Role of hydraulic diffusivity in the decrease of weathering rates over time

•Assessing plagioclase weathering rates at the spring watershed scale.•Considering rock structure in the assessment of the reacting surface area.•Bridging solid-state and solute-flux rates through change in hydraulic diffusivity.

SummarySprings emerging within massifs of crystalline rocks were monitored for discharge rate (Q), and the Q values combined with geomorphic and hydrographic parameters in a hydrologic model to calculate hydraulic conductivity (K) and effective porosity (ne) of the spring watersheds. The spring waters, several borehole waters and rain water were analyzed for major dissolved compounds, strontium and isotopes (δ18O, δ2H, δ13C and 87Sr/86Sr). With a shift to less negative values, δ18O and δ2H were fitted by a line approximately parallel to the GMWL, but no significant dependence on altitude was found. The δ18O and δ2H values correlate better with those of precipitation amount. The 87Sr/86Sr ratios in drilled well waters correlate positively with the depth of water circulation reported in the borehole logs. The corresponding regression equations were used to extrapolate the depth of hydraulic circuits within the spring watersheds. The previous data, together with groundwater travel times calculated by a water balance model, and with reactions of granite/metassediment plagioclase and biotite precipitating halloysite, gibbsite and vermiculite, were assembled in a mass balance model to calculate solute-flux weathering rates of plagioclase (WPl). The WPl's were described as a function of D∝K/ne, where D is the hydraulic diffusivity. The discrepancies between the WPl values and solid-state rates, based on the differences between elemental, isotopic and mineral compositions measured in present-day regoliths and in the assumed protolith, were assigned to a decrease in D over time, from values in the protolith to values in the weathered aquifer.