Portal de Periódicos da CAPES

The enigmatic Devonian fossil Prototaxites is not a rolled‐up liverwort mat: Comment on the paper by Graham et al. ( AJB 97: 268–275)

2010; Wiley; Volume: 97; Issue: 7 Linguagem: Inglês

10.3732/ajb.1000047

1537-2197

T.N. Taylor, Edith L. Taylor, Anne‐Laure Decombeix, Andrew B. Schwendemann, Rudolph Serbet, Ignacio H. Escapa, Michael Krings,

Geology and Paleoclimatology Research

It should not be surprising that there are numerous examples of fossil organisms for which there are no known direct modern analogues. The seed ferns of the Paleozoic and Mesozoic, arborescent sphenophytes, and a number of Silurian and Devonian organisms sometimes placed in the artificial group Nematophytales (19 or Nematophyta (27 represent examples of such organisms. These organisms either combine morphological and anatomical features not found in the modern biota or are constructed differently from any other living or fossil life form (10. The nematophytes are undoubtedly among the latter, and the biology and systematic affinities of these organisms have remained controversial since their first discovery more than 150 years ago (16; 20. In nematophytes that are structurally preserved, it can be seen that they are composed entirely of tubes of various size, shape, and orientation (29; some of the tubes are characterized by relatively complex cross walls with a roofed central pore resembling a dolipore or parenthesome (22; 16; 28, fig. 6.9). One of the most unusual of these organisms is Prototaxites, a fossil that was initially described as partially degraded gymnosperm wood based on silicified specimens from the Gaspé of Canada (8, 9; 5. Fossils interpreted as Prototaxites have been reported as compressed, coalified remains (4, but the most interesting are silicified specimens that occur as "logs," which may be more than 1 m in diameter and 8 m long (16. Hypotheses as to the affinities of these fossils have included seed plants (9, several algal types (e.g., 17; 24; 21, an early, terrestrial evolutionary dead end (19; 1, a lichen-like association of a fungus and an autotrophic carbon source (25, and some type of terrestrial saprotrophic organism, i.e., a fungus, with affinities perhaps closest to the Basidiomycota (7; 15; 16. Likewise, biomarkers and carbon-isotope signatures, while not identifying the organism, do suggest some type of heterotrophic nutritional mode (4. The most recent suggestion is that Prototaxites represents extensive mats of a liverwort similar to modern Marchantia that were rolled up by wind, gravity, or water to form the so-called logs that are found silicified in Silurian–Devonian rocks (13. Based on a number of factors, including the anatomy of Prototaxites trunks, their mode of preservation as fossils, and the environment in which they lived, we take exception with the opinion of 13 and with some of the methodology used to develop their interpretations as to the affinities of Prototaxites. The most authoritative report on the history, description, and structural organization of Prototaxites is that of 16. Of particular importance is his detailed description of the "tissue," technically pseudotissue, system of Prototaxites, which was determined from analyzing numerous specimens in various planes of section. As 16 notes, there are three basic types of tubes, which he termed hyphal elements, based on his interpretation of this organism as a fungus and following the organization of trimitic hyphal systems in modern basidiomycetes: (1) skeletal hyphae (18–50 μm in diameter), which are thick walled, long (>2.5 mm), straight or flexuous, aseptate, and unbranched; (2) which 16 termed generative hyphae, which are 12–42 μm wide, thin walled, septate with open or occluded pores, and highly branched; and (3) binding hyphae, which are small (5–7 μm wide), thin walled, and septate with a pore. Coltricioid hyphae are formed from generative hyphae and are arranged in clusters that can be followed through several areas that 16 termed growth increments. There are several fundamental problems with the anatomical comparisons that are presented in 13 between Prototaxites and rolled-up liverwort mats. The most obvious of these is the orientation of tubes in Prototaxites vs. rhizoids in Marchantia. 13 illustrated a transverse section of a rolled liverwort (fig. 3D) next to a section of Prototaxites (fig. 3E) that they cite (p. 272) as showing "open tubes of similar diverse orientation." This section of Prototaxites, however, is clearly oblique, and the tubes are not open but sectioned through the lumen of each tube in various orientations due to the obliquity of the section. There is evidence from numerous specimens of Prototaxites, including the type material, that the largest tubes (skeletal hyphae of 16 are almost exclusively oriented parallel to the long axis of the specimen (e.g., 8; 3; 22; 6; 16. This is particularly obvious in radial sections such as the illustrations of the neotype of P. loganii in 16, pls. IIIa, Vb). It is also obvious from transverse sections that the largest tubes are oriented in cross section (Fig. 1B) (16, pl. II; fig. 14b; pls. VIII-8 and 10). In fact, the overall vertical orientation of the tubes is one of the reasons that 8 first interpreted them as tracheids in gymnosperm wood. Because of this arrangement, the structure of Prototaxites cannot be adequately explained as a rolled mat of liverworts where the tubes would correspond to the liverwort rhizoids. Figure 1A depicts our interpretation of the rolling mat hypothesis as postulated by 13 