2) Evolutionary History

Is there a Western Indian Ocean “Coral Triangle”?

David Obura with Melita A. Samoilys, Johann R.E. Lutjeharms*, Juliet Hermes, Chris Reason, Raymond Roman, Charine Collins, Lorenzo Alvarez-Filip, Denis Macharia

* Professor Johann Reinder Erlers Lutjeharms, the world's pre-eminent expert on the oceanography of the Agulhas Current, sadly passed away on 8 June 2011. This talk is dedicated to him.

Marine Science for Management programme, WIOMSA. 2008-11

26 ICRI General Meeting Indian Ocean Day. 13 December 2011, La Reunion www.cordioea.org 1 What do I mean by a “coral triangle”?

Geo-physical integrity (geology, oceanography)

Ecological integrity (connectivity, productivity, diversity)

Historical integrity (evolutionary dynamics)

Human affairs (politics, culture, history)

2 Outline

1) TODAY

Biodiversity patterns - 2) EVOLUTIONARY HISTORY corals Geology 3) SO WHAT? The South Equatorial • Plate tectonics Current and the • Mascarene-Reunion hotspot Conclusions Mozambique channel Marine climate and habitats Relevance •Paleoclimate and marine carbonatesContinental shelf area and shallow seas •Connectivity

Coral phylogeny

3 1) Realms (MEOW) Spalding et al. 2008. Bioscience 57:373-383

2) Provinces

Scleractinian corals - Global Marine Species Assessment (GMSA) IUCN Red List of Threatened Species

Results 3) Ecoregions 1.All West Indo-Pacific provinces group together, except the Andaman Seas, which group with the Central Indo- Pacific provinces. 2.Within the West Indo-Pacific, the Western Indian Ocean Province (#20) ecoregions are the core/central cluster 3.Within the Western Indian Ocean Province (#20) the East African Coast (95) and West & North Madagascar (#100) are the most closely related ecoregions.

4 Survey data, coral species assemblages at 24 sites in the WIO - 413 coral species • Central cluster of sites - mixed EA coral coast and W&N Madagascar x sites. Highest diversity in the region, x x • Comorian and Seychelles islands

There is no biological boundary x between the mainland and islands!!!

* Core high biodiversity region in the northern Mozambique channel

Concentric/radial pattern around the northern Mozambique channel

x Oceanographic 1) South Equatorial Current - East-west transport of corals features that across the Indian Ocean support these from the Indonesian region. patterns

2) Mozambique channel gyres & eddies result in high connectivity across the northern Mozambique Channel and potential for accumulation of genetic and larval material

3) Currents flow north and south from the northern Mozambique Channel, with some eddies and return flow in the north (Seychelles/Chagos) and QuickTime™ and a south (Mauritius/Reunion) JVT/AVC Coding decompressor are needed to see this picture. Anti-cyclonic eddies – counter-clockwise rotation, positive sea level anomalies, warm core eddies.

Cyclonic eddies – clockwise rotation, negative sea level anomalies, cold core eddies. x

x x

x * Outline

1) TODAY

Distinct regional faunal identity, Biodiversity patterns - encompassing the WIO, 2) EVOLUTIONARY HISTORY northwestcorals Indian Ocean and Red Sea/Gulfs regions Geology 3) SO WHAT? The South Equatorial • Plate tectonics HighCurrent diversity and core the region • Mascarene-Reunion hotspot Conclusions centredMozambique on the NMC channel (corroborated by other studies) Marine climate and habitats Relevance •Paleoclimate and marine Currents that distinguish the NMC as an accumulation carbonatesContinental shelf area and center, and source for other shallow seas regions within the WIO •Connectivity

J. Biogeography - MS Coral phylogeny submission

200 168 120 100 Plate tectonics

Age of the ocean floor

mya 0

The Mozambique channel 67 has the oldest coastlines in the Indian Ocean

9 200 250 mya The Tethys Sea

mya 250 opening of Tethys Sea, eastern shore of Pangea 175 equatorial opening as breakup of Pangaea into Gondwana and Laurasia

