SEARCH

SEARCH BY CITATION

Data S1 Sample locations and chloroplastic microsatellite data: the excel files contains 370 lines, each corresponding to one individual. The columns correspond to: country, population code, population name, individual laboratory code, locus NTCP9, locus CCMP3, Chlorotype code. Each locus column corresponds to one allele coded with three digits.

Data S2 Sample locations and nuclear microsatellite data: the excel files contains 673 lines, each corresponding to one individual. The columns correspond to : population code, group code, individual collection code, population name, individual laboratory code, locus mCIRVp28, locus mCIRVp14, locus mCIRVp159, locus mCIRVp181, locus mCIRVp113, locus mCIRVp175, locus mCIRVp191, locus mCIRVp134, locus mCIRVp168, locus mCIRVp70, locus mCIRVp08. Each locus column corresponds to the two alleles coded each with three digits.

Fig. S1 Model checking for scenario 3 (a) and scenario 4 (b) using the ABC approach (Cornuet et al. 2008, 2010). Observed data (the biggest dot) are plotted against points generated using priors (5000 small size dots) or posteriors (1000 bigger size dots) within a space describes by PCA analysis axes 1 and 2. Percentage variation explained by each axis is indicated in brackets.

Fig. S2 Density function for parameter prior (grey) and posterior (black) distributions for scenario 3. Ne and N2 define population sizes and Te defines expansion time expressed in generation number. _micA is the average mutation rate and Pmean is the average parameter of the generalized stepwise model geometric distribution. Note that the effective priors of N2 and Ne are not uniform because of the constraint Ne > N2 we applied to these parameters. We also represented Ne/N2, which represents the inverse of population expansion.

Fig. S3 Density function for parameter prior (grey) and posterior (black) distributions for scenario 4. N3, Nr and N3 define population sizes. T1 defines bottleneck time and T2 defines expansion time expressed in generation number. _micA is the average mutation rate and Pmean is the average parameter of the generalized stepwise model geometric distribution. Note that the effective priors of Nr, Nbe, N3, T1 and T2 are not uniform because of the constraints (Nbe > Nr, Nr < N3 and T1 < T2) we applied to these parameters.

Table S1 Prior setting for demographic scenario of the shea tree in West Africa using the approximate Bayesian computation approach (Cornuet et al. 2008, 2010); N, N1, N2, N3, Ne, Nbe, Nr define population sizes in the four scenarios and tb, te, t1, t2, times of demographic event occurrence expressed in generation number. μmean is the average mutation rate and Pmean is the average parameter of the generalized stepwise model geometric distribution. Locus specific mutation parameters (μi and Pi), considered as nuisance parameters, were kept at default values

FilenameFormatSizeDescription
MEC_5249_sm_DataS1.xls69KSupporting info item
MEC_5249_sm_DataS2.xls142KSupporting info item
MEC_5249_sm_FigS1a.pdf22KSupporting info item
MEC_5249_sm_FigS1b.pdf23KSupporting info item
MEC_5249_sm_FigS2.pdf22KSupporting info item
MEC_5249_sm_FigS3.pdf28KSupporting info item
MEC_5249_sm_TableS1.doc39KSupporting 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.