a radical rethinking of regional

plate tectonics

. It also suffers logical flaws: if land bridges were

responsible, a greater variety of animals would have crossed and the

timing of arrivals would be correlated with the putative maximum

extent of the land bridge, neither of which results is supported by

data

1,23

. We further note that any islands the Davie Ridge may have

generated during the Cenozoic

would have been small and separated

by open-water gaps several tens to hundreds of kilometres apart

(Fig. 1). Consequently, substantial tracts of ocean (.230 km) sepa-

rated them from the nearest land in eastern Mozambique and western

Madagascar; thus, over-water dispersal was unavoidable.

The over-water dispersal mechanism was first mooted by Simpson

nearly seventy yearsago

. He proposeda ‘sweepstakes’ process by which

small mammals—potentiallywith low metabolicrates and/or a habit of

seasonal torpor—were unwittingly rafted to the island on large logs or

vegetation mats washed off eastern Africa, either down large rivers or

from the coastal strip

. Key predictions of Simpson’s argument, con-

firmed by thecolonization-history andgeological datadescribed above,

are the limited number of families that live on the island today, in

particular the absence of large-bodied forms (for example antelopes,

apes, elephants or lions), and the seemingly random distribution of

apparent arrivals in Madagascar from ,60 to ,20 Myr ago.

The main criticism of this hypothesis is that inferred currents and

prevailing winds, based on modern observations, are in the opposite

direction to those required. As elucidated in ref. 9, if today’s surface-

water currents in the region are used as a guide (Fig. 1), the strong

south-southwest-directed coast-parallel flow of the Mozambique

Current would have acted as barrier to eastward transport. Rafts

off the shore of Africa would have been entrained in the southward

flow and thus could never have beached on Madagascar. Instead they

would either return to the African shore or be transported north or

south, but never substantially to the east. No quantitative attempts

have been made to estimate how currents in the region may have been

different in the Cenozoic or what implications this may have for

ocean dispersal routes. Past currents remain the major unknown in

this controversial issue and many factors, including changing palaeo-

geographic and palaeo-oceanographic setting, must be considered.

Department of Earth Sciences, University of Hong Kong, Pokfulam Road, Hong Kong, China.

Earth and Atmospheric Sciences Department and the Purdue Climate Change Research

Center, Purdue University, West Lafayette, Indiana 47907, USA.

Vol 463

4 February 2010

doi:10.1038/nature08706

653

The Indian Ocean basin has altered considerably as a result of plate

tectonics. For instance, since 60 Myr ago Australia and India have

respectively migrated ,2,200 and ,4,000 km northwards. Globally,

six major ocean gateways have either opened (Tasman–Antarctica,

South America/Antarctica, Atlantic–Arctic) or closed (Panama

Isthmus, Indonesian, Africa–Arabia–Eurasia). Critically, Africa and

Madagascar have both moved some 15u(,1,650 km) closer to the

equator

. Thus, the potential for surface-water flow in the southwest

Indian Ocean to have been markedly different when the faunas are

thought to have crossed to Madagascar must be considered.

Today, the entire southern Indian Ocean, from Africa to Australia,

lies within a large anticyclonic ‘supergyre’

. Northern Madagascar

lies at the boundary between this gyre and a cyclonic gyre occupying

the northern Indian Ocean

26,27

. The southern Indian Ocean gyre’s

strongest flows are to the southwest of Madagascar, where they form

the Agulhas current. A relatively strong eddying flow in the

Mozambique Channel between Madagascar and Africa connects

the gyre’s equatorial currents with the vigorous transports of the

Agulhas system

. Another important feature of the southern Indian

Ocean gyre concerns the fact that Australia spans the latitudes

Madagascar does. As a consequence, the zonally integrated wind stress

curl, which drives the circulation, is maximized near the southern tip

of Africa

To test the hypothesis that changes in the ocean circulation around

Madagascar during the Cenozoic explain the observed pattern of

land-mammal migrations to the island, an independent knowledge

of surface ocean currents and surface wind stress is required. The only

means of obtaining this information (as there are no independent

proxy records for these variables) is from palaeo-oceanographic

modelling. We performed and analysed a suite of experiments using

a fully and interactively coupled ocean–atmosphere general circula-

tion model (the Community Climate System Model, version 3

(CCSM3), of the US National Center for Atmospheric Research) with

Eocene conditions (Methods). In such models, climate, wind and

surface-water currents are predicted and, importantly, are free to

evolve according to the equations of motion and thermodynamics.

Thus, we have not specified the winds or sea surface temperatures

that drive them, nor the ocean currents.

In all of our simulations, the large-scale ocean current systems in the

Eocene epoch were robustly different from modern observed and

modelled circulations in several crucial ways relevant to the rafting

problem. Africa and Madagascar were .10usouth of their current

positions, which placed Madagascar in the convergence zone at the

heart of the subtropical gyre (Fig. 2). Furthermore, because Australia

was also much further south, the northern part of that continent did

not impede the accumulation of wind stress curl at the latitudes of

Madagascar; therefore, the strongest current system in the entire

Southern Ocean was just to the east of Madagascar (Fig. 2a). This

vigorous eddy off the coast of Madagascar caused water-mass

trajectories throughout the region to converge.

As shown in Fig. 2a, the strong anticlockwise gyre directed much of

the flow along the African coast eastward towards Madagascar rather

than southward through the Mozambique Channel, towards the

Agulhas Current, as occurs today. Peak simulated eastward velocities

across the Mozambique Channel usually occur in January (Fig. 2b),

and monthly mean velocities of .10 cm s

occur for three to four

months within each century of simulation. Trajectories starting in the

region of northeast Mozambique and Tanzania sporadically experi-

ence enhanced eastward velocities of .20 cm s

and could therefore

have crossed the necessary distance in 25–30 days (Fig. 2c). This vigo-

rous eastward flow was not constant, occurring for only three or four

weeks within any century of simulation. Hence, on the long timescales

of relevance to this problem (tens of millions of years), these velocities

would have occurred many times, but not routinely. Because these are

rare events, it is likely that even faster eastward currents occurred,

albeit less frequently. Furthermore, flow through the Mozambique

Channel was probably strongly eddying, as it is today, but our simula-

tions do not have sufficient resolution to capture the vigorous, high-

kinetic-energy properties of these eddies, which today have velocities

of ,100 cm s

(ref. 27). Capturing such mesoscale variability would

Seychelles

Madagascar

Mascarene

basin

Africa

Comoros

islands

Réunion

Madagascar

basin

Western Somali

basin

Davie Ridge

highs

Land 0–200 m 200–2,000 m 2,000–4,000 m >4,000 m

Mozambique

basin

Madagascar

Ridge

Mozambique

Plateau

Mozambique

Channel

10° S

20° S

30° S

0°

60° E30° E 40° E 50° E

13.00° S

41.40° E

19.70° S

41.90° E

1,000

Depth (m)

