Evidence in Cerro Pampa Volcanic Rocks for Slab-Melting Prior to Ridge-Trench Collision in Southern South America

• Published in: The Journal of geology 1993-11, Vol.101 (6), p.703-714
• Main Authors: Kay, S. Mahlburg, Ramos, V. A., Marquez, M.
• Format: Article
• Language: English
• Subjects: Basalt; Chemistry; Contaminants; Geology; Lava; Magma; Melting; Pampas; Plate tectonics; Plateaus; Rocks; Tectonics; Volcanic rocks
• ISSN: 0022-1376; 1537-5269
• DOI: 10.1086/648269

Late Miocene ($\approx 12 Ma$) hornblende-bearing andesitic to dacitic ($63 to 68% SiO_{2}; 1.2 to 1.9% K_{2}O$) "adakite" flows at the small Cerro Pampa center in Patagonia (47°55'S, 71°25') have some of the clearest slab-melt geochemical signatures yet seen in a Phanerozoic center on continental crust. These magmas formed in response to melting of the hot, thin slab that was subducting beneath South America prior to the collision of the Chile rise at$\approx 6 Ma$or at$\approx 10 Ma$. Their N-MORB-like$^{87}Sr/^{86}Sr$(0.70285-0.70309) and low$^{206}Pb/^{204}Pb$(18.44-18.59) ratios show that they could have been generated by$\approx 3-5\%$partial melting of eclogite facies N-MORB oceanic crust. Low FeO/MgO (0.9-1.3) ratios and high Cr (>85 ppm) and Ni (>43 ppm) concentrations indicate some interaction with mantle peridotite. Low$\epsilon Nd$(+ 6.9 to + 5.5), high$^{207}Pb/^{204}Pb$(15.57-15.58) ratios, and high Ba, Cs, U, and Th concentrations compared to N-MORB modeled melts indicate some upper crustal contamination. In comparison with previously proposed Patagonian slab-melts, Cerro Pampa magmas require less mantle contamination than those at Cook volcano (54°S) and less crustal contamination than those in the northern Austral Volcanic Zone (49°S to 52°S). These differences fit a ridge-trench collisional (slab-window) model that explains the properties of slab-melts formed before (Cerro Pampa) and after (Austral Volcanic Zone) ridge collision.