Assimilation of Crustal Rocks

Because the various surface and crustal enviromnents are characterized by different and distinctive isotope compositions, stable isotopes provide a powerful tool for discriminating between the relative role of mantle and crust in magma genesis. This 

---

Based on major and trace element variations, it has been argued that fractional crystallisation was the dominant evolutionary mechanism for Ustica magmas (Cinque et al. 1988). The lack of correlation between silica and Sr isotopic ratios has been interpreted as evidence that assimilation of crustal rocks played a minor role during magma evolution. The variable contents of K20, P2Os incompatible elements and radiogenic Sr in the mafic rocks most likely reflect the occurrence of various types of primary melts at Ustica. 

The 5 0 values of rocks from the Dufek Massif reported by Ford et al. (1986) in Fig. 13.47 range narrowly from +5%o to +7%c with one low value of less than +4%c. These values are normal for mafic igneous rocks that crystallized from mantle-derived magma without extensive assimilation of crustal rocks or hydrothermal alteration. The 6 0 values of rocks from the Forrestal Range vary from normal values of +6%o to low values between 0%c and +l%o. Several prominent rock types in the Forrestal Range have distinctive 8" 0 values (Ford et al. 1986) ... 

The assimilation of crustal rocks by hot mantle-derived basalt magma is facilitated in cases where the temperature of the crust approaches the melting point of granitic rocks. Under these circumstances a partial melt may form by the additional heat emanating from the basalt magma that is rising from depth. Such silica-rich partial melts move with the basalt magma toward the surface and are mixed with it while both are in transit. 

There is little doubt that the Tuscany mafic magmas have been subject to fractional crystallisation, mixing and crustal assimilation (e.g. Conticelli 1998). However, their high MgO, Ni and Cr concentrations, whose values are close to those of primary mantle melts, exclude that the mafic magmas with different enrichments in potassium and incompatible elements can be derived from each other by any common evolution process. Therefore, it has been concluded that the variable petrological and geochemical compositions of mafic rocks in Tuscany basically result from anomalous and heterogeneous mantle sources (Peccerillo et al. 1987). 

Interaction between magma and wall rocks has been a common process in the Roman Province. However, the core problem is the question of how much this interaction has modified the pristine compositional characteristics of the mafic parent magmas. A particularly important issue is whether the crustal-like geochemical and isotopic signatures of Roman mafic vol-canics can be explained solely by some form of crustal assimilation. These problems have been discussed at length not only for the Roman Province but for all the potassic volcanoes occurring across the Italian peninsula and in the Aeolian arc. 

De Astis et al. (2000) and Calanchi et al. (2002b) noticed that calc-alkaline and HKCA basalts at Vulcano and Panarea have distinct trace element ratios (e.g. La/U, Rb/Zr, Zr/Nb) compared to the associated sho-shonitic and KS mafic volcanics. However, the rocks of the Calabro-Peloritano basement underlying the Aeolian volcanoes show compositions that resemble the calc-alkaline rather than shoshonitic and KS rocks this was interpreted to exclude a derivation of potassic rocks from calc-alkaline parents via crustal assimilation. The same conclusion was drawn by Frez-zotti et al. (2004), who modelled magma contamination processes using melt inclusions entrapped in metamorphic xenoliths as contaminants. 

The problem of crustal contamination is particularly acute for low mg continental flood basalts and smaller volume continental tholeiitic basalts, both of which have low trace-element concentrations (see Sections 3.03.3.2.3 and 3.03.3.3). The issue is less critical for many smaller volume continental rocks, such as kimberlites and alkali basalts, which have much higher abundances of many trace elements. As a result of their high strontium and neodymium content, for example, the isotopic compositions of these elements in kimberlites and alkali basalts are relatively insensitive to modification during crustal contamination. Conversely, the osmium and lead concentration of basaltic magmas are so low that these isotope systems are particularly vulnerable to modification by interaction with cmstal rocks (McBride et al, 2001 Chesley et al, 2002) hence these systems provide relatively sensitive indicators of crustal assimilation. 

Mantle-derived magmas were contaminated by mixing with crustal melts or by assimilating crustal rocks from the walls of the fractures through which they moved toward the surface. 

Fig. 13.47 The 5 0 (SMOW) values of whole-rock samples of the Dufek intrusion vary widely. The 6 0 values of the lower part of the intrusion in the Dufek Massif range in most cases fiom greater than -h5% to less than +7%o which is normal for mantle-derived mafic igneous rocks. However, the 6 0 of the upper part of the intrusion in the Forrestal Range decrease from normal values to low values between 0% and +l%

---