Tracking Changes in the Spring Barley Gene Pool in Poland during 120 Years of Breeding

Abstract

This study was undertaken to investigate the diversity and population structure of 83 spring barley (Hordeum vulgare L.) cultivars, which corresponded to 120 years of this crop's breeding in Poland. The analysis was based on 11,655 DArTseq-derived SNPs evenly distributed across seven barley chromosomes. Five groups were assigned in the studied cultivars according to the period of their breeding. A decrease in observed heterozygosity within the groups was noted along with the progress in breeding, with a simultaneous increase in the inbreeding coefficient value. As a result of breeding, some of the unique allelic variation present in old cultivars was lost, but crosses with foreign materials also provided new alleles to the barley gene pool. It is important to mention that the above changes affected different chromosomes to varying degrees. The internal variability of the cultivars ranged from 0.011 to 0.236. Internal uniformity was lowest among the oldest cultivars, although some highly homogeneous ones were found among them. This is probably an effect of genetic drift or selection during their multiplications and regenerations in the period from breeding to the time of analysis. The population genetic structure of the studied group of cultivars appears to be quite complex. It was shown that their genetic makeup consists of as many as eleven distinct gene pools. The analysis also showed traces of directed selection on chromosomes 3H and 5H. Detailed data analysis confirmed the presence of duplicates for 11 cultivars. The performed research will allow both improvement of the management of barley genetic resources in the gene bank and the reuse of this rich and forgotten variability in breeding programs and research.

Keywords: DArTseq; SNP; barley; breeding; cultivars; diversity.

Figure 1

Circular overview of seven H. vulgare chromosomes based on DArTseq data acquired for 83 spring cultivars. (a) DArTseq loci distribution; (b) Average polymorphism information content (PIC) distribution; (c) Average observed heterozygosity (Ho) distribution. A sliding window approach with 500 kb windows, printed for 250 positions along the full length of barley chromosomes based on the genome assembly: IBSC_v2 [21] was applied.

Figure 2

Summary of polymorphism information content (PIC) values. (a) Range of relative frequencies for all analyzed DArTseq loci in 83 spring barley cultivars; (b) Mean PIC value including chromosomal location of studied DArTseq loci. Letters above the bars in the graph indicate homogeneous groups determined by Tukey’s post hoc test.

Figure 3

Summary of the diversity coefficient values across barley chromosomes for 83 cultivars based on DArTseq data. Letters above the bars in the graph indicate homogeneous groups determined by Tukey’s post hoc test.

Figure 4

Summary of the diversity coefficient values for cultivar groups assigned based on the breeding date. Letters above the bars in the graph indicate homogeneous groups determined by Tukey’s post hoc test.

Figure 5

Summary of the diversity coefficients values for cultivar groups, assigned based on the breeding date considering chromosome localization. Letters above the bars in the graph indicate homogeneous groups determined by Tukey’s post hoc test. (a) observed heterozygosity (uHo); (b) expected heterozygosity (uHe); (c) inbreeding coefficient (F).

Figure 6

Heterogeneity level of 83 spring barley cultivars expressed by observed heterozygosity value based on SNPs derived from DArTseq analysis.

Figure 7

Summary of changes in the number of unique alleles during more than 120 years of breeding and cultivation of spring barley in Poland. Colors indicate groups and dashed lines connect compared periods. Above the axis, information about the new breeding objectives is placed.

Figure 8

Graphical presentation of the Principal Coordinate Analysis results for DArTseq data of 83 spring barley cultivars. Results in the first three coordinates’ system. Each point denotes one tested cultivar. Numbering according to Table 6. Rotable 3D figure can be found in the supplementary materials (Figure S1).

Figure 9

The results of 100,000 iterations of STRUCTURE software [22] for 83 spring barley cultivars based on DArTseq-derived SNPs data with K values K = 11 based on ad hoc measure ∆K [23,24], where K is the number of ad hoc clusters; each vertical bar represents one cultivar that is marked by order number according to Table 6. The length of the colored segment shows the estimated proportion of membership of each gene pool in the cultivar genetic makeup.

Figure 10

Proportion of 11 gene pools in five breeding periods of spring barley based on population structure analysis. (a) Cultivars bred before 1945; (b) cultivars bred between 1945 and 1969; (c) cultivars bred between 1970 and 1989; (d) cultivars bred between 1990 and 1999; (e) cultivars bred after 2000.

Figure 11

Circular overview of seven H. vulgare chromosomes. (a) Transformed FST¹⁰ for cultivars bred before 1945 and after 2000; (b) Average polymorphism information content (PIC) distribution in cultivars bred before 1945; (c) Average polymorphism information content (PIC) Distribution in cultivars bred after 2000; (d) Number of unique SNPs in cultivars bred before 1945; (e) Number of unique SNPs in cultivars bred after 2000. A sliding window approach with 500 kb windows, printed for 250 positions along the full length of barley chromosomes based on the genome assembly: IBSC_v2 [21].

Figure 12

Identity by descent (IBD) based clustering of spring barley cultivars with cutoff at 0.95. Accession numbers according to Table 6.