Pb- and Sr-isotopic data for subduction-related Late Cretaceous and Cenozoic granitoids of the coastal region and lowermost Pacific slope of the Western Cordillera of central Peru suggest that these granitoids are derived from an undepleted mantle source (OIB–type or enriched sub-continental). A depleted MORB–type mantle did not have a major role in the genesis of the magmas. The mantle source appears to have changed slightly with time. The data suggest that the asthenospheric wedge was isotopically heterogeneous during the early stages of Coastal batholith emplacement. During this time, the subduction slab geometry changed from steeply to moderately dipping, and the stress regime in the overriding plate changed from extensional to compressional. Subduction-related mantle convection caused a progressive homogenization of the asthenospheric wedge between 100 and 80 Ma. Since then, the isotopic composition of the mantle component has not varied significantly. Those parts of the metasomatized asthenosphere that did not melt (or where trapped magmas did not escape upward) did not participate significantly in the convective cells and were dragged along the slab to greater depth. Evidence for participation of a radiogenic crustal component in the evolution of these Peruvian granitoids is always present. Significant crustal contamination or assimilation cannot be ruled out, since the isotopic contrast between lower- and upper-crustal rocks and mantle-derived magmas is generally low, but contamination by highly radiogenic, in situ, upper-crustal rocks appears to be sporadic in time and space. The radiogenic crustal component is shown to be mainly a component recycled by subduction. We assume that it corresponds to fluids derived from the subducting slab and associated subducted sediments. This crustal component appears to have changed with time. The variations of Pb- and Sr-isotopic compositions are consistent with a two-stage model: (1) During the early stages of Coastal batholith emplacement (100 to 90 Ma) the crustal component was mainly derived from tectonic erosion of the Lower Cretaceous accretionary prism. The sediments in the accretionary prism were derived from the Brazilian Shield and had highly radiogenic Pb- and Sr-isotopic compositions similar to modern sediments of the Barbados ridge. (2) After 84 Ma, the crustal component was Pacific oceanic sediments (and a slab component) with Pb- and Sr-isotopic compositions like those of modern Pacific sediments. These results suggest that participation of recycled oceanic sediments in the genesis of Andean calc-alkaline magmas is the rule, although in many cases the evidence is obscured by interaction of mantle-derived magmas with the continental crust.

A detailed synthesis of the chronology and spatial distribution of magmatic activity along a transect of the Andes of central Peru (around 11°S latitude) and some new K-Ar radiochronological data are presented. This huge data set is compared with independent geophysical data and reconstructions dealing with the interaction dynamics between the oceanic Nazca plate (previously Farallon) and the South American continent. The data strongly suggest that magmatic activity has been discontinuous. All periods of high convergence rate (> 10 cm/yr) between Nazca (Farallon) and South American plates are characterized by important magmatic activity. In contrast, the intervals of magmatic quiescence or low magmatic activity are systematically associated with low convergence rate. Two main exceptions to these rules are: (1) The last period of Coastal Batholith emplacement, between 75 and 59 Ma, seems to be associated with a lower convergence rate (5 to 7 cm/yr) and a very weak compressive deformation. (2) Magmatic activity is absent in central Peru during the last 3 m.y. in a context of high convergence rate. This has been interpreted as a consequence of the lack of an asthenospheric wedge above an abnormally flat slab. The change from a “normal” slab dip (≈30°) to the “flat” slab is associated with the subduction of the Nazca ridge and not with a rejuvenation of the slab. The formation of the Copara-Casma (Aptian-Albian) aborted marginal volcanic basin appears to be contemporaneous with the rifting of the South Atlantic Ocean, an extensional tectonic regime and a steeply dipping slab. The beginning of emplacement of the Coastal Batholith (Albian to Paleocene) is synchronous with the onset of active spreading in the Atlantic Ocean, the change to a compressive regime along the Andean margin, and a decrease of slab dip. Channeling of magmas through deep-seated lithospheric structures along the axis of the Casma basin may have played a role during the emplacement of the Coastal Batholith. From Albian to Mio-Pliocene times, the eastward migration of the trenchward magmatic front did not exceed 50 km. The main change in the geometry of the magmatic belt is its rapid broadening, from a narrow (≈40 km) to a wide (>150 km) belt in the early and mid-Eocene. This change is contemporaneous with the main period of compressive deformation and corresponds to a decrease in the dip of the slab. It is not associated with a rejuvenation of the slab nor with a higher trenchward motion of the overriding plate. Rather, it appears to be related to an increase in the rate of convergence and to a change from a very oblique to a nearly normal convergence with respect to the continental margin. Tectonic erosion at the trench cannot be invoked at least for the last 40 m.y. but probably occurred during the first episodes of emplacement of the Coastal Batholith. No conclusion may be drawn for the period between ca. 75 to ca. 45 Ma.

Cordilleran I-type granitoids with a mantle signature are characteristic products of oceanic-plate subduction at continental margins. I-type granitoids with a crustal signature occur in a variety of tectonic settings, including that of subduction. S-type granites with crustal signatures are characteristic of collisional settings, but also occur in the same range of settings as crustal I-types. The role of the tectonic setting is subordinate to that of the composition of the source region in determining the typology of crustal granites, which is a function of the proportion of mantle-derived to crustal material mobilized during magma genesis. The production of melts with similar proportions of mantle to crustal components is triggered by a variety of tectonic processes in different tectonic settings. Crustal heterogeneity is probably the main factor contributing to the diversity of crustal granites.