Breeding methods have, naturally, been much influenced by the outbreeding nature of the crop. Rye has a gametophytic two-locus incompatibility system (Lundquist 1956). Early breeding approaches could best described as forms of simple recurrent selection. With improved genetic knowledge more sophisticated methods appeared. Rye shows inbreeding depression but inbred lines of acceptable vigor can be isolated and used in the construction of synthetic (and even hybrid) varieties, following suitable progeny tests for combining ability. Unlike those of wheat and barley breeding, the objectives of rye breeding have not been dominated by aspects of disease resistance, at least at the beginning. Improvement of grain yield, improved stability, fast growing, fine stemmed, resistance to powdery mildew and brown rust, protein content and quality together with cold tolerance and shorter straw, have been the aims of recent breeding (Tab. 7.1). Less problematic food and baking qualities and resistance to deformity largely determine the value of bread rye varieties. Varieties with plenty of leaf mass are particularly suited for use as green fodder.
DIPLOID RYE
Population breeding comprises the development of open-pollinated and synthetic varieties. In both cases the variety constitutes a panmictic population. The population is produced by random fertilization, at least in the final generation of seed production. The gametophytic self-incompatibility prevents self-fertilization under open pollination. It helps to avoid inbreeding depression
1. Population breeding: In population breeding, only self-incompatible varieties have usually been produced. An open-pollinated variety is the direct outcome of population improvement. When a breeding population has reached a per se performance level comparable to that of existing cultivars, it may be considered as a new variety. Depending on the local statutory regulations it can be released and registered as variety.
Various selection procedures are being described for population improvement in rye (Ferwerda 1956, Laube and Quadt 1959, Sengbusch 1940, Vettel and Plarre 1955, Wolski 1975, Geiger 1982). Self-incompatible and self-fertile rye populations can be considered. Improvement of self-compatible populations aims at improving either the per se performance, the potential of the population for synthetic variety production, or both. In self-fertile material, selection exclusively aims at improving the potential of the population for hybrid variety production
For all approaches the intra- and/or inter-population general combining ability of the parental units, i.e. plants, clones, pairs of plants or pairs of clones, has to be improved. In addition, for hybrid breeding the mutational load of the populations has to be diminished in order to reduce inbreeding depression in line establishment. As given in Tables 7.1.1.1 and 7.1.1.2 the different selection procedures can be operated in various ways according to the experimental facilities. Moreover, the procedures can be used simultaneously or successively in a given selection scheme. It follows the generalized scheme of population improvement of Halauer and Miranda (1981). The breeding procedure is divided into different selection cycles. Each cycle includes a parental unit (plants, clones, pairs of plants, or pairs of clones to be evaluated), a selection unit (plants, clones, or progeny that provide the data used as the basis for selection), and a recombination unit (plants, clones, or progeny that finally recombined to form the improved population).
Cloning of parental material is necessary to achieve sufficient seeds for progeny tests in multi-location yield trials, or even for micro-plots. The number of clones per genotype can be limited or even few when CMS testers are involved.
Selfing of heat-treated self-incompatible S0 plants generally yields few seeds only. In order to increase the number of seeds for yield trials, the S1 lines have to be multiplied by free pollination under isolation cages and/or in isolation.
Considering a wide range of experiments, it is difficult to predict an optimum number of parental, selection and recombination units and an optimum allocation of testing facilities. Even if just one selection procedure is taken into consideration, the optima may vary considerably depending on the underlying genetic and environmental factors
2. Synthetics: In rye the term synthetics is used to designate varieties that are produced by crossing inter se a number of selected parents with subsequent multiplication by open pollination under isolation. The parents can be clones, inbred families, or other genotypes. Long-term maintenance of clones was not feasible under practical conditions. However, their gametic arrays can be reproduced from S1 families, which can be established by selfing under heat treatment (Wricke, 1978).
The potential of a genotype as constituent of a synthetic variety is indicated by its general combining ability. Population improvement and the formation of synthetics ought to be organized as an integrated program. When a high testing accuracy is practiced in population improvement, a selected fraction of the recombined material may serve as varietal parent without additional testing. The optimal number of parents is determined by almost the same parameters as the optimum population size in recurrent selection. Since synthetic breeding is not directed on selection with varieties but on creating better parents, intra-varietal genetic variance is of minor concern, and the number of varietal parents can be chosen smaller than in long-term population improvement. Studies on rye revealed that the genetic variance among synthetics sharply decreases as the number of parent’s increases (Geiger et al., 1981). If several unrelated, well combining, and high performing plant populations are to the breeder’s disposal, the question arises whether a synthetic variety should be composed of parents from one single or from a certain number of such populations. Only in case a population is outstanding in both, per se performance and variance of general combining ability, it is likely to be better suited as source for synthetic than any population set. In all other situations, the optimum has to be determined by predicting the expected performance of the best synthetic of each set of populations. Frequently, two or more populations turn out to be more promising than just one.