150 greatest influence of Tethys - 70 mya. Shaded areas are shallow seas - drowned shallow seas between Laurasia continental margins and Gondwana -> shallow carbonates and main coral province globally 60-45 (Eocene) progressive obstruction by India, and by Africa/Middle East moving northwards 15 (Miocene) effective closure of Tethys

10 Quarternary 2.6 Neogene 23-2.6 my Paleogene 66-23 my Eocene 55 Miocene Miocene 24 mass extinction Cretaceous - 0

- 35 mya - 5 mya - Tertiary

Tethys Sea decline

Coral Triangleregion into SE Asia,formingdiversity the high started to form as Australia bumped From 24 mya,Indo the reefs mostand diverse coralfauna largestgrowththe with the planet, on the most extensive provinceof reef From 70 -

35 mya,35 Tethys the wasSea

- Malayan region

Plate tectonics animation, Legend

23 my 23

- 67 mya to present TECTONIC PLATES Varied colour - craton/continental crust

Paleogene 66 Paleogene outlines

HOTSPOTS Red - Large Igneous Provinces (hotspot activity)

By L.A. Lawver, M.F. Coffin, L.M.

Gahagan, D.A. Campbell, and J.-Y. 2.6 my 2.6

- Royer

Neogene 23 Neogene Sponsors: Conoco, Elf, Exxon-Mobil, Norsk Hydro, and Statoil.

Quarternary 2.6-0 12 i - Tethys closing ii - northward movement

13 Cretaceous Maastrichtian

PLATES 2000 UTIG Cretaceous Maastrichtian

PLATES 2000 UTIG Paleogene Early Danian

PLATES 2000 UTIG Paleogene Danian

PLATES 2000 UTIG Paleogene Late Danian

PLATES 2000 UTIG Paleogene Early Selandian

PLATES 2000 UTIG Paleogene Selandian

PLATES 2000 UTIG Paleogene Late Selandian

PLATES 2000 UTIG Paleogene Early Thanaetian

PLATES 2000 UTIG Paleogene Thanaetian

PLATES 2000 UTIG Paleogene Early Ypresian

PLATES 2000 UTIG Paleogene Ypresian

PLATES 2000 UTIG Paleogene Late Ypresian

PLATES 2000 UTIG Paleogene Early Lutetian

PLATES 2000 UTIG Paleogene Lutetian

PLATES 2000 UTIG Paleogene Late Lutetian

PLATES 2000 UTIG Paleogene Early Bartonian

PLATES 2000 UTIG Paleogene Bartonian

PLATES 2000 UTIG Paleogene Late Bartonian

PLATES 2000 UTIG Paleogene Early Priabonian

PLATES 2000 UTIG Paleogene Priabonian

PLATES 2000 UTIG Paleogene Late Priabonian

PLATES 2000 UTIG Paleogene Early Rupelian

PLATES 2000 UTIG Paleogene Rupelian

PLATES 2000 UTIG Paleogene Late Rupelian

PLATES 2000 UTIG Paleogene Early Chattian

PLATES 2000 UTIG Paleogene Chattian

PLATES 2000 UTIG Paleogene Late Chattian

PLATES 2000 UTIG Neogene Early Aquitanian

PLATES 2000 UTIG Neogene Aquitanian

PLATES 2000 UTIG Neogene Late Aquitanian

PLATES 2000 UTIG Neogene Early Burdigalian

PLATES 2000 UTIG Neogene Burdigalian

PLATES 2000 UTIG Neogene Late Burdigalian

PLATES 2000 UTIG Neogene Early Langhian

PLATES 2000 UTIG Neogene Late Langhian

PLATES 2000 UTIG Neogene Early Serravallian

PLATES 2000 UTIG Neogene Serravallian

PLATES 2000 UTIG Neogene Late Serravallian

PLATES 2000 UTIG Neogene Early Tortonian

PLATES 2000 UTIG Neogene Tortonian

PLATES 2000 UTIG Neogene Late Tortonian

PLATES 2000 UTIG Neogene Early Messinian

PLATES 2000 UTIG Neogene Late Messinian

PLATES 2000 UTIG Neogene Early Zanclean

PLATES 2000 UTIG Neogene Late Zanclean

PLATES 2000 UTIG Neogene Piacenzian

PLATES 2000 UTIG Quaternary

Restart animation PLATES 2000 UTIG Island height and mass Mascarene- decrease with age and 10-20 m crust subsidence, 21-30 m ending in coralline Reunion hotspot atolls, banks and Saya de Malha, 45 my submerged platforms A hotspot is a stationary point (guyots) where magma from the Earth’s mantle penetrates the oceanic crust above, forming island chains (e.g. Hawai’i). St. Brandon, 30 my?

