The fossil Decastronema kotori is a mineral replication of the exopolymer product of an ancient cyanobacterium. This fact raises a question regarding the nature of the process and the timing of sheath calcification and replacement.
An impressive fossil record shows that the potential for preservation of uncalcified cyanobacterial sheaths and envelopes is greater then that of the cells that produced them. The prokaryotic cells, enclosed during lifetime by an osmotically responsive plasma membrane, confined by cross-linked peptidoglycan walls, tend to collapse, shrivel and deform after death, while their sheaths contract in an orderly manner without change in shape as they dehydrate (Golubic, 1980). Thus, Proterozoic cyanobacteria, embedded and preserved in silica, consist mainly of extracellular envelopes and sheaths (Knoll & Golubic, 1979; Schopf & Klein, 1992; Seong-Joo & Golubic, 1998; Golubic & Seong-Joo, 1999), that sometimes retain pigmentation that may have screened UV-rays (Golubic & Hofmann, 1976). This pigmentation may have been a significant protective agent during Proterozoic and Archaean times (Garcia-Pichel & Castenholz, 1991; Garcia-Pichel, 1998). However, when embedded in a carbonate, Proterozoic microbial fossils are rarely preserved in as much detail (see Knoll & Semikhatov, 1998) and suffer more commonly from diagenetic recrystallization that obliterates the original structure (Hofmann, 1976).
In contrast, most Phanerozoic fossils exist as skeletal body parts preserved in carbonates. Well preserved non-mineralized cyanobacterial fossils embedded in silica are exceptionally rare (e.g. Taylor et alii, 1997; Chacón-Baca et alii, 2002; Kerp, 2002). Therefore, it is probable that the overall change in the fossil record from Proterozoic to Phanerozoic was to some extent influenced by a concurrent change in the predominant sedimentary environment.
Protists, animals and calcareous algae produce elaborate skeletal morphologies through enzymatically controlled intracellular or intercellular processes. Such precise cell-control systems are unknown in prokaryotes. Nevertheless, calcification can be promoted or inhibited by prokaryotic metabolic activities and products (Chafetz & Buchinski, 1992; Défarge & Trichet, 1995). These biogenic influences can modify the mineralogy, size and arrangements of the precipitates, resulting in species-specific patterns (Golubic & Campbell, 1981; Couté, 1982; Obenlüneschloss, 1991; Freytet & Verrecchia, 1998). Among the possible mechanisms for the preservation of Decastronema by calcification are the following:
- The sheaths were calcified during the life of the organism, along the lines of the model described for Scytonema in the waters of the Everglades, Florida (Merz, 1992; Merz-Preiss, 2000). Some , mainly freshwater and subaerial species, calcify (e.g. Scytonema julianum), while others (e.g. S. myochrous) do not (Fig. 7), even when growing together under the same conditions (Golubic et alii, 2000). De Castro selected as his model the modern S. myochrous, which was found calcified under supersaturated slightly brackish conditions on Andros Island (Monty, 1967). Such calcification is often localized on segments of filaments and so may contribute to their fragmentation. Although this type of primary calcification may have occurred in Decastronema, no evidence for it has been retained.
- Decastronema may have become calcified post mortem in conjunction with bacterial sheath degradation, as discussed by Arp et alii (1999), with the possible involvement of sulfate reduction and micrite replacement of EPS (Visscher et alii, 2000; Paerl et alii, 2001). However, in Decastronema the accuracy and precision with which the micritic grains outline the layers of EPS is markedly different from the clumpy, irregularly distributed micritic peloids associated with bacterial decay as illustrated by Sprachta et alii (2001) and Dupraz et alii (2004).
- In Decastronema the close relationship between the patterns of distribution of micritic grains and the structural properties of the sheath is in general agreement with the concept of 'organomineralization' sensu Trichet & Défarge (1995), in which an organic product guides the precipitate at the stage of crystal nucleation, while the preceding chemical induction of precipitation may be either abiotic or biogenic (Gautret et alii, 2004). Briggs (2003) discusses similar arrangements of uniform crystal size in diagenetic permineralization and preservation of soft-bodied tissues as analogous to the pixel resolution of computer images. This analogy is applicable to the wall texture of Decastronema. However, the permineralizations described by Briggs took place in the presence of high concentration of decaying organic matter. That environment makes it difficult to separate the effects of bacterial activities (induction) from those of bacterial products (templating). Morse & Wang (1996) offer theoretical models of such processes and suggest that the formation of distinct fields of uniform-sized crystals may be either flux- or substrate-controlled. Both types of control may apply for the mineralization of cyanobacterial sheaths. Braissant et alii (2003) showed that varying concentrations of polysaccharide and amino acids to form precipitates either with or without bacteria produced an array of discrete forms of crystals. Morover, control of crystal size has also been achieved with artificial polymers (D'Souza et alii, 1999).
We conclude that in Decastronema, templating on the polysaccharide matrix of sheaths is the mechanism most likely to have determined the sites of crystal nucleation, for the patterns of calcification correlate closely with the sheath architecture. We suggest that calcification was a part of post-burial diagenetic mineralization (Turner et alii, 2000; Pratt, 2001). In view of the known durability of cyanobacterial sheaths, it is conceivable that the pattern of grains may have persisted through more than one diagenetic recrystallization. This supposition is supported indirectly by the fact that a subset of fossil Decastronema from the same site showed permineralization with iron rather than by calcium enrichment (Figs. 4D-F).
The above interpretation does not preclude primary calcification that for Cretaceous microfossils may have been calcitic, because the ocean chemistry of the time was different from that of the modern ocean and calcification then favored calcite over aragonite (Stanley & Hardie, 1998). Primary calcite incorporated into the sheaths of Decastronema may have provided some initial stability to the organo-mineral relationship that was modified subsequently.
Riding (1994) noticed a positive correlation throughout the Phanerozoic between increases in the occurrence of calcified cyanobacteria and rises in abiotic carbonate precipitates. He proposed a model that distinguishes episodes of enhanced vs. reduced cyanobacterial calcification. The occurrences of Decastronema do not appear to conform to that model's prediction for these calcified filaments were most numerous during late Cretaceous times, a period during which the rate of calcification in cyanobacteria is said to have decreased. At that time massive skeletal calcification in plankton (coccolithoforids and foraminifera; Knoll, 2003) and benthos (e.g. rudists), may have competitively lowered environmental carbonate saturation levels. However, such changes in seawater chemistry would have mattered less if the calcification of Decastronema occurred post-depositionally.