Triploblastic Animals More Than 1 Billion Years Ago: Trace Fossil Evidence from India

Adolf Seilacher, Pradip K. Bose, Friedrich Pflüger

Science 1998 282: 80-83.

Some intriguing bedding plane features that were observed in the Mesoproterozoic Chorhat Sandstone are biological and can be interpreted as the burrows of wormlike undermat miners (that is, infaunal animals that excavated tunnels underneath microbial mats). These burrows suggest that triploblastic animals existed more than a billion years ago. They also suggest that the diversification of animal designs proceeded very slowly before the appearance of organisms with hard skeletons, which was probably the key event in the Cambrian evolutionary explosion, and before the ecological changes that accompanied that event.

A. Seilacher, Geologisch-Paläontologisches Institut der Universität, 72076 Tübingen, Germany, and Department of Geology and Geophysics, Yale University, Box 208109, New Haven, CT 06520, USA. P. K. Bose, Department of Geological Sciences, Jadavpur University, Calcutta 700032, India. F. Pflüger, Kantstrasse 34, 72762 Reutlingen, Germany, and Department of Geology and Geophysics, Yale University, Box 208109, New Haven, CT 06520, USA.

There are two seemingly contradictory views about the early history of metazoans. The "Cambrian explosion" scenario is based on Cambrian shelly fossils and Burgess-type lagerstätten. This scenario suggests that animal phyla originated rather suddenly ~540 million years ago (Ma) during the Proterozoic-Phanerozoic transition. This view was first modified by the discovery of Ediacaran fossils of late Proterozoic (Vendian) age in many parts of the world. Although most of these fossils are nonmetazoan [according to the Vendobionta hypothesis (1)], the presence of true, but soft-bodied, triploblasts in Ediacaran biota is now documented by worm burrows, by radular markings and body impressions (2) of early mollusks, and by phosphatized embryos (3). These discoveries lengthened the paleontological record of animals to ~580 Ma. The alternative "slow burn" scenario suggests that animals developed more slowly (4)--according to some molecular analyses (5), beginning more than 1 billion years ago (6).

These two scenarios are not incompatible, because they are based on different concepts. In paleontology, phyla are defined as bauplans that are typified by modern representatives. Because hard skeletons are an essential element of construction in many of these bauplans, it is reasonable to assume that many of them had their origin in the Cambrian skeletonization event. Molecular analyses, in contrast, include any ancestor in the lineage, irrespective of its morphology. Also, the calculated age of the last common ancestor is highly model dependent.

Paleontologists can document the hypothesized pre-Ediacaran animals in the following ways. (i) If it is true that early metazoans were minute larviform planktics (7), their remains could be preserved in cherts; so far, however, these cherts have yielded only unicellular organisms. (ii) Larger, but soft-bodied, forms can be expected in black shales, in which only macroalgae have been found so far. (iii) Clastic sediments should preserve the trails and burrows of benthic animals. Pre-Ediacaran trace fossils [summarized in (8)], however, are scarce and have not been very convincing. The worm burrows described here suggest that a preservational bias may also be involved in the Precambrian trace fossil record.

Trace fossils are best preserved in sedimentary rocks that have not been metamorphosed. One suitable site for trace fossils is the Vindhyan Supergroup in the Son valley, central India, where continuous sequences of nonmetamorphosed Proterozoic sandstones, shales, and algal limestones can be studied in continuous outcrops (Fig. 1). The Chorhat Sandstone, from which our trace fossils originate, appears to have been deposited in a shallow marine setting. It mainly consists of centimetric storm sands that are topped by oscillation ripples. However, unlike in younger tempestites, event layers have not become cannibalistically amalgamated but remain separated even when there are no mud drapings between them. The reason for this strange behavior is the presence of microbial mats (9). These mats transformed the upper millimeters of the sand into a leathery, erosion-resistant coating, so that the next storm sand could, on its sole, form a direct replica of the previous ripples. If the second storm bed was no thicker than the preexisting ripple relief, it filled only the former troughs with starved ripples of a different direction [palimpsest ripples (10)].

  

Fig. 1. The Vindhyan Supergroup, central India, with stratigraphic context, location of metazoan burrows (stars), and published ages with references. Ages are millions of years ago and calculated by the potassium/argon method (K/Ar) or by fission tracks (F-T).

 



Fig. 2. Worm burrows from the Mesoproterozoic Chorhat Sandstone of central India support the claim of molecular biologists that triploblastic metazoans existed twice as long ago as the assumed Cambrian evolutionary explosion of modern animal phyla. A remnant of the original sand veneer cover (arrow) shows that the burrows are not modern artifacts (Yale specimen YPM 37665). (Photograph by W. Sacco)

The matground state of Proterozoic sea bottoms (as opposed to the bioturbational mixground state of later times) has not only been responsible for strange sedimentary structures but may also have led to the strange life-styles (11) and preservation (12) of Ediacaran organisms and the scarcity of trace fossils in the Proterozoic rocks.

