Supplementary Table 1. Diversity of SSR markers in three barley samples: (i) 84 landraces, (ii) 40 old cultivars, and (iii) 68 modern cultivars. Unique alleles in brackets. Rare alleles <5 % frequency

Landraces (84) / Old cultivars (40) / Modern cultivars (68)
SSR / Chr. / Nr alleles / Overall
PIC / Frequent alleles / Rare alleles / Frequent
Alleles / Rare alleles / Frequent alleles / Rare alleles
Bmac0399 / 1H / 26 / 0.909 / 8 / 14 (9) / 7 / 5 (1) / 8 / 8 (2)
Bmag0211 / 1H / 7 / 0.741 / 5 / 1 / 5 / 0 / 5 / 2 (1)
scssr10477 / 1H / 17 / 0.798 / 5 / 8 (4) / 5 / 5 (1) / 5 / 6 (2)
Bmag0382 / 1H / 6 / 0.465 / 2 / 4 (3) / 2 / 1 / 2 / 1
HvHVA1 / 1H / 2 / 0.224 / 2 / 0 / 2 / 0 / 2 / 0
scssr02748 / 1H / 4 / 0.444 / 4 / 0 / 4 / 0 / 4 / 0
Bmac0134 / 2H / 22 / 0.866 / 4 / 15 (8) / 4 / 3 / 4 / 9 (3)
HVM36 / 2H / 9 / 0.756 / 6 / 3 (2) / 5 / 0 / 6 / 1
scssr07759 / 2H / 8 / 0.505 / 3 / 2 (1) / 2 / 2 / 3 / 3 (2)
scssr00334 / 2H / 4 / 0.060 / 1 / 3 (1) / 1 / 0 / 1 / 2
Bmag0125 / 2H / 10 / 0.749 / 5 / 4 (1) / 5 / 0 / 5 / 4 (1)
EBmac0415 / 2H / 6 / 0.504 / 3 / 3 (2) / 3 / 1 / 3 / 1
scssr08447 / 2H / 5 / 0.356 / 3 / 2 (1) / 3 / 1 / 3 / 1
HvL TPPB / 3H / 8 / 0.500 / 3 / 5 (3) / 2 / 1 / 3 / 1
scssr10559 / 3H / 5 / 0.601 / 4 / 0 / 4 / 1 (1) / 4 / 0
Bmac0067 / 3H / 24 / 0.808 / 6 / 13 (12) / 5 / 1 / 6 / 5 (5)
scssr25691 / 3H / 6 / 0.355 / 2 / 1 (1) / 2 / 3 / 2 / 3
scssr05281 / 3H / 1 / 0 / 1 / 0 / 1 / 0 / 1 / 0
HVM62 / 3H / 8 / 0.712 / 6 (1) / 2 (1) / 5 / 0 / 5 / 1
scssr25538 / 3H / 5 / 0.598 / 3 / 2 (2) / 3 / 0 / 3 / 0
HVM40 / 4H / 8 / 0.671 / 4 / 3 (3) / 4 / 0 / 4 / 1 (1)
scssr20569 / 4H / 5 / 0.628 / 4 / 1 (1) / 3 / 0 / 4 / 0
scssr18005 / 4H / 6 / 0.237 / 3 / 2 (2) / 3 / 1 (1) / 3 / 0
Bmag0353 / 4H / 15 / 0.833 / 7 (1) / 5 (3) / 5 / 2 / 6 / 5 (3)
scssr14079 / 4H / 3 / 0.240 / 2 / 0 / 2 / 1 (1) / 2 / 0
EBmac0701 / 4H / 12 / 0.798 / 5 / 6 (3) / 5 / 2 / 5 / 4 (1)
Bmag0419 / 4H / 11 / 0.768 / 6 / 4 (1) / 6 / 2 / 6 / 4
Af459084 / 4H / 33 / 0.812 / 3 / 25 (16) / 3 / 6 (2) / 3 / 8 (3)
scssr02306 / 5H / 5 / 0.589 / 3 / 1 (1) / 3 / 0 / 3 / 1 (1)
scssr07106 / 5H / 6 / 0.588 / 3 / 2 (2) / 3 / 0 / 3 / 1 (1)
scind02587 / 5H / 2 / 0.363 / 2 / 0 / 2 / 0 / 2 / 0
scssr02503 / 5H / 2 / 0.142 / 2 / 0 / 1 / 0 / 2 / 0
scind16991 / 5H / 4 / 0.179 / 2 / 1 (1) / 2 / 1 / 2 / 1
scssr05939 / 5H / 3 / 0.246 / 2 / 1 / 2 / 0 / 2 / 1
scssr10148 / 5H / 9 / 0.584 / 3 / 4 (2) / 2 / 0 / 3 / 4 (2)
Abc02203 / 5H / 4 / 0.488 / 3 / 1 (1) / 3 / 0 / 3 / 0
scssr03907 / 5H / 21 / 0.866 / 6 / 12 (6) / 6 / 3 / 6 / 8 (1)
scssr09398 / 6H / 12 / 0.763 / 5 / 4 (3) / 4 / 1 / 5 / 4 (2)
scind60002 / 6H / 7 / 0.602 / 4 / 3 (2) / 4 / 1 / 4 / 0
Bmac0018 / 6H / 6 / 0.617 / 3 / 3 (2) / 3 / 0 / 3 / 1
Bmag0009 / 6H / 6 / 0.708 / 5 / 1 (1) / 5 / 0 / 5 / 0
scssr05599 / 6H / 5 / 0.559 / 4 / 0 / 4 / 0 / 4 / 1 (1)
scssr00103 / 6H / 5 / 0.719 / 4 / 1 / 4 / 1 / 4 / 1
HVM4 / 7H / 23 / 0.872 / 6 / 15 (9) / 5 / 3 / 6 / 7 (2)
scssr07970 / 7H / 6 / 0.646 / 4 / 2 (1) / 3 / 0 / 4 / 1
HvCMA / 7H / 3 / 0.405 / 2 / 1 / 2 / 0 / 2 / 1
scssr15864 / 7H / 5 / 0.343 / 3 / 1 (1) / 3 / 1 (1) / 3 / 0
scind00149 / 7H / 1 / 0 / 1 / 0 / 1 / 0 / 1 / 0
Bmac0581 / 7H / 14 / 0.518 / 2 / 9 (4) / 2 / 4 (2) / 2 / 5 (1)
bmag120 / 7H / 14 / 0.830 / 6 / 6 (1) / 6 / 3 / 6 / 6 (1)
HVM5 / 7H / 24 / 0.897 / 8 / 13 (6) / 8 / 6 (1) / 8 / 6 (2)
Scssr04056 / 7H / 11 / 0.783 / 5 / 4 (1) / 5 / 1 / 5 / 5 (2)
Mean diversity / 9.11 / 0.56 / 198 (2) / 213 (124) / 184 / 63 (11) / 196 / 124 (40)

