The current 2015-16 spring and winter barley recommended lists on the Agriculture and Horticulture Development Board (AHDB) Cereals & Oilseeds classify varieties against 18 different characteristics with some varieties scoring high for some and low for others. Breeders choose parental combinations that complement each other for different aspects of these 18 characters and, as barley is a natural self-pollinator, make crosses between the parents by emasculating one and pollinating it with the other to generate F1 hybrid seed. The F1 seed is predicted to segregate out individual lines that combine the best attributes of both parents and breeders have developed a variety of selection methods to enable them to identify the best lines for possible commercial development. This sounds simple but the possibilities are considerable as the number of homozygous inbred lines that can be produced from a cross is 2n, where n = the number of genes segregating. If the 18 characteristics listed above were just controlled by 18 independent genetic loci, all of which were segregating in a cross, over 250,000 different homozygous allelic combinations could be produced. Crossing an elite line to another elite line gives a higher probability of producing superior progeny than crossing an elite line with an unadapted line, e.g. a wild barley or an old land-race. Breeders therefore tend to cross within their breeding gene-pools and avoid bringing in other material unless no useful variation can be found within it, e.g. if a new source of disease resistance is needed to help protect the crop from disease epidemics. This crossing within the gene pool is illustrated by the pedigree of UK winter barley malting cultivars from Maris Otter, first recommended in 1965, to SY Venture.

The characters listed on the AHDB Cereals and Oilseeds recommended list establish the usefulness of a potential cultivar in agriculture, its Value for Cultivation and Use (VCU), but do not determine the identity of the potential cultivar. Identity is established through a separate set of morphological characteristics that are largely independent of VCU characteristics and is used to establish Distinctness, Uniformity, and Stability (DUS). A number of DUS characteristics have been described, illustrated and studied genetically. Breeders use this information to submit sample stocks of breeding lines for official testing that will pass DUS as well as VCU tests.

Barley breeding is almost entirely a matter of deriving improved inbred cultivars. Some F1 hybrid cultivars have been released in Europe by Syngenta, having been developed by a cytoplasmic male sterility system coupled with a dominant fertility restorer gene. Whilst these lines offer a yield advantage over inbred lines they have been developed for the six-rowed winter feed barley market. The seed harvested from an F1 hybrid crop will have the properties of an F2 population and thus will segregate if used for malting and potentially lead to a very uneven performance. Data from UK recommended list trials shows that the F1 hybrid Volume is the highest yielding cultivar on the 2015/16 list but its yield advantage is just 2% greater than the best inbred two- and six-rowed varieties (KWS Infinity and Daxor respectively). F1 hybrids do appear to offer greater yield stability than conventional inbred lines and may therefore be useful under more variable climates.

The principal inbreeding methods that are utilised in barley breeding are pedigree selection, single seed descent, and doubled haploid (DH) production. Briefly, pedigree inbreeding involves selection of the best single plants from the F2 generation, evaluating their progeny in a row or very small plot in the F3 generation and repeating this plant to progeny cycle a number of times to generate homozygous lines that can then be advanced and tested in multi-site trials. Some breeders, especially those working within a narrow gene-pool, can advance directly from the F3 stage to a yield trial through the judicious use of DNA markers (see below). Breeders can submit their best lines to official trials between three and five years after making the cross by combining this strategy with out of season nurseries in the opposite hemisphere in ‘shuttle breeding’ to grow two generations in one year. Shuttle breeding is not an option in the winter crop due to its vernalisation requirement.

Doubled Haploids, the process of forming individuals homozygous at all genetic loci in a single generation, tend to be used more for winter than for spring barley as the technique provides an alternative opportunity to shorten the breeding in the absence of the shuttle breeding option. Doubled haploids can be produced by pollination of F1 plants with Hordeum bulbosum, a related species that results in haploid embryo formation. The embryos are rescued and grown in tissue culture to provide haploid plants that are treated with the spindle destroying agent, colchicine, to provide homozygous diploid plants. This is a labour intensive method, which limits the numbers of DH lines that can be produced each year. Anther and Isolated Microspore Culture (AC and IMC, respectively) are two other tissue culture techniques where pollen grains can change from a gametophytic (pollen) to a sporophytic (embryo) pathway when subjected to specific environmental stresses and supplied with the appropriate medium. Spontaneous doubling of the haploid embryo occurs in the vast majority of cases again leading to homozygous diploid plants. AC tends to be the most commonly used method although IMC potentially has higher throughput and is less labour intensive. IMC is more genotype dependent currently and that limits its application in routine breeding.

Single seed descent (SSD) is a specialized form of inbreeding where one grows the F2 generation under a high density and restricted water and nutrient supply to provide single-stemmed plants with few grain. One seed is taken from each stem and re-sown under the same conditions to start the cycle again. This shortens the generation time and can be practiced in any room with heating and lighting so that at least 3 generations of a spring crop can be grown in one year and over two of a winter crop. Practically, a breeder might use SSD for one year to generate near-homozygous lines and then sow out in the field and practice an abbreviated form of pedigree selection. Such a scheme is an alternative for spring barley and winter barley where resources do not permit shuttle breeding or tissue culture respectively.

After a minimum of 4-5 years from crossing, breeders enter the best lines (<10) from a crossing cycle into the official trialing and testing procedures that enable registration of that line in the countries they are testing in. It takes a minimum of two years to establish DUS and VCU and lines are discarded during this process. In the UK, the recommended list (RL) testing represents a further round of testing after registration (National Listing in the UK) and a line may receive a provisional recommendation after one year’s RL evaluation. Another year is required to generate a sufficient volume of certified seed for limited planting by farmers so the overall time taken from crossing to beginning to make a commercial impact is in the order of 8 years.

