Table 3

Tests of significance of migration and population size differences
A. linearis (1) - A. oblonga (1) - North (1) - South (2)
A. apiculata (2) A. macroclinidioides (north) (2)
Model log(P) df 2LLR P log(P) df 2LLR P log(P) df 2LLR P
θAθ1θ2m1 = m2 2.4387 1 6.4799 0.011 −3.7251 1 8.0867 0.004 4.9933 1 2.6006 0.107
θAθ1θ2m1m2=0 5.6746 1 0.008 0.929 −2.4051 1 5.4467 0.020 2.1984 1 7.7043 0.006
θAθ1θ2m1=0 m2 2.4387 1 6.4799 0.011 −1.9985 1 4.6336 0.031 5.7494 1 0.0043 0.948
θAθ1θ2m1=0 m2=0 2.4394 2 6.4784 0.039 −3.3128 2 7.2621 0.026 2.1981 2 7.7057 0.021
θAθ1=θ2m1m2 1.4532 1 8.4509 0.004 −2.6265 1 5.8896 0.015 5.5204 1 1.3799 0.240
θA=θ1=θ2m1m2 0.2299 2 10.8975 0.004 −6.3303 2 13.2971 0.001 4.156 2 1.4858 0.476
θAθ1=θ2m1=m2 1.4449 2 8.4674 0.014 −10.614 2 21.8637 0.000 4.7742 2 4.5953 0.100
θAθ1=θ2m1=0 m2=0 −5.1614 3 21.68 0.000 −145.77 3 292.18 0.000 0.9275 3 10.8436 0.013
θA=θ1=θ2m1=m2 0.178 3 11.0012 0.012 −10.805 3 22.2464 0.000 3.3505 3 4.8739 0.181
θA=θ1=θ2m1=0 m2=0 −34.912 4 81.1808 0.000 −279.85 4 560.328 0.000 −1.3107 4 18.0984 0.001
θA=θ1θ2m1m2 1.6998 1 7.9577 0.005 −3.4594 1 7.5554 0.006 4.8486 1 0.3983 0.528
θA=θ1θ2m1=m2 0.7965 2 9.7642 0.008 −4.5652 2 9.7669 0.008 4.0603 2 2.9642 0.227
θA=θ1θ2m1=0 m2=0 −34.889 3 81.1346 0.000 −270.87 3 542.386 0.000 −1.0241 3 12.4439 0.006
θ1θA=θ2m1m2 1.5984 1 8.1604 0.004 −3.9078 1 8.4522 0.004 4.6988 1 0.536 0.464
θ1θA=θ2m1=m2 0.4587 2 10.4398 0.005 −4.8235 2 10.2835 0.006 4.0149 2 3.1385 0.208
θ1θA=θ2m1=0 m2=0 −25.328 3 62.0126 0.000 −11.575 3 23.7873 0.000 1.1371 3 8.6762 0.034

16 nested models (no migration and equal population size) were compared to the full model (significant migration and different population sizes; θAθ1θ2m1m2).

“log(P)” is the posterior probability of the model given data, “2LLR” = 2 × (Log(P)nested model-Log(P)full model), “df” is the difference in number of parameters between nested and full model, and the P-value is the probability of achieving the test statistic (2LLR) by chance under the null model. The models with P < 0.05 represent rejection of the models and they are indicated in boldface. θ1, θ2, θA are effective population sizes of population (1), population (2), and ancestral population, respectively. m1 and m2 are the gene flow from (2) to (1) and (1) to (2), respectively.

Mitsui and Setoguchi

Mitsui and Setoguchi BMC Evolutionary Biology 2012 12:254   doi:10.1186/1471-2148-12-254

Open Data