## Table 2 |
|||||

Statistical support for alternative hypotheses on Stenodactylus phylogeny |
|||||

ML framework^{1} |
Bayesian framework^{2} |
||||

Tree |
-log likelihood |
AU P |
SH P |
HME |
log_{10 }BF |

Unconstrained tree | 15180.095955 | −15175.4907 | |||

Monophyly of Stenodactylus+Pseudoceramodactylus |
15180.877129 | 0.461 | 0.839 | −15175.0633 | 0.647 |

Monophyly of Tropiocolotes |
15183.765511 | 0.161 | 0.495 | −15175.6215 | −0.530 |

Monophyly of Stenodactylus+Pseudoceramodactylus and Tropiocolotes |
15183.696084 | 0.153 | 0.492 | −15177.9132 | 1.589 |

Monophyly of African species | 15192.711115 | 0.029 |
0.123 | −15190.3457 | 7.221 |

Monophyly of S. petrii |
15189.220967 | 0.036 |
0.210 | −15181.4116 | 2.578 |

All topological tests are done versus the unconstrained (best) tree. Values in bold indicate statistically significant results.

^{1}ML: Maximum likelihood; AU: Approximately Unbiased test (Shimodaira, 2002); SH: Shimodaira
& Hasegawa (1999) test. *P* < 0.05 suggests that the two solutions are significantly different.

^{2}HME: The harmonic mean of sampled likelihoods as estimated by Tracer. BF: Bayes Factor.
A log_{10} Bayes factor > 2 indicates decisive evidence for statistically significant difference
between solutions.

Metallinou * et al.*

Metallinou * et al.* *BMC Evolutionary Biology* 2012 **12**:258 doi:10.1186/1471-2148-12-258