Article: A phylogenetic classification of the graptoloids
Richard A. Fortey and Roger A. Cooper
Graptolite classification has traditionally been based upon grade groups reflecting general levels of evolutionary complexity. This has been acknowledged as unsatisfactory and the present classification, used widely in the Western world, is a hybrid between this system and what are claimed to be 'natural' groups. A phylogenetic classification, in which taxa are based upon common ancestry, produces a more objective classification with taxa that are diagnosable. We attempt a reclassification of the planktic graptolites (excluding retiolitids) using phylogenetic methods, and employ a cladistic representation of relationships to portray character distribution and decide taxonomic levels. The Graptoloidea are defined as a monophyletic group by retention of a nematophorous sicula in the adult. Hence the Anisograptidae are transferred from the Dendroidea to the Graptoloidea. The higher Graptoloidea fall into two natural suborders: the Virgellina suborder nov. and the Dichograptina Lapworth, which replace the four existing suborders. The Virgellina includes mostly scandent graptolites of monograptid, diplograptid, dicranograptid, nemagraptid, and phyllograptid groups. These are united by having a virgellar spine on the sicula. Recognition of superfamilies within the Virgellina is based upon proximal structure and development type. Seven groups at family level are recognized and it is only within these groups that thecal morphology—long used as the primary criterion for diplograptid genera—is likely to be useful for distinguishing genera. Within the Suborder Dichograptina are two super-families: the Glossograptacea is characterized by having isograptid symmetry and contains the families Isograptidae Harris, Pseudisograptidae Cooper and Ni, Corynoididae Bulman, and Glossograptidae Lapworth; the Dichograptacea contains three families, of which the Sigmagraptidae Cooper and Fortey is distinguished by its slender and asymmetrical proximal end, the Sinograptidae by prothecal morphology, while the Dichograptidae is subdivided by the progressive delay and suppression of dichotomies and some generic synonymy is recommended.Although our scheme embraces all graptoloids, problems remain with determining the interrelationships between some higher order taxa, and with the pendent graptoloids. The Anisograptidae is retained as a paraphyletic group.