MASSIVE SILICIC VOLCANISM OF THE WESTERN TRANS-MEXICAN VOLCANIC BELT: INFERENCE ON THEIR GENESIS FROM GEOCHEMICAL AND ISOTOPE DATA

In the western Trans-Mexican Volcanic Belt (WTMVB) a large amount of silicic rocks (~1,300 km3) were emplaced since 7.5 Ma in the rear half of the arc. We present fission tracks and Ar/Ar ages, geochemical, and isotope data and discuss the significance of this unusual occurrence of silicic volcanism in a volcanic arc. The first rhyolitic complexes and associated pyroclastic deposits were emplaced between 7.5 and 5 Ma, after a major pulse of mafic volcanism at 11-8 Ma. These rhyolites have an aggregate volume of about 370 km3 and they virtually represent the only volcanic products of the WTMVB during this period. Since 5 Ma rhyolites coexisted with alkaline to transitional basalts, and represents about 40% of the total volume of volcanism. Basalts are mostly emplaced along WNW-ESE trending normal faults and their eruption mostly coincide with the beginning of two extensional pulses. In the Guadalajara area this bimodal volcanism starts with the eruption of an ignimbrite with evidence of mixing between mafic and silicic magmas, suggesting that the arrival of mafic magma in the upper crust during extensional phase triggered these pyroclastic eruptions. Rhyolites of the 7.5-5 Ma episode are characterized by high, although variable, LILE/HFSE values as well as by spider diagrams with negative spikes at Nb, P and Ti. Incompatible trace element contents show a large range which encompasses that shown by the 11-8 Ma basalts. Indeed, rhyolites are characterized by the same Rb/Sr, Ba/Nb and K/Rb values as 11-8 Ma basalts. Available geochemical data do not support the genesis of the rhyolites by simple fractional crystallization neither by crust assimilation coupled with fractional crystallization processes from basaltic magmas. On the other hand, direct or indirect slab participation via slab melting or fluid transport to the mantle, does not seem to completely justify the chemical characteristic shown by our rocks. We favor a model in which silicic volcanism was the result of ponding of mantle melts in the lower crust, water release, and melting of the crust induced by a strong decrease in magma production rate (and crustal temperature). The latter was a consequence of the decreasing in the Rivera-North America convergence rate between 8 and 5 Ma. New Sr and Nd isotope data performed on these rocks will better constrain this hypothesis.