GEOCHEMISTRY OF STRONTIUM IN LAYERED INTRUSIONS: A PETROGENETIC ASPECT (THE EXAMPLE OF THE IOKO-DOVYREN AND SOME OTHER COMPLEXES)

Abstract

It is commonly believed that layered intrusion sections recorded the process of crystal fractionation in a magma chamber. The primary magma must have some initial Sr inventory uniformly distributed in the liquid. The amount of Sr extracted from the magma by plagioclase, which is the main Sr concentrator, depends on the partition coefficient of Sr between plagioclase crystals and the liquid, which is always much higher than one. Thus, plagioclase crystallization must always deplete the residual liquids and late differentiates in Sr. The gabbro section of the Ioko-Dovyren massif (northern Baikal region) does not show such a tendency. Two levels are distinguished on the basis of Sr content of plagioclase (SrPl), and within them the Sr concentration of rocks is approximated by a linear function of the abundance of plagioclase with constant SrPl (about 250 and 350 ppm, respectively). There is some doubt, therefore, that the section was formed through the crystal fractionation of a single portion of " picritoid" magma. This doubt is strengthened by the data on other objects. The 6000-m-thick section of the Freetown layered complex (Sierra Leone) shows an almost perfect linear correlation between the amount of plagioclase and Sr in the rock; i.e., all the rocks are as if mixtures in varying proportions of mafic minerals that are essentially free of Sr and plagioclase with constant SrPl (650 ppm). The nearly perfect linear trend of the Middle Banded series of the Stillwater complex (USA) corresponds to an SrPl of about 200 ppm. In the 7000-m-thick layered series of the Kiglapait complex (Canada , there is a distinct linear trend within the lower 90% of the section at a constant SrPl of about 450 ppm. The section of the Kivakka layered intrusion (northern Karelia) shows a nearly linear trend corresponding to an SrPl of approximately 400 ppm. In all the cases, Sr distribution does not obey the regularities predicted by the magmatic model. Consequently, a different process is recorded. The fact of the constancy of component concentration in a phase occurring in a varying amount can be adequately explained thermodynamically: it implies the constancy of the chemical potential of the component in the section. The whole population of layers was in mutual chemical equilibrium and composed together a common equipotential system (reservoir), whose parts were formed simultaneously rather than sequentially as follows from the traditional interpretation. It is reasonable to suppose the initial mantle rather than magmatic genesis of at least some layered complexes.