Reply to Comment on “Steady-state 226Ra/ 230Th disequilibrium in mantle minerals: implications for melt transport rates in island arcs” by R. George, M. Reagan, S. Turner, J. Gill, B. Bourdon

Abstract

We agree with George et al. that high 226Ra/ 230Th and correlated Ba/Th, 238U/ 230Th, and other elemental and isotopic ratios are the result of fluid addition to the wedge. The Feineman and DePaolo [Earth Planet. Sci. Lett. 215 (2003) 339–355] manuscript addresses more specifically the pathway followed by the fluid en route to the surface. In particular, we address whether the 226Ra/ 230Th data directly yield the total time available for the fluid to migrate from the point of origin in the slab to the melting region in the wedge, and ultimately to the surface with the melt. All of the processes involved in generating arc volcanism would of necessity have to take place very rapidly within the constraints proposed by George et al. The alternative is that the upward fluid movement is hindered by the fluid flow regime and solid mantle flow, allowing more time for reaction with the mantle wedge. There is evidence in the U-series data for more than one time scale, and our model helps to explain how this could come about. The extent to which grainscale 226Ra/ 230Th disequilibrium affects the inferred melt transport time is yet another issue, and this could be achieved either with the small-degree melt mechanism we discussed or by additional effects associated with melt and fluid migration at later stages in the melt generation process. We agree that the final melt transport event is likely to happen quickly. One of the most puzzling features of volcanic arcs is that the volcanic front emerges considerably behind the presumed location of primary water release from the subducting slab (depth to slab is ∼120 km beneath the volcanic front, as opposed to the predicted ∼80 km). This offset between the expected site of fluid release and the location of the volcanic front can be attributed to coupled transport of the fluid—in the form of hydrous minerals, such as amphibole and/or phlogopite in the wedge [J. Geophys. Res. 97 (1992) 2037–2070] or redirection of the melt due to flow and stresses in the mantle matrix as the magma slowly percolates upward [J. Geol. 95 (1987) 285–307]. In either case, the fluid and/or melt is required to ascend slowly, at least for some part of its journey, to explain the location of the volcanic front. Furthermore, 238U/ 230Th disequilibria in arc lavas seem to suggest a relatively long interval for fluid addition (∼30–60 ka) [Science 292 (2001) 1363–1366; Earth Planet. Sci. Lett. 179 (2000) 581–593; Geochim. Cosmochim. Acta 61 (1997) 4855–4884], which is difficult to reconcile with the 226Ra/ 230Th disequilibrium observed in the lavas. For these reasons, we sought to find a means whereby apparently short-lived isotopic disequilibria could be maintained during complex fluid and melt migration in the mantle wedge. Below, we address a few of the main issues presented by George et al. [Earth. Planet. Sci. Lett. this issue] in their comment.