2,000

3,000

~750 km

Macua

seamount

Sakalave

seamounts

Paisley

seamount

30° E

Seychelles

Madagascar

10° S

20° S

Latitude

Longitude

30° S

0°

40° E 50° E 60° E

East

Madagascar

Current (S)

East African

Coastal Current

Agulhas

Current

South

Equatorial

Current

Mozam-

bique

Tanzania

Kenya

South

Africa

Mozambique

Current

Current ow

Figure 1

Key geographical features/phenomena of the southwest Indian

Ocean region. a, Simplified bathymetric map. b, Present-day surface-water

circulation based on refs 25 and26 (M, Mauritius; R, Re

´union).

c, Bathymetric cross-section along the Davie Ridge. Even in the Eocene,

when parts of the feature may have been sub-aerial, the deep troughs

separating the peaks would have posed formidable barriers

LETTERS NATURE

Vol 463

4 February 2010

654

only increase the estimated probability of the occasional very vigorous

eastward transport from Africa to Madagascar, and would enhance

transport across the Mozambique Channel in jet-like currents.

Consequently, our speed estimate is almost certainly lower than the

true maximum eastward rafting velocity.

An additional consideration is the fact that tropical storms are

known to generate the large, floating tree ‘islands’, as well as asso-

ciated precipitation, that might make a successful ocean voyage of

this type possible

. It is therefore noteworthy that preliminary ana-

lysis of modelled tropical cyclone activity indicates that this region

was a locus of such activity, as it is today, and that the tropical cyclone

season encompasses the period of highly favourable ocean currents

(January). Thus, successful rafting may have involved the fortuitous

coincidence of transient storms and ocean current activity.

Thus, all signs point to the Simpson sweepstakes model being

correct: ocean currents could have occasionally transported rafts of

animals to Madagascar from Africa during the Eocene. Specifically,

transport should have been from northeast Mozambique and

Tanzania to the north coast of Madagascar. Given the slow tectonic

drift of the island, this configuration probably continued at least

through the Oligocene epoch. However, by the early Miocene,

Madagascar breached the margin of the subtropical and equatorial

gyres. Thereafter, currents were perennially directed westward

towards Africa, making the ocean journey for mammals to

Madagascar much more difficult, if not impossible.

METHODS SUMMARY

We used the current version of the fully coupled (ocean/atmosphere/land/

vegetation/sea ice) global climate model CCSM3. The atmospheric resolution

was set at ,3.75u?3.75u(T31 spectral resolution); the oceans had a nominal 3u

longitudinal resolution (and variable latitudinal resolution) and 25 vertical

levels. This model has been applied to a wide variety of modern

and palaeo-

climate studies, for example of the Holocene epoch and the Last Glacial

Maximum

and the Eocene

We carried out a suite of fully coupled simulations for a simulated time of

more than 3,000 yr without any acceleration until they had clearly reached equi-

librium. A description of the suite of simulations is found in ref. 30; results from

Longitude

Latitude

Longitude

Current strength (cm s–1)

Mad Mad

Africa Africa

0°

–70

10° S

15° S

20° S

25° S

30° S

35° S

40° S

20° E 30° E 40° E 50° E 60° E 20° E

–30 –10 10 30

20 cm s–1 30° E 40° E 50° E 60° E

10° S

15° S

20° S

25° S

30° S

35° S

40° S

–30 10 50

90° E 180°

Volume ux (Sv)

Figure 2

Eocene ocean currents. a, Simulated annual mean vertically

averaged (barotropic) currents (volume flux) and surface ocean currents

(streamline vectors). The barotropic currents delineate the average positions

of the major Eocene ocean gyres, with Madagascar at the heart of the

strongest gyre on Earth, as described in the text. The modern location of

Madagascar is shown in red outline, showing how the interplay of

continental position and the gyres controls dispersal pathways. b, Ensemble-

averaged monthly mean ocean surface currents for January, the month in

which climatological ocean current directions were optimal for transport

towards Madagascar. c, During sporadic events, as typified by this ensemble

average of the four optimal ocean current events evaluated from model

output saved at a temporal resolution of three days, rapid transport directly

to Madagascar from Africa was possible at rates of .20 cm s

. Currents are

shown using vectors (with the scale shown), and the magnitude of the

east–west current strength is shown in colour. The simulations are described

in Methods. Mad, Madagascar.

NATURE

Vol 463

4 February 2010 LETTERS

655

the particular simulation, with a concentration of atmospheric CO

1,120 p.p.m., used in this study have not been previously described. Here we

have concentrated on one simulation appropriate for mid-Eocene to late-Eocene

conditions, although none of the results we have discussed are sensitive to the

particulars of that choice, as the relevant boundary conditions and main palaeo-

currents remained largely unchanged between the end of the Cretaceous period

and the early-to-mid Miocene, when current systems shifted towards their

modern state

Full Methods and any associated references are available in the online version of

the paper at www.nature.com/nature.

Received 28 May; accepted 25 November 2009.

Published online 20 January 2010.

1. Yoder, A. D. & Nowak, M. D. Has vicariance or dispersal been the predominant

biogeographic in Madagascar? Only time will tell. Annu. Rev. Ecol. Evol. Syst. 37,

405

–

431 (2006).

2. Simpson, G. G. Mammals and land bridges. J. Wash. Acad. Sci. 30, 137

–

163 (1940).

3. Heaney, L. R. Is a new paradigm emerging for oceanic island biogeography?

J. Biogeogr. 34, 753

–

757 (2007).

