Tracing the Sources of Cave Sulfates: A Unique Case from Cerna Valley, Romania

Document Type


Publication Date



Sulfur isotopes, Cave sulfates, Thermo-mineral springs, Sulfuric acid speleogenesis

Digital Object Identifier (DOI)



In order to reliably distinguish between different genetic processes of cave sulfate formation and to quantify the role of thermo-mineral waters on mineral deposition and cave morphology, it is critical to understand sulfur (S) sources and S transformations during hydrological and speleogenetic processes. Previous work has shown that sulfuric acid speleogenesis (SAS) often produces sulfate deposits with 34S-depleted isotopic signatures compared to those of the original source of S in sulfate rocks. However, 34S-depleted isotopic composition of S-bearing minerals alone does not provide enough information to clearly distinguish SAS from other speleogenetic processes driven by carbonic acid, geothermal heat, or other processes. The isotopic composition (δ18O and δ34S) of sulfate minerals (mainly gypsum) from seven caves of the Cerna Valley (Romania) defines three distinct populations, and demonstrates that the δ34S values of SAS-precipitated cave sulfates depend not only on the source of the S, but also on the H2S:SO42− ratio during aqueous S species reactions and mineral precipitation. Population 1 includes sulfates that are characterized by relatively low δ34S values (− 19.4 to − 27.9‰) with δ18O values between 0.2 and 4.3‰ that are consistent with oxidation of dissolved sulfide produced during methane-limited thermochemical sulfate reduction (TSR) that presently characterizes the chemistry of springs in the upper Cerna Valley. Population 2 of cave sulfates has 34S-enriched δ34S values (14.3 to 19.4‰) and more 18O-depleted δ18O values (from − 1.8 to − 10.0‰). These values argue for oxidation of dissolved sulfide produced during sulfate-limited TSR that presently characterizes the chemistry of springs further downstream in the Cerna Valley. The δ18O values of cave sulfates from Population 1 are consistent with oxidation under more oxic aqueous conditions than those of Population 2. δ34S values of cave sulfates within Population 3 (δ34S: 5.8 to 6.5‰) may be consistent with several scenarios (i.e., pyrite oxidation, oxidation of dissolved sulfide produced during methane-limited TSR coupled with O2-limited oxidation during SAS). However, comparatively 18O-enriched δ18OSO4 values (11.9 to 13.9‰) suggest the majority of this sulfate O was derived from atmospheric O2 in gas-phase oxidation prior to hydration. Thus, the combined use of oxygen- and sulfur-isotope systematics of sulfate minerals precipitated in a variety of cave settings along Cerna Valley may serve as an example of how more complex cave systems can be deconvoluted to allow for more complete recognition of the range of processes and parameters that may be involved in SAS.

Citation / Publisher Attribution

Chemical Geology, issues 3-4, p. 105-114

Was this content written or created while at USF?