Results from nested model testing as implemented in “Load-trees” mode of IMa for the divergence of blue-footed and Peruvian boobies.

Test one was between the full six parameter model of divergence (θ1, θ2, θA, m1, m2, t) and the strict allopatric model of divergence (θ1, θ2, θA, m1 = 0, m2 = 0, t). Test two was between the full, six parameter model of divergence and a model of divergence without gene flow from blue-footed boobies into Peruvian boobies (θ1, θ2, θA, m1, m2 = 0, t). Critical χ2 value = 5.99. Significance (P<0.05) indicates that the reduced model can be rejected in favour of the model of divergence with gene flow.