Deep mantle components and geodynamic control on intraplate magmatism

As shown in the previous sections, the compositional characteristics of the parental basic magmas for the three investigated provinces are compatible with their segregation from the lithospheric mantle between ca. 30 and 110 km depth. The solidus condition of magma sources appears fairly close to the regional geotherm inferred for the central Mediterranean area, thus suggesting that partial melting processes could be easily triggered by local decompression effects. The intraplate volcanic activity is therefore compatible with a limited extensional/transtensional regime as in fact observed in impactogenic rift systems oblique to orogenic chains.

In the investigated provinces, the alkaline and deeper magmas show a more marked HIMU-like signature compared to the subalkaline basalts; this suggests a more intensive and probably more recent enrichment, of the deeper lithospheric mantle sources (> 60-70 km, i.e. in the Thermal Boundary Layer; Wilson et al., 1995) by the HIMU metasomatizing agents. On the other hand, the subalkaline magmas generated from shallow lithospheric mantle sources display a variable influence of EM components, particularly in Sardinia.

Mantle xenoliths associated to alkaline lavas of the three volcanic provinces generally represent shallow portions of the lithospheric mantle column (< 40-60 km depth, i.e. in the MBL) according to thermobarometric estimates and rheologic characteristics (Siena and Coltorti, 1993; Verde, 1996; Beccaluva et al, 2005b). They usually reflect reaction(s) between sub-lithospheric metasomatizing agents and previously depleted lithospheric mantle, inducing variable enrichments of the most incompatible elements (e.g. LREE, LFSE, etc.). The resulting isotopic signatures conform to those of their host magmas, being dependent on the variable contributions of the DM, HIMU, and EM components. In particular, a prevailing HIMU imprint is recorded for both Veneto and Iblean mantle materials, whereas a predominant EMI component, in addition to HIMU, is observed for the Sardinian mantle.

On the basis of the available data Beccaluva et al. (2005b) proposed a distinct compositional evolution of the European (Sardinia) and African/Adriatic (Iblei and Veneto) lithospheric mantle: the European lithosphere is characterized by DM, enriched by prevailing EMI metasomatic components whose effects have been recorded at least since the mid-Mesozoic (as recorded in European peridotite massifs), whereas the addition of HIMU seems to have been effective in both European and African lithosphere since the Late Cretaceous (Wilson and Bianchini, 1999).

From the above, two main questions remain open: 1) which is the ultimate origin of geochemical components that metasomatized intraplate magma sources? 2) Is there any influence of the neighbouring subduction zones on intraplate magmatism?

Mantle sources of Cenozoic intraplate magmatic events in the Central Mediterranean area invariably involve OIB-type (HIMU, EMI, EMII, Zindler and Hart, 1986) geochemical components, whose genesis require recycling via ancient subductions (and long-term isolation) of crustal s.l. material deep in the mantle (Weaver, 1991; Carlson, 1995; Hofmann, 1997).

The mantle region in which crustal material is preferentially recycled and stored, as a result of subduction processes, is the zone between the Upper and the Lower Mantle (ca. 410-660 Km depth) where most slabs flatten. The exclusive OIB signatures recorded in the studied intraplate magmas imply that: 1) any geochemical influence from the neighbouring Cenozoic subduction zones has to be excluded; 2) the old “refertilized” deep mantle regions were repeatedly reactivated releasing the OIB metasomatic agents, which in turn may rise in the overlying lithospheric mantle.

However, subducted slabs of the Mediterranean orogens seem to play a significant role in the genesis of anorogenic intraplate magmas shortly after the end of an orogenic cycle as observed in Sardinia (Beccaluva et al, submitted). In this province the anorogenic volcanic fields lies above the ca. 800 km long Ionian slab, subducted since the Eocene and currently flattened at ca. 600 km depth beneath the Tyrrhenian and Sardinia. The down going slab may have triggered convective instabilities in the surrounding mantle ultimately favouring melting processes at shallower levels.

This interpretation may also stand for the Veneto Volcanic province close to the collisioned Alpine subduction(s) during Paleogene and for the Iblean volcanism neighbouring the southern edge of the still active Eolian arc subduction.

Laboratory and theoretical models (Kincaid and Griffith, 2003; Faccenna et al., 2010) have shown the importance toroidal/vertical mantle flows around the edges of a subducting slab. Moreover convective instabilities around subducted slabs have recently been referred to as “splash plumes” (Davies and Bunge, 2006) involving localized mantle upwellings and rising metasomatic agents from the top of the Mantle transition Zone (410-660 km depth) to the overlying upper mantle/lithosphere (Wilson and Downes, 2006; Lustrino and Wilson, 2007; Bianchini et al., 2010a; 2010b).

In this framework it is plausible that the subduction processes of the Alpine (s.l.)-cycle generated “splash plume” instabilities, favouring intraplate magma generation in the surrounding mantle. Therefore, the studied intraplate volcanism may represent a far field consequence and a dynamic (not compositional) response to the Cenozoic subductions throughout the Mediterranean realm. This hypothesis could explain the timing of intraplate magmatism often nearly coeval or postdating the orogenic (subduction related) magmatism.