. 2013 Jun 5;13:114. doi: 10.1186/1471-2148-13-114

Table 2.

Comparison of BEAST dating results employing a relaxed-clock approach with combined tip (interior) and root (exterior) calibration

Scenario	Root height prior (min.) [ka]	Median posterior substitution rate [per site and 10⁶ years]	Divergence time estimates [ka]
Scenario	Root height prior (min.) [ka]		Red / arctic fox	Red fox diversification	Nearctic lineage	North America, eastern lineage	North America, widespread lineage	North America, mountain lineage	Japan, Honshu / Kyushu	Japan, Hokkaido II
Scenario 1	1750	33.2%	1,894 (1,750–2,412)	173 (92–316)	68 (39–93)^a	42 (18–82)^a	42 (18–82)^a	52 (20–104)	24 (9–52)^a	27 (11–60)^a
Scenario 2	5100	25.6%	5,325 (5,100–5,847)	235 (118–423)	128 (73–206)^a	57 (23–121)^a	73 (28–164)	57 (23–121)^a	64 (46–94)^a	45 (22–69)^a
Scenario 3	500	41.9%	586 (500–874)	129 (78–208)	61 (30–92)^a	32 (13–61)^a	32 (13–61)^a	39 (14–74)	21 (8–51)^a	32 (13–64)^a

All three scenarios were using the same tip dates, but varying root heights as exterior calibration points (red/arctic fox divergence time estimate). Scenario 1: uniform prior for root height of 1.75-4 million years (My) according to the 95% credibility interval in Perini et al. [80] (Figure 2). Scenario 2: uniform prior for the root height of 5.1-6 My according to the 95% credibility interval in Nyakatura and Bininda-Emonds [81]. Scenario 3: lognormal root height prior based on the first appearance of the red fox in the fossil record (0.5-1 My ago) [30,31], setting the minimum age of the root height to 0.5 My. The 95% interval of the lognormal prior included a period of up to 5.9 My.

^aLess than 95% posterior support for the divergence from the next most closely related sequence, but at least 95% support for the most basal divergence within the lineage. Note also that, despite uncertainty regarding the phylogenetic placement of these groups, their inferred age was relatively similar across calibration scenarios.

ka thousand years.