Search

To study barley adaptation and improvement in the Mediterranean basin, a collection of 188 entries comprising landraces and old genotypes and current modern varieties from the Mediterranean basin and elsewhere was tested on moisture-contrasted environments in seven Mediterranean countries, during 2004 and 2005 harvest seasons. The experimental design consisted of an unreplicated trial for all entries, augmented by four repeated checks to which a partial replicate containing a quarter of the entries was added. Best Linear Unbiased Predictions (BLUPs) representing adjusted genotypic means were generated for individual trials using a mixed model. BLUPs were used for genotype by environment interaction analysis using main effect plus genotype by environment interaction (GGE) biplots of yield ranked data and for comparisons of landraces, old and modern genotypes using analysis of variance. Mean yields ranged from near crop failure to 6 t/ha. Local landraces were better adapted to environments yielding below 2 t/ha, thus breeding has mostly benefited environments yielding above 2 t/ha where modern genotypes out yielded landraces and old cultivars by 15%. Current barley selection is leading to specifically adapted genotypes. [ABSTRACT FROM AUTHOR]

An association panel consisting of 185 accessions representative of the barley germplasm cultivated in the Mediterranean basin was used to localise quantitative trait loci (QTL) controlling grain yield and yield related traits. The germplasm set was genotyped with 1,536 SNP markers and tested for associations with phenotypic data gathered over 2 years for a total of 24 year × location combinations under a broad range of environmental conditions. Analysis of multi-environmental trial (MET) data by fitting a mixed model with kinship estimates detected from two to seven QTL for the major components of yield including 1000 kernel weight, grains per spike and spikes per m(2), as well as heading date, harvest index and plant height. Several of the associations involved SNPs tightly linked to known major genes determining spike morphology in barley (vrs1 and int-c). Similarly, the largest QTL for heading date co-locates with SNPs linked with eam6, a major locus for heading date in barley for autumn sown conditions. Co-localization of several QTL related to yield components traits suggest that major developmental loci may be linked to most of the associations. This study highlights the potential of association genetics to identify genetic variants controlling complex traits.

Population structure and genome-wide linkage disequilibrium (LD) were investigated in 192 Hordeum vulgare accessions providing a comprehensive coverage of past and present barley breeding in the Mediterranean basin, using 50 nuclear microsatellite and 1,130 DArT((R)) markers. Both clustering and principal coordinate analyses clearly sub-divided the sample into five distinct groups centred on key ancestors and regions of origin of the germplasm. For given genetic distances, large variation in LD values was observed, ranging from closely linked markers completely at equilibrium to marker pairs at 50 cM separation still showing significant LD. Mean LD values across the whole population sample decayed below r (2) of 0.15 after 3.2 cM. By assaying 1,130 genome-wide DArT((R)) markers, we demonstrated that, after accounting for population substructure, current genome coverage of 1 marker per 1.5 cM except for chromosome 4H with 1 marker per 3.62 cM is sufficient for whole genome association scans. We show, by identifying associations with powdery mildew that map in genomic regions known to have resistance loci, that associations can be detected in strongly stratified samples provided population structure is effectively controlled in the analysis. The population we describe is, therefore, shown to be a valuable resource, which can be used in basic and applied research in barley.