Limits...
Spatial structure and climatic adaptation in African maize revealed by surveying SNP diversity in relation to global breeding and landrace panels.

Westengen OT, Berg PR, Kent MP, Brysting AK - PLoS ONE (2012)

Bottom Line: Environmental association analysis was used to detect SNPs associated with three climatic variables based on the full 43,963 SNP dataset.Controlling for population history in a linear model, we identify 79 SNPs associated with maximum temperature during the growing season.The associations located in genes of known importance for abiotic stress tolerance are interesting candidates for local adaptations.

View Article: PubMed Central - PubMed

Affiliation: Centre for Development and the Environment, SUM, University of Oslo, Oslo, Norway. ola.westengen@sum.uio.no

ABSTRACT

Background: Climate change threatens maize productivity in sub-Saharan Africa. To ensure food security, access to locally adapted genetic resources and varieties is an important adaptation measure. Most of the maize grown in Africa is a genetic mix of varieties introduced at different historic times following the birth of the trans-Atlantic economy, and knowledge about geographic structure and local adaptations is limited.

Methodology: A panel of 48 accessions of maize representing various introduction routes and sources of historic and recent germplasm introductions in Africa was genotyped with the MaizeSNP50 array. Spatial genetic structure and genetic relationships in the African panel were analysed separately and in the context of a panel of 265 inbred lines representing global breeding material (based on 26,900 SNPs) and a panel of 1127 landraces from the Americas (270 SNPs). Environmental association analysis was used to detect SNPs associated with three climatic variables based on the full 43,963 SNP dataset.

Conclusions: The genetic structure is consistent between subsets of the data and the markers are well suited for resolving relationships and admixture among the accessions. The African accessions are structured in three clusters reflecting historical and current patterns of gene flow from the New World and within Africa. The Sahelian cluster reflects original introductions of Meso-American landraces via Europe and a modern introduction of temperate breeding material. The Western cluster reflects introduction of Coastal Brazilian landraces, as well as a Northeast-West spread of maize through Arabic trade routes across the continent. The Eastern cluster most strongly reflects gene flow from modern introduced tropical varieties. Controlling for population history in a linear model, we identify 79 SNPs associated with maximum temperature during the growing season. The associations located in genes of known importance for abiotic stress tolerance are interesting candidates for local adaptations.

Show MeSH
Structure results for the combined African panel and association panel.Plots of STRUCTURE results for the combined African panel and Association Panel showing: a) the Ln (probability of the data) for the values of K from 1 to 6; b) the similarity coefficient for nine different runs per K; and c) the delta K value.
© Copyright Policy
Related In: Results  -  Collection


getmorefigures.php?uid=PMC3472975&req=5

pone-0047832-g003: Structure results for the combined African panel and association panel.Plots of STRUCTURE results for the combined African panel and Association Panel showing: a) the Ln (probability of the data) for the values of K from 1 to 6; b) the similarity coefficient for nine different runs per K; and c) the delta K value.

Mentions: STRUCTURE analysis of the combined African panel and AP dataset (26,900 SNPs) confirms earlier findings about the structure of the breeding material and firmly clusters the African panel within the tropical and subtropical group. The likelihood value increases continuously with no obvious inflection point (Fig. 3a). The similarity coefficient based on comparison of Q matrices from different runs with the same value of K [28] is close to 1 (>0.99) for K values 2 and 3 while it drops significantly for K = 4 and the variation between runs remains high also for higher values of K (Fig. 3b). The highest delta K value (according to [31]) is observed for K value 3 (Fig. 3c). The three clusters identified correspond well with those identified in earlier assessments of the AP; the non-stiff stalk (NSS), stiff-stalk (SS) and tropical and subtropical (TS) subpopulations (Table S5). By plotting the relationship of NSS and TS Q group membership from Flint-Garcia et al. [14] vs. the corresponding Q group memberships from our analysis, we obtain strong correlations (Pearson's R2 = 0.94 and 0.96 (p<0.01); Fig. S2d,e), slightly higher than those of Hamblin et al. [27] for similar plots of Q memberships from SNP vs. SSR data.


