Info

Figure 1. Schematic summary of preservation resolution, temporal distribution, and environmental distribution of various taphonomic pathways (Butterfield, 2003).

lack of phosphatization of more labile substrates such as cellular and subcellular structures. Some phosphatized biotas in Cambrian successions (Zhang and Pratt, 1994; Bengtson and Yue, 1997; Dong et al., 2004) may be considered as transitional taphonomic windows that bridge the end members of the Doushantuo and Orsten biotas, because they show evidence of both cellular and cuticular phos-phatization. When considered together, the two phosphatization windows seem to be open only in the Ediacaran and early Paleozoic (Donoghue et al., 2006).

Beecher's trilobite-type preservation represents pyritization of relatively recalcitrant tissues, for example chitinous arthropod cuticles (Briggs et al., 1991) and cellulose-based cell walls (Yuan et al., 2001; Grimes et al., 2002). The preservational resolution of this taphonomic window is limited by crystal size of authigenic pyrite and controlled by the balance between organic degradation through bacterial sulfate reduction (as a source of sulfide) and authigenic precipitation of pyrite. The temporal and environmental occurrences of this taphonomic window are not well characterized, but are potentially widespread given that pyrite formation is sensitive to local geochemical conditions rather than global secular trends (Rickard et al., 2007). Possible Proterozoic examples of this type of preservation include pyritized chuarid vesicles (Yuan et al., 2001) and pyritized tubes (Cai and Hua, 2007) in Ediacaran successions of South China.

The four preservational pathways discussed above are collectively known as permineralization - the preservation of soft tissues with three-dimensional detail by authigenic minerals (Briggs, 2003). In contrast, Burgess Shale-type preservation is characterized by two-dimensional compression and preservation of more recalcitrant tissues as carbonaceous films (Butterfield, 1995; Gaines et al., 2005), typically on bedding planes of fine-grained sediments deposited in deep-water environments below fair weather wave bases. Several mechanisms have been proposed to explain Burgess Shale-type preservation, although these need not to be mutually exclusive. Some argue that authigenic aluminosilicate minerals may have played a role in delaying organic degradation, therefore promoting organic preservation (Butterfield, 1995; Orr et al., 1998). But recent investigation of the Chengjiang biota - an early Cambrian example of Burgess Shale-type preservation - seems to indicate that pyrite mineralization was at least partly responsible for the exceptional preservation of soft tissues (Gabbott et al., 2004; Zhu et al., 2005). These authors argue that the degradation of more labile tissues by sulfate reduction bacteria provides a source of hydrogen sulfide (H2S), which in the presence of reactive iron (Fe) would promote pyrite mineralization and the preservation of more recalcitrant soft tissues. Burgess Shale-type preservation is most common in the Cambrian, but also occurs in the Proterozoic (Xiao et al., 2002) and post-Cambrian Paleozoic (Butterfield, 1995). The secular trend of Burgess Shale-type preservation may be controlled by several factors, two of primary importance may be clay mineral geochemistry (Butterfield, 1995; Orr et al., 1998) and bioturbation (Allison and Briggs, 1993; Orr et al., 2003).

Finally, the Ediacara-type preservation is a non-actualistic taphonomic window, characterized by the casting and molding of macroscopic organisms in siliciclastic rocks, including sandstones. Microbial mats, through the formation of a "death mask" on degrading Ediacara organisms and the promotion of mineralization beneath microbial mats, may be responsible for the casting and molding of Ediacara fossils in South Australia (Gehling, 1999). Other Ediacara fossils were probably "masked" by volcanic ashes (Narbonne, 2005) or event deposits of fine-grained silts (Xiao et al., 2005). Ediacara-type preservation mostly occurs in Ediacaran rocks, although rare occurrences in Phanerozoic and Mesoproterozoic successions have been reported (Jensen et al., 1998; Hagadorn et al., 2000; Zhang and Babcock, 2001; Fedonkin and Yochelson, 2002).

From an astrobiological perspective, the Bitter Spring-, Doushantuo-, and Orsten-type preservation pathways are most important because of their likelihood of capturing morphological information about microscopic organisms consisting of more labile organic structures. In the following section, we will take a closer look at the Doushantuo-type preservation at its type locality.

