Geoscientist Online

John C W Cope* has difficulty understanding the International Commission on Stratigraphy…

The International Commission on Stratigraphy (ICS) decrees that thou shalt worship only one boundary. That is what it exists to do, on the commonsense grounds that stratigraphers everywhere should be using the same language. To draw such stratigraphical time-planes in the rocks, it has become accepted practice over the past 30 years to define them (hopefully permanently) in terms of a selected point in a reference section - a "Global Stratotype Section and Point" (GSSP). GSSPs have been used to define System, Series and Stage boundaries, and ICS hopes that all such boundaries will be agreed by the end of 2008.

It was a relatively easy matter for the Silurian-Devonian boundary (the first GSSP to be proposed and ratified by ICS, which hammered in the "golden spike" at the onomatopoeic town of Klonk (Czech Republic) at the base of the Monograptus uniformis zone¹. The considerable faunal provincialism of the Ordovician has however created a serious obstacle to such correlation, and the System has lagged behind.

The ICS Ordovician Subcommittee decided to produce a new ‘global’ standard chronostratigraphy from scratch, working from System level down to Series and then Stage. The initial proposal, to divide the Ordovician like Caesar's Gaul into three - Lower, Middle and Upper Series² - received majority support and was ratified in February 1996 with a 90% majority vote. The boundaries between these Series were to be defined by widespread (if not always cosmopolitan) species of graptolites and/or conodonts. The base of the Middle Series of the Ordovician would be drawn at the base of the Tripodus laevis conodont zone, and the base of the Upper Series at the base of the Nemagraptus gracilis graptolite zone. These boundaries received an 82% majority vote (February 1996) from the Subcommission.

It was, however, conveniently forgotten that the Silurian System (only about half the length of the Ordovician) had been successfully divided into four Series, all internationally accepted and without any subsequent problems. One could argue that with a total of c. 57 million years to play with, five or six Ordovician Series would be reasonable. This tripartite division was to prove crucial in what followed.
 

Concentrate – here comes the science bit

The next decision decreed that each Series should be divided into two Stages. The first proposed was the Darriwilian Stage, its basal GSSP being drawn at the first appearance of the graptolite Undulagraptus austrodentatus (in a section at Huangnitang, Zhejiang Province, China). So in 1997 a global chart for the Ordovician showed three Series, but only one Stage in the next column, hanging in space. The next Stage was not accepted until 1999. Its basal boundary also defined the base of the Ordovician System (at the base of the Iapetognathus fluctivagus conodont zone, Green Point, Newfoundland). This was approved in 1999 with the agreed name of Tremadocian.

The revision of the Correlation Charts for the Ordovician of the British Isles³ tried to make the divisions of the British Ordovician comparable with those in other parts of the world. The Llandeilo Series was largely subsumed as a stage within the Llanvirn Series, while the nine Stages of the Caradoc Series (its base now lowered to the base of the Nemagraptus gracilis zone, the base of the Global Upper Ordovician) were grouped into four new Stages, with former divisions retained but demoted to Substages. The hope was that, as Britain was the historical and nomenclatorial type area for the Ordovician, the revised British chronostratigraphical scheme could form the basis of a new internationally agreed primary chronology. But it was not to be.

The next step internationally was the approval of the base of the upper Stage of the Lower Ordovician Series. This was overwhelmingly accepted as the first appearance of Tetragraptus approximatus approximatus (Diabasbrottet section, Hunneberg, Sweden). But no name for the Series was forthcoming. Its base was clearly close to the base of the British Arenig Series but, under the new Series divisions, the British Arenig was partly in the Lower and partly in the Middle Series (whereas under the British scheme it formed the upper Series of the Lower Ordovician Subsystem). Thus, a new name was needed for this interval. In 2001 a GSSP for the base of the Upper Series of the Ordovician was approved at Fågelsång, near Lund, Sweden at the first appearance of Nemagraptus gracilis; but no names for any stages were proposed, and the subdivision of the Upper Series was not decided.

