(Table 1) Chemical composition and bulk densities of dolerites from ODP Hole 504B lower dikes

Chemical interactions between seawater and the oceanic crust have been widely investigated during recent years. However, most of these studies concern the uppermost volcanic part of the crust. The contribution of the underlying sheeted dike complex to the global budget of the oceans is inferred solely from some ophiolite studies and from the 500-m high-level dike section of DSDP/ODP 504B which was drilled in 1981. Hole 504B is the only place where a continuous and long (1260 m) section in the sheeted dike complex has been cored, and it is now regarded as a reference section for the upper oceanic crust. Many petrological and chemical data from these dolerites are available, including the relative proportions of veins, extensively altered adjacent rocks, and less altered 'host-rocks'. For these three reasons, considering the entire dike section penetrated by Hole 504B is a unique chance to study chemical fluxes related to hydrothermal alteration of this part of the oceanic crust. The calculation of any chemical flux implies knowledge of the chemical composition of the fresh precursor (protolith). Previously, mean compositions of glasses (=P1a) or basalts from the Hole 504B volcanics have been used as protoliths. In this paper, we calculate and discuss the use of various protoliths based on dolerites from Hole 504B. We show that the most adequate and realistic protolith is the mean of individual protoliths that we calculated from the acquisition, by automatic mode, of about 1000 microprobe analyses in each thin-section of dolerite from the Hole 504B lower dikes. Consequently, PFm is further used to calculate chemical fluxes in the dike section of Hole 504B. The chemical compositions of the host-rocks adjacent to alteration halos tend to converge to that of PFm with depth, except for Fe2O3t and TiO2. Because the volume percent of alteration halos increases with depth, the total fluxes related to these halos increase with depth. This explains why the mean flux (host-rocks+halos+veins) of the upper dikes is roughly similar to the mean flux of the lower dikes. During the alteration of the entire Hole 504B dike section, the dolerites gained relatively large quantities of Fe2O3t (+4.0 g/100 cm**3) and released much SiO2 (-6.8 g/100 cm**3), CaO (-5.8 g/100 cm**3), and TiO2 (1.6 g/100 cm**3), and minor Al2O3 (-0.7 g/100 cm**3) and MgO (-0.7 g/100 cm**3). We show the importance of the choice of the protolith in the calculation of chemical budget, particularly for elements showing low flux values. In Hole 504B, the Mg uptake by the volcanics during low temperature alteration added to the Mg release by the dikes gives a net flux of -0.07x10**14 g/year. We propose that part of the Mg uptake by the oceanic crust, which is necessary to compensate the rivers input (-1.33x10**14 g/year), occurs in the underlying gabbros and/or in sections which are altered such as Trinity and Troodos ophiolites. Compared with ophiolites, fluxes calculated for elements other than Mg for the entire crust are generally similar (in tendency, if not in absolute value) to that we obtained from Hole 504B.