Supplementary Table 2.Strong marker-trait associations(P<0.001) for powdery mildew linked to candidate genes reported in the literature using a mixed model with relative kinship (K) orStructure matrix of cluster membership probabilities (Q) for five cluster soluation and Bin positions based on the SteptoexMorex bin map ( Resistance genes are described in (Chelkowski et al. 2003). * refers to QTLs described in the literature without an associated candidate gene: Rbgqs (Shtaya et al. 2006), qtl (Thomas et al. 1995), and QTL (Heun 1992).

Markers associated with powdery mildew (Q) / positive markers (K) / candidate gene?
Chromosome / Position (cM) / No markers / Trials / Bin
1H / 4.07-10.07 / 2 / 1 / 1 / Mlra
1H / 18.93-48.1 / 10 / 3 / 2 / 2 / Mla1,Mla2,Mlk,Mla12,Mla13,Mla14 (alleles at the Mla locus)
1H / 95.52 / 1 / 1 / 8 / 2
1H / 142.33 / 1 / 1 / 11 / Mli,MlGa
1H / 179.1-180.95 / 5 / 1 / 14
2H / 27.96 / 1 / 1 / 3 / 1 / MlHb
2H / 39.96 / 1 / 1 / 4 / MlHb
2H / 48 / 2 / 2 / 5 / MlHb
2H / 74.87-81.97 / 3 / 1 / 8 / 1 / Rbgp1*
2H / 145.66-147.02 / 3 / 1 / 13
2H / 167.71-172.63 / 9 / 1 / 15 / MlLa
3H / 0-18.53 / 0 / 0 / 1 / 2 / Rbgq2*
3H / 75.38 / 1 / 1 / 3 / Rbgq2*
3H / 111.53 / 1 / 1 / 6 / Rbgq2*
3H / 158.22 / 2 / 1 / 11
3H / 209.7 / 1 / 1 / 15
3H / 222.8-223.6 / 3 / 2 / 16
4H / 69.49 / 2 / 1 / 6 / Mlg
4H / 149.35 / 1 / 1 / 10 / mlo
5H / 21.16 / 1 / 1 / 1 / QTL*, qtl1*
5H / 33.2-33.66 / 2 / 1 / 2 / 1 / qtl1*
5H / 64.19 / 2 / 2 / 4 / qtl1*
5H / 74.48 / 1 / 1 / 6 / 1 / Mlj
5H / 100.55-100.57 / 2 / 1 / 8 / Rbgq3*
5H / 122.05 / 1 / 1 / 9 / Rbgq3*
6H / 72.53 / 2 / 1 / 7 / Mlh
7H / 38.28 / 2 / 2 / 2 / mlt
7H / 125.73-129.79 / 2 / 1 / 7
7H / 143.51 / 1 / 1 / 8
7H / 203.12-208.19 / 11 / 2 / 12 / Mlf