3. Phosphatization in the Neoproterozoic Doushantuo Formation

The late Neoproterozoic phosphorite of the Doushantuo Formation at Weng'an, South China, hosts some of the best preserved microfossils, providing both cellular and subcellular insights into a variety of eukaryote organisms, including animals and algae (Xiao et al., 2004; Hagadorn et al., 2006). Because the Doushantuo Formation at Weng'an hosts the earliest record of animal life in the form of fossilized metazoan embryos (however, see Bailey et al., 2007 for an alternative interpretation and Xiao et al., 2007 for a rebuttal), past investigations have focused primarily on their evolutionary significance (Xiao et al., 1998). Only a few studies were designed specifically to understand the taphonomy of the Weng'an biota (Xiao and Knoll, 1999; Dornbos et al., 2005, 2006). The lack of in-depth understanding of Doushantuo taphonomy not only makes the biological interpretation of certain Doushantuo fossils controversial (Chen et al., 2000, 2004; Xiao et al., 2000; Bengtson and Budd, 2004), but also weakens the role of the Doushantuo biota as a model to guide further exploration of Doushantuo-type preservation in other ages and on other planets.

Previous examination (Xiao and Knoll, 1999) using scanning electron microscopy provided evidence that the preservation of Doushantuo microfossils critically depends on two mineralization processes: (1) encrustation through mineral nucleation and precipitation on organic substrates, such as cell membranes, cell walls, and mucus strands; and (2) impregnation of phosphatic minerals within organic substrates. Recent analysis provides further evidence to substantiate these processes.

In our investigation, we extracted Doushantuo microfossils using the standard acid digestion method (dissolution in 10% acetic acid for 3-6 days). Extracted specimens were gold-palladium (Au-Pd) sputter-coated to ~20 nm in thickness and then observed using scanning electron microscopy. We recognize four different styles of phosphate mineralization, each characterized by distinct crystal size, orientation, and organization. These are interpreted as different phosphatization processes related to availability of nucleation sites and degradation of organic substrate. The following sections explore these four styles of phosphate mineralization (Figs. 2-11), including (1) perpendicularly oriented, prismatic apatite crystals; (2) tangentially oriented, bladed apatite crystals; (3) randomly oriented, equant apatite crystals; and (4) phosphatic filaments, rods, and granules.

Figure 2. Phosphatic encrustation on Megasphaera inornata. (A) Eccentric egg cell and phosphatic envelope with uneven thickness. Boxed areas magnified in (B, D, and F). (B) Botryoidal cements on cell surface and in space between cell and envelope. (C) Magnification of (B) showing crystal terminations (lower right) and inward growing crystals (upper left). (D-E) Botryoids with inward growing crystals. Arrowed area in (D) magnified in (E). (F-H) Isopachs with inward growing crystals (H) and randomly oriented crystals (G). Arrowed areas in (F) magnified in (G) and (H).

Figure 2. Phosphatic encrustation on Megasphaera inornata. (A) Eccentric egg cell and phosphatic envelope with uneven thickness. Boxed areas magnified in (B, D, and F). (B) Botryoidal cements on cell surface and in space between cell and envelope. (C) Magnification of (B) showing crystal terminations (lower right) and inward growing crystals (upper left). (D-E) Botryoids with inward growing crystals. Arrowed area in (D) magnified in (E). (F-H) Isopachs with inward growing crystals (H) and randomly oriented crystals (G). Arrowed areas in (F) magnified in (G) and (H).

Figure 3. Phosphatic encrustation on Megasphaera inornata (A-E) and Parapandorina rhaphospissa (F-H). (B, D, and H) are magnified views of boxed areas in (A, C, and F), showing isopachous cements with inward growing crystals on inner surface of envelope. (E and G) are magnified views of cell surface in (C and F), showing distal views of hexagonal crystal terminations.

Figure 3. Phosphatic encrustation on Megasphaera inornata (A-E) and Parapandorina rhaphospissa (F-H). (B, D, and H) are magnified views of boxed areas in (A, C, and F), showing isopachous cements with inward growing crystals on inner surface of envelope. (E and G) are magnified views of cell surface in (C and F), showing distal views of hexagonal crystal terminations.

Figure 4. Phosphatic encrustation on Parapandorina rhaphospissa (A-B), Megaclonophycus onustus (C-F), and a strongly degraded spheroidal fossil (G-H). (B and D) are magnified views of boxed areas in (A and C), showing isopachous cements on both inner and outer surfaces of envelope/membrane, with crystals growing centrifugally from the original organic envelope that may be represented only by the gap between centrifugally growing cements. (D) is further magnified in (E and F), showing botryoidal cements within cells as well as isopachous cements on membrane. (H) is magnified view of (G), showing crystal terminations and botryoidal cements on strongly degraded cellular content.

