Destruction of Cratonic Lithosphere: the North China Craton

Cratons, by definition, are regions of prolonged stability (> 1 Ga) within the continents and are characterized by cold, thick lithosphere whose density is offset by its refractory composition, giving rise to chemical buoyancy [the 'tectosphere' hypothesis of Tom Jordan]. The thick, chemically buoyant mantle keel beneath Archean cratons is considered to be mechanically strong, in part because it is cold and thick, but also because it is likely to be anhydrous, and thus contributes significantly to the relative stability of cratons. Indeed, in nearly all cratons the keel has persisted, despite repeated tectonism at the craton margins. For example, the Archean Tanzanian craton has withstood several major orogenies on its margin and is presently surrounded by the East African rift, yet still retains its thick lithospheric keel, as revealed by seismic data and studies of mantle xenoliths. In contrast, the North China craton appears to be a unique example of destruction of an Archean keel. Here, a thick lithospheric keel developed in the Archean and persisted at least through the Paleozoic, as Ordovician kimberlites carry ancient, refractory peridotites and diamonds. In the Mesozoic, the eastern block of the craton became destabilized, as witnessed by the onset of voluminous Mesozoic magmatism. Today, the eastern block is characterized by high heat flow, active seismicity and is underlain by thin and relatively hot mantle lithosphere (sampled as xenoliths in Tertiary basalts) that is chemically indistinguishable from modern convecting mantle. Late Jurassic high Mg andesites, dacites and adakites from the northern portion of the North China craton have unusual compositions that are interpreted to result from melting of ancient, mafic lower crust from the North China craton that had foundered into the mantle. The ages of these rocks therefore date the onset of the destruction of the keel. The processes by which the keel was destroyed remain enigmatic. Continent-continent collision on the southern and northern margins, subduction and slab roll-back on the eastern margin, have all been implicated.