4. Thiel, M. & Haye, P. A. The ecology of rafting in the marine environment. III.

Biogeographical and evolutionary consequences. Oceanogr. Mar. Biol. 44,

323

–

429 (2006).

5. Tattersall, I. in Elwyn Simons: A Search for Origins (eds Fleagle, J. G. & Gilbert, C. C.)

397

–

408 (Springer, 2008).

6. Tattersall, I. Historical biogeography of the strepsirhine primates of Madagascar.

Folia Primatol. (Basel) 77, 477

–

487 (2006).

7. Masters, J. C., de Wit, M. J. & Asher, R. J. Reconciling the origins of Africa, India

and Madagascar with vertebrate dispersal scenarios. Folia Primatol. (Basel) 77,

399

–

418 (2006).

8. McCall, R. A. Implications of recent geological investigations of the Mozambique

Channel for the mammalian colonization of Madagascar. Proc. R. Soc. Lond. B 264,

663

–

665 (1997).

9. Stankiewicz, J., Thiart, C., Masters, J. C. & de Wit, M. J. Did lemurs have

sweepstake tickets? An exploration of Simpson’s model for the colonization of

Madagascar by mammals. J. Biogeogr. 33, 221

–

235 (2006).

10. von der Heydt, A. & Dijkstra, H. A. Effect of ocean gateways on the global ocean

circulation in the late Oligocene and the early Miocene. Paleoceanography 21,

PA1011 (2006).

11. Goodman, S. M., Ganzhorn, J. U. & Rakotondravony, D. in The Natural History of

Madagascar (eds Goodman, S. M. & Benstead, J. P.) 1159

–

1186 (Chicago Univ.

Press, 2003).

12. Yoder, A. D. et al. Single origin of Malagasy Carnivora from an African ancestor.

Nature 421, 734

–

737 (2003).

13. Poux, C. et al. Asynchronous colonization of Madagascar by the four endemic

clades of primates, tenrecs, carnivores, and rodents as inferred from nuclear

genes. Syst. Biol. 54, 719

–

730 (2005).

14. Vences, M. Origin of Madagascar’s extant fauna: a perspective from amphibians,

reptiles and other non-flying vertebrates. Ital. J. Zool. (Modena) 71 (suppl.),

217

–

228 (2004).

15. Nagy, Z. T., Joger, U., Wink, M., Glaw, F. & Vences, M. Multiple colonization of

Madagascar and Socotra by colubrid snakes: evidence from nuclear and

mitochondrial gene phylogenies. Proc. R. Soc. Lond. B 270, 2613

–

2621 (2003).

16. Krause, D. W. Fossil molar from a Madagascan marsupial. Nature 412, 497

–

498

(2001).

17. Krause, D. W. et al. Late Cretaceous terrestrial vertebrates from Madagascar:

implications for Latin American biogeography. Ann. Mo. Bot. Gard. 93, 178

–

208

(2006).

18. Stuenes, S. Taxonomy, habits, and relationships of the subfossil Madagascan

hippopotami Hippopotamus lemerlei and H.madagascariensis.J. Vertebr. Paleontol.

9, 241

–

268 (1989).

19. Rabinowitz, P. D., Coffin, M. F. & Falvey, D. The separation of Madagascar and

Africa. Science 220, 67

–

69 (1983).

20. Ali, J. R. & Aitchison, J. C. Gondwana to Asia: plate tectonics, paleogeography and

the biological connectivity of the Indian sub-continent from the Middle Jurassic

through latest Eocene (166

–

35 Ma). Earth Sci. Rev. 88, 145

–

166 (2008).

21. Godinot, M. Lemuriform origins as viewed from the fossil record. Folia Primatol.

(Basel) 77, 446

–

464 (2006).

22. Rabinowitz, P. D. & Woods, S. The Africa

–

Madagascar connection and

mammalian migrations. J. Afr. Earth Sci. 44, 270

–

276 (2006).

23. Bassias, Y. Petrological and geochemical investigations of rocks from the Davie

Fracture Zone (Mozambique Channel) and some tectonic implications. J. Afr.

Earth Sci. 15, 321

–

339 (1992).

24. Krause, D. W., Hartman, J. H. & Wells, N. A. in Natural Change and Human Impact in

Madagascar (eds Goodman, S. D. & Patterson, B. D.) 3

–

43 (Smithsonian Inst.

Press, 1997).

25. Kappeler, P. M. Lemur origins: rafting by groups of hibernators? Folia Primatol.

(Basel) 71, 422

–

425 (2000).

26. Schott, F. A., Xie, S. P. & McCreary, J. P. Jr. Indian Ocean circulation and climate

variability. Rev. Geophys. 47, RG1002 (2009).

27. de Ruijter, W. P. M., Ridderinkhof, H. & Schouten, M. Variability of the southwest

Indian Ocean. Phil. Trans. R. Soc. A 363, 63

–

76 (2005).

28. Yeager, S. G., Shields, C. A., Large, W. G. & Hack, J. J. The low-resolution CCSM3.

J. Clim. 19, 2545

–

2566 (2006).

29. Otto-Bliesner, B. L. et al. Last glacial maximum and Holocene climate in CCSM3.

J. Clim. 19, 2526

–

2544 (2006).

30. Liu, Z. et al. Global cooling during the Eocene-Oligocene climate transition. Science

323, 1187

–

1190 (2009).

Acknowledgements M. Nowak, W. de Ruijter, I. Tattersall and A.Yoder supplied

reprints. J. Aitchison, R. Corlett and A.Switzer are thanked for sharing information.

M.H. is supported by US National Science Foundation (NSF) grant 0927946-ATM

and uses the US National Center for Atmospheric Research CCSM, which is

supported by the NSF. M.H. acknowledges conversations with P. Koch and D. Raup

on vicariance biogeography. All computing was performed at the Rosen Center for

Advanced Computing, which is part of Information Technology at Purdue, Purdue

University.

Author Contributions J.R.A. initiated the study and was primarily responsible for

the geologically related aspects. M.H. carried out the palaeo-oceanographic

modelling and its interpretation. Both authors contributed to the writing of the

paper.