Figure 4. Phosphatic encrustation on Parapandorina rhaphospissa (A-B), Megaclonophycus onustus (C-F), and a strongly degraded spheroidal fossil (G-H). (B and D) are magnified views of boxed areas in (A and C), showing isopachous cements on both inner and outer surfaces of envelope/membrane, with crystals growing centrifugally from the original organic envelope that may be represented only by the gap between centrifugally growing cements. (D) is further magnified in (E and F), showing botryoidal cements within cells as well as isopachous cements on membrane. (H) is magnified view of (G), showing crystal terminations and botryoidal cements on strongly degraded cellular content.

Figure 5. Phosphatic encrustation on organic filaments (mucous strands, bacterial filaments, or fungal hyphae). (B and D) are magnified views of (A and B), showing external and cross section views of phosphatic filaments. Crystals are oriented radially and coarsen centrifugally.

3.1. PERPENDICULARLY ORIENTED, MICROMETRIC, PRISMATIC APATITE CRYSTALS: PHOSPHATIC ENCRUSTATION ON ORGANIC AND INORGANIC SUBSTRATES

3.1.1. Description

We begin by examining one Doushantuo specimen of Megasphaera inornata (Fig. 2A), interpreted as a phosphatized animal egg cell encased within an egg envelope. At a closer look, there is abundant evidence for encrusting apatite crystals of micrometric size (Figs. 2C, E, H). These crystals are often oriented perpendicular to encrusted surfaces, which can be the egg cell surface (Fig. 2B, lower right) or the outer envelope (Fig. 2E). The egg cell is eccentrically located within the envelope and entirely covered with apatite botryoids approximately 10-20 |im in size (Fig. 2B, lower right). There is evidence that the botryoids overgrow on each other and sometimes aggregate to form cauliflower-like structures (Fig. 2B). The botryoids are made of prismatic apatite crystals that are radially oriented. Thus, they appear as hexagonal terminations when viewed distally on botryoid surface (Fig. 2C, lower right). The crystals are prismatic euhedra, typically ~0.09-0.85|im in width and ~0.40-2.25|im in length (Fig. 12).

Similarly, the envelope is covered with both botryoidal (Figs. 2D-E) and isopachous apatite cements (Fig. 2H). Thus, the apparent thickness of the phosphatic

Figure 6. Phosphatic encrustation on a tubular microfossil (possibly Sinocyclocyclicus guizhouensis; A-D) and a problematic microfossil (E-H). (B-D) are successively magnified views of (A), showing phosphatic encrustation (with perpendicularly oriented crystals) overlying small randomly oriented crystals (D), interpreted as phosphatic impregnation of organic tube walls. (F-H) are successively magnified views of (E), showing phosphatic encrustation (H, distal view) mantling small randomly oriented crystals (G).

Figure 6. Phosphatic encrustation on a tubular microfossil (possibly Sinocyclocyclicus guizhouensis; A-D) and a problematic microfossil (E-H). (B-D) are successively magnified views of (A), showing phosphatic encrustation (with perpendicularly oriented crystals) overlying small randomly oriented crystals (D), interpreted as phosphatic impregnation of organic tube walls. (F-H) are successively magnified views of (E), showing phosphatic encrustation (H, distal view) mantling small randomly oriented crystals (G).

Figure 7. Phosphatization of algal cell walls. (B and C) are successively magnified views of (A), showing fractured algal thallus with cellular preservation. Small apatite crystals nucleated on cell walls. (E and F) are successively magnified views of (D), showing poorly organized crystals on cell walls.

Figure 7. Phosphatization of algal cell walls. (B and C) are successively magnified views of (A), showing fractured algal thallus with cellular preservation. Small apatite crystals nucleated on cell walls. (E and F) are successively magnified views of (D), showing poorly organized crystals on cell walls.

envelope depends on the degree of encrustation and may be significantly different from the thickness of the original organic envelope. In the specimen illustrated in Fig. 2, because of the eccentric location of the egg cell and asymmetrical encrustation, its phosphatic envelope is strongly uneven in thickness. Indeed, the phos-phatic envelope splits into two parts, creating a cavity between what appear to be layers of phosphatic encrustation (Fig. 2A, upper right). Phosphatic encrustation occurs on all surfaces - the inner and outer envelope surfaces, as well as the walls defining the cavity. These apatite crystals are also approximately 0.09-0.85 |im in width (Fig. 12). They terminate and increase in size toward the cavity (Figs. 2D-E),

Figure 8. Phosphatic infilling of Archaeophycus venustus cells. (B) is magnified view of (A), (D and E) are successive magnifications of (C), and (G and H) are magnified views of (F), showing tangentially oriented crystals.