By the complex theoretical and practical situation, synthetic rye varieties have not gained acceptance at the seed market. If improvement of both, hybrid and population varieties, is conducted at the same breeding company, and integrated approach would be desirable to make more use of genetic resources, labor supply, and technical equipment, a general scheme for simultaneous development of hybrid and synthetic varieties in rye was proposed by Geiger (1982) (Fig. 7.1.2.1).
3. Hybrid breeding: The main goal of hybrid breeding is a stable grain yield. This includes tolerance to drought and nutrient stress because rye is widely grown on poor, sandy soils where it has a higher relative performance than wheat and triticale. Caused by hybrid breeding, rye can now compete with these cereals even on more fertile, productive soils. Since hybrids are genetically more uniform than population varieties, breeding for disease resistance is urgently needed, especially for those diseases that cannot be prevented by chemical means (Miedaner and Geiger, 1999; Miedaner et al., 1995). Concerning quality, a large grain weight and resistance to pre-harvest sprouting are the most important features.
Hybrid breeding started in 1970 in Germany. The first three hybrid varieties “Forte”, “Aktion”, and “Akkord” were released in Germany in 1984. In multilocal testing trials they surpassed standard “population” varieties with approximately 10 % in yield. Moreover, they showed shorter straw, better resistance to preharvest sprouting, but lower thousand-kernel weight and unchanged lodging resistance.
Today, more than 20 hybrid varieties are on the official list, occupying more than 60 % of the total rye acreage. Some of these hybrids also are registered and distributed in Austria, Denmark, France, UK, Scandinavian countries and Netherlands. In Sweden, one and in Poland, two independent hybrid rye breeding programs are running at present with the first released Polish hybrids in 1999. In addition, in Russia, several programs are conducted in different areas of the country. Outside Europe, the only hybrid-rye breeding program is situated at the University of Sidney, Australia.
Rye is the only cross-pollinated species among the small grain cereals. Selfing is naturally prevented by an effective gametophytic self-incompatibility mechanism. Self-fertile forms have been found in several populations and are routineously used for developing inbred lines. Selfing results in strong inbreeding depression and hybrids display a high heterosis. For grain yield, new data show a relative midparent heterosis of about 100 %. Heterosis can be exploited only, when pre-selected inbred lines of different gene pools are crossed to an F1 hybrid. The use of a cytoplasmic-genic male sterility is necessary. Male sterility is mainly caused by introgression in the “Pampa“ cytoplasm of Argentinean rye that is environmentally highly stable. Pollen fertility in the hybrids is restored by the use of dominant nuclear-coded restorer genes from European or exotic populations.
Systematic search for gene pools with maximal heterosis revealed that two German populations “Petkus“ and “Carsten“ were particularly well matching. Inbred lines from the two pools are used successfully for the development of hybrids. A simplified scheme of hybrid rye breeding can be taken from Fig. 7.1.3.1. Seed-parent lines are developed from “Petkus“ and pollinator lines from “Carsten“ gene pool. Intensive selection for line performance is practiced in selfing generations S1 and S2.
Selfing is done by hand under isolation bags. After one or two stages of selection, the seed-parent lines are transferred into the CMS-inducing “Pampa” cytoplasm by repeated backcrossing yielding in backcross progenies. BC1 and BC2 are subsequently crossed to parents from the opposite pool to select for testcross performance, i.e. combining ability for grain yield. The reverse procedure is used for the pollinator lines. They are grown between isolation walls in adequate plots and crossed to CMS single crosses as testers. The testcrosses are evaluated in multi-environmental trials with two replications.
Commercial hybrids are produced between CMS single crosses as seed parents and restorer synthetics as pollinator parents. The latter are mostly composed of two inbred lines crossed by hand and further multiplied by random open pollination. This complex type of hybrid ("topcross hybrid") needs several stages/years for production (Fig. 7.1.3.1). The main advantages are a cheaper and more stable seed production with less risk for the breeder, a higher vigor of the hybrid seed, and an extended period of pollen shedding by the pollinator synthetic.
Substantial progress in grain yield has been achieved in hybrid rye breeding. In 1999, the best hybrid surpassed the best open-pollinated check by 20 % in the official German trials. Over the years, progress was significantly higher for hybrid than for population breeding. This clearly reflects that hybrid breeding is genetically the most efficient method available. Progress was also made for other important traits, such as lodging resistance, leaf rust resistance, and bread-making quality. Particularly the lodging resistance has to be improved, when rye intensively is produced and yields higher than 10 t/ha are common. Hybrids, therefore, became more attractive to farmers than the traditional populations. Progress in hybrid breeding steadily increased this superiority since the 1990th (Geiger 1990).
