Miocene Slab Detachment in Central Mexico: Causes and Consequences

I propose that the lower part of the Farallon slab subducted beneath Mexico has detached twice during the Miocene. Initiation of slab detachment is a consequence of the incoming of increasingly buoyant oceanic crust at the paleotrench off Baja California and, eventually, the capture by the Pacific plate of the microplates left over by the Farallon plate that produced a retrograde motion of the shallow part of the slab. Our first integration of the geology of the whole central Mexico reveals a number of features that support the detachment model and allow inferring the location through time of these events. A main consequence of detachments is the influx of hotter and geochemically enriched sub-slab material into the slab-free area. The process is thus expected to generate a pulse of volcanism and a geochemically heterogeneous mantle wedge. A first detachment is inferred to have begun shortly before the contact between the Pacific and North America plate at 28.5 Ma. A tear in the slab propagated to the SSE, producing a transient thermal anomaly, ignimbrite flare up, and extension at ~24-20 Ma in the southern Sierra Madre Occidental. A second detachment event is inferred as a result of the end of subduction off the southern half of Baja California at 12.5 Ma. In this case the slab detached from the southern Gulf of California toward the ESE, paralleling the southern Mexico trench system. This event caused an eastward migrating mafic pulse of volcanism, presently observed in the northern part of the Trans-Mexican Volcanic Belt (TMVB), between 11 and 7 Ma. New geochronological and geochemical data suggest the continuation of the detachment to the SSE in Veracruz state (Sierra de Tantima, Alamo, Poza Rica, Palma Sola and Los Tuxtlas volcanic fields). The occurrence of OIB volcanism in the TMVB since the end of Miocene is explained by the influx of enriched asthenospheric material into the gap formed by the detachment, enhanced by the corner flow induced by the continuing subduction. Slowing convergence of the Rivera and Cocos plate with respect to North America after ~10 Ma is a likely consequence of the loss of slab pull following the last detachment event. Our case study suggests that mantle dynamic is driven from above (subducting plate) and that slab detachment may produce a plume-like thermal/magmatic effect in the upper plate.