Sulfur isotope values in the sulfidic Frasassi cave system, central Italy: A case study of a chemolithotrophic S-based ecosystem

Abstract

Sulfide oxidation forms a critical step in the global sulfur cycle, although this process is notoriously difficult to constrain due to the multiple pathways and highly reactive intermediates involved. Multiple sulfur isotopes (δ³⁴S and Δ³³S) can provide a powerful tool for unravelling sulfur cycling processes in modern (and ancient) environments, although they have had limited application to systems with well-resolved oxidative S cycling. In this study, we report the major (δ³⁴S) and minor (Δ³³S) isotope values of sulfur compounds in streams and sediments from the sulfidic Frasassi cave system, Marche Region, Italy. These microaerophilic cave streams host prominent white biofilms dominated by chemolithotrophic organisms that oxidize sulfide to S⁰, allowing us to estimate S isotope fractionations associated with in situ sulfide oxidation and to evaluate any resulting isotope biosignatures. Our results demonstrate that chemolithotrophic sulfide oxidation produces ³⁴S enrichments in the S⁰ products that are larger than those previously measured in laboratory experiments, with ³⁴ε_S0–H2S of up to 8‰ calculated. These small reverse isotope effects are similar to those produced during phototrophic sulfide oxidation (⩽7‰), but distinct from the small normal isotope effects previously calculated for abiotic oxidation of sulfide with O₂ (∼−5‰). An inverse correlation between the magnitude of ³⁴ε_S0–H2S effects and sulfide availability, along with substantial differences in Δ³³S, both support complex sulfide oxidation pathways and intracellular recycling of S intermediates by organisms inhabiting the biofilms. At the ecosystem level, we calculate fractionations of less than 40‰ between sulfide and sulfate in the water column and in the sediments. These fractionations are smaller than those typically calculated for systems dominated by sulfate reduction (>50‰), and contrast with the commonly held assumption that oxidative recycling of sulfide generally increases overall fractionations. The relatively small fractionations appear to be related to the sequestration of S⁰ in the biofilms (either intra- or extra-cellularly), which removes this intermediate substrate from fractionation by further disproportionation or oxidation reactions. In addition, the net ³³λ_H2S–SO4 values calculated in this system are larger than data published for systems dominated by reductive sulfur cycling, partially due to the isotopic imprint of chemolithotrophic sulfide oxidation on the aqueous sulfide pool. These distinct isotopic relationships are retained in the sedimentary sulfur pool, suggesting that trends in ³⁴S and ³³S values could provide an isotopic fingerprint of such chemolithotrophic ecosystems in modern and ancient environments.

Introduction

Sulfur (S) plays an important role in global biogeochemical cycling on Earth. It is nearly ubiquitous in natural systems, where it can exist in multiple redox states (S²⁻ to S⁶⁺) and participate in numerous geochemical and biochemical processes. Sulfate (SO₄²⁻) is the most abundant soluble form of sulfur in modern aqueous systems, primarily sourced via fluvial runoff from land. This sulfate provides an important substrate for anaerobic respiration in marine sediments, with dissimilatory sulfate reduction (DSR) accounting for up to 50% of total carbon remineralization in marine sediments (Jørgensen, 1982). Hydrogen sulfide (H₂S) is a product of DSR, and can accumulate to appreciable (mM) concentrations in anoxic waters and sediments. In modern marine settings, the majority of H₂S is recycled within the sediment, and oxidized back to sulfate. The remainder ∼10–20% is buried as pyrite and other iron sulfides (Jørgensen, 1990, Canfield and Teske, 1996).

Oxidative recycling of sulfide is governed by a complex series of heterogeneous biological and abiotic pathways. Sulfide oxidizes abiotically by reaction with Fe(III), Mn(IV), or by rapid reaction with molecular oxygen. Sulfide oxidation is also an important energy-yielding metabolism in a range of prokaryotic organisms that are diverse and widespread in natural ecosystems. These organisms include photoautotrophs that use reduced sulfur compounds as electron donors for anoxygenic photosynthesis, and chemolithotrophs that can oxidize reduced sulfur aerobically with O₂ or anaerobically with NO₃⁻. Notably, chemolithotrophic S-oxidizing organisms play an important role in anoxic marine sediments where chemical oxidants are either absent or present at very low concentrations (e.g., Bruchert et al., 2003, Pellerin et al., 2015). In addition to sulfide oxidation, these organisms can perform a wide array of oxidation reactions that involve highly reactive intermediate sulfur compounds, including thiosulfate (S₂O₃²⁻), sulfite (SO₃²⁻), and elemental sulfur (S⁰). Thiosulfate has been implicated as an important product of sulfide oxidation, although it is quickly recycled in anoxic marine sediments (Jørgensen, 1990). Of the intermediate compounds, only S⁰ builds up to appreciable concentrations in most natural environments (Troelsen and Jørgensen, 1982).

