Accelerated weathering of silicate rock dusts predicts the slow-release liming in soils depending on rock mineralogy, soil acidity, and test methodology

Abstract

The ongoing acidification of soil poses a significant threat to the proper functioning of various ecosystems worldwide. Silicate rock dusts (RD) are increasingly amended to acid soils to restore their pH, but the acid neutralising capacity (ANC) and dissolution rate of these products are highly variable and lack proper assessment protocols. It is expected that pH-dependent RD ANCs and dissolution rates dictate the pH increase in soils depending on the initial pH and pH buffer power of the soil. This study addressed these questions by comparing and validating three accelerated weathering tests for their capacity to predict the gradual liming effects in a two-year outdoor mesocosm. Five commercial RDs (two basalts, phonolite, foidite, and trachy-andesite) were tested in four acidified forest soils varying in initial pH, in texture and associated pH buffer power. First, RD dissolution was measured in aqueous batch renewal systems during one year at various starting pH (3.5, 4.5, and 5.5) and constant temperatures (20 °C, 37 °C, and 65 °C). These showed that the ANCs of RDs exhibit a fast fraction (half-life < 1 day) followed by a slower fraction. Second, titration tests of RDs at fixed solution pH, i.e., pHstat, performed between pH 3.5 and 4.5 revealed surface normalised dissolution rates that decreased factors 10–100 per unit pH increase, the slope depending on the RD’s mineralogy. Indeed, the rate ranking of observed surface area normalised dissolution rates confirmed that of literature-based values, weighed by the XRD-derived mineralogical fractions. Third, lime-calibrated agitated soil-RD suspension tests were conducted for two months, illustrating the dependency of RD ANC on soil context, and yielding ANCs that were factor 2–5 times larger than in the batch renewal test because of pH buffering effects of soil particulates. Finally, an outdoor soil mesocosm on the four acid soils amended with RDs at 12 Mg/ha and 340 Mg/ha was established and soil pH was monitored for 2 years. The observed pH trends corresponded well for both the low dose (R2a = 0.67) and the high dose (R2a = 0.93) with model predictions made using the pH- and temperature-dependent RD dissolution rates (pHstat test) and the soil pH buffer power and RD slow ANC-fraction (soil-RD suspension test). This model predicts half-lives of RD dissolution ranging between <1 month to >100 years depending on the starting pH of the soil, its pH buffer power, the RD mineralogy (XRD based) and its specific surface area. This study shows that the dissolution and ANC of RD can be most pragmatically predicted with a series of lime-calibrated soil-RD suspension tests of maximally two months.