Figure 9. Phosphatic infilling in Meghystrichosphaeridium reticulatum vesicle (A-C), and phosphatic encrustation (E, white arrow) and impregnation (F and H) of Archaeophycus venustus (D-H). (B and C) are successive magnifications of (A), showing tangentially oriented crystals. (E and F) are successive magnifications of (D), and (H) is magnified view of (G), showing randomly oriented sub-micrometric crystals.

Figure 9. Phosphatic infilling in Meghystrichosphaeridium reticulatum vesicle (A-C), and phosphatic encrustation (E, white arrow) and impregnation (F and H) of Archaeophycus venustus (D-H). (B and C) are successive magnifications of (A), showing tangentially oriented crystals. (E and F) are successive magnifications of (D), and (H) is magnified view of (G), showing randomly oriented sub-micrometric crystals.

Figure 10. Phosphatic rods (B, arrow) and granules (D-E and G-H) on a tubular microfossil (possibly Sinocyclocyclicus guizhouensis, A-B), an unidentified algal fossil (C-E), and animal embryo Parapandorina rhaphospissa (F-H). (B) is magnified view of (A), (D and E) magnified views of (C), and (G and H) magnified views of (F). Note granular texture in (D-E and G-H).

Figure 10. Phosphatic rods (B, arrow) and granules (D-E and G-H) on a tubular microfossil (possibly Sinocyclocyclicus guizhouensis, A-B), an unidentified algal fossil (C-E), and animal embryo Parapandorina rhaphospissa (F-H). (B) is magnified view of (A), (D and E) magnified views of (C), and (G and H) magnified views of (F). Note granular texture in (D-E and G-H).

Figure 11. Granules on animal egg/embryo fossil Megasphaera ornata (A-C) and an algal fossil (possibly Paramecia incognata, D-H). (B and C) are successive magnifications of (A). (E and G) are magnifications of (D), and are further magnified in (F and H), respectively. Granular texture best seen in (F and H).

Figure 11. Granules on animal egg/embryo fossil Megasphaera ornata (A-C) and an algal fossil (possibly Paramecia incognata, D-H). (B and C) are successive magnifications of (A). (E and G) are magnifications of (D), and are further magnified in (F and H), respectively. Granular texture best seen in (F and H).

Figure 12. Crystal aspect variation by morphology. (A) Length-width plot of four crystal morphologies. Examples of crystal morphologies with marked dimensions and color coding are shown to the right in figure key. (B) Length-width plot of perpendicular prismatic apatite crystals. (C) Length-width plot of comparatively smaller crystal morphologies (tangential blades, random equant crystals, granule aggregates, and granules). (D) Bootstrapped mean crystal length versus mean crystal width, with 95% confidence intervals. (E) Bootstrapped length:width ratio with 95% confidence intervals.

Figure 12. Crystal aspect variation by morphology. (A) Length-width plot of four crystal morphologies. Examples of crystal morphologies with marked dimensions and color coding are shown to the right in figure key. (B) Length-width plot of perpendicular prismatic apatite crystals. (C) Length-width plot of comparatively smaller crystal morphologies (tangential blades, random equant crystals, granule aggregates, and granules). (D) Bootstrapped mean crystal length versus mean crystal width, with 95% confidence intervals. (E) Bootstrapped length:width ratio with 95% confidence intervals.

toward the egg cell (Figs. 2F, H), and away from the envelope, indicating that they grew as cavity infillings and surface encrustations.

Similar isopachous and botryoidal encrustation occurs in virtually all Doushantuo animal eggs/embryos (Figs. 3-5), tubular fossils (Figs. 6A-D), and multicellular algae (Fig. 7). Apatite crystals are perpendicularly oriented on egg envelope surfaces (Figs. 3B, D, H, 4B), egg/embryo cell surfaces (Figs. 3E, G, 4E), filaments (possibly mucus strands, bacterial filaments, or fungal hyphae; Figs. 5B, D), the surface of a problematic fossil (Fig. 6H), and algal cell walls (Fig. 7C).

