Paricutin re-examined: a classic example of crustal assimilation in calc-alkaline magma

Abstract

Following its birth on the 20th of February 1943, the Mexican volcano Paricutin discharged a total of 1.38 km3 of basaltic andesite and andesite before the eruption came to an end in 1952. Until 1947, when 75% of the volume had been erupted, the lavas varied little in chemical or isotopic composition. All were basaltic andesites with 55 to 56% SiO2, δ 18O of +6.9 to 7.0, and 87Sr/86Sr ratios close to 0.7038. Subsequent lavas were hypersthene andesites with silica contents reaching 60%, δ18O values up to +7.6, and 87Sr/86Sr of 0.7040 to 0.7043. The later lavas were enriched in Ba, Rb, Li, and K2O and depleted in MgO, Cu, Zn, Cr, Ni, Sr, and Co. The isotopic and other chemical changes, which appeared abruptly over a few months in 1947, are interpreted as the result of tapping a sharply zoned and density stratified magma chamber. Xenoliths of partially fused felsic basement rocks in the lavas have silica contents greater than 70%, δ 18O of +5.6 to 9.9 and 87Sr/86Sr between 0.7043 and 0.7101. In many respects they resemble samples of basement rocks collected from nearby outcrops. Three analysed samples of the latter have silica contents of 65 to 67%, δ 18O of +7.7 to 8.6, and 87Sr/86Sr between 0.7047 and 0.7056. These new data provide strong support for the original interpretations of Wilcox (1954), who explained the chemical variations by a combination of fractional crystallization and concurrent assimilation of up to 20 weight % continental crust. Except for a few trace elements, particularly Ba, Sr, and Zr, the chemical and isotopic compositions of the xenoliths and basement rocks that crop out nearby match the type of contaminant required to explain the late-stage lavas. Some of the discrepancies may be explained by postulating a contaminant that was older and richer in Ba, Sr, and Zr than those represented by the analysed xenoliths. Others can be attributed to chemical changes accompanying disequilibrium partial melting, contact metamorphism, and meteoric-hydrothermal alteration of the country rock. Many of the xenoliths show evidence of having been affected by such processes. The lavas were erupted from a zoned magma chamber that had differentiated by liquid fractionation prior to the eruption. The order of appearance of the lavas can be explained in terms of withdrawal of stratified liquids of differing densities and viscosities.