TETRAPLOID RYE
Most cultivated rye contains seven pairs of somatic chromosomes. Artificially produced tetraploid rye with 14 pairs of chromosomes is grown for seed production in limited amounts in Europe. As forage crop, it is grown in large scale either individually or in mixtures with other forage crops in USA and other countries. Those tetraploid ryes often show perennial growth habit by introgression of S. montanum.
With the 1937 discovery of the colchicine method for inducing polyploidy, development of polyploid crops was regarded as an unconventional technique to penetrate yield and other barriers in plant breeding. The two more or less universal effects of chromosomal doubling are increased cell size and decreased fertility. Consequently, crops that benefit most from increased cell size and suffer least from reduced fertility are inherently predisposed to benefit from polyploid breeding. Crops most amenable to improvement through chromosome doubling should (1) have a low chromosome number, (2) be harvested primarily for their vegetative parts, and (3) be cross-pollinating. Two other conditions, the perennial habit and vegetative reproduction, have a bearing on the success of polyploid breeding by reducing a crop's dependence on seed production. In the 1950 -1970's, induced autopolyploidy in rye was considered to be an important breeding method. Rye showed good prerequisites for an autopolyploid crop. Russia, Poland, Germany, and Sweden spent considerable attention to the production of tetraploid rye.
In order to overcome low fertility, seed shriveling and aneuploid offspring – all three features were believed to be influenced by irregular meiotic chromosome pairing (Tab. 7.2.1) – two major approaches of chromosomal pairing regulation were investigated. (1) The research group belong H. Rees at Aberystwyth (UK) (Hazarika and Rees 1967) favored an increased quadrivalent formation with convergent or parallel centromere co-orientation (Fig. 7.2.1) as a mean of reduced aneuploidy and, consequently improved fertility. A “disjunction index” (number of pollen mother cells without univalents and trivalents divided by the total number of pollen mother cells) was used as measure for meiotic pairing regularity. They even demonstrated a positive correlation between the disjunction index and fertility.
(2) Alternatively, Sybenga (1964) proposed a preferential bivalent pairing in order to reduce meiotic irregularities by induced allopolyploidization of an autotetraploid. Beside a genetic control of diploid-like chromosome pairing, a complex system of experimentally induced reciprocal translocations was intensively discussed. Despite strong efforts and tests over decades no practical benefit could be achieved.
Schlegel (1976) introduced a method for mass production of autotetraploid rye by so-called valence crosses. Under microplot isolators non-emasculated tetraploid genotypes (clonal plants) were crossed by spontaneous pollination with diploid genotypes (clonal plants). The tetraploids used as mother plants showed a recessive pale grain character, while the diploids used as male plants showed dominantly green seeds. In this way, green xenia could be selected among the pale grains after harvest of the mother plants. All the green xenia must be hybrids, either triploid or tetraploid. Microscopic chromosome counting can differentiate them. The triploids are resulting from fusion of a diploid female gamete and a haploid male gamete, while the tetraploids are derived from a diploid female and an unreduced male gamete (Schlegel and Mettin, 1975).
In this way, diploid genotypes showing high chiasma frequencies per PMC (>17 Xta/PMC) and other useful agronomic characters could be transferred into the so-called half-meiotic tetraploids. The method proved to be efficient for broadening genetic variability of tetraploid breeding material without deleterious effects of colchicine treatment. Until this time all tetraploids available worldwide based on colchicine-induced genotypes in a very limited number.
On the other hand, the higher chiasma frequency of the diploid parent (Fig. 7.2.2) did not significantly influence the number of multivalents of the tetraploids, despite the missing correlation between the quadrivalent frequency, aneuploid frequency and fertility, at least in advanced breeding strains. However, depending on the karyological structure of the diploid parent, a more or less strong modification of the bivalent frequencies could be observed, i.e. as bigger the differentiation of the chromosome structure between the parental genotypes as higher was the number of bivalents per PMC. Although just minor changes of the karyotype have been considered, they contributed to a preferential bivalent pairing, obviously within the parental diploid genomes of both female and male donors.
Summarizing the data over 50 years research, neither an increased chiasma frequency and/or higher frequency of quadrivalents per PMC nor a diploidization mechanism substantially contributed to breeding progress. Interchromosomal substitutions as a source of irregular chromosome pairing and aneuploidy, i. e. aneusomy, could be excluded as well (Schlegel et al., 1985). Gradual selection for yield, fertility and low degree of shriveling was more successful than experimentally induced genetic or cytological changes.