It appears that botryoidal and isopachous cementation is selective with respect to encrusted substrate. Botryoidal cements tend to occur on egg/embryo cell surfaces or cell interiors. In some Megasphaera specimens, the egg cell was significantly reduced (Figs. 3A, C) or strongly degraded beyond recognition (Fig. 4G). However, their cell surface, regardless of the degree of degradation, is often completely covered with botryoidal apatite cements. Similar botryoidal cements also occur on the cell surfaces of Parapandorina rhaphospissa (Figs. 4A, 5A), interpreted as blastula-stage embryos (Xiao and Knoll, 2000). These botryoidal cements are sometimes continuous with phosphatic filaments (Fig. 5). The cells of Megaclonophycus onustus, interpreted as possible blastula-stage embryos (Xiao and Knoll, 2000), are also covered with cements (Figs. 4C-E), although these are more isopachous than botryoidal. However, in the same specimen, abundant botryoidal cements occur within cells, as revealed by natural fractures of the cells (Fig. 4F). Additionally, some botryoidal cements also occur on the inner surfaces of egg envelopes (Figs. 2C-E).

In contrast, isopachous cements preferentially occur on egg envelopes, both on the inner and outer surfaces, with perpendicularly oriented crystals growing inward (Figs. 3B, D, H) or outward (Fig. 4B), respectively. Occasionally a thin veneer of isopachous cement overlies phosphatic substrate consisting of smaller, randomly oriented crystals (Figs. 6C, F, 9E). Finally, some algal cell walls appear to be phosphatized by poorly oriented, relatively small (sub-micrometric) apatite crystals (Fig. 7).

3.1.2. Interpretation

Phosphatic encrustation is interpreted as a relatively late diagenetic process. This is supported by the following observations: (1) it consists of relatively larger (micrometric) crystals and (2) it mantles early diagenetic phosphate that consists of smaller (sub-micrometric) and randomly oriented crystals (Figs. 6C, F, 9E). It is interesting to note that, with some exceptions, botryoidal cements tend to occur on animal egg/embryo cell surfaces and inner surfaces of egg envelopes, whereas isopachous cements tend to occur on envelopes or mantle early diagenetic phosphate. It is likely these two types of cements represent two generations of cementation that were controlled by nucleation processes (e.g., nucleation on small isolated particles vs. surfaces). We hypothesize that more recalcitrant substrates (e.g., envelope) can maintain their integrity and provide coherent surfaces on which late diagenetic, isopachous cements nucleate and grow. In contrast, more labile substrates

(e.g., cytoplasm, cell membranes) were easily degraded into organic particles or macromolecules that served as isolated nucleation sites for the growth of early diagenetic botryoidal cements.

3.2. TANGENTIALLY ORIENTED, SUB-MICROMETRIC, BLADED APATITE CRYSTALS: PHOSPHATIC INFILLING OF INTRACELLULAR SPACE

3.2.1. Description

Tangentially oriented, sub-micrometer-sized apatite crystals occur in a number of Archaeophycus venustus (=Paratetraphycus giganteus) specimens, interpreted as algal fossils (Zhang et al., 1998). Archaeophycus venustus cells are polyhedral in shape and approximately 10-30 |im in size (Figs. 8, 9D-H). They are often packed into sarcinoidal clusters. The cells are phosphatized by tangentially oriented, sub-micrometric (<0.20 |im in width, ~0.20-0.55 |im in length; see Fig. 12), apatite crystals - many of which with long axes parallel to the cell surface (Figs. 8B, E, H). One polyhedral cell has two of its facets exposed (Fig. 8H), and it can be observed from this cell that exposed crystals on both facets are tangentially oriented. Tangentially oriented crystals also occur in the acritarch Meghystrichosphaeridium reticulatum (Xiao and Knoll, 1999), whose vesicle surface is defined by sub-micrometric crystals that lie parallel to its vesicle surface (Figs. 9A-C). With a mean width of ~0.10 |im (ranging from 0.07-0.15 |im), tangentially oriented crystals are more slender than the perpendicular crystals in botryoidal and isopa-chous cements (Fig. 12). Moreover, they are often bladed and less euhedral than the prismatic crystals in botryoidal and isopachous cements.

