The southwards drift of Madagascar lasted for 50 million years, from the Middle Jurassic to the Early Cretaceous, and was governed by an almost 1,000 km-long north–south fracture zone which, east of the present Davie Ridge crests, is associated with a number of deep north-west to south-east trending faults.
This oblique rather than simple tectonic regime rippled and sheared the continental crust west of Madagascar, eroded and erased mountains, emerged and submerged volcanic islands and created sedimentary basins which were only fully connected when the sea finally inundated all the area between Madagascar and Mozambique in the Miocene.
Mapping Mounds and Evolutionary Bridges
In Figure 1 (below) the present situation is compared to that close to the Jurassic-Cretaceous boundary, around 145 million years ago. Triangles represent horst structures mapped from seismic and their position is projected back to that time. Although this is an illustration, the cluster of high mounts formed in the middle of the strait between Madagascar and Africa seem to have created a network of scattered islands that subsided later in Miocene times (Figure 1b). Did some of those islands, not just the Davie Ridge crests, remain above sea level until the Eocene? Rock analysis supports this. Detailed geological descriptions of dredges and cores along the Davie Ridge were published in the late ’80s and early ’90s and rock samples were recently correlated with seismic, gravimetry and airmag.
Understanding the surface environment has important implications for biodiversity. Zoologists still question whether ‘evolutionary bridges’ allowed mammalian groups to pass from Africa to Madagascar in Eocene times. The presence on Madagascar of several terrestrial mammalian groups with poor dispersion capacity across wide water barriers led several authors to suggest the existence of land-bridge connections between Africa and Madagascar during Middle Eocene to Early Miocene, (47–26 Ma), rather than assuming the rafting on vegetation mats hypothesis. This debate moved recently from classical taxonomy to genetic molecular comparative studies.
Plate reconstructions brought arguments constraining the ‘evolutionary bridges’ scenario, but rock descriptions from the Davie Ridge crests and flanks suggest there were long emersion periods. This in turn could indicate that until the Late Cretaceous and probably Eocene, the Rovuma Basin, today off the Mozambique and Kenyan coast, was somehow isolated from the Angoche Basin, now found off the Mozambique coast, and from the Morondava Basin, which is now off the Madagascan west coast, by a number of islands.
During the Palaeogene these islands could have restricted the eastward Mozambique palaeocurrent at latitudes south of today’s 19°S. It is possible that the Middle Eocene circum-Antarctic current between Australia and Antarctica, although it disrupted ocean currents worldwide and caused global cooling, could not strongly affect sea and land north of 19°S in the Mozambique Channel. Under these conditions the presence of a series of small, scattered Mesozoic-Cenozoic islands or remnants of them in the middle of the channel, combined with the eastward Mozambique palaeocurrent and wind circulation in the Palaeogene, favours the migration of mammalian groups from Africa to Madagascar.
Evidences of Emersion
Late Jurassic to Miocene formations and structures in the Mozambique Channel are visible on seismic. However, knowledge of the nature of the rocks and their crystals and the erosion that had a hand in shaping the peaks is not available without sampling, dredging, coring or drilling.
Dredge sampling of the sedimentary blanket of the Davie Ridge has revealed that the volcanic crests of the northern part of the ridge (around 14°S) remained close to or above sea-level for long periods and were covered by Miocene breccia, conglomerates and carbonates (Figure 2, above). Meanwhile, the crystalline basement of the crests at the median parts of the ridge (between 15° and 16°S) has been shown to have been eroded and there are outcrops of arkosic sediments, recrystallised pelites and flysch-like black psammitic beds, Upper Eocene bioturbated calcareous sandstones and high energy calcareous breccia below Late Miocene oozes (Figure 3, below). The southern crests of the ridge (between 17° and 19°S) were covered by highly altered basaltic breccia cemented in a carbonate matrix and by Middle Eocene platform carbonates below Early Miocene limestone.
The Paleocene is absent on the ridge and was not found by DSDP Leg 25, implying that a major unconformity has to be considered between the Cretaceous and the Paleocene. The local distensions which accompanied the drift of Madagascar and the lithospheric fissuring were both transgressive southwards, inducing either the first Cretaceous pelagic sedimentation or hypabyssal to subaerial volcanic activity. The northern part of the ridge around 14°S remained close to sea level until the Miocene. Volcanic activity was present at the northern and southern parts of the ridge although in the middle, around 16°S, the eroded crystalline basement was not penetrated by volcanics, and Late Cretaceous sedimentation was prominent on the eastern flanks. Eocene outcrops are better exposed at the southern parts of the ridge between 17°S and 19°S.