and shows that the normal orientation of the rhizoids on the substrate surface of the thallus (thalli) is quite different from the orientation of the tubes in Prototaxites. (A) Schematic drawing of a rolled mat of liverworts comparing the expected orientation of the rhizoids to the orientation of tubes in Prototaxites. (B) Cross section of Prototaxites loganii from the Lower (Emsian) Devonian of New Brunswick, Canada showing the anatomy across the border of a growth increment (dotted line). Note the continuity of the pseudotissue and the similar orientation of all the large (skeletal) tubes. The border of the growth increment is characterized by skeletal tubes that have higher density, reduced diameter, and thicker walls. Scale bar = 2 mm. Center of the log is toward the bottom. Slide #26505 (Delevoryas Collection), Paleobotanical Collections, University of Kansas. Another major obstacle to the rolling mat hypothesis concerns what 16 and others have referred to as the growth increments in Prototaxites. Transverse sections of large specimens indicate a series of concentric rings that look superficially similar to growth rings of a woody plant with secondary xylem. 16 suggests that an amphigenous hymenium (i.e., a hyphal layer containing the sexual spore-producing cells in Ascomycota and Basidiomycota that extends over the entire surface of the spore-producing body) periodically developed on the surface of the Prototaxites organism and that this alternation between layers of the so-called vegetative hyphal system and amphigenous hymenophoral strata produced in time a pattern of multiple concentric growth increments. We have reproduced (Appendix S1, see Supplemental Data at http://www.amjbot.org/cgi/content/full/ajb.1000047/DC1) the figure from appendix S7 of the 13 paper at http://tinyurl.com/38dx4qq) at a high-enough resolution so that it is possible to trace the boundary (i.e., the hymenophoral strata of 16 of several of the growth increments at high magnification. To improve the ease of tracing these increments, we have added colored dots showing that each increment marked in the central part of the trunk is concentric, not spiraled, although the growth of the increments is somewhat eccentric, as 16 noted. This significant feature of Prototaxites refutes the hypothesis that the logs formed by mats of liverworts rolling up, as the accumulation of new tubes would have resulted in a spiral pattern of increments in transverse section. This figure (their appendix S1) also indicates several regions where growth increments are narrow or tightly compacted. We interpret these as areas in which growth of the hymenium was somehow disturbed or interrupted, perhaps by other closely spaced Prototaxites individuals or some other impediment to the circumferential growth of the organism, for example, wildfire (e.g., 12. Even in specimens of Prototaxites that have eccentric growth increments, the pattern of growth shows closed and complete boundary layers that are telescoped in width, rather than being spiral. In contrast, rolled mats of Marchantia (13, fig. 3C, D) clearly show dense layers formed by the mat itself, alternating with areas of less density that are filled with rhizoids extending at several different angles from the bottom of the mat. 13, p. 5) also suggested that the borders of the growth increments in Prototaxites are "explainable as the remains of resistant marchantioid lower epidermis." Even a cursory examination of transverse and longitudinal sections of Prototaxites clearly indicates that the growth increments are characterized not only by changes in the density of the largest tubes, but also by significant changes in their diameter and in the thickness of their walls (their fig. 1B; 16, pl. II). In the Marchantia mats, by contrast, the rhizoids may be more numerous close to the lower epidermis of the thallus, but there is no evidence that they also have a smaller diameter and thicker walls in this zone. 13, fig. 4B, p. 270) compared the narrow, thin-walled rhizoids of Marchantia with the narrow, thin-walled tubes (binding hyphae of 16 in Prototaxites and noted that the diameter and wall thickness of all the Marchantia rhizoids "closely matched" the tubes in their single specimen of Prototaxites. Unfortunately, no cell measurements were given for the fossil specimen, and only a single range of all sizes was given for the extant liverworts. Both types of thin-walled tubes recognized in Prototaxites (binding and generative hyphae of 16, however, are described as profusely branched, which is not the case for rhizoids. More generally, 13 only illustrated individual types of tubes, thus providing no evidence that the complex arrangement observed in the fossil can be achieved. For example, fig. 3D (13 can only explain one of the different types of tubes present in the section of Prototaxites shown (13, fig. 3E). Finally, it is highly probable that regardless of the structure and organization of whatever the liverwort mat might have been, during the downhill rolling phase, there would have been bits and pieces of other organisms, weathered components of the substrate, and other contaminants within the rolls. Extensive and compact mats of marchantioid liverworts today are found in a variety of habitats, including burnt soil and sandy beds of streams and gullies. These mats are typically firmly connected to the substrate by the innumerable small rhizoids of the individual thalli that extend between and around litter and sediment particles and into even the smallest substrate fissures. If such