la Reunion - Piton de la Fournaise

Mauritius, 8 my

83 Reunion, 2 my

Cretaceous-Tertiary 23 my 23

- mass extinction

Deccan Traps Paleogene 66 Paleogene superplume event

A superplume is an extraordinary extrusion of magma, forming Large

Igneous Provinces (e.g. Deccan Traps, 2.6 my 2.6 - India), sometimes associated with major geological and evolutionary

tipping points (e.g. K-T extinction). Neogene 23 Neogene

Dyment 2007, Quarternary 2.6-0 Sheth 2009 84 Quarternary 2.6 Neogene 23-2.6 my Paleogene 66-23 my - 0

Mascarene Deccan Traps 66 Nazareth 31 Lakshadweep60 Saya de45 Malha Cargados Carajos Maldives 57 Chagos 49 Mauritius 8 Deccan Deccan Traps to Reunion Reunion 2 - 33 - Reunion hotspot activity

hotspot passed over the spreadingridge Malha as the Chagos andSaya de Split between basalts granitics/Deccan Seychelles

i - hotspot activity (red lines) ii - position of Seychelles (moves) iii - position of the hotspot (stationary)

86 Cretaceous Maastrichtian

PLATES 2000 UTIG Cretaceous Maastrichtian

PLATES 2000 UTIG Paleogene Early Danian

PLATES 2000 UTIG Paleogene Danian

PLATES 2000 UTIG Paleogene Late Danian

PLATES 2000 UTIG Paleogene Early Selandian

PLATES 2000 UTIG Paleogene Selandian

PLATES 2000 UTIG Paleogene Late Selandian

PLATES 2000 UTIG Paleogene Early Thanaetian

PLATES 2000 UTIG Paleogene Thanaetian

PLATES 2000 UTIG Paleogene Early Ypresian

PLATES 2000 UTIG Paleogene Ypresian

PLATES 2000 UTIG Paleogene Late Ypresian

PLATES 2000 UTIG Paleogene Early Lutetian

PLATES 2000 UTIG Paleogene Lutetian

PLATES 2000 UTIG Paleogene Late Lutetian

PLATES 2000 UTIG Paleogene Early Bartonian

PLATES 2000 UTIG Paleogene Bartonian

PLATES 2000 UTIG Paleogene Late Bartonian

PLATES 2000 UTIG Paleogene Early Priabonian

PLATES 2000 UTIG Paleogene Priabonian

PLATES 2000 UTIG Paleogene Late Priabonian

PLATES 2000 UTIG Paleogene Early Rupelian

PLATES 2000 UTIG Paleogene Rupelian

PLATES 2000 UTIG Paleogene Late Rupelian

PLATES 2000 UTIG Paleogene Early Chattian

PLATES 2000 UTIG Paleogene Chattian

PLATES 2000 UTIG Paleogene Late Chattian

PLATES 2000 UTIG Neogene Early Aquitanian

PLATES 2000 UTIG Neogene Aquitanian

PLATES 2000 UTIG Neogene Late Aquitanian

PLATES 2000 UTIG Neogene Early Burdigalian

PLATES 2000 UTIG Neogene Burdigalian

PLATES 2000 UTIG Neogene Late Burdigalian

PLATES 2000 UTIG Neogene Early Langhian

PLATES 2000 UTIG Neogene Late Langhian

PLATES 2000 UTIG Neogene Early Serravallian

PLATES 2000 UTIG Neogene Serravallian

PLATES 2000 UTIG Neogene Late Serravallian

PLATES 2000 UTIG Neogene Early Tortonian

PLATES 2000 UTIG Neogene Tortonian

PLATES 2000 UTIG Neogene Late Tortonian

PLATES 2000 UTIG Neogene Early Messinian

PLATES 2000 UTIG Neogene Late Messinian

PLATES 2000 UTIG Neogene Early Zanclean

PLATES 2000 UTIG Neogene Late Zanclean

PLATES 2000 UTIG Neogene Piacenzian

PLATES 2000 UTIG Quaternary

Restart animation PLATES 2000 UTIG What influence have the banks had on WIO currents and evolutionary processes?