The Roles of Compression and Erosion
With a 1,000 km drift of Madagascar southwards along the fracture zone between 165 and 120 Ma, an average drift (half spreading rate) of 22 mm per year should yield extremely high convergent rates on the transpressional segments of the fracture zone. This amount of shortening should be sufficient to uplift large blocks, tens of kilometres across, over thousands of metres, a process which has been examined in depth at the San Andreas fault. Consequently, the origin of some push-up blocks east of the Davie Ridge may be related not only to listric but also to convergent wrench faults and flower-like structures if a wedge tectonic system was present. Similar extensional and compressional structures were recognised north-east of the Davie Ridge on the Madagascar side, between 12° and 14°S as well as west of the Davie Ridge on the African side, at around 15°S.
The shaping of the mountains depends as much on the destructive forces of erosion as on the constructive power of tectonics. In Cretaceous times erosion of the mountains is thought to have removed large quantities of rock but the resultant uplift replaced about 80% of the mass removed. From seismic lines we estimate that the missing parts of folds together with vertical displacement suggest that the removed mass represents about one second of folded Jurassic and Karoo, that led to about 0.8 seconds, or around 1,500m, of isostatic uplift (Figure 3). A very general estimation indicates that 100m could be removed through erosion every 10 million years at an average low rate of 0.01 mm/yr.
The process of erosion around the volcanic islands was later influenced by deep faulting and sliding, affecting unstable slopes at depths that can exceed five kilometres. The occurrence of such large slides on the flanks of the Davie Ridge is connected with dissolution features of the carbonate terraces and platforms, and may be influenced by Cenozoic high-amplitude fluctuations of the eustatic level.
The Role of Magma Underplating
Part of this uplift may be related to magma underplating as well as to compression and erosion. Up and down movements of several hundreds of metres at plate edges can be explained by the mantle influence. Several cases of continental ‘bobbing’ during Cretaceous times have been observed around the globe. Mantle is, in fact, able to lift the surface by hundreds of metres, easily enough to cause an apparent drop in sea level.
In the case of the Mozambique Channel, basalts were emplaced after cessation of the southwards movement of Madagascar, apparently at a time of quiet when India separated from Madagascar at around 83 Ma. These basalts have alkaline affinities and were emplaced along pre-existing zones of weakness of the fracture zone under hypabyssal to subaerial conditions and in some cases suffered hydrothermalism. The crustal thickness between Africa and Madagascar seems to have inhibited the emplacement of huge volumes of magma along the fracture zone but in turn this could have favoured a Cretaceous uplift.
The Role of Eustasy
A sea level drop of a few hundred metres due to Middle Cretaceous eustatic changes would narrow the channel only slightly to reveal few islands. However, the combined effects of tectonic activity, sediment accumulation/compaction and eustasy on accommodation changes can be important on a regional scale, with rates varying from several to tens of metres per million years. The Cenomanian-Turonian interval was punctuated by large and rapid sea level drops. High-magnitude and short duration changes (tens of metres in <100,000 years) are known, resulting in ‘absolute’ rates of sea level change in the order of 150–1,000 m/Ma.
The rise in sea level due to sea floor rise as a result of oceanic crust global events did not compensate for uplift along the Davie Ridge. From the Eocene and through the early Miocene, water retreat left deposits of limestone and chalk, while basalts were eroded and brecciated and the fragments were deposited in a carbonate matrix, dated by nannoplanktons to be Middle Eocene. The presence of limestones and chalks at the southern parts of the ridge can be explained by this mechanism. It is important to note that these rocks were dredged at present depths of between 2,450 and 600m, which do not represent the shallow depths of deposition.
From the Miocene these islands started to sink to finally form present-day seamounts, atolls and islands.
The Cretaceous ocean floor rise, although helped by an increase in carbon dioxide, a Middle Cretaceous important greenhouse episode and the deposition of limestone, did not compensate for the large uplift due to compression and folding which has been observed in several parts of Gondwana, including the Cretaceous Mozambique Channel. Many geological processes have taken place between East Africa and Madagascar since Jurassic times, and although today the channel is a huge marine environment, these processes can be observed in the rock history. Even if the amplitude of lithospheric thinning was the guiding mechanism for volcanic activity along the Davie Fracture Zone, both the basement reactivation and the volcanic activity along the fracture zone were controlled by the amplitude of uplift and erosion. From a petroleum exploration standpoint, the size of this uplift may have enormous consequences for the Mesozoic depositional environments of the Mozambique Channel.
The Majunga and Rovuma Basins were probably isolated from the Angoche and Lacerda Basins and from the Morondava Basin with distinctly different sedimentary and diagenetic histories until the Middle Miocene, which has important implications for hydrocarbon exploration.
The presence of a series of small, scattered Mesozoic-Cenozoic islands in the middle of the channel, across which mammalian groups could migrate from Africa to Madagascar, brings some new support onto the land-bridges theory. There was perhaps an initial period of around 20 million years, until the Jurassic–Cretaceous boundary, when biotic exchange would have been present between western and eastern Gondwana, followed by a second period during the Eocene that preserved emerged parts of the Davie Ridge and of its eastern slopes.
It is clear that the Mozambique Channel still holds a number of secrets and there is scope for considerable further research.