Undersampling Taxa Will Underestimate Molecular Divergence Dates: An Example from the South American Lizard Clade Liolaemini

Table 6

Comparison of node age estimates obtained using three fossil calibration sampling schemes in MrBayes. Values represent mean and 95% posterior credibility intervals (in parentheses) for the age of each node. All analyses implemented the independent gamma rates (IGR) relaxed clock model, a continuous uncorrelated model. This model explained the data better than the other two relaxed clock models currently implemented in MrBayes. The prior parameter specifying the rate of increase over time in variance of effective branch lengths was drawn from either a uniform or exponential distribution as either explained the data equally well according to Bayes Factors calculated in Tracer. Clade names correspond to names in Figure 1.

Assumed model	Clade name
	Liolaemini	Phymaturus	Liolaemus (genus)	Eulaemus (subgenus)	Liolaemus (subgenus)
IGR-exponential; All calibrations	121.0; (102–143)	42.1; (31–55)	60.3; (50–71)	38.0; (29–46)	53.4; (44–64)
IGR-exponential; Iguania ancestor only	143.4; (121–177)	56.8; (36–77)	92.3; (76–113)	73.5; (58–91)	79.7; (64–99)
IGR-exponential; Iguania ancestor and Pristidactylus	132.9; (96–170)	50.9; (30–77)	81.9; (53–108)	63.2; (42–84)	70.7; (47–94)
IGR-uniform; All calibrations	114.7; (87–148)	39.8; (26–53)	57.3; (43–72)	36.2; (27–45)	50.5; (38–64)
IGR-uniform; Iguania ancestor only	147.5; (123–174)	57.0; (36–77)	93.2; (73–116)	73.9; (54–95)	80.9; (62–102)
IGR-uniform; Iguania ancestor and Pristidactylus	N/A*	53.2; (34–72)	86.5; (63–112)	68.2; (49–90)	74.6; (54–99)

∗Divergence times for Liolaemini were not included because the clade credibility tree did not recover this clade as monophyletic.