Supertrees, Components and Three-Item Data

Abstract

Supertree construction is explored from the perspective of binary and threeitem data. Binary data (components) code groups and subgroups, three-item data code relationships. Data are corroborative, consistent or conflicting. For either components or three-item statements, corroborative data support the same group; consistent data support different, non-conflicting groups; and conflicting data suggest alternative solutions. Binary data, when analyzed using parsimony, are unable to resolve simple cases of conflict. The lack of resolution is a result of the nature of the data and the peculiarities of “optimization”. Three-item data resolve most cases of conflict. Problems in systematic data analysis might be improved by investigations relating to the data, their meaning and their representation rather than exploring more methods.