Journal of Mammalian Evolution

, Volume 2, Issue 2, pp 133-152

First online:

The pitfalls of molecular phylogeny based on four species, as illustrated by the Cetacea/Artiodactyla relationships

  • Hervé PhilippeAffiliated withLaboratoire de Biologie Cellulaire (URA CNRS 1134 D), Université Paris-Sud
  • , Emmanuel DouzeryAffiliated withLaboratoire de Paléontologie, Institut des Sciences de l'Evolution (URA 327 CNRS), USTL

Rent the article at a discount

Rent now

* Final gross prices may vary according to local VAT.

Get Access


We study the reliability of phylogeny based on four taxa, when the internal, ancestral, branch is short. Such a quartet approach has been broadly used for inferring phylogenetic patterns. The question of branching pattern between the suborders Ruminantia and Suiformes (order Artiodactyla) and the order Cetacea is chosen as an example. All the combinations of four taxa were generated by taking on and only one species per group under study (three ingroups and one outgroup). Using real sequences, the analysis of these combinations demonstrates that the quartet approach is seriously misleading. Using both maximum parsimony and distance methods, it is possible to find a quartet of species which provided a high bootstrap proportion for each of the three possible unrooted trees. With the same set of sequences, we used all the available species simultaneously to construct a molecular phylogeny. This approach proved much more reliable than the quartet approach. When the number of informative sites is rather low, the branching patterns are not supported through bootstrap analysis, preventing us from false inference due to the lack of information. The reliable resolution of the phylogenetic relationships among Ruminantia, Suiformes, and Cetacea will therefore require a large number of nucleotides, such as the complete mitochondrial genomes of at least 30 species.

Key Words

molecular phylogeny quartets mitochondrial DNA Cetacea Artiodactyla