Fig. S1 Posterior distributions and PPD from IM analysis of the GOR and CLE populations. (A) Marginal posterior probability distributions (MPD) and joint posterior probability distributions (JPD) of effective population sizes, &thgr;i, the proportion of immigrants, mi, and the effective number of immigrants, Mi, in population i. Effective population size was about three times greater in GOR (&thgr;GOR) than CLE (&thgr;CLE), Pr(&thgr;S > &thgr;N|X) = 0.92, P.O.R. = 12:1. The proportion of immigrants in GOR (mGOR) was six times greater than CLE (mCLE), Pr(mS > mN|X) = 0.93, P.O.R. = 14:1. More than 25 times more immigrants arrive in GOR than CLE each generation, Pr(MS > MN|X) = 0.97, P.O.R. = 32:1. (B) PPD calculated from the island model calibrated with COI using only samples collected from the two populations that flank the Blanco Transform Fault (CLE, GOR). Arrows and subscript ‘obs’ indicate observed values computed from the data. Summary statistics: &pgr;, theta estimated from the average number of pairwise differences; &thgr;, Watterson's theta; D, Tajima's D; FST, Wright's FST. S, summary statistics computed from GOR; N, summary statistics computed from CLE. P values are the proportion of the PPD that are more extreme than the observed value (see Methods for details).

Fig. S2 Posterior distributions from MIGRATE-N analysis. Posterior distributions on the diagonal are estimates of theta, whereas posterior distributions off the diagonal are estimates of immigration rates (mij is migration from population i into j forward in time). Blank boxes are parameters that were not estimated (i.e. constrained to be zero). Posterior distributions for demographic parameters estimated from the southern populations (GOR and JDF) were insensitive to the prior parameterizations (&thgr;MAX = 0.10, mMAX = 15 000). Estimates of demographic parameters from the AXI and EXP populations were sensitive to the prior parameterizations, and should not be trusted since we have no prior information on these parameters. However, given the stability of the immigration estimates for different prior parameterizations in the Blanco Transform Fault region, we are confident that our inferences for these parameters are not greatly affected by the prior information contributing to estimates of the other demographic parameters.

Table S1 Summary statistics and Wright–Fisher model tests

Table S2 AMOVA tables*

Please note: Blackwell Publishing are not responsible for the content or functionality of any supplementary materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.

MEC_3609_sm_FigS1.eps1343KSupporting info item
MEC_3609_sm_FigS2.eps3643KSupporting info item
MEC_3609_sm_TableS1-S2.doc58KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.