Supplementary Table 3. Correlation of powdery mildew scores in the four trials under study with the first 5 PCOs. PCO analysis was based on simple matching using 1130 DArT® markers. Significant correlations at P<0.001 (***) and P<0.05 (*) are shown.

trial / PCO1 / PCO2 / PCO3 / PCO4 / PCO5
ITA_4D / -0.092 / -0.238*** / 0.062 / 0.239*** / -0.111
ITA_4W / -0.130 / -0.120 / 0.063 / 0.351*** / -0.196*
MOR_4D / -0.455 / -0.442*** / -0.241*** / -0.154* / -0.213*
MOR_4W / -0.515 / -0.333*** / -0.222*** / -0.120 / -0.235***

Supplementary Figure 1. Goodness of fit for various clustering models versus increasing number of populations(k) determined by STRUCTURE

Supplementary Figure 2. Linkage Disequilibrium (r2) as a function of genetic distance for 811 mapped DArT® loci, across the whole of the genome. Only values with P<0.001 of significance are plotted.

Supplementary Figure 3. Linkage Disequilibrium (r2) across the barley chromosomes using 811 mapped DArT® loci with marker allele frequencies higher than 0.1. LD heatmaps built using the software Haploview (Barrett et al. 2005). Increasing shades of grey indicate a higher degree of correlation. (a) LD heatmaps for the seven barley chromosomes. (b) LD heatmap for chromosome 1H. Short range LD(blocks of markers in high LD) ranges between 0 and 3 cM, although long range high LD values can also be observed, presumably due to population structure effects.

1

Supplementary Figure 4. Scatter plots of all pairwise combinations of the first five scores from a PCO analysis of our association mapping population using simple matching from data obtained by genotyping with 1130 polymorphic DArT® markers.

1

Supplementary Figure 5. Whole genome scans for powdery mildew using distinct statistical models. Significances of the calculated p-values are shown as –log10(p-value). (a) Results from the four trials without correcting for population stratification are plotted together. This figure illustrates the effect of ignoring population structure and family relationships. (b) For comparison, the genome scans using STRUCTURE matrix of probabilities, Q, for each one of the trials.

Supplementary Figure 6. Model comparison with powdery mildew. Cumulative percentages of observed p-values are presented for a naive model, the model including Q and the model with K for each one of the trials. Approaches that appropriately control type I errors should approximate to a uniform distribution of the p-values. The naive approach does not involve any correction for population structure.

References

Barrett, J. C., B. Fry, J. Maller, and M. J. Daly, 2005 Haploview: analysis and visualization of LD and haplotype maps. Bioinformatics. 21: 263-265.

Chelkowski, J., M. Tyrka, and A. Sobkiewicz, 2003 Resistance genes in barley (Hordeum vulgare L.) and their identification with molecular markers. J.Appl.Genet. 44: 291-309.

Heun, M., 1992 Mapping quantitative powdery mildew resistance of barley using a restriction fragment length polymorphism map. Genome 35: 1019-1025.

Shtaya, M. J. Y., T. C. Marcel, J. C. Sillero, R. E. Niks, and D. Rubiales, 2006 Identification of QTLs for powdery mildew and scald resistance in barley. Euphytica 151: 421-429.

Thomas, W. T. B., W. Powell, R. Waugh, K. J. Chalmers, U. M. Barua et al. 1995 Detection of quantitative trait loci for agronomic, yield, grain and disease characters in spring barley (Hordeum vulgare L.). Theor.Appl.Genet. 91: 1037-1047.

1