Spatial structure and climatic adaptation in African maize revealed by surveying SNP diversity in relation to global breeding and landrace panels.

Westengen OT, Berg PR, Kent MP, Brysting AK - PLoS ONE (2012)

Structure results for the combined African panel and association panel.Plots of STRUCTURE results for the combined African panel and Association Panel showing: a) the Ln (probability of the data) for the values of K from 1 to 6; b) the similarity coefficient for nine different runs per K; and c) the delta K value.
© Copyright Policy
Related In: Results  -  Collection

Show All Figures
getmorefigures.php?uid=PMC3472975&req=5

pone-0047832-g003: Structure results for the combined African panel and association panel.Plots of STRUCTURE results for the combined African panel and Association Panel showing: a) the Ln (probability of the data) for the values of K from 1 to 6; b) the similarity coefficient for nine different runs per K; and c) the delta K value.
Mentions: STRUCTURE analysis of the combined African panel and AP dataset (26,900 SNPs) confirms earlier findings about the structure of the breeding material and firmly clusters the African panel within the tropical and subtropical group. The likelihood value increases continuously with no obvious inflection point (Fig. 3a). The similarity coefficient based on comparison of Q matrices from different runs with the same value of K [28] is close to 1 (>0.99) for K values 2 and 3 while it drops significantly for K = 4 and the variation between runs remains high also for higher values of K (Fig. 3b). The highest delta K value (according to [31]) is observed for K value 3 (Fig. 3c). The three clusters identified correspond well with those identified in earlier assessments of the AP; the non-stiff stalk (NSS), stiff-stalk (SS) and tropical and subtropical (TS) subpopulations (Table S5). By plotting the relationship of NSS and TS Q group membership from Flint-Garcia et al. [14] vs. the corresponding Q group memberships from our analysis, we obtain strong correlations (Pearson's R2 = 0.94 and 0.96 (p<0.01); Fig. S2d,e), slightly higher than those of Hamblin et al. [27] for similar plots of Q memberships from SNP vs. SSR data.

Bottom Line: Environmental association analysis was used to detect SNPs associated with three climatic variables based on the full 43,963 SNP dataset.Controlling for population history in a linear model, we identify 79 SNPs associated with maximum temperature during the growing season.The associations located in genes of known importance for abiotic stress tolerance are interesting candidates for local adaptations.

View Article: PubMed Central - PubMed

Affiliation: Centre for Development and the Environment, SUM, University of Oslo, Oslo, Norway. ola.westengen@sum.uio.no

ABSTRACT

Background: Climate change threatens maize productivity in sub-Saharan Africa. To ensure food security, access to locally adapted genetic resources and varieties is an important adaptation measure. Most of the maize grown in Africa is a genetic mix of varieties introduced at different historic times following the birth of the trans-Atlantic economy, and knowledge about geographic structure and local adaptations is limited.

Methodology: A panel of 48 accessions of maize representing various introduction routes and sources of historic and recent germplasm introductions in Africa was genotyped with the MaizeSNP50 array. Spatial genetic structure and genetic relationships in the African panel were analysed separately and in the context of a panel of 265 inbred lines representing global breeding material (based on 26,900 SNPs) and a panel of 1127 landraces from the Americas (270 SNPs). Environmental association analysis was used to detect SNPs associated with three climatic variables based on the full 43,963 SNP dataset.

Conclusions: The genetic structure is consistent between subsets of the data and the markers are well suited for resolving relationships and admixture among the accessions. The African accessions are structured in three clusters reflecting historical and current patterns of gene flow from the New World and within Africa. The Sahelian cluster reflects original introductions of Meso-American landraces via Europe and a modern introduction of temperate breeding material. The Western cluster reflects introduction of Coastal Brazilian landraces, as well as a Northeast-West spread of maize through Arabic trade routes across the continent. The Eastern cluster most strongly reflects gene flow from modern introduced tropical varieties. Controlling for population history in a linear model, we identify 79 SNPs associated with maximum temperature during the growing season. The associations located in genes of known importance for abiotic stress tolerance are interesting candidates for local adaptations.

Show MeSH