When studying pre-Phanerozoic trace fossils, it is crucial to distinguish true biogenic sedimentary structures from pseudotraces whose shapes are purely (or mainly) formed from physical processes. Among these structures, syneresis cracks (10) are the most deceiving. They open because of early diagenetic dewatering (within the sediment rather than at the surface). In Proterozoic times, the ubiquity of mucus glue caused not only muds but also silts and sands to shrink in a cracking mode. Of the various crack patterns observed in the Chorhat Sandstone, some are particularly intriguing. Syneresis cracks tend to start as isolated fusiform fissures. After such incipient cracks had been filled with coarser sediment from the layer above, subsequent compaction caused the less compactible fillings to bulge out laterally. As a result, the casts look much like the almond-shaped resting traces of bivalves (Lockeia). Longer incipient cracks could also be mistaken for the bases of spreite burrows (Diplocraterion), if they are preserved as grooves on a sandstone top. An even more intriguing crack pattern results from the combination of syneresis with oscillation ripples. In this situation, cracks originate in the ripple troughs, whose contours they follow with a strikingly sinusoidal course. Such structures have repeatedly been described as the body casts, or traces, of early worms, particularly if they bear parapodia-like corrugations on their flanks (13). The name "Manchuriophycus" (referring to an earlier interpretation as fossil algae) is used for this sediment structure, which is particularly common in Proterozoic matground environments.

Following an earlier report (14), we found the structures described here (Figs. 2 to 4) in considerable numbers on top of sandstone beds along the road east and west of Chorhat village. Chorhat burrows are too irregular to be syneresis cracks, too sharply delineated to be wrinkles, and too large to be attributed to protists or fungal rhizoids. Also, diameters vary, but they remain constant within systems. When we first found these structures in 1996, we suspected that they were modern artifacts caused by plant roots (in which secondary branchings should be narrower) or by termites, because they occurred only on weathered surfaces. We revisited the site in 1997, hoping to find grooves that pass into a ridge (such reversal of relief commonly occurs when soft sediment is burrowed along a sand-mud interface, and this phenomenon would be impossible in modern rock perforations). Instead, we found other evidence disproving a modern origin. (i) The margins of the burrows are sometimes elevated above the surrounding surface (Fig. 3a). This could be due to the pushing up of soft sediment or to a diagenetic halo that was more resistant to weathering. (ii) In one place (Fig. 4), the burrows became visible only on the weathered rock surface; the burrows are absent in adjacent areas of the same bedding plane that had been less exposed or protected by an overlying sandstone bed. (iii) One of the burrows [arrow in Figs. 2 and 4 (inset A)] is still partly covered by a small remnant of the original sandstone veneer. (iv) Cross sections (Fig. 5) reveal that this veneer has a larger amount of dark matrix between the sand grains.

Fig. 3. Chorhat worm burrows on a weathered top surface. The patterns are unlike syneresis cracks, and (a) elevated rims provide evidence that these burrows are not modern artifacts. (b) Elevated crack fillings suggest the thickness of veneer that has been weathered away (casts in Tübingen collection, GPIT 1847/1-2). (Photograph by W. Gerber)

Fig. 4.The burrows [block diagram of Fig. 1 drawn from the outcrop situation and the cast in the Tübingen collection (GPIT 1847/3)] become visible only after the top millimeter of the sandstone bed has been weathered away. Fresh surfaces lack such burrows except for less distinct collapse structures. Therefore, we assume that the wormlike maker was an undermat miner [part of the burrow that is still covered by the original sand veneer is indicated by the arrow (see Fig. 2)]. Panels 1 to 4 show further examples, which were drawn from casts and field photographs, and insets A to D show different blocks of the bed.


Fig. 5. (bottom right). A thin section of the trace-bearing bed (Fig. 4, inset A). The grains are less closely packed (and have a larger amount of dark matrix) in the upper 2 mm, where they were microbially bound. The section broke at about the level of the burrows during preparation

This evidence demonstrates that these are original burrows, but they are in an unusual mode of preservation. We suggest that the veneer represents the microbially bound biomat, which served as a food source as well as an oxygen mask for the wormlike animals that were exploiting its decaying base. They probably wedged themselves through the sediment by peristalsis, because the burrows are not actively backfilled. The open tunnels later collapsed, as indicated by faint depressions on the unweathered bed surface (Fig. 4, inset D).

Often such undermat burrows cannot be seen on unweathered sandstone surfaces, where trace fossils would be best preserved in later rocks. If the biomat veneer had the same structure as the underlying sand, they might not even weather out.