Author Information Reprints and permissions information is available at

www.nature.com/reprints. The authors declare no competing financial interests.

Correspondence and requests for materials should be addressed to M.H.

(huberm@purdue.edu).

LETTERS NATURE

Vol 463

4 February 2010

656

METHODS

Our simulations incorporate the detailed boundary conditions (that is, topo-

graphy, vegetation and bathymetry) developed previously for Eocene condi-

tions

30–32

. The latitudinal position of the gyres and their strengths are very

robust to boundary-condition changes appropriate to the Palaeogene as they

are essentially Sverdrupian responses to wind stress changes as modulated by the

slowly varying palaeogeographies

31–33

As described in ref. 9, the rafting problem involves estimating the surface

ocean currents with the potential additional influence of surface winds.

Currents below the surface are not relevant because typical rafts, such as a tree

trunk launched into the ocean from a river mouth during a flood, would not have

‘keels’ that extended below several metres. Similarly, it is unlikely they had

substantial protruding elements that could have formed ‘masts’, and would thus

have been unaffected by wind shear. Here we assume that the raft simply acted

like a drifter embedded in the ocean surface current (the surface model level

extends down to 8 m). Analysis of the modern simulation using this same model

(not shown) produces ocean currents in agreement with observations and the

prior work on Madagascar

The Eocene simulation focused on in this study was carried out with an atmo-

spheric CO

concentration of 1,120 p.p.m. and all other boundary conditions set at

near-modern, pre-industrial values. It is a branch simulation from a simulation with

a higher concentration of atmospheric CO

that was integrated for a simulated time

of more than 3,000 yr. After this spin-up time, the 1,120-p.p.m. simulation was

integrated for 3,500 yr. We analysed the monthly mean output from the last 400yr

of simulation to generate Fig. 2a, b. Output at 3-d resolution for the final century of

integration was used to generate Fig. 2c. Although the magnitude of the optimal

eastward flow is somewhat dependent on the averaging period (peak velocity

increases from ,13 cm s

when monthly mean values are used to ,23 cm s

when

3-d means are used), the qualitative aspects of the flow are not sensitive to averaging

length. In the sampling of both the mean configuration and the extreme events, flow is

always westward from Madagascar towards Africa because this is the direction deter-

mined by the large-scale gyres in the modern era. Preliminary results from a com-

parable Miocene (15 Myr ago) simulation show similar results to the modern-day

one. All file processing and graphics were performed using the US National Center for

Atmospheric Research Command Language (http://ncl.ucar.edu).

31. Huber, M., Sloan, L. C. & Shellito, C. J. in Causes and Consequences of Globally Warm

Climates in the Early Palaeogene (eds Wing, S. L., Gingerich, P. D., Schmitz, B. &

Thomas, E.) 25

–

47 (GSA Special Paper 369, Geological Society of America,

2003).

32. Huber, M. & Nof, D. The ocean circulation in the southern hemisphere and its

climatic impacts in the Eocene. Palaeogeogr. Palaeoclimatol. Palaeoecol. 231, 9

–

(2006).

33. Huber, M. et al. Eocene circulation of the Southern Ocean: was Antarctica kept

warm by subtropical waters? Paleoceanography 19, PA4026 (2004).

doi:10.1038/nature08706

Media File: Selection of URLs related to the Madagascar study

Data

February 2010

Jason R Ali · Matthew Huber

Download

Media File: Kate Wong (Scientific American, 2010, v 302, i 4, p 11)

Data

February 2010

Jason R Ali · Matthew Huber

Download

Media File: David W Krause (Nature, 2010, v 463, p 613–614)

Data

February 2010

Jason R Ali · Matthew Huber

Download

New records, diagnostics and preliminary checklist of the superfamily Pentatomoidea (Hemiptera: Heteroptera) from the Comoro Islands

Article

Jul 2024

The following species are recorded from the Comoro Islands for the first time: 12 species of Pentatomidae: Pentatominae—Anoano pronotalis Cachan, 1952 (from Mayotte), Antestiopsis clymeneis cf. galtiei (Frappa, 1934) (Mayotte), Bagrada hilaris (Burmeister, 1835) (Mayotte), Bathycoelia rodhaini Schouteden, 1913 (Mayotte), Coquerelia ventralis Horváth, 1904 (Mayotte), Eurysaspis transversalis Signoret, 1851 (Anjouan, Grande Comore, Mayotte), Gadarscama ebenaui Reuter, 1887 (Anjouan, Grande Comore, Mohéli, Mayotte), Lerida annulicornis (Signoret, 1861) (Anjouan, Mayotte), Neoacrosternum validum (Horváth, 1904) (Grand Comore, Mayotte), Nezara viridula (Linnaeus, 1758) (Mayotte), Piezodorus hybneri (Gmelin, 1790) (Mayotte), and Stenozygum mirabile (Signoret, 1861) (Mayotte); two species of Plataspidae—Brachyplatys hemisphaerica (Westwood, 1837) (Mayotte) and Coptosoma maculatum Westwood, 1837 (Mayotte); and one species of Scutelleridae—Hotea denticulata Stål, 1865 (Mayotte). The endemic pentatomine species Bathycoelia cuneifera Bergroth, 1893, syn. nov., is recognized as a junior subjective synonym of the widely distributed African species Bathycoelia thalassina (Herrich-Schäffer, 1844). The lectotype of B. cuneifera is designated herein. The checklist of the Pentatomoidea of Comoro Islands includes six species of Cydnidae, 21 species of Pentatomidae, two species of Plataspidae, one species of Scutelleridae and 1 species of Tessaratomidae. Of these species, six are endemic to the archipelago, 14 species are subendemic (shared with Madagascar and/or Aldabra Atoll), and nine species are widespread, occurring also on the African mainland (including the cosmopolitan pest Nezara viridula and two species of Bathycoelia Amyot & Serville, 1843 from the Guinean area, all three possibly alien to the archipelago). We document an additional eight species (six species of Pentatomidae, one species of each Scutelleridae and Tessaratomidae) that are identified to generic level only (two recorded from Grand Comore, seven from Mayotte, one of them occuring in both islands). Mayotte is the best sampled island, with 28 recorded species, followed by Grande Comore (11 species), Anjouan (six species), and Mohéli (four species). In addition the following new records are provided: Bathycoelia rodhaini from Cameroon, Central African Republic, Gabon and Zambia, Bathycoelia thalassina from Central African Republic and Ethiopia, and Piezodorus hybneri from Angola (Cabinda) and Sierra Leone. The type locality of Eurysaspis transversalis Signoret, 1851, India: Puducherry, is considered an error as the species occurs only in Madagascar and Comoros. The following plant associations are recorded from Mayotte: Bagrada hilaris on Cleome viscosa (Cleomaceae), Eurysaspis transversalis on Senna singueana (Fabaceae), and Piezosternum sp. on Cucumis sativus (Cucurbitaceae) and Mikania capensis (Asteraceae).

