Cladistics and Parsimony

Cladistic parsimony chooses the evolutionary tree that explains the observed characters with the fewest changes, treating homoplasy as the cost to be minimized.

Definition

Maximum parsimony is an optimality criterion that selects the phylogenetic tree minimizing the total number of character-state changes required to explain the data.

Scope

This topic covers the parsimony optimality criterion, the counting of character-state changes (tree length), search strategies for finding shortest trees, consensus methods for summarizing multiple equally parsimonious trees, and the assumptions and criticisms of parsimony as a basis for phylogenetic inference.

Core questions

• How is the length of a tree under parsimony calculated?
• How are shortest trees found when exhaustive search is impossible?
• How are multiple equally parsimonious trees summarized?
• Under what conditions can parsimony be misled?

Key theories

Minimum-change principle

Parsimony prefers the tree requiring the fewest evolutionary changes, on the rationale that ad hoc hypotheses of homoplasy should be minimized.

Long-branch attraction

Parsimony can be statistically inconsistent, grouping rapidly evolving lineages together because of convergent change rather than shared ancestry, a failure mode known as long-branch attraction.

Clinical relevance

Parsimony-based trees remain widely used for morphological and combined datasets, including fossil-bearing analyses, that inform our understanding of trait evolution relevant to comparative biology and biomedicine.

History

Parsimony became the dominant computational method of the cladistic era, implemented in widely used software from the 1980s onward; Felsenstein's demonstration that parsimony can be inconsistent under unequal rates spurred the rise of model-based alternatives.

Debates

Whether parsimony or likelihood is preferable

Advocates of parsimony emphasize its minimal assumptions and transparency, while critics point to its statistical inconsistency under heterogeneous rates and favor explicitly model-based methods.

Key figures

• Willi Hennig
• Joseph Felsenstein
• James Farris

Seminal works

• hennig1966
• wiley2011
• felsenstein2004

Frequently asked questions

Why minimize the number of evolutionary changes?

Parsimony treats each extra change as an additional ad hoc assumption of homoplasy, so the shortest tree is taken to be the best-supported explanation of the data.

What is long-branch attraction?

It is an artifact in which two unrelated, fast-evolving lineages are grouped together because chance convergent similarities are mistaken for shared ancestry, a known weakness of parsimony.