Table 2.

Molecular clock estimations for each geographically isolated lineage of Collembola in SVL, the QAR, and the QMM

	Geographical separation			Posterior age estimates (3.54% My−1)		5- to 3-My calibration
				Divergence time (My)	95% HPD	ucldMEAN	SD	Divergence rate My−1, %
A. subpolaris (3)
	Beardmore Gl.	vs.	Shackleton Gl.	8.48	6.24 to 11.06	0.024	0.005	4.8
	Shackleton Gl. West	vs.	Shackleton Gl. East	5.27	3.76 to 7.01	0.016	0.004	3.2
G. hodgsoni (4)						0.0153	0.005	3.06
	SVL	vs.	QAR	3.52	2.56 to 4.64
	Towle Gl.	vs.	All other SVL	2.78	2.00 to 3.66
	Mt. Gran	vs.	Widespread coastal	2.20	1.49 to 3.03
C. nivicolus (3)						0.0089	0.003	1.78
	Towle Gl.	vs.	All others	1.69	1.14 to 2.30
	Springtail Pt.	vs.	Mackay Gl. coastal	1.29	0.79 to 1.84
A. monoculata (2)						0.0184	0.006	3.68
	North SVL	vs.	South SVL	5.54	3.88 to 7.48

Divergence times ±95% HPD are estimated utilizing the mitochondrial clock rate of 3.54% divergence per My (27). Clock rates were then separately estimated based on calibration to the 5- to 3-My period of WAIS collapse. All estimations were performed on the alignment containing one representative sequence (658 bp) for each lineage (number of lineages is shown in parentheses), implemented in BEAST 2.4.7.