Rampant Dispersal Without Gene-Flow: Reproductively and Geographically Isolated Lineages of the Supertramp Lizard Lamprolepis smaragdina Permeate the Lesser Sunda Archipelago

Article

Full-text available

Jan 2025
J BIOGEOGR

Aim We investigated the biogeographical patterns of a widespread arboreal lizard found throughout Wallacea and the western Pacific islands. Wallacea contains longstanding oceanic currents that flow from the Pacific Ocean into the Indian Ocean, known as the Indonesian Throughflow. Using genetic data, we aim to estimate the timing and sequence of island colonisation in the Lesser Sunda archipelago of southern Wallacea, and to determine the proportion of dispersal events that may have been aided by ocean currents. Levels of gene flow are also explored in the context of island arrangement, past connectivity (ex. Pleistocene land bridges), and ocean currents. Location Southern Wallacea (Indonesia, Timor‐Leste). Taxon Emerald Tree Skink ( Lamprolepis smaragdina ). Methods We estimated a mitochondrial DNA phylogeny from southern Wallacean populations to identify divergent populations and to guide sample selection for genomic‐scale data collection. These genomic data were used to estimate (1) phylogenetic relationships of island populations using both concatenated Maximum Likelihood, summary coalescent and multispecies coalescent methods, (2) levels of genetic connectivity between lineages using genetic clustering and migration analyses, and (3) to reconstruct their dispersal history using ancestral range reconstruction analysis. Results We found that the Emerald Tree Skinks of southern Wallacea diverged from a population on Peleng Island ~1.25 Ma and that Lesser Sunda populations diverged between ~1 and 0.4 Ma. MtDNA data are insufficient to estimate the relationships of each population and find one island to be paraphyletic and two to be polyphyletic. Genomic data find that Timor Island is polyphyletic while all other islands are recovered as monophyletic. Gene flow is minimal or absent across the archipelago, even between islands that become periodically land bridged. More than half of the estimated over‐water dispersal events occurred perpendicular to or against ocean currents. Main Conclusions Despite this species' over‐water dispersal abilities island populations are genetically isolated from one another, even on adjacent islands that become periodically land bridged. This suggests that species formation is complete. The inferred dispersal history involved many paths that would be difficult to complete given the direction of ocean currents, indicating that historical ocean currents are more dynamic than currently understood.

Global biogeographic patterns of the genus Indigofera (Fabaceae: Indigofereae)

Article

Full-text available

Jan 2025

Indigofera L. is the most diverse genus in the tribe Indigofereae and the third largest genus in the Fabaceae with over 750 species distributed in most tropical to temperate areas of the world. The centre of diversity of the genus lies on the African continent, accounting for over two-thirds of its global diversity. Diversification of Indigofer a’s four main clades began in the Early Miocene, with the Tethyan and Cape Clades having slightly older crown ages than the Pantropical and Paleotropical Clades. Using biogeographical models under a newly assembled, more comprehensively-sampled ITS phylogeny, we confirmed that the tribe Indigofereae originated in the Africa-Madagascar centre, with the genus Indigofera resolving with a mainland African origin c. 38 Mya. Long-distance dispersals were rare, but did occur from Africa to South America and Africa to Australia during the late Miocene. Multiple dispersals from Africa into Asia took place via the arid north-east African and Arabian corridors during the same period. At least four Indigofera dispersals to Madagascar took place during the Miocene. The Cape Clade is a narrowly endemic lineage, mostly restricted to the Greater Cape Floristic Region (GCFR) of southwestern South Africa, while the Tethyan, Pantropical and Paleotropical Clades are more widely dispersed across multiple continents. Both the Cape and Tethyan Clades display strong signals for arid origins in the southern Namib and Richtersveld regions of Southwestern Africa. This study emphasizes the importance of the global Succulent biome in shaping biogeographical patterns of Indigofera globally, particularly in terms of biome conservatism and how it created corridors facilitating trans-continental dispersal.

Madagascar (Primates in History, Myth, Art, and Science)

Chapter

Full-text available

Apr 2024

Lemurs, as well as being endemic to Madagascar, are the flagship group for conservation efforts on the island. Lemur morphology and behavior likely played a role in inspiring legends, beliefs, taboos, and practices that survive today. Legends suggest that giant lemurs left traces in the folklore prior to disappearing before the 18th century. Stories are based on the human-lemur bond, which can be a real kinship, and they legitimate a taboo that forbids killing and consuming certain species. Beliefs and taboos are partially explained by morphological traits, but they may also have origin in disease prevention. The lemur decline is caused by forest loss and fragmentation related to slash-and-burn agriculture, logging, and mining, with hunting another major threat. Stronger commitment to support the network of protected areas and resource management by locals is crucial to protect lemurs from disappearing, along with the legends and practices that involve them.

