Incipient chemical weathering at bedrock fracture interfaces in a tropical critical zone system, Puerto Rico

Abstract

Abstract The processes that control chemical weathering of bedrock in the deep critical zone at a mm-scale are still poorly understood, but may produce 100 s of meters of regolith and substantial fluxes of silicate weathering products and thus may be important for modeling long-term, global CO2. Weathering controls are also difficult to ascertain, as laboratory determined dissolution rates tend to be 2–5 orders of magnitude faster than field determined dissolution rates. This study aims to establish (i) the incipient processes that control the chemical weathering of the Bisley bedrock and (ii) why weathering rates calculated for the watershed may differ from laboratory rates (iii) why rates may differ across different scales of measurement. We analyzed mineralogy, elemental chemistry, and porosity in thin sections of rock obtained from drilled boreholes using Scanning Electron Microscopy (SEM) with energy dispersive spectrometry, electron probe microanalysis, and synchrotron-based Micro X-ray Fluorescence (µXRF) and X-ray Absorption Near Edge Structure (XANES). Weathering ages were determined from U-series isotope analysis. Mineral specific dissolution rates were calculated from solid-state mineralogical gradients and weathering ages. Mineralogical and elemental transects across thin sections and SEM images indicate that trace pyrite is the first mineral to dissolve. Micro-XRF mapping at 2 µm resolution revealed sulfate in pore space adjacent to dissolving pyrite, indicating that the incipient reaction is oxidative. The oxidative dissolution of pyrite produces a low pH microenvironment that aids the dissolution of pyroxene and chlorite. The rate-limiting step of weathering advance, and therefore the creation of the critical zone in the Bisley watershed, is pyrite oxidation, despite the low abundance (∼0.5 vol%) of pyrite in the parent rock. The naturally determined dissolution rates presented here either approach, converge with, or in some cases exceed, rates from the literature that have been experimentally determined. The U-series weathering age data on the mm-scale integrates the weathering advance rate over the ∼4.2 ± 0.3 kyrs that the weathering rind took to form. The weathering advance rate calculated at a watershed scale (from stream chemistry data) represents a contemporary weathering advance rate, which compares well with that calculated for the weathering rind, suggesting that the Bisley watershed has been weathering at steady-state for the last ∼4 kyrs.