"The Western Mediterranean: a natural laboratory for geochemical processes" by Massimiliano Melchiorre, Institut Ciències de la Terra Jaume Almera (CSIC), will be held at room AULA of the Institut de Ciencies del Mar (ICM - www.icm.csic.es).

The geodynamic evolution of the Western Mediterranean is responsible for several geologic and tectonic processes that, during the last 30 My, led to widespread volcanic activity and metamorphism of crustal rocks, as well as exhumation of ultramafic massifs. The Mediterranean volcanic activity is usually classified as orogenic and anorogenic, characterized by subductive or intraplate geochemical imprints, respectively The geochemical features of major and trace elements, and the isotopic ratios of ~600 samples of both orogenic and anorogenic rocks were merged in a unique database that was processed through a statistical approach. We performed a factor analysis using the Principal Component Analysis method that allowed to reduce the original 36 geochemical parameters expressed as oxides, elements or isotopic ratios to seven factors, accounting for ~84% of variance. Combining these factors in binary diagrams allows to clearly separating the anorogenic and orogenic fields. Anorogenic samples usually fall in a narrow compositional range, while orogenic rocks are characterized by a greater variability and by alignment along different trends. These different trends account for large heterogeneities of the sublithospheric Mediterranean mantle due to extensive recycling of geochemically different lithologies at least since Paleozoic times, thus supporting the existence of different mantle reservoirs as responsible for the Mediterranean volcanism. Furthermore we tried to test the double subduction polarity model recently proposed for the Western Mediterranean from a petrologic point of view. The model results compatible with the volcanic petrology of the last 30 My. Rocks affected by polimetamorphic histories are widespread along the Western Mediterranean. Integrating geochronological analyses of in-situ zircons with the petrological study of their microstructural growing microdomains –through petrography and phase diagram modelling– yields to a more precise interpretation and understanding of any geological process than the study of ages from separate zircons. We therefore performed in-situ U-Pb zircon dating from metapelitic granulites exposed in contact with the ultramafic massif of Beni Bousera (N Morocco, western Mediterranean). Results scatter from Paleoproterozoic (1508 ± 23 Ma) to Miocene (22.9 ± 0.7 Ma), though the majority of ages ranges between 100 and 400 Ma with four main groups clustering around 105, 193, 264-286 and 319-390 Ma. The Middle Permian (~ 286-264 Ma) and earliest Miocene (22.9 ± 0.7 Ma) episodes are constrained from a thermo- barometric point of view: the former represents the zircons’ formation at depth (higher pressure), and the latter the exhumation process (lower pressure). Through the petrological, geophysical and geodynamical studies in the area we propose a four-steps geodynamic model for the Beni Bousera orogenic peridotite in which our dating fit perfectly, since the end of Paleozoic (namely the Middle Permian) to present, supporting an evolution linked to a subduction-related scenario since Late Cretaceous times.