At younger ages, these banks may have been emergent land masses higher than Reunion and 100s of km in extent, and may have formed a more continuous land mass than at present.Emergent land Legend (m below sea level) at -125 m SL Yellow-red: 125 - 0 Greens: 500 - 125 One of the biggest question marks Blues: > 500 Almost no records of Tertiary Paleoclimate & carbonates carbonate deposits in WIO region.

Closest are in Somalia (later mention 23 my 23 - End of the Tethyan era was warm, associated of Acropora fossils). with lower pH of seawater - less formation of carbonates, and favouring foraminifera, During this warm period, corals may bivalves. have found refuge at higher latitudes

Paleogene 66 Paleogene (cooler, less acidic) Mass extinction at K-T boundary = fewer species to fossilize The northern Mozambique Channel was at higher latitudes - may have “Paleogene gap” in fossil record. remained in a consistent climatic

zone.

2.6 my 2.6

- Neogene 23 Neogene

Quarternary 2.6-0

Continental shelves, banks and sea level change

Paleogene Neogene

23 my 23

- Paleogene 66 Paleogene

Yellow - red < 125 m Bathymetry less than 125 m below current sea level Approximate reconstruction of Tertiary sea levels

Steep, simple continental shelves indicate little change in coastline shapes - stable

refuges from change elsewhere

2.6 my 2.6 - At lower sea level stands, the Mascarene banks would have been islands several 100s km wide -

fluctuating connectivity/isolation and a Neogene 23 Neogene possible diversity pump

Quarternary 2.6-0 Quarternary 2.6 Neogene 23-2.6 my Paleogene 66-23 my connectivity Ocean currents and - 0

• • • • • contained sea isolating semiWIO asa the Bank formation possibly the north connectivity Tethyswith and Early limited in this period, by India and banks E Asia central bumping and IO into India migrating across Eocene,55 - W intoflowblocked WIO

- 35 mya35

• • • • •

level fluctuations E E Species diversification in ‘coral triangle’ starts Isolation from diminishing Tethys Miocene, 20 - - W connectivity alteredW connectivity bybanks and sea being W establishedconnectivity - - 10 mya10

Huberand 2008 Ali

Coral species phylogeny

Acropora

23 my 23 - First Acropora fossil from Somalia, Late

Paleocene (Carbone et al. 1994) Paleogene 66 Paleogene

Eocene fossils from European locations, of 9 of the 20 extant species groups (shown in • most diverse family (Acroporidae) and bold, Wallace and Rosen 2006) genus with > 120 species Acropora species groups • highest diversity and endemism in the Coral Triangle, suggesting that this may have been its center of origin, as

well as current center of diversity

2.6 my 2.6 - These are indicative of a Tethys Sea origin and diversification, long before the Coral Triangle

Neogene 23 Neogene region was formed

Quarternary 2.6-0 Regional endemics Other groups with apparent Tethyan origins and diversification

Siderastrea Stylophora

• Three branching morphs are distinguished (Stefani et al. 2010), • phylogenetic tree suggests Indo- with most nominal spp split Pacific species (S. savignyana) is among two of these. differentiated into two separate species, by samples from • Most diverse in the WIO (Flot et Oman/Kenya, versus al. 2011) Taiwan/Australia. Possible that these reflect an • Ancestral species in the Red Sea older paleo-geology (Tethyan) and • Oman/Kenya ancestral, and close (Chen and colleagues, paleo-oceanography (isolation of to the Atlantic species unpublished) WIO)?

