APPLICATION OF PHYLOGENETIC PRINCIPLES TO TESTING EVOLUTIONARY SCENARIOS: A COMMENT ON KACZMARSKA ET AL. “MOLECULAR PHYLOGENY OF SELECTED MEMBERS OF THE ORDER THALASSIOSIRALES (BACILLARIOPHYTA) AND EVOLUTION OF THE FULTOPORTULA”1

While rigorous techniques have usually been used to generate phylogenetic trees from molecular data, morphological analysis has sometimes been more informal. A recent example was a study of the evolution of the fultoportula in the diatom order Thalassiosirales ( Kaczmarska et al. 2006 ). Phylogeny was inferred using modern phylogenetic principles applied to nuclear SSU rDNA sequences, but inferences about morphological character evolution were made using noncanonical reasoning and evolutionary scenario building. The preferred hypothesis posited that marginal fultoportulae evolved from the marginal ridge of Lithodesmiales. A related hypothesis suggested that fultoportulae in the valve center were not homologous with those near the valve margin. Shared symplesiomorphies, shared homoplasies, gaps in the fossil record, and subtle morphological differences between central‐ and marginal‐area fultoportulae were offered as the primary evidence for these scenarios. The literature has demonstrated such arguments to be either irrelevant or logically weaker than inferences made under the tests of similarity, conjunction, and congruence. Five prior hypotheses about the origin and evolution of the fultoportula were examined in this study using these tests. The hypothesis that the areola evolved into the multistrutted process, which evolved into the fultoportula, was best supported.