Evaluating the earliest traces of Archean sub-seafloor life by NanoSIMS

The Paleoarchean sub-seafloor has been proposed as an environment for the emergence of life with titanite microtextures in pillow lavas argued to be the earliest traces of microbial micro-tunneling (Furnes et al. 2004). Here we use a nano-scale ion microprobe (NanoSIMS) to evaluate possible geochemical traces of life in 3.45 Ga pillow lavas of the Barberton Greenstone Belt, South Africa. We investigated both surface and drill core samples from the original "Biomarker" outcrop in the Hooggenoeg Fm. Pillow lava metavolcanic glass contain clusters of segmented microcrystalline titanite filaments, ~4μm across and <200μm in length. Their size, shape and distribution have been directly compared to those found in recent oceanic crust. Thus it has been argued that they are the mineralized remains of tunnels formed by microbes that etched volcanic glass in the Archean sub-seafloor (Furnes et al 2004; Banerjee et al. 2006). Elemental mapping by NanoSIMS was undertaken to investigate reports of enrichments in carbon (possibly also nitrogen) along the margins of the microtextures previously interpreted as decayed cellular remains. We mapped for ¹²C^-, ²⁶CN^-, ³²S^- along with ¹⁶O^-, ²⁸Si^-, ²⁴Mg⁺,²⁷Al⁺, ⁴⁰Ca⁺, ⁴⁸Ti⁺ and ⁵⁶Fe⁺ in chlorite and quartz hosted examples. The ¹²C^- or ²⁶CN^- linings were not found along the margins of the microtextures in neither the original, nor the drill core samples, despite NanoSIMS being a more sensitive and higher-spatial-resolution technique than earlier microprobe X-ray maps. The absence of organic linings in these samples excludes a key line of evidence previously used to support the biogenicity of the microtextures. Sulfur isotopes ³²S and ³⁴S were measured by NanoSIMS on two types of sulfide: i) small sulfides (1-15μm) intimately associated with the microtextures and; ii) larger sulfides (10-60μm) that cross-cut the microtextures and are disseminated near a quartz-carbonate vein. The sulfide inclusions in the microtextures have strongly depleted δ³⁴S_VCDT values of -39.8 to +3.2‰ (n= 32). The magnitude, range and spatial heterogeneity of these δ³⁴S values are consistent with an early microbial origin (McLoughlin et al. 2012). In contrast, sulfides cross-cutting the microtextures related to later veining have positive δ³⁴S of +6.7 to +18.0‰ (n=20). These data can be compared to magmatic sulfides (δ³⁴S = +3±3‰), Archean seawater (δ³⁴S ca. +5‰) and Archean sedimentary sulfides (δ³⁴S = +8 to -23‰). We propose that the Hooggenoeg sulfides probably formed during early fluid-rock-microbe interaction involving sulfate-reducing microbes (c.f. Rouxel et al. 2008). The pillow lavas were then metamorphosed, the glass transformed to a greenschist facies assemblage and titanite growth encapsulated the microbial sulfides. In summary, the extreme sulfur isotope fractionations reported here independently point towards the potential involvement of microbes in the alteration of Archean volcanic glass. In situ sulfur isotope analysis of basalt-hosted sulfides may provide an alternative approach to investigating the existence of an Archean sub-seafloor biosphere that does not require the mineralization of early microbial microborings with organic linings.