Rye (Secale cereale L.) originates from an area extending from central and eastern Turkey through nortwesten Iran to Transcaucasia. It was probably not cultivated continuously in its area of origin. When rye came to northern Europe as a weed, it was able to exploit its advantages of superior winter hardiness and modest nutrient requirements and became the dominant cereal from the early Middle Ages^[1]. Initially, rye was grown in a vast region from northern France to eastern Siberia. Nowadays, rye-growing is still centered in Europe, where 86% of grains are produced, with Germany, Poland, the Russian Federation, Denmark, and Belarus as the main producers in 2019^[2]. In rye as a cross-pollinating crop, population and hybrid cultivars are available. Hybrid breeding started in the 1970s at the University of Hohenheim, Germany, and in 1985 the world’s first three hybrid cultivars were released^[3]. This breeding method revolutionized rye-growing, because hybrid breeding provided about 20% higher grain yield and a threefold higher gain from selection, compared to population rye, over a 30-year period^[4]. Nowadays, about 80% of the German rye-growing area is planted with hybrid rye cultivars, and hybrid rye is available in all other main production countries, as well as in Austria, the USA, and Canada.

1. Basics of hybrid rye breeding

Hybrid rye breeding leads to considerably higher grain yield and a higher revenue to the farmer. The basis of hybrid seed production is the cytoplasmic-male sterility (CMS)-inducing Pampa (P) cytoplasm derived from an Argentinean landrace^[5] and restorer-to-fertility (Rf) genes^[6]. To ensure a maximal heterosis, two gene pools are used, that are in Central Europe the Petkus (seed parent) and the Carsten (pollen parent) gene pool. The pools were derived from two breeding companies that used different landraces for starting their selection work in the 19^th and the beginning 20^th century. The basis of hybrid breeding are inbred lines that are produced in rye by subsequent selfing with the use of a dominant single-gene mutation leading to full self-fertility (sf).

Breeding starts with crossing of parents (Figure 1), followed by two selfings of the progenies and subsequent per se performance test of S₂ lines for highly heritable traits (e.g., vitality, heading date, plant height, disease resistances, lodging tolerance, 1000-grain weight, falling number). The selected fraction is crossed in the seed parent pool to a CMS donor to start the backcrossing procedure that is necessary to result at the end in near-isogenic lines with the CMS-inducing P cytoplasm and the normal cytoplasm. During backcrossing selection for non-restorer is performed. In contrast, in the pollen parent pool a strict selection for Rf genes has to practised. Here, only two selfings are done, because afterwards the best selected lines are crossed to a synthetic. In parallel to selfing and backcrossing two stages of yield tests (YT) are made by crossing the inbred lines with testers from the opposite pool. These testcrosses (TC) are planted in 3-5 locations in two replicates to select for general combining ability (GCA) for grain yield and the other mentioned traits. The finally selected inbred lines are used to compose a topcross hybrid with two genetically different seed parent lines (A-P, B-N) and a restorer synthetic composed of two S₂ lines (Syn_Rf). This complex hybrid is necessary due to the high inbreeding depression that rye is suffering as an obligate cross-pollinating crop.

 

Figure 1. Scheme for hybrid rye breeding

Hybrid rye production requires strict logistics for seed multiplication, because rye is an effective wind pollinator. All stands of seed production must be strictly isolated and checked for outcrossings several times. Mostly the line multiplications (A-P × A-N, B-N) and the production of basic seed (A-P × B-N) are done in foreign countries with no commercial rye growing (France, Italy, Hungary). The lines are planted in stripes with a 3:1 proportion for females:males. The multiplication of the synthetics and the production of certified seed (A-P • B-N × Syn_Rf) is done in the area of marketing of the rye hybrid cultivars. The certified seed is produced in a technical mixture of 90-95% CMS seed parent and 5-10% restorer synthetic. This approach cheapens production, but currently requires the presence of synthetics to escape the strong inbreeding depression of lines.

