Uranium-series investigations of magma-genesis and differentiation at island arc and continental settings

I present combined U-series, Sr-Nd-Pb isotope, and major/trace element data to constrain mass transfer processes and the timescales over which they occur beneath the Lesser Antilles and Philippine arc systems and in continental basalts.; Isotopic and element data from the Bataan and Bicol arcs, Philippines, suggest the fraction of sediment added to the mantle is limited, despite differences in sediment supply. It follows that most sediments are accreted, reside in the subarc lithosphere, or are recycled into the mantle. U-Th isotope data from the same rock suites can be explained by 230Th ingrowth models where the tectonic convergence rate controlls the extent of ( 230Th/238U) disequilibria. Thus in many cases 230Th disequilibria may not have chronologic significance and instead reflect the competing effects of U enrichment and partial melting.; The same spectrum of data from the Lesser Antilles island arc indicates ∼0.2--1.8% bulk sediment and 0.8--2.4% fluid contribution in the northern and central segments of the arc respectively. Combined with the U-Th isotope data this allows ∼100--140 ka since U enrichment of the mantle wedge.; Differentiation timescales of Taal and Mayon lavas were constrained using combined U-Th-Ra-Pa disequilibria. Ra and Pa disequilibria from Taal volcano are not "concordant", and suggest Ra data reflect open system behavior. Thus based on Pa data, differentiation from basalt to dacite is constrained at ∼30 ka. In contrast 231Pa disequilibria from Mayon volcano are constant within analytical uncertainties, and variations in 226Ra disequilibria vs. Th suggest a much shorter differentiation timescale of ∼1000--2200 yrs across the basaltic andesite compositional window.; U-Th-Ra-Pa data from three continental basalts are used to demonstrate overprinting of mantle produced 226Ra disequilibria. 231Pa disequilibria correlate with the degree of melting. However 226Ra disequilibria are constant, regardless if the lava is tholeiitic or alkali basalt. This observation could be explained by residual phlogopite/amphibole or differences in crustal storage times, yet there is no consistent evidence for either of these scenarios. The constant (226Ra/ 230Th) can be explained by crystal-melt diffusive interaction. Thus 226Ra disequilibria from melting may be overprinted, even in the simplest scenarios of melt generation and transport.