干旱区生态环境知识资源中心

Magnesium-isotope time series MC-ICP-MS analyses from a Pleistocene speleothem - collected in a limestone cave in NW Africa (Morocco) - are reported and discussed in a process-oriented context. In addition, high-resolution C, O and Sr-isotope data, and Mg and Sr element abundances were compiled from the same stalagmite. Sub-samples were collected along the stalagmite growth axis and a second data set was drilled perpendicularly within one growth interval (Hendy test). The analytical results show clearly co-variant, systematic and cyclical fluctuations for all proxies collected along the growth axis and - with respect to the analytical error - invariant data within one growth increment. Magnesium-isotope ratios (delta Mg-26) fluctuate between -4.39 parts per thousand +/- 0.02 2 sigma arid -4.17 parts per thousand +/- 0.05 2 sigma and are within the range of published results of speleothem calcite from limestone caves. The difference of 0.22 parts per thousand is significantly beyond the error of the external reproducibility of 0.03 parts per thousand 2 sigma for 626 Mg. Considering the analytical error, neither a purely kinetic nor an equilibrium fractionation process explains the observed isotope pattern. For the time being, it is suggested that two external factors drive the speleothem Mg-isotope cyclicity: (1) climate-driven (arid versus humid) variances in the precipitation rate of a carbonate phase from meteoric water within the karstic system prior to entering the cave system; and (2) changing rates of silicate (aeolian matedal) versus carbonate weathering. Both of these processes fractionate the Mg-isotopic composition of runoff/seepage water. There is evidence that the magnesium-isotope system, applied to spelcothem archives, bears significant information concerning continental climatic variability in arid zones and deserves further research. (C) 2007 Elsevier B.V. All rights reserved.