Missing data and rhynchosaur phylogeny

Abstract

Consensus methods have been widely applied to the problem of summarising relationships that are common to a set of equally parsimonious trees. Unfortunately, the most commonly used consensus methods frequently fail to represent all such relationships. In phylogenetic palaeontology, inclusion of poorly known fossils with abundant missing data can lead to a plethora of equally parsimonious trees and increased loss of resolution/information in consensus trees. In such cases, recently developed consensus methods, and a method of eliminating taxa, termed safe taxonomic reduction, can be used to circumvent these difficulties. These methods are illustrated through their application to cladistic data for the Rhynchosauria. In particular we investigate the use of disqualifier faithful trees as a framework for representing consensus relationships and describe a method of determining the possible positions of taxa that are excluded from a disqualifier faithful tree.

Key words:

• Rhynchosauria
• consensus trees
• missing data
• parsimony
• phylogeny

Notes

Correspondence to: Mark Wilkinson, Tel.: [phone] ext. 5976, E‐Mail: [email protected].