Iguanas rafted more than 8,000 km from North America to Fiji

Article

Mar 2025
P NATL ACAD SCI USA

Founder-event speciation can occur when one or more organisms colonize a distant, unoccupied area via long-distance dispersal, leading to the evolution of a new species lineage. Species radiations established by long-distance, and especially transoceanic, dispersal can cause substantial shifts in regional biodiversity. Here, we investigate the occurrence and timing of the greatest known long-distance oceanic dispersal event in the history of terrestrial vertebrates—the rafting of iguanas from North America to Fiji. Iguanas are large-bodied herbivores that are well-known overwater dispersers, including species that colonized the Caribbean and the Galápagos islands. However, the origin of Fijian iguanas had not been comprehensively tested. We estimated the phylogenetic relationships and evolutionary timescale of the iguanid lizard radiation using genome-wide exons and ultraconserved elements (UCEs). Those data indicate that the closest living relative of extant Fijian iguanas is the North American desert iguana and that the two taxa likely diverged during the late Paleogene near or after the onset of volcanism that produced the Fijian archipelago. Biogeographic models estimate North America as the most probable ancestral range of Fijian iguanas. Our analyses support the hypothesis that iguanas reached Fiji via an extraordinary oceanic dispersal event from western North America, and which spanned a fifth of the earth’s circumference (>8,000 km). Overwater rafting of iguanas from North America to Fiji strengthens the importance of founder-event speciation in the diversification of iguanids and elucidates the scope of long-distance dispersal across terrestrial vertebrates.

Dispersal Biogeography

Chapter

Jan 2025

Land Bridges and Rafting Theories to Explain Terrestrial-Vertebrate Biodiversity on Madagascar

Article

Jan 2025
ANNU REV MAR SCI

Madagascar's celebrated land-vertebrate assemblage has long been studied and discussed. How the ancestors of the 30 different lineages arrived on the island, which has existed since 85 Mya and is separated from neighboring Africa by 430 km of water, is a deeply important question. Did the colonizations take place when the landmass formed part of Gondwana, or did they occur later and involve either now-drowned causeways or overwater dispersal (on vegetation rafts or by floating/swimming)? Following a historical review, we appraise the geological–geophysical evidence and the faunal-suite colonization record. Twenty-six of the clades are explained by temporally stochastic overwater dispersals, spanning 69–0 Mya, while two others are considered Gondwanan vicariant relicts. Due to a lack of information, the remaining two groups cannot be evaluated. The findings thus appear to resolve a debate that has rumbled along, with sporadic eruptions, since the mid-1800s.

Archaeology of Madagascar

Chapter

May 2024

Sean Hixon

Off the east coast of Africa lies the “Great Red Island” of Madagascar, with a history that has left the island rich in superlatives: It is Earth’s oldest island and among the hottest of the biodiversity hotspots. Definitive European accounts of the island extend over five hundred years into the past, but our knowledge of the island’s human history extends further via the archaeological record. Basic questions on the earliest human settlement of the island remain unresolved. However, archaeological traces of how people subsisted on endemic taxa in the island’s diverse environments are relatively clear, and traces of introduced plants and animals reflect connections across the Indian Ocean. How past people thrived on the island is closely tied with century-old environmental history narratives regarding extinction and deforestation, which remain relevant during ongoing attempts to conserve the island’s biological and cultural diversity. Durable elements of the archaeological record also reflect past resource extraction, connections with far-reaching trade networks, and the rise of an empire that ruled much of the island by the late 19th century. Though the Portuguese captain Diogo Dias visited Madagascar in 1500, the island’s recent history stands out due to its limited period of colonial control (French: 1895–1960). The fantastical stories of Madagascar’s man-eating trees and elephant-hunting birds no longer capture the Western imagination, yet the island’s diverse cultural heritage and human-environment interactions draw the attention of researchers and the curious public both within Madagascar and abroad.

Morphologie et phylogénomique au service de la biodiversité : systématique de Dombeya section Dombeya

Thesis

Full-text available

Dec 2023

Lucile Jourdain Fievet

The Dombeya Cav. genus (Malvaceae, ex-Sterculiaceae) comprises around 150 emblematic species from the western Indian Ocean. This genus is found in Africa, the Comoros, the Mascarenes and in many habitats of Madagascar; in each of these areas of distribution, Dombeya has a specific endemism rate of almost 100%. This ecological and specific diversity seems to be linked to major morphological variations and original life-history traits. With an estimated age of less than 10 Ma, this specific diversity and morpho-ecological complexity could be the result of rapid radiation. The taxonomy of the genus is still debated and still requires major taxonomic revision. Our study is part of this approach and focuses on a particular section of the genus: the autonymic section Dombeya, which occurs in all the genus' distribution areas. The section is characterised by a unique feature within the subfamily: a biparous cyme inflorescence with scorpioid terminal units. Using morphometric, morphological and genetic approaches, we were able to test species hypotheses previously formulated in the literature. The floral organs have proved to be a good object for differentiating taxonomic entities: three major geographical groups, African, Madagascan and Mascarenan, have thus been defined. A molecular approach using genome skimming was then used to extract the largest molecular matrix produced in the genus, which was used for phylogenetic analyses. By combining morphometry, a global morphology approach and phylogeny at the section level, two distinct taxonomic entities emerged within Dombeya sect. Dombeya. The monophyly of the section is not confirmed and it seems that the scorpioid biparous cyme character was acquired twice independently in Africa and Madagascar. The Malagasy species appear to be closely related to the Mascarenan species, which are thought to have derived from them, and form at least 3 distinct taxonomic units, the delimitation of which is debated.

Genetic analysis of Octopus cyanea reveals high gene flow in the South-West Indian Ocean

Article

Full-text available

Apr 2024

Mary A. Kishe

Octopus cyanea (Gray, 1849), abundant in the South‐West Indian Ocean (SWIO), constitutes a vital resource for both subsistence and commercial fisheries. However, despite this socioeconomic importance, and recent indications of overfishing, little is known about the population structure of O. cyanea in the region. To inform sustainable management strategies, this study assessed the spatio‐temporal population structure and genetic variability of O. cyanea at 20 sites in the SWIO (Kenya, Tanzania, Mozambique, Madagascar, Mauritius, Rodrigues, and the Seychelle Islands) by complementary analysis of mitochondrial DNA (mtDNA) noncoding region (NCR) sequences and microsatellite markers. MtDNA analysis revealed a shallow phylogeny across the region, with demographic tests suggesting historic population fluctuations that could be linked to glacial cycles. Contrary to expectations, NCR variation was comparable to other mtDNA regions, indicating that the NCR is not a hypervariable region. Both nuclear and mtDNA marker types revealed a lack of genetic structure compatible with high gene flow throughout the region. As adults are sedentary, this gene flow likely reflects connectivity by paralarval dispersal. All samples reported heterozygote deficits, which, given the overall absence of structure, likely reflect ephemeral larval recruitment variability. Levels of mtDNA and nuclear variability were similar at all locations and congruent with those previously reported for harvested Octopodidae, implying resilience to genetic erosion by drift, providing current stock sizes are maintained. However, as O. cyanea stocks in the SWIO represent a single, highly connected population, fisheries may benefit from additional management measures, such as rotational closures aligned with paralarval ecology and spanning geopolitical boundaries.