3.2.2. Interpretation

Tangentially oriented apatite blades exclusively occur in algal and acritarch fossils, but not in phosphatized animal embryo cells or embryonic envelopes, which are typically characterized by botryoidal and isopachous cements, respectively. The tangential orientation of the crystals indicates that they did not nucleate on cell walls; instead, their orientation seems to be constrained by cell walls. We hypothesize that these tangential crystals grew on floating nuclei within algal cells or acritarch vesicles and were pushed against the cell/vesicle walls, essentially making an internal mold of the cells or vesicles.

It is uncertain why tangentially oriented crystals are not present in animal embryo cells or envelopes. It is possible that nucleation sites within algal cells and acritarch vesicles were abundant, so that randomly nucleated crystals tend to be smaller and their tangential orientation was constrained by the relatively recalcitrant algal cell walls and acritarch vesicle walls. In contrast, the space between animal egg/embryo cells and egg/embryonic envelopes is usually significant (partly because of shrinkage of egg/embryo cells), and nucleation sites may have been relatively rare in this space, thus nucleation was focused on egg/embryo cell surfaces and egg/embryonic envelopes. The ample space between egg/embryo cells and encasing envelopes allowed apatite crystals to grow larger than tangentially oriented crystals.

3.3. RANDOMLY ORIENTED, SUB-MICROMETRIC, EQUANT APATITE CRYSTALS: PHOSPHATIC IMPREGNATION

3.3.1. Description

Some Doushantuo fossils were preserved through the precipitation of randomly oriented, sub-micrometric, equant apatite crystals. Such sub-micrometric crystals occur on the tube walls of Sinocyclocyclicus guizhouensis (Fig. 6D), the phosphatic wall of a spherical fossil (Fig. 6G), and cell surfaces of Archaeophycus venustus (Figs. 9F, H). In all cases, the randomly oriented crystals are subsequently mantled by isopachous cements of larger and perpendicularly oriented apatite crystals. Like the tangentially oriented crystals described above, the randomly oriented crystals are less euhedral and smaller (~0.08-0.30 |im in width, ~0.10-0.35 |im in length; see Fig. 12) than the perpendicularly oriented crystals, but they are equant rather than bladed.

3.3.2. Interpretation

Clearly, precipitation of the randomly oriented crystals predates the perpendicularly oriented crystals. Thus, randomly oriented crystals have the greatest potential to replicate the most labile organic structures. We hypothesize that the tube walls and cell walls were impregnated with microcrystals after only minimal degradation. In the impregnation process, crystal orientation and morphology are constrained by available "interstitial" space within the organic structures being impregnated. Of course, subsequent crystal overgrowth after complete degradation of organic structure is likely, so that the "interstitial" space is unlikely to have been faithfully replicated by the sub-micrometric crystals.

3.4. PHOSPHATIC FILAMENTS, RODS, AND GRANULES: EVIDENCE FOR MICROBIAL ACTIVITIES?

3.4.1. Description

Phosphatic filaments, described above (Fig. 5), are typically 10-20 |im in diameter and up to 100 |im in length, and often form pillars or networks in the space between shrunken eggs/embryos and envelopes. At closer look, they consist of radially oriented crystals that form botryoidal or isopachous cements. Crystals in the axial region (~0.1 |im in width) are much smaller than in the peripheral region (0.3-2.0 |im in width). Similar phosphatic filaments have been found in Phanerozoic phosphatized biotas (Bengtson, 1976; Conway Morris and Chen, 1992; Ding et al., 1992; Müller and Hinz-Schallreuter, 1993; Martill and Wilby, 1994; Duncan and Briggs, 1996; Duncan et al., 1998; Yue and Bengtson, 1999). Some of these (e.g., Fig. 9D of Yue and Bengtson, 1999) have an axial lumen surrounded by isopachs or botryoids of radially oriented and outward growing crystals. The Doushantuo filaments do not have an axial lumen (Fig. 5D); instead, the axial region is characterized by much smaller crystals.

One specimen of the tubular microfossil Sinocyclocyclicus guizhouensis (Liu et al., 2008) bears rare phosphatic rods (Fig. 10A-B). The rods are slightly curved, about 0.2 ||m in diameter and up to approximately 1.3 ||m in length. Furthermore, these phosphatic rods consist of small granular sub-structures that are approximately 0.05-0.08 |im in size. It is uncertain whether they have a central lumen but they appear to have distally closed ends.

