Fluid inclusions in upper mantle xenoliths from Northern Patagonia, Argentina: evidence for an upper mantle diapir

Three generations of fluid inclusions can be recognized in upper mantle xenoliths from alkali basalts of the Somoncura Massif, Northern Patagonia, Argentina. The first (''early'', ''primary'') one consists of dense CO2 inclusions which were trapped in the mantle-crust boundary zone (22-36 km minimum trapping depth). Their co-genetic relationship with silicate melt inclusions enables us to constrain their minimum trapping temperature at 1200 degrees C, indicating a high temperature event in a cooler environment. The ''late'' (''pseudosecondary'' and ''secondary'') generations of fluid inclusions were classified in accordance with their homogenization temperature to liquid CO2 (L1) and vapor CO2 (L2) phase. The minimum trapping depth for the first of the late inclusions (L1) is about 16 km. In spite of the uncertainties related to this value, L1 inclusions indicate that the upper mantle rocks, of which samples were delivered by the basalts, had some residence time in the middle crust where they experienced a metasomatic event. The fact that this event did not destroy the earlier inclusions, places severe constraints on its duration. The second late inclusions (L2) are low-pressure CO2 inclusions with a minimum-trapping depth of only 2 km, presumably a shallow magma chamber of the host basalts. The succession of fluid inclusions strongly points toward a fairly fast uprising upper mantle underneath Northern Patagonia. The petrology and mineral chemistry of the peridotitic xenoliths support this view. Extensive partial melting and loss of these melts is indicated by the preponderance of harzburgites in the upper mantle underneath Northern Patagonia, a fairly unusual feature for a continental upper mantle. That depletion event as well as several metasomatic events - including those which left traces of fluid inclusions - are possibly related to a high-speed diapiric uprise of the upper mantle in this area. The path can be traced from the garnet peridotite stability field into the middle crust, a journey which must have been unusually fast. Differences in rock, mineral, and fluid inclusion properties between geographic locations suggest a diffuse and differential type of diapirism. Future studies will hopefully help to map the full extent and the highs and lows of this diapir and elucidate questions related to its origin and future.