Likelihood and parsimony diverge at high taxonomic levels (Letter to the Editor)

Smith, Martin R. (2013) Likelihood and parsimony diverge at high taxonomic levels (Letter to the Editor). Cladistics, 29 (5). p. 463. ISSN 1096-0031 DOI

[img] PDF (This is a RoMEO yellow journa)
Restricted to Registered users only

Download (197kB)
Official URL:


Sir, Rindal and Brower (2011) present the surprising conclusion that likelihood and parsimony criteria produce the same “optimal” tree topologies, based on a meta-analysis of studies from Molecular Phylogenetics and Evolution. I enlarged this dataset, which is heavily biased towards low taxonomic levels, by adding studies that addressed relationships between higher taxa (Table S1). Whereas parsimony and likelihood tree topologies are often identical at species level, major differences occurred in one-third of studies above the familial level (Fig. 1). Discrepancies become larger and more common at higher taxonomic ranks (Spearman correlation coefficient: taxonomic rank versus proportion of studies reporting major differences in tree topology, 0.89; major or minor differences, 1.00).

Item Type: Article
Uncontrolled Keywords: 2013AREP; IA65;
Subjects: 04 - Palaeobiology
Divisions: 04 - Palaeobiology
Journal or Publication Title: Cladistics
Volume: 29
Page Range: p. 463
Identification Number:
Depositing User: Sarah Humbert
Date Deposited: 25 Feb 2013 16:15
Last Modified: 25 Nov 2013 17:30

Actions (login required)

View Item View Item

About cookies