The Chorhat burrows, unlike other claimed trace fossils of similar age (8, 15), can only be explained as the works of macroscopic wormlike animals. Because they always follow a bedding plane, which is only a few millimeters below the original surface, we suppose that their makers were undermat miners. From the common occurrence of branchings, and of occasional collapse structures (Fig. 4, inset D), one may assume that the tunnels did not become actively backfilled. Apart from branching and tier-keeping, their makers had not yet developed the systematic search behaviors (meandering and so forth) that one observes in later undermat miners. Burrowing was probably performed by peristaltic movements, which would suggest the presence of a coelom. Because pore water below biomats tends to be in a reduced state, we must also consider that these animals used the mat not only as a food source but also as an oxygen mask.

In any case, these burrows suggest the existence of fairly large (up to 5-mm-wide) triploblastic metazoans as early as 1.1 billion years ago, in agreement with molecular data. The burrows also indicate that animal body plans changed very little before the explosive emergence of new designs in the Precambrian-Cambrian transition and before the onset of a more competitive style of Darwinian evolution.

REFERENCES AND NOTES

  1. A. Seilacher, J.Geol. Soc. London 149, 609 (1992).
  2. M. A. Fedonkin and B. M. Waggoner, Nature 388, 868 (1997) .
  3. S. Bengtson, ibid. 391, 529 (1998) ; S. Xiao, Y. Zhang, A. H. Knoll, ibid., p. 553; C.-W. Li, J.-Y. Chen and T.-E. Hua, Science 279, 879 (1998) [Abstract/Full Text].
  4. R. A. Fortey, D. E. G. Briggs, M. A. Wills, Biol. J. Linn. Soc. 57, 113 (1996) .
  5. G. A. Wray, J. S. Lewinton, L. H. Shapiro, Science 274, 568 (1996) [Abstract/Full Text].
  6. F. J. Ayala and A. Rzhetsky, Proc. Natl. Acad. Sci. U.S.A. 95, 606 (1998) [Abstract/Full Text].
  7. E. H. Davidson, K. J. Peterson, R. A. Cameron, Science 270, 1319 (1995) [Abstract].
  8. T. P. Crimes, in The Paleobiology of Trace Fossils, S. K. Donovan, Ed. (Wiley, New York, 1994), pp. 105-133.
  9. A. Seilacher and F. Pflüger, in Biostabilization of Sediments, W. E. Krumbein, D. M. Paterson, L. J. Stal, Eds. (Bibliotheks- und Informationssystem der Carl von Ossietzky Universität Oldenburg, Oldenburg, Germany, 1994), pp. 97-105; F. Pflüger, Neues. Jahrb. Geol. Palaeontol. Abh. 195, 75 (1995).
  10. A. Seilacher, Fossil Art (Royal Tyrell Museum of Paleontology, Drumheller, Alberta, Canada, 1997).
  11. ___, Geol. Soc. Am. Abstr. Progr. 32, 193 (1997).
  12. J. G. Gehling, Mem. Geol. Surv. India 20, 181 (1991).
  13. H. J. Hofmann, Bull. Geol. Surv. Canada 189, 146 (1971).
  14. S. Sarkar, S. Banerjee, P. K. Bose, Neues Jahrb. Geol. Palaeontol. Monatsh. 7, 432 (1996).
  15. The corrugated structures from the Koldaha Shales (14) may well be impressions of segmented algae (A. H. Knoll, personal communication).
  16. A. P. Vinogradov et al., in Absolyutnoye Datirovanie. Tektono-Magmaticheskih Siklovi Etapov ordeneriya po dannim 1964 (Nauka, Moscow, 1966).
  17. A. P. Srivastava and G. Rajagopalan, J. Geol. Surv. India 32, 527 (1988).
  18. A. I. Tugarinov, L. L. Shanin, G. A. Kazakov, M. M. Arakelyanta, Geokhimiya 6, 652 (1965).
  19. A. P. Vinogradov and A. I. Tugarinov, Proc. Int. Geol. Congr. 10, 652 (1964).
  20. G. Rajagopalan and P. K. Maithy, Group Discussion on Vindhyans (Jadavpur University, Calcutta, 1993), pp. 35-36.
  21. A. P. Srivastava and G. Rajagopalan, in Nuclear Tracts, K. K. Sharma, Ed. (Government of India, New Delhi, 1986), pp. 41-52.
  22. A. P. Srivastava, G. Rajagopalan, K. K. Nagpaul, Indian J. Earth Sci. 12, 89 (1985).
  23. Trips to India were made possible by the Crafoord Prize (1992) of the Swedish Academy and through support by Coy Squires (Houston). We also thank E. Klitzsch, H. Luginsland, S. Sarkar, and E. Seilacher for assistance in the field. Photographs were taken by W. Gerber (Fig. 3) and W. Sacco (Fig. 2), and word processing was completed by E. Seilacher.

24 April 1998; accepted 29 July 1998