Testing the quality of the fossil record by groups and by major habitats

The evolution of life is a form of history and, as Karl Popper pointed out, that makes much of palaeontology and evolutionary biology metaphysical and not scientific, since direct testing is not possible: history cannot be re‐run. However, it is possible to cross‐compare three sources of data on phylogeny ‐ stratigraphic, cladistic, and molecular. Three metrics for comparing cladograms with stratigraphic information allow cross‐testing of (1) the order of branching with the stratigraphic order of fossils, and of (2) the relative amount of cladistically‐implied gap in proportion to known fossil record.

Results of the metrics, based upon a data set of 376 cladograms, show that there are statistically significant differences in the results for echinoderms, fishes, and tetrapods. Matching of rank‐order data on stratigraphie age of first appearances and branching points in cladograms, using Spearman Rank Correlation (SRC), is poorer than reported before, with only 148 of the 376 cladograms tested (39%) showing statistically significant matching. Tests of the relative amount of cladistically‐implied gap, using the Relative Completeness Index (RCI), indicated excellent results, with 288 of the cladograms tested (77%) having records more than 50% complete. Assessment of the Stratigraphic Consistency Index (SCI), the relative number of nodes in a cladogram having younger taxa located above them than immediately below them showed that 79% of cladograms tested scored over 0.500. This indicates that node order was generally consistent with stratigraphic order of first finds, contrary to the findings of the rather cruder SRC metric.

Group‐by‐group results were variable. SRC results, based on significances of age‐clade rank order equivalence, were best for tetrapods, moderate for echinoderms, and worst for fishes. Best RCI values were for fishes, and poorer for tetrapods and echinoderms. SCI values were best for echinoderms, poorer for tetrapods, and poorest for fishes. Arbitrary scoring of the three metrics suggests provisionally that the echinoderms and tetrapods have a better fossil record than fishes.

Comparison of marine groups (echinoderms + fishes) versus continental groups (tetrapods) showed that many more continental cladograms (50%) showed significant SRC matching of age and clade order than marine cladograms (25%). Continental cladograms also yielded higher SCI values, with 87% achieving values equal to, or higher than, 0.5, while for marine cladograms, the value was 72%. Marine cladograms, on the other hand, scored better for relative completeness, with 79% showing RCI values greater than 50%, compared to 74% for continental cladograms. The aggregate results suggest that continental cladograms are better than marine cladograms by two metrics (SRC, SCI) to one (RCI).

Key words:

• Cladistics
• phylogeny
• stratigraphy
• evolution
• Popper
• fossil record