Scleractinian corals - diversification during the Cenozoic 23 my 23 - Atlantic/Caribb Europe/Mediter West Indian Indo-West

ean ranean Ocean Pacific Paleogene 66 Paleogene

1 Eocene diversification after K/T extinction

2 Miocene (early)

Formation of Coral

Triangle

2.6 my 2.6 - 3 Miocene (mid) Tethys Sea closed

Neogene 23 Neogene 4 Miocene (end) Isthmus of Panama formed

Quarternary 2.6-0 5 Today 162 What are the regional Atlantic/Caribb Europe/Mediter West Indian Indo-West affinities of the Indian ean ranean Ocean Pacific Ocean-restricted corals in the WIO?

Coral species samples from 24 sites in the WIO

Total of 413 species:

375 (90.8 %) are broadly distributed in the Indo-Pacific

38 (9.2 %) are found only in the West Indo-Pacific Realm most species are 9.2% 90.8% widespread in the WIO - regionally distinct fauna (24) Red Sea/northern core region species on mainland spp (5) and islands (7)

163 Outline 1) TODAY 2) EVOLUTIONARY HISTORY 3) SO WHAT? DistinctBiodiversity regional patterns faunal identity, - GeologyTethys as a center of origin for modern encompassingcorals the WIO, •coralPlate lineages tectonics Conclusions northwest Indian Ocean and • Mascarene-Reunion hotspot The WIO contains distinct evolutionary RedThe Sea/Gulfs South Equatorial regions Relevance lineages from the Tethys Current and the Marine climate and habitats HighMozambique diversity core channel region centred on the NMC •FluctuatingPaleoclimate connectivity/isolation and marine of the (corroborated by other studies) carbonatesContinentalWIO as a result of the Mascarene shelf area banks and shallowand islands seas Currents that distinguish the • Connectivity NMC as an accumulation point, The northern Mozambique Channel as a and source for other regions Coralstable phylogenyrefuge in a dramatically changing Ocean basin within the WIO

J. Biogeography - MS Dominance of West-Pacific fauna is a submissions recent function of E-W transport in the SEC, masking a regionally-specific fauna

Book proposal - WIOMSA/independent publication Conclusions - conservation biology

“Museum” hypothesis - the Mozambique channel •has the oldest coastlines in the Indian Ocean/WIO •stable with respect to climatic zone (northward migration of Refuge/preservation of Africa/Madagascar) genetic material/species - •stable with respect to habitats and SL change remnant lineages, endemic - the WIO species •has a relict fauna from the Tethys Sea

“Diversity pump” - Mascarene-Reunion hotspot/tectonics • diversification following K-T extinction in the shallow Tethys Speciation of WIO • relative isolation as India migrated northwards regional fauna - • fluctuating isolation due to banks/islands endemics

“Species sink” - ‘recent’ oceanography (20 my) •accumulation of species from West-Pacific/Eastern IO •retention of genetic material in Mozambique channel Homogenization of fauna gyres/eddies with broader Realm/Biome 165 Relevance Geo-physical integrity (geology, oceanography) Ecological integrity (connectivity, productivity, diversity) Historical integrity (evolutionary dynamics) Human affairs (politics, culture, history)

Marine Ecoregions of the World (MEOW)

EastRealm Africa - Marine West Indo-Pacific Ecoregion- Tethyan (EA origins,Coastal deepMarine evolutionary Protected Area history Network Initiative, WWF) Network (RAMP-COI) & Province - Western Indian OceanWIOMER - WWF - Tethyan origins - continental coasts and core region = refuge - banks and connectivity - diversity pump

Ecoregion Northern Mozambique channel - refuge over multiple scales of evolutionary time - species accumulation - high connectivity, source region - core ecoregion for the WIO

Other ecoregions Mascarene banks and islands, southern Mozambique channel, northern Monsoon coast Coping with Global Change and threats ...

Regional marine science and conservation in an evolutionarily consistent framework ...

x Tectonics

Geology

Oceanography Phylogeny Biogeography

This knowledge supports ... Current status •Policy cohesion (Nairobi Convention, ICRI, etc) Many new •Scientific integration (WIOMSA, ASCLME, etc) questions •Increasing and sustaining financing posed ...