Studies of the stable isotope ratios of sulfur compounds (³²S and ³⁴S) have long played an important role in constraining complex biogeochemical sulfur cycling in modern environments (e.g., Kaplan et al., 1963, Habicht and Canfield, 2001). By proxy, the distribution of sulfur isotopes preserved in geologic materials, as pyrite and sulfate evaporites or carbonate associated sulfate, can reveal information about how the sulfur cycle was operating on the early Earth. The record of δ³⁴S preserved in ancient marine sediments has been used to constrain the early history of Earth surface oxidation and to infer the evolution of various sulfur metabolisms on Earth (Schidlowski et al., 1983, Canfield and Teske, 1996). More recently, the inclusion of minor sulfur isotopes, notably ³³S, in such studies has proven a valuable tool in unravelling complex biogeochemical sulfur cycling in both modern and ancient systems (Johnston et al., 2005b, Canfield et al., 2010, Li et al., 2010, Zerkle et al., 2010, Kamyshny et al., 2011).

Interpretation of sulfur isotope signatures in the environment and in the rock record is based on decades of research into the magnitude and controls on sulfur isotope fractionation by pure cultures and mixed populations of sulfur cycling organisms in laboratory experiments. DSR has received the most attention, due to its importance in the marine S cycle and its dominance of the resulting isotopic signatures (e.g., Kaplan and Rittenberg, 1964, Canfield, 2001, Habicht and Canfield, 2001). DSR generally produces sulfides depleted in ³⁴S by more than 40‰, although the fractionations vary with the organism, the sulfate concentration, temperature, and electron donor availability (Harrison and Thode, 1958, Kaplan and Rittenberg, 1964, Detmers et al., 2001, Habicht et al., 2002, Bruchert, 2004, Canfield et al., 2006, Sim et al., 2011b, Bradley et al., 2015). Biological disproportionation of S intermediates can also produce large isotope effects, with product H₂S generally depleted in ³⁴S by 5–7‰, and product SO₄²⁻ generally enriched in ³⁴S by 17–21‰ (Canfield and Thamdrup, 1994).

Studies of sulfur isotope fractionation during oxidation reactions are limited, presumably due to the complexity of the different reaction processes and the high reactivity of the product intermediates. Experimental studies suggest that abiotic oxidation of sulfide with molecular oxygen can enrich the reactant sulfide in ³⁴S by up to 5‰ (Fry et al., 1986), while phototrophic sulfide oxidation can cause ³⁴S depletions of up to ∼4‰ for δ³⁴S, with small changes in Δ³³S signatures (Zerkle et al., 2009, and references therein). Measurements of S isotope fractionations produced by chemolithotrophic organisms utilizing oxygen or nitrate to oxidize reduced S are particularly limited, presumably because many of the environmentally-relevant S-oxidizing organisms are difficult to cultivate, and only a few strains have been successfully isolated. Previous laboratory experiments with S-oxidizers have yielded inconsistent results, with fractionations varying from −6‰ to +5‰ (for ³⁴ε, as defined in Section 2.4), depending on the substrate oxidized and the growth stage in batch cultures (Table 1).

Fractionations in δ³⁴S between sulfide and sulfate in modern environments and in ancient sediments are often greater than what is typically expressed by DSR alone, from 50‰ to 70‰ (Fry et al., 1991, Canfield and Teske, 1996, Canfield, 2001, Neretin et al., 2003). Large fractionations of up to 70‰ have only recently been measured in incubations with natural populations and pure cultures of sulfate reducers (Canfield et al., 2010, Sim et al., 2011a). These large fractionations during DSR could single-handedly explain the S isotope values in some natural systems (Wortmann et al., 2001, Canfield et al., 2010, Li et al., 2010). In other systems, large fractionations seem to require sulfate reduction followed by the recycling of sulfide by oxidation and disproportionation reactions (Canfield and Thamdrup, 1994, Canfield and Teske, 1996, Habicht et al., 1998, Zerkle et al., 2010). Despite the relatively smaller fractionations inferred for sulfide oxidation processes in comparison, models suggest these processes can have important consequences for the overall isotopic signatures preserved in natural systems, especially for Δ³³S (Zerkle et al., 2009). Additionally, recognizing signatures of S oxidation processes in the rock record is important for testing hypotheses concerning the advent of oxidative sulfur cycling and the evolution of Earth surface redox (Johnston et al., 2005b, Bailey et al., 2013, Lepland et al., 2014).

In this study, we investigate the major and minor sulfur isotope values (δ³⁴S and Δ³³S) of sulfur compounds associated with streams, biofilms, and sediments in the sulfidic Frasassi cave system of central Italy. Visible white biofilms in the Frasassi cave streams are up to 5 mm thick, and span the sharp redox interface that occurs within fast-moving cave streams or at the sediment–water interface of more stagnant waters (Macalady et al., 2008, Jones et al., 2015). These biofilms are overwhelmingly dominated by sulfide-oxidizing organisms (⩾90% filamentous Gamma- and Epsilonproteobacteria) that harness the chemical energy of sulfide and oxygen from the cave waters to grow chemolithotrophically (Macalady et al., 2008, Hamilton et al., 2014). Furthermore, a wide range of physicochemical conditions (e.g., temperature, H₂S and O₂ concentrations) exist within the cave streams due to complex hydrology and varying degrees of meteoric dilution of the persistently sulfidic aquifer. This system therefore provides an ideal natural laboratory to estimate the S isotope fractionations associated with in situ sulfide oxidation by chemolithotrophs, and to examine how these fractionations vary across a range of environmental parameters. In addition, this is the first study of multiple sulfur isotope values associated with S species in a natural sulfidic system characterized by a complete absence of light, and as such will provide valuable insight into sulfur cycling in aphotic ecosystems.