Even by 2004 there was little to show, except that the boundary between the Lower and Middle Ordovician Series was now to be drawn at the base of either the Baltoniodus? triangularis or the Cooperignathus aranda conodont zones. A decision has yet to be made on which is preferable.

By 2005, the Hirnantian Stage had been accepted as the upper stage of the Upper Series, with the GSSP being selected at the first appearance of the graptolite Normalograptus extraordinarius (at Wangjiawan, Yichang, China). The base of the international Hirnantian Stage was in fact higher that the British Hirnantian Stage base (named after Cwm Hirnant in mid-Wales). The British Hirnantian has not been properly defined and has hitherto been based upon the appearance of the (diachronous) Hirnantia fauna. Clearly, if this Stage is to continue to be used in Britain, it will need re-defining in the same terms as the international Stage, and the Rawtheyan Stage extended upwards to the new base.

If such a short period as the Hirnantian (< 1 Ma) could be accepted as a Stage it was apparent that a further subdivision of the Upper Series would be needed, making three Stages instead of two (as originally proposed) in the Upper Ordovician Series. This new line was to be drawn at the first appearance of the graptolite Diplacanthograptus caudatus at Black Knob Ridge, Oklahoma.

At this time, names were proposed for three of the four un-named Ordovician Stages. The upper Stage of the Lower Series was to be the Floian, while the two Stages beneath the Hirnantian were to be called Sandbian and Katian. These names have now been formally ratified4 and leave only the lower Stage of the Middle Series un-named.

Which puts us – where, exactly?

So - what have 12 years of deliberation achieved - a triumph of international cooperation? Maybe; but is it really much of a triumph? The practical difficulties of using this scheme become readily apparent and it seems clear that nowhere in the world can all the new divisions be unequivocally recognised.

The claim⁵ is that this scheme offers geologists ‘the best of both worlds, because the regional classifications remain unchanged and can be used where they work best in describing regional geology. At the same time, the global units, which are based on cosmopolitan index species, allow for precise and reliable global correlations’. But is this not an admission that the new scheme offers no advance? Geologists will continue to use the older, local schemes because they offer a familiar database and much greater precision. The new scheme is only likely to satisfy those whose attempt at a correlation is at the level of time-resolution typical of the new Global Stages - to the nearest 10 million years or so.

In most parts of the Phanerozoic, chronostratigraphical precision is steadily being improved. It reaches the highest precision with Milankovitch cyclicity, while in the Mesozoic, Jurassic ammonite horizons may allow correlations accurate to 50-100 ka ^6,7. In the Ordovician, things appear to be going backwards.

The British Ordovician is presently divided into five Series and 15 Stages. How is a global scheme of three Series and seven Stages an advance? The British Ashgill Series is divided into four Stages: the Pusgillian, Cautleyan, Rawtheyan and Hirnantian. The last is recognised in the global scheme as a Stage, but the base of the caudatus zone (defining the base of the underlying Katian — i.e. the proposed global Primary Standard Stage) correlates with a level more than half way down our Caradoc Series and over six British Stages beneath the base of the Hirnantian. Is this progress?

The table clearly shows that British Series are at the same sort of level of precision as Global Stages. The British Stages must by and large be the equivalent of any future Global Substages. But what the current British Substages' equivalent be at global level?

The ICS may have achieved a scheme that may look good internationally, but it does nothing to advance stratigraphical precision. Locally, Secondary Standard Series, Stages and Substages⁸ will undoubtedly continue to be used, not only in Britain, but throughout the world, basically because they offer a greater degree of precision, are well engrained in the literature, can be readily used in the field and are more fully understood.

The idea that the Ordovician System fell naturally into three Series doomed attempts to provide a usable Global Standard from the start. It is not possible to shoehorn nature into preconceived subdivisions. As someone once said ‘a camel is a horse designed by a committee’. I submit that the new Ordovician chronostratigraphical scheme is such a beast.

