PLAIN LANGUAGE SUMMARY:
The character of palaeotology has changed considerably in the past 25 years. Much more emphasis is now placed on investigating patterns in the fossil record using numerical and statistical techniques to consider such topics as rates of evolution in the fossil record. Description of fossils is still an important part of palaeontology, but has not been subject to such radical changes. The advent of the microcomputer has greatly expanded the ability of palaeontologists to handle large amounts of data and ask very different questions from the qualitative (non-numerical) ones that dominated palaeontology before.
This paper presents a computer program called ADAPTS (Analysis of Diversity, Asymmetry of Phylogenetic Trees, and Survivorship) that was developed to analyse stratigraphic range data, which is the raw material that palaeontologists use to ask questions about rates of evolution in the fossil record. The stratigraphic range of a group (taxon) is the time from when it first appears in the fossil record (the oldest occurrence) until the last time it appears (the youngest occurrence). A rate is simply the amount of change in a given period of time. So, if you look at how the number of groups (taxa) alive every million years changes over fifty million years, you are considering changes in the diversity. Similarly rates of origination and extinction can be studied with the same sort of approach.
Survivorship is a way of looking at how long taxa live for in the fossil record. To do this a life table is required. To construct a life table you must first decide on the age class size that you will use. An age class is just a length of time, say 0-5 years old in humans. You would then see how many people fell into that particular class, and all the other classes from 0-5 upwards. Then you would see how many people died in each age class and this allows you to get some idea of the probability of an individual dying, or living to a certain age.
For fossil taxa the same sort of thing can be done, only across entire species and millions of years. The shape of the curve that appears when a life table is made into a graph can also distinguish whether taxa in a group are more or less likely to become extinct as they get older.
If a phylogenetic tree is available for the group, then aspects of that tree can also be analysed. A phylogenetic tree is a bit like a family tree, but instead of human relatives in a family the branches are taxa, such as genera of dinosaurs. It has been proposed that differential rates of extinction and origination in the fossil record will be recorded as asymmetry in the shape of the tree. One part of the tree will have many more branches than the other, generating a "lop-sided" tree. ADAPTS has a test which examines whether such unbalanced patterns are the results of unbalanced numbers of originations and/or extinctions and whether there are biases for parent or offspring taxa to preferentially go extinct first or give rise to new species. If evolution was "random" in this respect then a 50:50 split would be expected and the tree would be balanced.
By bringing all of these methods together in one package that interfaces with spreadsheets, this program should be of use to palaeontologists in general, many of whom are interested in rates of evolution. By automating these routines hopefully a lot of time consuming, and frankly boring, calculations can be done easily with this software tool. It also means that it is possible to carry out calculations across different intervals of time, say looking at rates for every two million years, rather than every one million, which can be used to check if the patterns are robust or just caused by timescale effects. The fact that several less used methods are available in the package may also serve to encourage workers to use tests that they have not applied before, due to the lengthy calculations involved, which could lead to the discovery of new patterns in the fossil record.
Alistair J. McGowan, Department of Geophysical Sciences, Henry
Hinds Laboratory, University of Chicago, 5734 South Ellis Avenue,
Chicago, Illinois. IL 60637. U.S.A.
Paul N. Pearson, Department of Earth Sciences, University
of Bristol, Wills Memorial Building, Queens Road, Bristol, BS8
1RJ, U.K.