Mass spectrometry
δ34S and δ18O were analysed using a Pyrocube elemental analyser (Elementar™), linked to an Isoprime 100 continuous flow isotope ratio mass spectrometer at Lancaster University, UK, with instrumental set up followed the protocols established in Wynn et al. (2015). For the finely ground soil and vegetation samples, only34S/32S was measured. For the BaSO4 extracted from drip water, rainfall and carbonate bedrock, both δ34S and δ18O were analysed.
Combustion of samples within tin capsules in the presence of vanadium pentoxide at 1050°C yielded SO2 for determination of δ34S-SO4 and pyrolysis within silver capsules in the presence of carbon black catalyst at 1450°C yielded CO for determination of δ18O-SO4. δ34S isotopic composition values were corrected against VCDT using within run analyses of international standard NBS-127 and SO5 (assuming δ34S isotopic composition of +21.1‰ and +0.5‰, respectively (Halas and Szaran, 2001). δ18O isotopic values were corrected to V-SMOW using NBS-127 and NISTSO6, assuming δ18O values of +9.3 ‰ and -11.4 ‰, respectively. Within-run standard replication (1 standard deviation) was <0.5‰ for both 34S/32S and18O/16O.