Key barley characters such as yield and end-use quality tend to be controlled by a number of different genes, each of which may interact with the growing environment. These characters therefore need to be assessed under relevant growing conditions and at multiple different sites to establish reliable data. It is only the later stages of a breeding programme when sufficient seed becomes available to do this. Early stage selection tends to concentrate upon general agronomic merit and resistance to pests and diseases but reduces the population size drastically. The consequence of such a process is to limit the possibility of major improvement in yield and quality unless some form of early selection can be included in the programme. Using DNA markers as surrogates for key aspects of characters like yield and quality is one method to ensure that selection for yield and end-use characters is practised as early as possible in a breeding programme to ensure that advances are made. Nearly all, if not all, barley breeders probably use molecular markers in some aspect of their breeding schemes. What is clear is that efficient genotyping can now be highly automated and high throughput but such an implementation is largely confined to the major commercial breeding companies that have invested in the technology for other, more profitable, crops than barley.

The simplest application of markers in breeding is in the characterization of breeding germplasm so that breeders can now make more informed choices about the crosses that they make. For instance the figure below shows that if one crossed the malting winter barley cultivar Flagon with the feed cultivar Accrue, then all the progeny would be expected to have the malting alleles at the QTL region highlighted. If, however, Flagon was crossed with the feed variety Retriever, then half of the progeny would have the feed alleles and half the malting alleles, assuming no recombination has occurred. Such information can guide the breeder on issues such as population size and the selection that might need to be applied to achieve the desired recombinations of characters.

Breeders then have to choose whether or not to deploy markers to select amongst progeny for characters – marker-assisted selection (MAS). Using MAS instead of phenotypic selection is advantageous when phenotypes show complex inheritance, interact with the environment or developmental stage, identification of relatively rare allelic combinations, and the pyramiding of resistant alleles at different genetic loci. Clearly, malting quality is a complex character and interacts with the environment so would be suitable for MAS but, despite much research on identifying markers associated with malting quality characteristics, the number of targets for MAS that are actively used is largely limited to selected candidate genes from amongst those listed inMQ & Other Major Genes Genome map001the map. The Scotch Whisky industry has taken steps to eliminate ethyl carbamate precursors from barley to ensure that levels of the carcinogen are well below regulatory limits. Research has led to a molecular marker that is diagnostic for non-production of epiheterodendrin that is being routinely used by European breeders to select lines for the distilling market and by the Institute of Brewing and Distilling (IBD) Malting Barley Committee to flag the varieties that they promote to growers for use by malt and grain distillers.

It is clear from the rapid pedigree breeding system described above that breeders are basing pure stock production on progenies originating from F4 plants at the very least. One would expect an average of 12.5% heterozygosity at each locus that was segregating in the original cross so the probability that a stock that can be randomly derived that is uniform and stable is low, leading to probable failure at the DUS testing stage. Markers can, however, greatly improve the chances of deriving a pure-bred stock, especially if the genes affecting the DUS characters are known and the associations of a range of DUS characters with DNA markers have been placed on a genetic map. Using this knowledge, breeders can identify mother plants that are homozygous at the key loci and thus can be expected to breed true and be far more liable to pass DUS tests. In addition, this strategy can be combined with a DNA fingerprinting analysis of the genetic background of advanced lines to identify lines that carry the appropriate parental combinations in genomic regions relevant to yield and end-use quality.

Breeding Progress
Much of the improvement in UK crop yield is due to improved cultivars. In the AGOUEB project, we grew a subset of cultivars representing 25 years of recent breeding history in the same fungicide-treated trial over a 3 year period at 7-8 UK sites each year and estimated a significant rate of breeding progress for yield at over 0.5% per annum. Despite the increase in genetic potential, analysis of official statistics shows that on farm average yield figures appear to have reached a plateau in recent years. The reasons behind this decline have yet to emerge but it is probable that on-farm yields would be in decline without the development of new cultivars.

Samples from the trials were micro-malted in collaboration with member companies of the MAGB and the data showed that there had been a significant increase in malt extract potential for spring barley, even when the analysis was restricted to just those cultivars that had received at least a provisional recommendation for malting use by the Institute of Brewing and Distilling. For winter barley, the scatter of points was considerable meaning that the relationship of malt extract with time was non-significant but it is noticeable that the mean malt extract was some 10 litre°/kg lower than that of the spring crop. This means that the spirit yield per tonne of malt used is likely to be significantly less from the winter crop compared to the spring as there was no significant relationship of the fermentability of malt extract over the time period surveyed nor any significant difference between the crop types for this character. Taken together, there has been an overall improvement in the predicted spirit production per tonne of malt from 1991 to 2013 (Bringhurst, 2015). Whilst the grain nitrogen content of both crop types has decreased significantly over the period, there were few significant trends amongst other malting quality parameters. There has been a significant decrease in wort viscosity in winter barley of approximately 0.3 milli-Pascals p.a. but it was still greater than the average viscosity of the spring barley malting cultivars. Whilst the wort β-glucan content has also decreased over the same time-period, the relationship with time was non-significant and so the improvements in this aspect of malt processability are probably not solely due to changes in β-glucan synthesis and degradation potential.