Some (but not all) Doushantuo fossils are covered with a granular texture (Figs. 10C-H, 11). Such granular texture occurs on the surface of multicellular algae (Figs. 10C-D), animal embryo cells (Fig. 10F), and egg envelopes (Fig. 11A). The texture consists of granules of relatively uniform size, on average about 0.05 |im in width and 0.09 |im in length (range of ~0.03-0.09 |im and ~0.06-0.15 |m, respectively). The granules forms aggregates about 0.30-0.70 |m in length and 0.08-0.15 |m in width (Figs. 11E-F, 12) or cover the surface of individual crystals (Figs. 11G-H). The occurrence of such granules on larger euhedral crystals (Fig. 11H) indicates that they postdate such euhedral crystals and are likely late diagenetic in origin. But it is uncertain whether the granular aggregates (Figs. 10D, 11F) and the granules on extremely small crystals (Fig. 11C) were also of late diagenetic origin, because they may or may not share the same origin with the granules on large euhedral crystals.

3.4.2. Interpretation

We interpret the phosphatic filaments, rods, and granules - in order of decreasing confidence - as possible evidence of microbial activities. The phosphatic filaments are almost identical in morphology to silicified microbial filaments in modern hot-spring environments (Jones et al., 1997, 2004; Renaut et al., 1998). Microbial filaments in modern hot-spring environments are rapidly encased by opal-A microspheres, often leaving an axial lumen (Fig. 10M of Jones et al., 1997) similar to those described by Yue and Bengtson (1999, their Fig. 9D). Jones et al. (2004) also showed that extremely thin mucus strands can serve as nucleation substrates for the nucleation of opal-A microspheres, which upon growth can be coalesced to form siliceous pseudofilaments. These siliceous filaments are plausible modern analogs for the phosphatic filaments from the Doushantuo Formation. Although Doushantuo filaments do not have a well defined axial lumen, the extremely small apatite crystals in the axial region (as compared with crystals in the peripheral regions) suggest the former presence of an organic or microbial filament, which either escaped from entombment or degraded after being entombed, and the former axial lumen was then filled with diagenetic phosphate. The phylogenetic affinity of the microbial filaments, however, is more difficult to determine. They can be bacterial filaments, fungal hyphae, or simply mucous strands produced by any microbes (Xiao and Knoll, 1999). Bailey et al. (2007) suggested that the filaments may represent symbiotic epibionts, but they can also be interpreted as saprophytic bacteria given that such filaments typically occur on shrunken and degraded cells (Fig. 5; Xiao and Knoll, 1999).

Doushantuo phosphatic rods (Fig. 10B) can also be interpreted as phos-phatized bacteria. Indeed, their morphology and granular texture is very similar to silicified bacterial rods reported in Jones et al. (1997, their Fig. 5G), except they are about five times smaller. Their smaller size (about 0.2 |im in diameter and 1.3 | m in length), however, does not preclude a bacterial interpretation (Southam and Donald, 1999).

Doushantuo phosphatic granules (0.09 |im in length and 0.05 |im in width; Figs. 11E-H), on the other hand, may approach the size limit of cellular life (Nealson, 1997; Southam and Donald, 1999). Structures of similar size and shape have been interpreted as nanobacterial fossils (Folk, 1999; Folk and Rasbury, 2002), but this interpretation has been met with skepticism because their extremely small size may not be sufficient to house necessary metabolic machineries (Nealson, 1997; Southam and Donald, 1999). A recent report describes granular-textured sheets from Triassic stromatolites (Perri and Tucker, 2007). These granular-textured surfaces in Triassic stromatolites are strikingly similar to those in the Doushantuo Formation (compare Figs. 11G-H with Figs. 2C, 3B, and 4C in Perri and Tucker, 2007), and a similar origin for both occurrences is plausible. Perri and Tucker (2007) interpreted the Triassic granular-textured sheets as mineralized extracellular polymeric substances (EPS), on the basis of their similarity to the sub-polygonal honeycomb structure of modern EPS. This interpretation may also be applicable to the Doushantuo Formation, but more research is needed to test the possibility that the granular texture in the Doushantuo Formation may be abiotic precipitation formed during acetic acid treatment in the laboratory or, less likely due to their irregular nature, artifacts resulting from excessive Au-Pd coating during sputtering.

Was this article helpful?

0 0

Post a comment