Roadmap | The review process | Home
   Cape Verde
Ancient continental lithospheric mantle beneath oceanic islands – Cape Verde Archipelago


Suzanne O’Reillya, Massimo Coltortib, Costanza Bonadimanb, William Griffina & Norman Pearsona


aGEMOC National Key Centre, Department of Earth and Planetary Sciences, Macquare University, NSW 2109, Australia,,

bDip. Scienze della Terra, Università di Ferrara, Via Saragat, 1, 44100 Ferrara, Italy;


This webpage is a summary of: Coltorti, M., Bonadiman, C., O’Reilly, S.Y., Griffin, W.L. and Pearson, N.J. Buoyant ancient continental mantle embedded in oceanic lithosphere (Sal Island, Cape Verde Archipelago), Lithos, 120, 223-233, 2010.


The Cape Verde Islands lie in the Atlantic ocean off West Africa (Figure 1), in a clearly oceanic setting. The lavas of some islands carry mantle-derived xenoliths of depleted peridotites (low in Ca, Al, Fe and other basaltic components), petrologically similar to those derived from cratonic lithospheric mantle. Oceanic lithospheric mantle, in contrast, consists mainly of less-depleted peridotites (lherzolites and harzburgites) formed by the extraction of mid-ocean ridge basalts. In situ Re-Os analyses of individual sulfide grains from depleted xenoliths from the island of Sal yield Re-depletion model ages ranging mainly from Neoproterozoic to Archean. Their age distribution mirrors the tectonic history of the western margin of the West African Craton and the corresponding continental margin of Brazil. These data suggest that part of the Cape Verde Archipelago is underlain by a fragment of ancient subcontinental lithospheric mantle (SCLM), left stranded in the oceanic lithosphere during the opening of the Atlantic Ocean (Coltorti et al., 2010). Contamination of magmas by this ancient continental root can explain the unusual isotopic characteristics of some Cape Verde lavas without recourse to recycled continental material in the sources of mantle plumes (O’Reilly & Griffin, 2010).


Figure 1. Locality map for the Cape Verde Islands relative to Africa, and for the position of Sal within the Cape Verde Island group.


The sulfide Re-Os age data from the Cape Verde mantle xenoliths (Figure 2) broadly reflect the timing of events characterizing the tectonic history of the Atlantic suture, and especially of the West African margin. The oldest Archean ages correlate with the ancient rocks of the West African Craton; the peak around 2 Ga correlates with the Birrimian volcanism and the Eburnian orogeny along the margins of the craton. The major Neoproterozoic peak (ca 1 Ga) matches the collision of the West African and Sao Luis Cratons and related volcanism. The youngest group of ages (550-800 Ma) corresponds to the breakup of the continent along the earlier suture, and the opening of the Iapetus Ocean. The dominance of Neoproterozoic ages is consistent with the position of the Cape Verde Archipelago closer to the inferred former suture between Africa and South America, where pre-existing SCLM would have been most strongly modified.


Figure 2. Age histogram (n=135) and cumulative probability plot of TRD (rhenium depletion) age (n=108) for mantle sulfides.


The broad spread of model ages between ca 1800-1300 Ma (ca 30% of the data) may reflect either reactions between successive sulfide generations (i.e. mixed ages), or episodic metasomatism of the SCLM. In any case, the existence of possible “mixed ages” requires the presence of an older (cratonic) component. This possibility is also suggested by recent global tomographic mapping which shows a high-velocity region below the Cape Verde Archipelago (Figure 3).


Figure 3. Seismic tomography (Vs) image of northern Africa and the adjacent Atlantic Ocean at 0-100 km (for description of the methodology and sources, see O’Reilly et al., 2009 and Begg et al., 2010) showing high-velocity regions in the ocean basin. (Note colour reversal with hot colours indicating high velocity).


The presence of a domain of Archean-Proterozoic SCLM beneath the Cape Verde Islands suggests that the opening of the ocean basin did not involve a clean break, with generation of new oceanic lithosphere directly overlying the underlying convective mantle. Depleted Archean to Proterozoic SCLM is buoyant relative to the convecting mantle (e.g., Poudjom Djomani et al., 2001; O’Reilly et al., 2001; Griffin et al., 2009), and detached fragments are thus unlikely to sink. Instead they may “surf the convecting mantle” (Alard et al., 2005) as the ocean opens. Such fragments of relict SCLM may be widespread in the ocean basins.

We suggest that during rifting, before true oceanic crust has been formed, rising MOR-like melts impregnated the pre-existing SCLM, causing a progressive heating and rheological weakening. A combination of rifting and ductile extension could produce an intermingling of old pieces of SCLM and new oceanic lithosphere. Geological and geochemical studies of the Lanzo peridotite massif (Piccardo et al., 2007; Rampone et al., 1998) provide a possible analogue. Lanzo may represent a transition between old SCLM and oceanic lithosphere. Extending SCLM was modified by magmas, causing progressive thermo-chemical erosion of the SCLM.

The presence of an SCLM remnant beneath the Cape Verde islands also can explain the complex isotopic signatures of the magmatic rocks. Geochemical signatures typical of ancient SCLM components have been recognised in some lavas on Sal (e.g., Holm et al., 2006) and the interaction of rising magmas (derived from the convecting mantle) with the SCLM could impart these signatures.

If such SCLM remnants are widespread in the ocean basins, they offer an alternative interpretation of the origin of EM1 and EM2 components (geochemical signatures in basalts interpreted as evidence of recycled crust) in ocean-island basalts. Rather than reflecting continental components subducted into the lower mantle and sequestered until tapped by rising plumes, these signatures may simply indicate contamination of rising magmas by relatively shallow SCLM remnants (O’Reilly et al., 2009).

The new Re-Os data resulting from this study, coupled with those from abyssal peridotites, suggest that the traditional picture of oceanic lithosphere as the residues of basalt extraction at mid-ocean ridges is probably an oversimplification. Instead, the development of ocean basins may involve disruption of continental lithosphere and incorporation of relict SCLM domains in oceanic regions (Figure 4). The repeated opening and closing of oceans, commonly along older sutures, must have left many SCLM remnants of different ages within the ocean basins. This process will produce significant compositional and geochronological heterogeneity in the oceanic lithosphere over time.


Figure 4. Cartoon showing possible oceanic rifting mechanism with listric faulting at cratonic margins and stranding of ancient subcontinental lithospheric mantle remnants in the ocean basin (after O’Reilly et al., 2009).



last updated 26th November, 2010