Fig. 2

Bayesian 50% majority-rule consensus tree based on concatenated COI and 16S sequence data (left topology); nodes displaying the posterior probability (PP) values (top) from the Bayesian inference (BI), bootstrap support (BS) values from the maximum parsimony (MP) analysis (middle) and from the maximum likelihood (ML) analysis (bottom). Only nodes with PP values ≥0.95 and/or BS values from ML and/or MP analyses ≥70% are annotated. Columns to the right of the mitochondrial tree represent reproductive mode found in an individual or within the population of the sequenced individual (left column) and assumed reproductive mode based on area of origin (right column). Empty lines indicate absence of gravid females in the population. Abbreviations: V = veliger stages, VP = population found with veliger stages, J = juveniles, JP = population found with juveniles. Right topology: ancestral range estimation, with pie charts at nodes representing the estimated proportions for each area of origin. Colour codes correspond to areas depicted in the map: A = Southern West Indian Ocean, B = Red Sea, C = Northern West Indian Ocean, D = Northern Indo-West Pacific, E = Australia and New Caledonia. A combination of letters represents a shared probability of ancestral range from the corresponding regions. Abbreviations for countries of sample origin: AUS = Australia, EGY = Egypt, FIJ = Fiji, IND = Indonesia, YEM = Yemen, MAU = Mauritius, MAD = Madagascar, MAY = Malaysia, MOZ = Mozambique, NEC = New Caledonia, OMA = Oman, PAN = Panama, THA = Thailand. Map made with Natural Earth. Free vector and raster map data @ naturalearthdata.com; for details refer to methods section