Table 3.
Ancestral state reconstructions (ASRs) for presence or absence of P1, presence or absence of P2, and molar type for Vombatiformes and selected vombatiform subclades (see Table 1 and Figs. 1, 2).
| Clade | P1 | P2 | molar type |
|---|---|---|---|
| Vombatiformes | present (p = 0.98) | present (p = 0.96) | bunolophodont (p = 0.96) |
| Thylacoleonidae | present (p = 0.99) | present (p = 1.00) | bunolophodont (p = 0.97) |
| Phascolarctidae+Vombatomorphia | absent (p = 0.98) | absent (p = 0.95) | selenodont (p = 0.88) |
| Phascolarctidae | present (p = 1.00) | absent (p = 1.00) | selenodont (p = 1.00) |
| Vombatomorphia | absent (p = 1.00) | absent (p = 1.00) | selenodont (p = 0.84) |
| Vombatoidea | absent (p = 1.00) | absent (p = 1.00) | selenolophodont (p = 1.00) |
| Vombatidae | absent (p = 1.00) | absent (p = 1.00) | selenolophodont (p = 1.00) |
| Diprotodontoidea | absent (p = 1.00) | absent (p = 1.00) | fully lophodont (p = 1.00) |
ASRs were inferred using the Mk1 (symmetrical) model in Mesquite version 3.61 on the majority rule consensus with all compatible partitions of the post-burnin trees from a Bayesian analysis of our 99 character morphological matrix under the Mkv model; the most likely optimisation is given for each clade, with the associated probability in brackets.