a mat is removed from the substrate, it usually comes off with considerable amounts of litter and sediment clinging to the underside. As a result, a similar mat in the Kettle Point Formation paleoecosystem (3, where 13 specimen originated, would certainly have picked up sediment, as well as parts of other plants growing between the liverworts, while being rolled up by wind or some other external force. The depositional environments described by 14 and 15 include numerous layers of vascular plants in the shales, with Prototaxites deposited in sandy streambeds. This indicates the presence of a rich, contemporaneous vascular flora, which would have been either caught up in the "rolling" process or possibly have impeded the movement of a structure as large as a Prototaxites "roll." The embedded plant fragments noted by 16 occur on the surface of the specimen, not within. These may have become attached to the organism after deposition in the streambed where some of the Prototaxites specimens are thought to have been fossilized (14. On the other hand, they may have actually been growing on Prototaxites. Both of these scenarios are not uncommon in fossil assemblages. 13, p. 273) stated that the stable carbon isotope values obtained in their experiments "span the range of δ13C values reported for Prototaxites fossils" in 4. However, their isotope values only partially overlap those of Prototaxites and fail to take into account important differences in atmospheric δ13C between the present day and the Devonian. When comparing carbon-isotope data across geologic time, it is necessary to convert numbers in δ-notation to discrimination (Δ), which is independent of the isotopic composition of the atmosphere (11. Discrimination values can be obtained by the equation:where δa is the composition deviation of the 13C in the atmosphere relative to the standard, and δp is the composition deviation of the 13C in the plant relative to the standard. Using Eq. 1, ancient δa13C values obtained from the literature (–6.2‰ for Prototaxites from ca. 375 Ma, –6.4‰ for Prototaxites from ca. 400 Ma, –8.0‰ for extant Marchantia; 26, and δp13C values from 4 and 13, we calculated carbon-isotope discrimination values for Prototaxites and Marchantia (Table 1). Discrimination values of both organisms span a broad range yet contain measurements substantially different from one another. 13 used the range of δ13C they obtained experimentally along with a mixing model to justify the statement that isotopic values of Marchantia span the range of those reported for Prototaxites. However, the isotopic value of the glucose used in their experiment (–10.2‰) indicates a C4-derived glucose. Thus, the isotopic values of the Marchantia specimens grown on an agar-mineral medium with added glucose are enriched in δ13C (less negative: –12.3‰) compared to the control (–20.6‰; Table 1). If this anomalous value is removed, Prototaxites fossils are more depleted in δ13C than both the Marchantia control and the Marchantia associated with cyanobacteria (Table 1). 13 did not demonstrate in a controlled experiment that the addition of cyanobacteria to Marchantia populations will cause their tissues to be more depleted in 13C. Isotopic values of liverworts growing in association with cyanobacteria are given in 13, p. 273) (Table 1), but the specimens growing on agar media cannot be used as a control for them because the liverworts associated with cyanobacteria were grown partially submerged in standing water and not on an agar medium. Furthermore, 13 suggested the presence of Devonian cyanobacteria growing with Prototaxites, but provided no concrete evidence of such an association (e.g., 18. Determining the affinities of enigmatic fossils is often a daunting task, and it is important that apparent similarities not be overly emphasized in equating these organisms with possible modern analogues. Numerous fossils have no close modern homologues or analogues, and the further back in geologic time, the truer this statement becomes. Modern laboratory manipulations should not alter the structure or organization of the organism without taking into account the known sedimentology and taphonomy of the site where the fossil was collected and the conditions under which the fossil was preserved. We do not know definitively what kind of organism Prototaxites represents, or where on the tree of life its affinities may reside, although at least two forms of evidence point toward fungal affinities (16; 4. Even among the authors of the present paper, there are various interpretations of what this organism represents, how it may have functioned, and how it maintained itself nutritionally. It seems probable from previous work, based on anatomy (16, isotopes (4, and the environment of deposition (14; 15 that Prototaxites was a terrestrial organism that grew upright and added new pseudotissue (i.e., tubes) in a circumferential pattern. Although there has been speculation as to the proximal and distal ends of the organism (e.g., 2; 23, 24, there is no compelling evidence based on structurally preserved specimens to suggest what these regions looked like. The reconstruction of Prototaxites as a rolled-up mat of liverworts, as proposed by 13, simply does not fit the available evidence about this organism gathered over the last 150 years. While there is still much to learn about this enigmatic organism, e.g., reproduction, habit of growth, it does not represent a mat of liverworts that rolled down some Devonian landscape. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.