De Castro (1975) recognized the unique architecture of the fossil, specifically the size, false branching and divergent layering of the wall, as characteristic of cyanobacterial sheaths. In search of the biological affinity of the fossil organism and its possible modern counterpart, De Castro referred to Monty's (1967) study of modern microbial mats on Andros Island, Bahamas and considered among others a mat-forming cyanobacterium identified as Scytonema myochrous. Earlier, Black (1933) had described from that island two species with similar characteristics, but from a predominantly freshwater habitats: Scytonema crustaceum and S. androsense. However, De Castro thought the fossil more closely related to the modern genus Tolypothrix because of the prevalence of single-sided false branching
We have studied the modern mat-building cyanobacteria on the mud flats of the west coast of Andros Island, which is influenced by marine tidal circulation. Two species of Scytonema are common there. One inhabits a narrow range of the intertidal zone on the seaward side of the barrier beach, whereas the other covers vast surfaces around mangrove bushes on the ponded mud flats behind the barrier beach. The morphology of the mud-flat-dwelling species of Scytonema is closer to that of the Cretaceous fossil. The similarities include divergent sheath layering as well as the prevalence of single sided false branching, which like that of Tolypothrix is related to its upright growth (Fig. 5A). The filaments consist of thick sheaths harboring much narrower cellular trichomes in their core. The outer margins of V-shaped layers show externally as barely visible rings. The relationship between cellular trichome and layered sheath visible in fractured specimens (Figs. 5B-C) is similar to the architecture of the walls of the Cretaceous fossil (compare Fig. 4I).
Thick envelopes and sheaths produced by copious amounts of extracellular polymeric substances (EPS) are common, particularly in subaerially growing cyanobacteria. However, the characteristic sheath architecture with open, upwardly divergent layering is particular to the section Myochrotes within the genus Scytonema (Bornet et Flahault, 1887). Trichomes of Scytonema have the highest rates of cell division and sheath excretion at their tips, thus forming apical meristems; both decrease basipetally as the cells grow older. The newly excreted portion of the sheath encloses several actively growing cells below the apex (Fig. 6A). Due to apical growth, cell division and the accompanying production of EPS move in an acropetal direction. Consequently, the envelopes assume the asymmetrical shape of inverted cones (Fig. 6B). The angle of these cones is a function of the rate at which the trichome elongates: it is more divergent when the growth is slow, and almost parallel in fast-growing aquatic forms (Bhâradwâja, 1934). The pressure of the growing trichomes causes consecutive bursting and opening of envelopes at the trichome tip so the continuing growth produces a series of characteristic, funnel-shaped collars (Fig. 7, below). This feature has a potentiality for preservation and thus for recognition in the fossil record (see Golubic & Barghoorn, 1976). In addition, nitrogen fixation in these organisms, performed by specialized cells called heterocysts, stimulates growth in the middle of the trichome, forming intercalary meristems. Differentiated heterocysts do not produce EPS and act as anchoring points for trichomes, restricting their gliding inside the sheaths (Golubic et alii, 1996). As a consequence, the trichomes protrude through the old sheath and continue to grow as false branches.
The presence of these same properties in the fossil Decastronema filaments, in particular the collar-like opening of sheath layers in conjunction with false branching, as well as an overall similarity in shape and size demonstrates unquestionably that De Castro's interpretation and identification of this organism as an ancient cyanobacterium is valid.
Using light microscopy, De Castro (1975) saw in longitudinal sections of Decastronema tubules, faint dark lines across the lumen. He interpreted these lines as the possible cross walls of the cellular trichomes, which would imply cell lengths of up to 30 µm. Our use of cross-polarized light microscopy and SEM images resolved these lines as grain boundaries between microsparitic grains that filled the tubes, often in a single row (Figs. 4D-E). At the level of resolution used in this study no remains of trichome cells or cross walls separating them were found.
The meristematic (i.e. rapidly dividing) cells in modern Scytonema are short and wide. As the rate of division slows, the cells grow longer and narrower, accompanied by a loss in turgescence. The consequence is a wide variability in the dimensions of the cells, which is consistent with the considerable variability of the internal diameters of Decastronema tubules.
The appearance of one-sided 'closure' reported in the original description (Radoičić, 1959) exists also in modern Scytonema, but is limited to growing tips and young false branches. This condition is explained by the life cycle and mode of reproduction of these filamentous cyanobacteria, during which trichomes fragment to produce and release short segments called hormogonia, which then open and leave the sheath. A similar feature was described as a 'terminal chamber' in Decastronema and interpreted as a possible heterocyst (De Castro, 1975, Pl. 6, Fig. 1). Actually, it represents a section through a curved or falsely branched filament of which a part was sectioned longitudinally and the other part transversally (Fig. 3F).
The habitat of the coastal Scytonema of Andros Island is an intertidal mudflat, flooded by meandering tidal creeks and periodically flushed by seawater during storms. Freshwater comes from rain and the drainage of water. Thus, salinity fluctuates from brackish to hypersaline. The organisms endure periodic shortages of water and an excessive solar illumination. Their thick sheaths may slow down the loss of water; they also contain the dark pigment scytonemin which provides a modicum of protection against damage from excessive light and UV radiation (Garcia-Pichel, 1998).
We consider this organism to be the closest modern counterpart of Decastronema. The fossil must have had ecological requirements very much like those of the modern species. Therefore we propose a similar habitat for the Cretaceous form.