2. Ergot as a challenge for hybrid rye production

Since Medieval times, the results of ergot infections have been documented. Rye and ergot were so closely interwoven that botanical drawings of rye in the 17^th century included ergot. Ergot refers to the blackish purple sclerotia of the fungus Claviceps purpurea (Fries ex Fries) Tulasne that extend far beyond the husks of rye (Figure 2a). They contain more than 80 ergot alkaloids, of which six alkaloids and their corresponding epimers were found to be the most important ones in Europe. Prolonged consumption of ergot-containing flour, which was common among the peasant population in the Middle Ages, leads to a serious disease, ergotism, that is characterized either by nervous disorders (fever, muscle spasms, tremor, paralysis, hallucinations) or by vasoconstriction effects (gangrene, violent burning, shooting pain in the fingers and toes, called “St. Anthony’s fire”).

The high effectiveness of non-adapted Rfp genes will play a prominent role in future hybrid breeding. The main task is to reduce the obvious yield penalty and other negative agronomic features. Even if the introgressed genomic section is significantly shortened, a significant yield loss remains that could not be further reduced using more closely linked markers. However, the effects of the same Rfp gene on different genotypes significantly differs. Therefore, a consequent selection in the elite gene pools for mitigating undesired side effects of Rfp genes seems feasible in the long run. This would replicate the coevolution that has occurred in adapted CMS/Rf systems over evolutionary time periods.

This coevolution could be enhanced in a breeding context, when environmentally stable adaptation genes exist in the elite gene pools that lower the negative agronomic effects of non-adapted Rf genes. Research activities should be started to identify and to exploit these genes. In conclusion, the development of commercial hybrid cultivars in P cytoplasm, with a similarly low or an even lower ergot severity than population cultivars, appears to be promising.

In hybrid rye breeding, modern genomic techniques are practised to increase the performance of hybrids and ensure a steady selection gain in future (Table 1). Marker-assisted selection is useful for all monogenically inherited traits. Marker-assisted backcrossing helps to shorten the backcrossing procedure in the seed-parent pool (rf. to Figure 1) by selecting against the CMS donor background. Using a low-density marker array in BC₃ already the same low donor background could be achieved like normally in BC₆. QTL mapping is an important tool when lines with special alleles are searched that are not available in the current breeding populations. With this tool, the genetic architecture of the trait can be analysed with high power, because even rare alleles are pushed to an allele frequency of 50% in bi-parental crosses.

Table 1. Genome-based techniques and their use in hybrid rye breeding (KASP=kompetitive allele-specific PCR)

Technique	Useful for …	Requirement
Marker-assisted selection (MAS)	Introgression of mono-genic resistances, Rf genes, dwarfing genes	Closely linked diagnostic KASP markers
Marker-assisted backcrossing (MAB)	Conversion of N lines: Selection against CMS donor background	Low-density marker array
Quantitative trait loci (QTL) mapping	Detection of rare alleles with large effects	Low-density marker array
Genome wide association mapping (GWAS)	Detection of common alleles with smaller effects	Medium-density marker array
Genomic selection (GS)	Selection of genome-wide marker effects for all traits	High-density marker array

Genomic selection (GS) is already practised in some breeding companies. Typically, GS is done after the first S₂ line per se performance (rf. to Fig. 1). Based on previous cycles that are used as training population, superior lines in GCA for grain yield and other traits are predicted solely on their genomic composition and only a reduced set of positively predicted lines enters the costly 2-year testcross procedure^[11]. After the yield test, a final selection can take place based on an index of the phenotypically measured and genomically predicted GCA values. Implementing GS allows to test about 50% more candidates with the same budget, thus considerably increasing selection gain. 

The great future challenges of climate change, such as increased drought stress tolerance, improved lodging tolerance, as well as the increasing need for resistant varieties can only be met effectively and as quickly as possible with hybrid breeding.