Early Paleogene oceans and climate: A fully coupled modeling approach using the NCAR CCSM

Chapter

Full-text available

Jan 2003

Attempts to understand early Paleogene climate using models have provided insights, but have consistently had difficulty explaining essential climatic features. A main limitation of those studies has been the lack of interaction between dynamical ocean and atmospheric models. After reviewing previous model studies and quantitatively defining the "low gradient problem," we present results from the first fully coupled general circulation model simulations in equilibrium with reconstructed early Paleogene topography, bathymetry, vegetation, and pCO2 (560 ppm). Although our results should be understood as preliminary, predictions of both Paleogene sea surface temperatures and salinities are, for the first time, made without assuming an arbitrary value of the oceanic or atmospheric heat transport. Model-predicted bottom water temperatures are 7 °C warmer than modern. Deep convection occurs in the North Atlantic and Tethys, forming warm (12-15 °C), salty water masses (35.5 ppt). Through most of the Southern Ocean, vigorous wind-driven upwelling and a stable water column caused by low salinities (∼32 ppt) produce a temperature inversion that may explain proxy interpretations of "warm salty deep water" formation in the tropics. Estimates of seawater oxygen isotopic ratios reveal substantially different patterns than usually predicted. Ocean heat transport is little-changed from modern values and warming of tropical temperatures (by ∼3 °C) is about half that occurring in high latitudes. The mean annual temperature and seasonality patterns produced by the simulation are biased toward colder and more seasonal values than those reconstructed from temperature proxies.

Origin of Madagascar's extant fauna: A perspective from amphibians, reptiles and other non‐flying vertebrates

Article

Full-text available

Jan 2005

Miguel Vences

The origins of the highly endemic and partly very diverse fauna and flora of Madagascar were longtime unsolved and are still strongly disputed, although dispersalists are recently prevailing over advocates of Gondwanan vicariance. Madagascar has been separated from other continents and continental fragments since the Mesozoic, and the assumption of vicariance origins requires very old ages of the lineages involved. Herein I review the recent progress of knowledge on the molecular phylogenetic relationships of the amphibians, reptiles, and other non‐flying vertebrates of Madagascar. Of 17 Malagasy clades for which relationships seem to be resolved with sufficient and unanimous support, the largest proportion have their sister‐group in Africa, and very few show a general area cladogram consistent with the succession of events in the fragmentation of the supercontinent Gondwana. A survey of pairwise sequence divergences between 25 Malagasy lineages and their non‐Malagasy relatives in the 16S and 12S rRNA genes resulted in values that in most cases were distinctly below the saturation plateaus of these genes. Multiple calibrations based on 34 largely independent data points indicate that such saturation would be strongly expected in the case of ages corresponding to the geographical isolation of Madagascar in the Jurassic and Cretaceous. A Cenozoic age for most Malagasy lineages is therefore most likely. Analysis of phylogenetic patterns of some taxa provides indications for a scenario in which the ancestors of the Malagasy clades first arrived by transmarine dispersal from Africa at the Malagasy west coast, and in a second step a subset of them underwent species‐rich radiations into the rainforests.

Indian Ocean circulation and climate variability

Article

Full-text available

Jan 2009

In recent years, the Indian Ocean (IO) has been discovered to have a much larger impact on climate variability than previously thought. This paper reviews climate phenomena and processes in which the IO is, or appears to be, actively involved. We begin with an update of the IO mean circulation and monsoon system. It is followed by reviews of ocean/atmosphere phenomenon at intraseasonal, interannual, and longer time scales. Much of our review addresses the two important types of interannual variability in the IO, El Niño-Southern Oscillation (ENSO) and the recently identified Indian Ocean Dipole (IOD). IOD events are often triggered by ENSO but can also occur independently, subject to eastern tropical preconditioning. Over the past decades, IO sea surface temperatures and heat content have been increasing, and model studies suggest significant roles of decadal trends in both the Walker circulation and the Southern Annular Mode. Prediction of IO climate variability is still at the experimental stage, with varied success. Essential requirements for better predictions are improved models and enhanced observations.

The Natural History of Madagascar

Book

Jan 2003

Eocene circulation of the southern ocean: Was antarctica kept warm by subtropical waters

Article

Jan 2004

Matthew Huber

The Ecology Of Rafting In The Marine Environment. Iii. Biogeographical And Evolutionary Consequences

