Stromatolites are laminated sedimentary structures that provide a controversial candidate signature for early life. In the following discussion, we therefore adopt the non-genetic term 'stromatoloid', thus making no assumptions about their biogenicity (cf. Buick et al., 1981). A stromatoloid can be defined non-genetically as "...an attached, laminated, lithified sedimentary growth structure, accretionary away from a point or limited surface of initiation" (Semikhatov et al., 1979), and in outcrop are macroscopically layered and wrinkled surfaces forming domes, cones and columns. Alternative biogenic interpretations which imply microbial mediation (e.g. Krumbein and Werner, 1983) are not favored by us, because they imply an origin which can be very difficult to establish in the early rock record.
In the Pilbara, nodular and wavy-laminated stromatoloids (Fig. 2g) have been described from a chert unit within the Dresser Formation at North Pole (Walter et al., 1980). Coniform stromatoloids have also been described from the Strelley Pool Chert (Lowe, 1980; Hofmann et al., 1999; Allwood et al., 2006, 2007). In both cases their interpretation as biogenic was initially based largely upon simple macro-morphological comparisons with modern day stromatolites, for example those found at Shark Bay in Western Australia.
Stromatoloids from this area have also been used to argue for the presence of oxidative photosynthesis by -3,500 Ma (e.g., Awramik, 1992; Schopf, 1999) as well as phototrophic behaviour (Hofmann et al., 1999; Allwood et al., 2006). It is important to note, however, that neither convincing microfossils nor wrinkle mat fabrics have ever been found in these stromatoloids. This anomaly is usually excused by the low preservation potential of stromatolitic micro-facies and by the fact that microfossils are only found in an estimated -1% of occurrences in Phanerozoic stromatolites. This line of argument is somewhat unsafe, however, given that the preservation potential of microbial remains in an Archean world lacking oxygen and supersaturated with respect to silica is expected to be much higher than in the Modern world with a highly oxygenated atmosphere, silica undersaturated seawater, and aggressive regenerative recycling of carbon. This lack of microfossil evidence should not therefore be dismissed lightly.
New nano-scale evidence for microbial processes in the form of aragonite nano-crystals intimately associated with organic carbon globules has recently
been found in younger 2.72 Ga stromatolites of the Tumbiana Formation in Western Australia; this has yet to be extensively sort in early Archean stroma-toloids (cf. Lepot et al., 2008).
Several authors have re-examined the Dresser Formation stromatoloids; Buick et al. (1981) attempted to define universal stromatolite biogenicity criteria and concluded that these stromatoloids were only "probable or possible" biogenic stromatolites. Lowe (1994) re-interpreted the Dresser stromatoloids as produced by soft sediment deformation of originally flat layers. Lowe (1994) also directly questioned his original biogenic interpretation of the Strelley Pool stromatoloids, instead concluding that they formed through evaporitic precipitation. Van Kranendonk (2006) showed that the Dresser Formation stromatoloids occur in the vents of barite dykes and suggested that they may have been constructed by hyperthermophilic microbes. The biogenicity of these stromatoloids is questionable, however, because their macro-morphology appears to be largely controlled by the thickness of precipitated barite crusts and draping chert layers. Their distribution most likely reflects the supply of supersaturated solutions from which they precipitated. Robust micro-textural and isotopic evidence for the involvement of any kinds of microbes in the growth of these stromatoloids is still lacking.
The intriguing Trendall locality within the Strelley Pool Chert exhibits coniform stromatoloids that were first described by Hofmann et al. (1999). These stromatoloids are notable for their diverse range of coniform and rare columnar morphologies, their greater variation in size, plus one example of putative 'branching' (Fig. 2d). A biological origin for these structures has been advanced, based largely upon morphological arguments (Allwood et al., 2006, 2007; Van Kranendonk et al., 2003; Hofmann et al., 1999), but also rare earth element data suggestive of a shallow marine setting (Van Kranendonk et al., 2003). In their recent work Allwood et al. (2006, 2007) present a depositional model that is taken to support a shallow water phototrophic origin for the stromatoloids. Our own field work, undertaken across the whole of the outcrop belt, however, leads us to contest this interpretation. In the West Strelley belt, for example, small unbranched coniform stromatoloids are common and these do not show depth controlled changes in morphology or distribution (Wacey et al., 2008b). We also find a close interrelationship between coniform stromatoloids and crystal fan arrays, upon which they can be seen to nucleate. Further, they intergrade with <-
Figure 2. (continued) Chert; MF = McPhee Formation; DF = Dresser Formation; (c) automontage image of putative filamentous microfossils from the 3,235 Ma Sulphur Springs volcanogenic massive sulfide deposit; (d) stromatoloids preserved in carbonate from the 'Trendall Locality' within the -3,400 Ma Strelley Pool Chert; (e) automontage image of microtubular structures from the -3,400 Ma Strelley Pool Sandstone; (f) putative cyanobacterium-like structure Archaeoscillatoriopsis disconformis Holotype from the 3,460 Ma 'Apex chert' (after Schopf 1993) re-imaged by us to show growth as a self organizing structure around a rhombic crystal (inset is an interpretative sketch in the style of Schopf (1993) which omits the lower structure and side branch seen in the main image, adapted from Brasier et al., 2002); (g) stromatoloid preserved in ferruginous carbonate from the -3,500 Ma Dresser Formation.
linguoid, lunate, sinuous and linear current ripples. We here argue that there was a strong chemical component in their accretion, with growth of carbonate crystals influenced by current velocities. In the absence of supporting micro-textural and biomarker evidence, the biogenicity of all of these stromatoloids remains to be demonstrated.
More generally, we caution that macroscopic self organising structures closely resembling stromatoloids are readily generated by abiogenic processes (see Brasier et al., 2006). These processes include diffusion limited aggregation of synthetic colloids in laboratory experiments (McLoughlin et al., 2008a), computer simulations using the Kardar Paris Zhang equation (Grotzinger and Rothman, 1996) and cellular automata (Wolfram, 2002). Given the absence of compelling microbial mat or microfossil remains in early Archean stromatoloids and the possibility that some of these deposits were formed from colloidal silica gel precursors (cf. McLoughlin et al., 2008a), questions remain as to whether, alone, Archean stromatoloids have anything to unambiguously tell us about microbes or early biology. We tend to agree with Schopf (2006), that unfortunately "it is perhaps impossible, 'to prove beyond question that the vast majority of reported stro-matolites...are assuredly biogenic'".
Was this article helpful?