Figure 2.6 Secular variations in δ³⁴S for carbonate‐associated sulfate (CAS: blue circles) and evaporitic sulfate (orange circles). Red circles indicate 10 million year average values (0–500 Ma), 50 million‐year average values (500–1000 Ma) and 100 million‐year average values (1100–2600 Ma) calculated from δ³⁴S_CAS.

      (Modified from Crockford et al., 2019.)

Kampschulte and Strauss (2004) published the first CAS‐based sulfur isotope time series of oceanic sulfate for the Phanerozoic. Clear differences were discernible such as high δ³⁴S_sulfate values in the early Paleozoic and minimum δ³⁴S_sulfate values during Permian times. While the overall temporal trend was comparable to observations published earlier and based on evaporite minerals, this new time series of the sulfur isotopic composition of Phanerozoic seawater sulfate showed substantially more internal structure because of its significantly better temporal resolution. Numerous studies on δ³⁴S_sulfate, published in the past 15 years, have provided a detailed view of the secular variations in the sulfur isotopic composition through time; these have most recently been compiled by Crockford et al (2019) extending the record back into the Proterozoic (Figure 2.6). A first‐order interpretation, based on isotope mass balance considerations, views these temporal variations as a reflection of changes in pyrite burial (f_sulfide), with high rates in the early Paleozoic, minimum rates in the late Paleozoic and its transition into the Mesozoic, and an evolution towards modern values. However, additional information can be gathered from a detailed multiple sulfur isotope record through time.

      Based on complimentary records of δ³⁴S_sulfate and Δ³³S_sulfate for Paleozoic and Mesozoic carbonate‐associated sulfate, and considering the temporal record of δ³⁴S_sulfide through time, Wu et al. (2010, 2014) proposed significant changes in the operational mode of the global sulfur cycle through the Phanerozoic. Most notably, these authors highlighted a change in the magnitude of isotopic fractionation between sulfate and sulfide from lower values in the Paleozoic to higher values in the Cenozoic. Moreover, they argue for a change in the isotopic composition of the input function (δ³⁴S_in) to the global sulfur cycle (i.e. the continental weathering signature), as previously discussed by Canfield (2013). Wu et al. (2014) attribute this to changes in the Earth surface sulfur pool, notably a rapid recycling of newly formed sulfate minerals. Discernible shorter‐term fluctuations (on the tens of million years scale) in the temporal records are interpreted as reflecting changes in the intensity of sulfide oxidation during cycling of sulfur and/or by rapid changes in sulfur influx to the oceans and its associated sulfur isotopes. Recently, Crockford et al. (2019) compiled and substantially extended the time series of δ³⁴S_sulfate and Δ³³S_sulfate for seawater sulfate (and δ¹⁸O_sulfate and Δ¹⁷O_sulfate) back in time into the late Archean, with new data mostly derived from gypsum or anhydrite. As in other previous studies, the observed large scale secular variations in the sulfur isotopic composition of seawater sulfate are attributed to temporal changes in burial/weathering of sedimentary sulfide. Moreover, these time series substantiate earlier suggestions (Melezhik et al., 2005) that the operational mode of the sulfur cycle as we know it today and, in particular, the continental weathering and riverine delivery of sulfate to the ocean only commenced in the early Proterozoic, postdating the first significant rise in atmospheric oxygen.

      Two additional aspects related to the study of carbonate‐associated sulfate are noteworthy as they likely reflect on future research. An increasing number of studies have focused on critical time boundaries such as the Permian–Triassic transition (Schobben et al., 2015, 2017), and careful work has resulted in high‐resolution profiles across prominent rock successions allowing for a renewed view on the causes and consequences of short‐term perturbations of global sulfur cycling. The second aspect pertains to an analytical improvement, notably the measurement of sulfur isotopes at very low sulfate concentrations (nanomole level) using multi‐collector induced coupled plasma mass‐spectrometry (MC‐ICP‐MS) (Craddock et al., 2008; Paris et al., 2013). Applications of the latter will allow for an even higher temporal resolution for rock successions across critical time boundaries than presently available.

2.5. PYRITE AND ORGANIC‐BOUND SULFUR AS RECORDERS OF MICROBIAL SULFUR CYCLING IN THE PAST

      Being a key element of life, sulfur is taken up by plants and microorganisms via assimilatory reduction of inorganic sulfate with little sulfur isotope fractionation associated. In contrast, dissimilatory sulfate reduction is an energy‐yielding process associated with the release of hydrogen sulfide. Kaplan and Rittenberg (1964) in their seminal paper on isotopic fractionation associated with dissimilatory sulfate reduction established the foundation for the application of sulfur isotopes as recorder of microbial activities in the geological past. Hartmann and Nielsen (1969) were the first to apply this new understanding that dissimilatory sulfate reduction is generally associated with a substantial isotopic discrimination against the heavy ³⁴S isotope in their study of marine sediments. Subsequently, a negative δ³⁴S value measured for a sedimentary pyrite or at least a sizeable apparent isotopic fractionation between the parental sulfate (preserved in the rock record as evaporite) and resulting sedimentary (iron) sulfide, as observed in natural marine settings as well as in the early laboratory experiments, were considered as evidence for the biogenicity of a sedimentary sulfide: it provided the basis for tracing microbial sulfur cycling through time.

      Since the 1970s, numerous studies explored the antiquity of microbial sulfur cycling by studying sedimentary rocks as far back as 3.8 billion years ago (Monster et al., 1979). Time series of δ³⁴S values for pyrite in sedimentary rocks were presented, among others, by Schidlowski et al. (1983), Hayes et al. (1992), Strauss (1999), and Canfield (2001a). Reviewing sulfur isotope research targeting sedimentary sulfides with the objective of identifying microbial sulfur cycling and trace it through time two milestone discoveries in the past 20 years by Sim et al. (2011) and Pellerin et al. (2019) are most notable.

In 1964, Kaplan and Rittenberg reported that a maximum sulfur isotopic fractionation of 46‰ associated with dissimilatory sulfate reduction, modern, and ancient marine sediments and sedimentary rocks frequently yielded a larger isotopic difference