SEARCHING AND PRESENTING THE FAMILY DATA

These data are organized in the different ranks of taxa, habitats, and stratigraphic ranges in which the species composing the families lived, originated, and became extinct. Users can search from lists within kingdom, habitat, age, chapter number and other classifications within The Fossil Record 2. They may also plot various patterns within any of these levels, selecting the maximum or minimum extent for each family range (e.g., The Fossil Record 2, uncertainty about the reliable range is available and is shown in the retrieved data).

However, it is important to consider very carefully which rank of taxa you select for such curves because different kinds of organisms diversify at different rates and over differing scales of morphological complexity. There is also considerable variation in the taxonomic judgements of different specialists, and in how to define the level of each hierarchy from the different cultures between, say, botanists and zoologists. Even if we limit the comparison to similar habitats there are substantial differences of scale and judgement.

There is another important set of factors that may influence these kinds of results: whether family data are a good proxy for diversity at the species level. Families originate and disappear only by virtue of the origination and extinction of their first and last constituent species. Do data categorised at other taxonomic levels give similar trends? There are several dangers here, not least those concerning which taxonomic hierarchy is to be searched. Left alone, these Fossil Record 2 searches may turn this into a tautology, but the advent of molecular and morphological cladistics of modern organisms may help break that circle for groups with modern representatives. Such problems also call into question the identification of an extinction event. Is it an artifact of classification? Because of these problems, this article is offered with words of extreme caution about how you use the fossil record and our database, some aspects of which have recently been debated on the Nature website.

To introduce a little of what can be done easily, the results from three searches are plotted in Figure 1, showing the number of extinctions, and the family diversifications of different groups. The red line shows the record of saurischian dinosaur family extinctions. The blue line shows the family diversity of the Artiodactyla. The green line shows a diversity record for the grass-like families of the monocotyledon Commelinales (after "The Thorne System" in Brummitt 1992). When you have created each of these curves by clicking on the links, you can plot them together on one screen by pressing the "Add previous data" button.

Several features emerge when the diversity curves are examined and compared to one another and to other trends. For

For example, not of least significance is that total mammal family diversity reached a maximum (172 families) at the middle Miocene and that the group has shown a mean trend toward extinction since then. There is another peak, comprising 148 families, at the approximate time of the Eocene-Oligocene boundary, but the stratigraphic scales used in this database are not accurate enough to make definite correlations with La Grande Coupure in Europe (Pomerol 1973) or the time interval when plants show maximum warmth in the London and Paris basins. Even then, the global ranges of The Fossil Record 2 are difficult to compare with these regional features. The plant family data show a different reaction from the Early Eocene to the present. Their number stays constant or continues to increase, albeit at a reducing rate. This can be seen in Figure 7 from the total plant family diversity plot in The Fossil Record 2. A number of mammal orders show peaks of family diversity around the Eocene--Oligocene boundary (e.g., Soricomorpha, Rodentia, Primates, Artiodactyla and Proboscidea). It is not our purpose here to give detailed interpretations of these interactions, but rather, to demonstrate the tools we have created that enable users to search different aspects of The Fossil Record 2 database for taxa of their choice.

One of our early objectives in planning this method of analysis was to compare the rates and times of diversification of plants and animals. The sheer quantity of data available and the many tens of thousands of curves that can be created for such comparisons overwhelmed us. By providing these tools to the palaeontological community our original objective can now be accomplished in principle, by allowing palaeontologists with different interests to compare taxa across their whole range of interests in systematics, ecology, and stratigraphy.