The Great Fossil Enigma (30 page)

Sweet knew that more and better data were required if provincialism was to be firmly established. There were, after all, only three complete conodont sequences known for the Ordovician, all in North America. He was, however, never one to give up on a good idea, and soon help arrived in the form of the meticulous Stig Bergström and Cooper's extraordinary Pratt Ferry residues. Bergström's head was already full of Swedish conodonts and he was intent on studying the conodonts of the Appalachians, suspecting that they might be familiar to him. Back in 1926, U.S. Geological Survey workers had discovered that rocks in the eastern trough of the Appalachian Valley contained fossils with a curiously European aspect. No one had looked at the conodonts. Cooper's Pratt Ferry material came from the southern end of the Appalachians, and in it Bergström had no difficulty finding conodonts he knew from Sweden that had never previously been found in North America. He thought this astonishing, as it was now possible to locate two sections of identically aged rocks in the United States, just twenty kilometers apart, that could not be correlated on the basis of conodonts. Yet one of these sections could be precisely correlated with rocks ten thousand kilometers away.
⁹
Could there be more dramatic proof of provincialism?

It was, however, the follow-up paper – Bergström and Sweet's 1966 coup de grâce, discussed in the
last chapter
, which sought to end old ways and, in some senses, re-establish a biological animal – which really showed the practical benefit of thinking about provincialism.
¹⁰
The quarter of a million fossils they isolated in that study told them that the midcontinent fauna was distributed in a certain way. They could now detect provincial encroachments in great detail and utilize the relative abundance of species to powerful stratigraphic effect. Their data suggested three partially open and shifting subprovinces, each perhaps controlled by water temperature and depth. As a result, a species dominant in one area might at the same time be rare in another. Across a network of geological sections, the European fauna could be seen to appear suddenly at different locations at different times – an invading population of animals, whose geographical advance could be read in detail in the sequential appearance of individual species. And here, of course, Bergström and Sweet were writing about
animals
rather than isolated fossils, as they were discussing their new statistical – or biological – species. The picture they created was not so much one of “herds” of conodont animals sweeping across an underwater equivalent of the African savannah as one in which the expansiveness and position of that savannah was shifting over time.

This changing complexity meant one could not identify a species at two different locations and say they represented identical moments in time, as stratigraphers might wish. Nevertheless, Bergström and Sweet saw in their data a way to use these long-ranging species as stratigraphic markers, and here they found another use for pollen expert Aureal Cross's relative abundance charts. The charts, which plot the relative abundance of particular species of plant, represented by pollen and found particularly in postglacial peat, were used to map environmental change since the Ice Age. Once this pattern of change was established through repeated study, it became possible to use it as a relative timescale. Bergström and Sweet took this idea and applied it using the ubiquitous, and often dominant, midcontinent species,
Phragmodus undatus.
The remains of this animal were found in a variety of rock types, suggesting that it was unaffected by environment and thus useful for correlation. So they plotted its relative abundance at every level in every section. The results showed similar, though by no means identical, logs for all sections. This permitted geological sections to be correlated with each other and with sections further afield. Bergström and Sweet also used the local histories of faunal incursions and fluctuations, recorded in these rock sequences, to define local time units. They could not, however, given the long-ranging nature of the species available to them, establish formal stratigraphic zones.

We can now see how fundamentally Bergström and Sweet's remarkable paper challenged everything to which Ziegler aspired just at that moment when he had achieved global success. Not only did it suggest he needed to document the evolution of assemblages, rather than isolated elements, but it also put into question the very notion of the universal conodont.

Bergström and Sweet continued to develop increasingly sophisticated and subtle interpretations of provincialism in their animals. Soon they were “tracing and matching faunal ‘tongues' that represent the shifting of provincial and subprovincial boundaries in time.” In his paper with Kohut, Sweet could talk of
Phragmodus undatus
retreating northwards while other conodonts were “more tolerant” of change and stayed put. Increasingly he began to think about environments, and as he did so, particular conodont species acquired lifestyles. One, for example, “seems to have flourished in a nearshore, shallow-water environment, perhaps on a tidal mud flat that was periodically exposed to the atmosphere.” In contrast
Phragmodus undatus
was “an inhabitant of deeper waters.” Gerald Webers, influenced by the Ohio workers, also began to think along these lines. In Ohio, at least, and in the Ordovician in particular, the universal conodont was dead.
¹¹
Necessity had also spawned a rather different approach to stratigraphy. There were no neat time markers here. Rather, time seemed to be marked by the very ebb and flow of life.

In the language of the 1970s, Sweet and Bergström were applying a generalized conceptual “model” – that of the province – and using its perceived attributes to interpret their data. It was a model so enshrined in geological practice that few questioned it. However, at a London conference in 1969, it became apparent that this universal concept was far from concrete. And as the delegates at the meeting deliberated its meaning, so Peter Sylvester-Bradley became increasingly depressed: “I am afraid that we can claim to have answered no problems in this symposium. Quite the reverse. We have dredged up from the stores of knowledge many old problems that had been shelved away and almost forgotten.” Those who had organized the conference had to agree: There was no agreement on what a province was, how it could be recognized, or even whether such things existed in the geological record.
¹²
This did not invalidate Bergström and Sweet's interpretation, but it did mean that the province remained hypothetical.