Article

Jul 2006
OCEANOGR MAR BIOL

Rafting of marine and terrestrial organisms has important ecological, biogeographical and evolutionary implications. Herein the general principles of rafting are described and how they contribute to population connectivity. Rafting dispersal has particular characteristics, which may differ substantially from those of species with planktonic larval dispersal. Dispersal distances achieved via rafting can vary considerably: journeys may be very short or in some cases extremely long, depending on currents and wind. Accumulation of rafts in convergence zones facilitates cohesion of travelling groups, possibly reducing the risk of founder populations being very small. This becomes particularly important over long distances where singular founder events could provoke strong reduction of the genetic variability in the founded population. The frequency of transport affects the degree of connectivity between local populations. Three important raftingroutes are distinguished: frequent, intermittent and episodic. Frequent rafting routes are found in bays, lagoons and estuaries, and they are typically facilitated by substrata of biotic origin (seagrass, saltmarsh vegetation, intermediate-sized algae and mangroves). Intermittent rafting routes are found along temperate continental shores where they are facilitated primarily by giant kelps. In the subtropics and the Arctic intermittent rafting routes facilitated by wood are particularly important. Episodic rafting routes, which often cross vast areas of open ocean (biogeographic barriers), are facilitated by volcanic pumice, ﬂoating trees and occasionally by giant kelps when these are pushed beyond intermittent routes by strong winds or currents. Dispersal events occur in a highly sporadic manner in this latter category of rafting route, but when they happen, large amounts of ﬂoating substrata and rafters may be dispersed simultaneously. Intervals between events can be decades, centuries or even millennia, and consequently populations resulting from these events may be isolated from each other for long time periods. Population connectivity on frequent, intermittent and episodic rafting routes is high, intermediate and low, respectively. Genetic studies support these predictions, and furthermore underline that rafting may contribute to population connectivity over a wide range of geographic scales, from <100 km up to >5000 km. Rafting also has a strong effect on evolutionary processes of the organisms dispersed by this means. It is suggested that local recruitment (consequence of direct development) contributes to enhanced rates of population diver- gence among local populations of common rafters, but occasionally high genetic diversity may result from secondary admixture. Isolation of colonisers after singular episodic rafting events facilitates allopatric speciation. Through these processes rafting dispersal may support local species richness and thus have an inﬂuence on local biogeography and biodiversity. Human activities affect rafting connections in the oceans either by reducing or enhancing the possibility of transport and landfall. In many cases it cannot be safely decided whether the appearance of a species in a new habitat is due to rafting or to other transport mechanisms, and genetic studies can help to identify the most likely causes. Future ﬁeld and laboratory studies on the ecology of potential rafters in combination with genetic studies on different spatial and temporal scales will contribute to a better understanding of the mechanisms of rafting dispersal, consideration of which is crucial in developing efﬁcient conservation measures in the marine environment.

Mammals and Land Bridges

Article

Jan 1940
J Wash Acad Sci

George Gaylord Simpson

The Africa-Madagascar connection and mammalian migrations

Article

Mar 2006
J AFR EARTH SCI

Madagascar separated from Africa in the Middle–Late Jurassic and has been in its present position relative to Africa since the Early Cretaceous (∼120–130my). Several Early Eocene to Late Oligocene (∼50–26my) terrestrial mammalian groups are observed on Madagascar that have a similar ancestral lineage to those found in Africa. These mammalian groups means of transport across the Mozambique Channel from Africa to Madagascar was either by traversing on exposed land masses across a land bridge or by swimming/rafting, since (1) Madagascar has been separated from mainland Africa for at least 70my before their arrival, and (2) it is unlikely that similar ancestral lineage’s evolved simultaneously in separated regions. No evidence has been found for a land bridge across the Mozambique Channel. The mammals thus either swam or have been swept away on vegetation mats from rivers flowing out of Mozambique or Tanzania.

Petrological and geochemical investigations of rocks form the Davie Fracture Zone (Mozambique Channel) and some tectonic implications

Article

Oct 1992
J AFR EARTH SCI

Yannis Bassias

The Davie Ridge is an important structural feature of the Mozambique Channel. It represents the actual configuration of an old fracture zone (Davie Fracture Zone) between Africa and Madagascar along which Madagascar drifted southward during Late Jurassic and Early Cretaceous times. This fracture zone is built of crystalline continental basement (granites, gneisses and metaakoses) locally covered by a deformed ≪flysch≫ sequence, by alkaline lavas, tuffs and breccias and by a thin succession of carbonate oozes ranging in age from retaceous to Pleistocene. The sedimentary terrigeneous products derived from the fractured and eroded African basement and were deposited during an initial rifting/transtension in the incipient Somali Basin. During the southwards displacement of Madagascar, the lowermost arkosic sediments suffered low grade metamorphism in a collisional setting. The peak metamorphic conditions reached 4.5-5 kb and 350-400°C. These conditions imply at least 10-12 km deep burial of sediments (in transtensional?) and subsequent uplift (in transpressional setting?) to the present surface. Deep fracturing along the Davie Fracture Zone allowed the emplacement of alkaline within plate basalts in a marginal basin setting. However, these basalts were emplaced only at the northern and southern extremities of the fracture zone, while the central parts remained, at least superficially, undisturbed. Thermobarometric calculations on fragments of the old African continental basement found in the meta-arkoses confirm that the crustal pile below the present central parts of the fracture zone underwent slow uplift and erosion. These features imply that, during Paleozoic and Mesozoic and before the symmetric rifting in the Somali and the Mozambique Basins, the crustal pile between Africa and Madagascar underwent slow uplift and erosion. Unlike the evolution of the Limpopo belt in southeastern Africa and the western India basement (where uplift patterns are steeper or disturbed), the crustal thickness between Africa and Madagascar seems to have inhibited the emplacement of huge volume of magma along the Davie Fracture Zone. Although the amplitude of lithospheric thinning was the guiding mechanism for the volcanic activity along the fracture zone, both the basement reactivation and the volcanic activity along the Davie Facture Zone went controlled by the amplitude of uplift and erosion.

Taxonomy, habits, and relationships of the subfossil Madagascan hippopotami Hippopotamus lemerlei and H. madagascariensis

Article

Sep 1989

Solweig Stuenes

The revised taxonomy of the subfossil hippopotami of Madagascar suggests that of the hitherto described four species, only two are valid. Lectotypes of these two species, Hippopotamus lemerlei Grandidier and H. madagascariensis Guldberg, are designated and the species are redescribed. Cranial features suggest differences in functional anatomy and ecology between these two species. The short and deep glenoid fossa together with the lateral wear facets of the incisors in H. lemerlei indicate restricted lateral movements of the mandible, while the long and shallow glenoid fossa and the horizontally worn incisors of H. madagascariensis demonstrate extensive lateral mobility. The long, narrow skull of H. lemerlei, with its elongated facial portion and short postorbital part, points to an amphibious mode of life; while the proportions of the more robustly built skull of H. madagascariensis indicate a mainly terrestrial habit. Tip-to-tip occlusion of the incisors, a posterior groove in the upper canines, and double-rooted first premolars are present in H. madagascariensis, but not in H. lemerlei. Since the two Madagascan species are closely related, these features cannot be used as distinguishing characteristics on the generic level. The ancestral form(s) of the Madagascan hippopotami is not known, but H. amphibius, with its well-developed aquatic adaptations, seems to be a possible ancestor. Like many other fossil hippopotami from isolated islands, the two Madagascan species are dwarfs.