Table 2

Bayes factors and Akaike weights reveal differences in model fitness for the different partitioning strategies applied to the concatenated, multilocus dataset
M1↓ a2ΔlnB10 Akaike Weight
M0→ MB1 MB2 MB3 MB4 MB5 MB6 MB7 MB8 MB9
MB1 --- 412.11 700.90 2371.08 2584.86 2462.04 2818.59 603.91 891.39 0.000
MB2 −412.11 --- 288.79 1958.97 2172.74 2049.93 2406.48 191.80 479.28 0.000
MB3 −700.90 −288.79 --- 1670.18 1883.96 1761.14 2117.70 −96.99 190.49 0.000
MB4 −2371.08 −1958.97 −1670.18 --- 213.77 90.96 447.51 −1767.17 −1479.69 0.000
MB5 −2584.86 −2172.74 −1883.96 −213.77 --- −122.82 233.74 −1980.95 −1693.46 1.000
MB6 −2462.04 −2049.93 −1761.14 −90.96 122.82 --- 356.56 −1858.13 −1570.65 0.000
MB7 −2818.59 −2406.48 −2117.70 −447.51 −233.74 −356.56 --- −2214.68 −1927.20 0.000
MB8 −603.91 −191.80 96.99 1767.17 1980.95 1858.13 2214.68 --- 287.48 0.000
MB9 −891.39 −479.28 −190.49 1479.69 1693.46 1570.65 1927.20 −287.48 --- 0.000

aPositive Bayes factors (2ΔlnB10) support model M0 over model M1 and negative values support model M1 over model M0. Bayes factor support values >10 are shown in bold.

Lamers et al.

Lamers et al. BMC Evolutionary Biology 2012 12:171   doi:10.1186/1471-2148-12-171

Open Data