At that time, the descriptive natural sciences were adjusting to the logic of the computer and the imagined objectivity of numbers. Its world was about to be redrawn in systems diagrams of the type used by programmers and systems analysts in the computer industry. Among those taking a lead was another of Sweet's former students, Tom Schopf, whose
Models of Paleobiology
performed as an evangelical tract for this new way. Schopf argued for a retreat from short-sighted realism and precision, suggesting that generalized theoretical models should be developed to predict, and be tested by, data. He was then at the University of Chicago, and was the latest in a line of academics there who sought to lift paleontology out of those habits of which Ager had been so critical.
¹³

The conodont, which in Bergström's and Sweet's minds seemed to exist as a great underwater swarm of expanding, contracting and shifting life, was then being captured by other modelers who made the animal's distribution rather more structured. The origins of this close-up view of conodont distribution lay in the paper Glenister and Klapper wrote on conodonts from the Canning Basin in Australia, which implanted Ziegler's Upper Devonian zonation there. When Glenister and Klapper came to make this correlation, they drew upon a recent study of the basin by Phillip Playford and D. C. Lowry of the Geological Survey of Western Australia. In order to understand these complex rocks, Playford and Lowry developed a model suggesting that the rocks there represented four different reef environments: the back-reef, reef, fore-reef, and inter-reef. To Glenister and Klapper, the distribution of the Australian conodonts suggested they were ecologically controlled, for, like the cephalopods, these fossils were found in the fore – and inter-reef areas but were rare in the reef itself and absent from the back-reef. They were also rare in beds containing brachiopods.
¹⁴

This understanding of ecological control became more specific when English emigrant Ed Druce investigated the conodont fauna of the Bonaparte Gulf Basin in the far north of Australia. Druce was a veteran of the field: “He enjoyed the subtle beauty of the outback and camping appeared to be part of his nature. His mobile conodont lab and field processing of samples meant that field seasons could easily be up to 3 months long. As long as the tea, beer and meat were there, the work would go on.” In the Bonaparte, he found conodont faunas restricted to particular parts of the Devonian reef complex. And although Glenister and Klapper had reported no conodonts from the back-reef, Druce found a fauna there dominated by
Pelekysgnathus
and other conodonts not utilized in Ziegler's standard. The fore- and inter-reef were in contrast populated with high numbers of
Palmatolepis, Polylophodonta
, and
Scaphignathus.
Here
Pelekysgnathus
was absent.
¹⁵
Although these findings were present to some degree in his published data, it was only later that he gave these distinctions clarity. Not long after completing this work, Druce entered the Canning Basin, collecting material from the Devonian reef complex for a doctorate supervised by Frank Rhodes, who had recently moved to the University of Michigan.

As Druce's study of the Bonaparte Gulf went to press, George Seddon was completing a four-year consultancy for
WAPET
on the Canning Basin.
¹⁶
Seddon was unusual in being a member of the Departments of Geology and Philosophy at the University of Western Australia. He would later be celebrated for the range and significance of his work, but in 1970, his mind was on conodonts.

Both Druce and Seddon sought to resolve the outstanding strati-graphic problems of a region that could be linked to Europe but within which many rocks remained uncorrelated because of complexities presented by the presence of reefs. Seddon's aim was not simply to superimpose Ziegler's system on these Australian rocks but to consider the significance of those conodonts that had gone unmentioned in the German scientist's work. Seddon soon came to understand, as Glenister and Klapper had before him, that the distribution of conodonts was controlled by the different environments that made up a reef. Like Druce, but independently, he found two distinct, environmentally controlled, faunas characterized by different form genera. One fauna was typified by Branson and Mehl's favorite,
Icriodus
, together with
Polygnathus
and
Pelekysgnathus.
The other was dominated by Ziegler's most useful
Palmatolepis
, along with
Ancyrodella
and
Ancyrognathus.
¹⁷
The
Palmatolepis
fauna contained examples of the
Icriodus
fauna but in lower abundance. With few exceptions, the
Icriodus
fauna did not include those key genera from the
Palmatolepis
fauna. Seddon imagined the existence of a one-way filter through which
Icriodus
and associated forms could pass but through which
Palmatolepis
could not, though at the time he did not identify what that filter might be. The
Icriodus
fauna was found in the fore-reef but close to the reef itself, while the
Palmatolepis
fauna occurred seawards, from the fore-reef to the inter-reef.
¹⁸
Palmatolepis
was, of course, the key to Ziegler's universal standard, but
Icriodus
was not. However, Seddon's discovery of a filtering mechanism enabled him to relate the two and determine the age of his rocks. To achieve this he created a parallel
Icriodus-based
chronology recognizing that this represented a rather specialist environment. Seddon's paper was published in 1970. It acknowledged Druce's helpful critique of his work, though it seems possible that Seddon knew little of Druce's thinking.

In May that year, both Seddon and Druce presented their findings at the Michigan meeting of the Pander Society. Druce could now talk generally about the depth control of his two faunas and locate parallel examples in the lowest Carboniferous (
figure 9.1a
).
¹⁹
The “exotic and bizarre,” he noted, were restricted to deeper waters but had evolved from shallow-water stock. These deeper-water forms were generally more diverse, and it was for this reason they had been so useful to Ziegler and Helms. Druce also noted the same one-way mixing of faunas. That Ziegler's system was again called into question was not lost on the audience, particularly on Ziegler himself. Druce advised his colleagues to record
Icriodus
and other shallow-water forms in order to counter these weaknesses in Ziegler's standard.