REE mobility due to alteration of Indian Ocean basalt: Reply

Abstract

The suggestion by Church that Ti would be more immobile than Lu and therefore a better normalizing element raises an interesting question. Are relative elemental mobilities the same for ocean-floor alteration as they are for continental weathering? Certainly the studies of elemental mobility by Nesbitt (1979) and Chesworth et ul. (1981) indicate that Ti, as well as Fe and A1 or Zr, is immobile during continental weathering. However, the chemical conditions for continental weathering can be quite different from those for ocean-floor weathering. Even different studies of ocean-floor weathering have found that different elements are immobile. A study of chemical trends in oceanic basalt with distance from spreading ridges by Hart (1970) found that Ti was added and A1 was immobile. Ludden and Thompson (1979) examined oceanic basalts from a dredging transect and concluded that the HREE were immobile and used Yb to normalize their chemical data. Staudigel and Hart (1983) summarized previous data and presented new data on the ocean-floor alteration of basaltic glass to palagonite that indicate Ti and Fe, but not Al, were immobile and that Ti is the best element for normalization. However, the relative mobilities of various elements during alteration might be expected to be somewhat different for crystalline basalt than for basaltic glass, since crystalline basalt contains minerals that do not always alter simultaneously. In order to be mobile an element must first be released from the mineral or minerals in which it occurs. The sequence of elemental release in site 249 basalts is dependent on the order in which the minerals alter. Both LREE-rich plagioclase and HREE-rich augite appear to be relatively fresh in least-altered basalt 33,CC. The opaques, determined by Erlank and Reid (1974) to consist of magnetite, ilmenite, and pyrite in least-altered sample 33 ,CC , appear ragged in transmitted light and might be releasing TiOz. The incompatible-element-rich groundmass is badly altered and replaced largely with secondary minerals. Reflected light study would be necessary to determine whether or not the ilmenite is altering and therefore releasing TiOz. There is, however, no relationship between the elements being released and the mobility of the elements once released. In conclusion, there is no way to be certain which element is most immobile in site 249 basalts, since no fresh basalt was obtained for comparison with the altered basalts. In addition, all of the samples have abundant secondary mineral growth in the matrix and sometimes in vesicles, and these secondary minerals will preferentially capture some elements relative to others. Regardless of whether the normalization is with Lu or Ti, the LREE are greatly enriched, the LREEIHREE ratio is considerably modified, KzO is gained, and MgO, CaO, and MnQ are lost. If Lu is immobile, the LREE and K,O gains are less pronounced. If Ti is immobile, so are Fe, Al, Si, and Na in these rocks.