The Ordovician System and Period was the solution proposed by Lapworth⁹ (picture) in 1879 to end the problem over the position of the boundary between the Cambrian and Silurian Systems. The dispute about the boundary had made enemies out of Sedgwick and Murchison who founded the two Systems, and Lapworth’s solution was gradually adopted by most of the world. Even the word "Ordovician" provided a deft compromise, for the country of the Ordovices lay to the north of that of the Silures and south of Cambria, or north Wales.

Despite international adoption problems remained, and the Geological Survey of the UK persisted until the 1990s) to treat the oldest of the British Ordovician Series, the Tremadoc, as Cambrian – following the line of chief palaeontologist, and later Director, Sir James Stubblefield. The faunas of the Tremadoc Series are unquestionably, however, of Ordovician rather then Cambrian affinity.

Subsequent attempts at refinement by subdividing the Ordovician into Series turned increasingly to time-correlations based on guide-fossils. Thus during the latter part of the 19th Century and into the first half of the 20th, the British Ordovician was split up into six Series: in ascending order Tremadoc, Arenig, Llanvirn, Llandeilo, Caradoc and Ashgill – some of which go back to Murchison. The first four were named after Welsh type localities, the fifth after the Welsh Borderland (South Shropshire) and the final one after a locality in the English Lake District. Their faunas were systematically collected and described in a series of important papers and monographs that spanned most of the 20th Century and further subdivided into Stages as knowledge developed.

Early on it became apparent that there were considerable differences between supposedly coeval faunas from different parts of the world. With the advent of plate tectonics came the idea of "allochthonous terranes", which made it easy to understand why faunas in, for instance, the North-west Highlands of Scotland, differed from those in Girvan (Southern Uplands), and why these, in turn, differed from those of England and Wales. In other countries separate schemes of Ordovician Series were built up. In North America the System is divided into Ibex, Whiterock, Mohawk and Cincinnati Series; while different (but chronologically parallel) schemes are used in Scandinavia and China because of faunal differences.

Nevertheless it remained possible to correlate from one faunal province to another by using the planktic graptolites and conodonts. Admittedly these correlations were approximate; but enough was known to recognise, for instance, that the base of the North American Ibex Series was older than that of the Tremadoc. In other words, for practical convenience, a different Cambrian-Ordovician boundary was in use in North America from that recognised in Britain.

References cited

1. McLaren, D. J. 1977. The Silurian-Devonian Boundary Committee. A final report. In Martinsson, A. (ed.) The Silurian-Devonian Boundary. Schweizerbart, Stuttgart, 1-34.
2. Webby, B. D. 1994. 1994. Chairman’s Report. Ordovician News, 11.
3. Fortey, R. A., Harper, D. A. T., Ingham, J. K., Owen, A. W., Parkes, M. A.,Rushton, A. W. A. & Woodcock, N. H. 2000. A revised correlation of Ordovician rocks in the British Isles. Geological Society, London, Special Reports, 24, 83 pp.
4. Bergström, S. M., Löfgren, A. & Maletz, J. 2004. The GSSP of the Second (upper) Stage of the Lower Ordovician Series: Diabasbrottet at Hunneberg, Province of Västergötland, southwestern Sweden. Episodes, 27, 267-272.
5. Bergström, S. M., Finney, S. C., Chen Xu, Goldman, G. & Leslie, S. A. 2006. Three new Ordovician global stage names. Lethaia, 39, 287-288.
6. Cope, J. C. W. 1993. High resolution biostratigraphy. In Hailwood, E.A. & Kidd, R.B. (eds) High resolution stratigraphy. Geological Society, London, Special Publications, 70, 257-265.
7. Callomon, J. H. 1995. 1995. Time from fossils: S.S. Buckman and Jurassic high-resolution geochronology. In: Le Bas, M. J. (ed.) Milestones in geology. Geological Society, London, Memoirs, 16, 127-150.
8. Cope, J. C. W. 1996. The role of the Secondary Standard in stratigraphy. Geological Magazine, 133, 107-110.
9. Lapworth, C. 1879. On the Tripartite Classification of the Lower Palaeozoic rocks. Geological Magazine, 6, 1-15