2004 Denver Annual Meeting (November 7–10, 2004)

Paper No. 12
Presentation Time: 1:30 PM-5:30 PM


EATON, J.K.1, VALLEY, J.W.1, VOGEL, T.A.2, PATINO, L.C.2 and ALVARADO, G.E.3, (1)Department of Geology and Geophysics, Univ of Wisconsin-Madison, 1215 West Dayton Street, Madison, WI 53706, (2)Department of Geological Sciences, Michigan State Univ, 206 Natural Science Building, East Lansing, MI 48828-1115, (3)Escuela Centroamericana de Geologia, Universidad de Costa Rica, Apdo. 35, San Jose, jake_eaton@hotmail.com

Central American silicic (>65 wt% SiO2) ignimbrites are volumetrically significant despite the absence of evolved crust. Oxygen isotopes, together with major and trace element data, provide a key constraint to understanding the evolution of these rocks. Oxygen isotope ratios of phenocrysts from Nicaraguan and Costa Rican ignimbrites show an increase in δ18O from north to south of 1.5‰, except for a low δ18O excursion in northern Costa Rica (Guanacaste). Magmatic whole rock δ18O (WR) values (calculated in equilibrium with Cpx, Opx, and Mt) increase from 5.2‰ in Nicaragua to 6.7‰ in central Costa Rica. Whole rock calculations are a function of assumed temperature; a decrease of 100°C in the input temperature would shift δ18O (WR) up by 0.5‰. However, this shift is systematic and does not affect the arc-wide trend. Thus, δ18O (WR) values in Nicaragua are lower than oceanic basalt but increase to more normal values in central Costa Rica. Fractionation of primitive magmas would produce a subtle increase in δ18O vs. SiO2 or a decrease vs. MgO. The silicic ignimbrites display no correlation between major elements and δ18O, suggesting the δ18O is not related by crystal fractionation and may be characteristic of the source rock. Certain incompatible trace element ratios (Ba/La, U/Th, Ba/Nb) used as indicators of fluid flux from the slab, show good, negative correlations with δ18O. Ce/Pb, used as an indicator of contribution from the mantle, displays a positive correlation with δ18O. These trace element trends mimic those seen in basalts and andesites from the active arc, which have been linked to relative input of fluids from the subducting slab (Ba/La, U/Th, Ba/Nb) and input from the mantle (Ce/Pb). In contrast, samples from Guanacaste are low in δ18O and show unexpected correlations, negative vs. SiO2 and positive vs. MgO, but do not correlate to these trace elements.

The along arc increase in δ18O, correlation with incompatible trace elements, and not with major elements are consistent with a model in which silicic magmas are generated by melting of previously emplaced subduction-related, mantle-derived plutons, generated from a metasomatized mantle source. Oxygen isotopes of Guanacaste ignimbrites deviate from this model and suggest that parent magmas assimilated material that interacted with heated surface waters.