Breeding for disease resistance in wheat
R.P. Singh, S. Rajaram

Diseases of wheat, mostly caused by fungal pathogens and a few by viruses and bacteria, are important production constraints in almost all wheat-growing environments (Rajaram and van Ginkel, 1996; McIntosh, 1998). The long-term success of breeding for disease resistance is influenced by the following factors:

• the nature of the pathogen and diversity of virulence in the population;
• availability, diversity and type of genetic resistance;
• screening methodology and selection environment for tracking resistance.

FACTORS IN BREEDING FOR DISEASE RESISTANCE

Nature of the pathogen

Globally important fungal diseases of wheat caused by obligate parasites include the three rusts, powdery mildew, the bunts and smuts, etc.; whereas those caused by facultative parasites include Septoria tritici blotch, Septoria nodorum blotch, spot blotch, tan spot, scab, etc. The obligate parasites are highly specialized, and significant variation exists in the pathogen population for virulence to specific resistance genes. Evolution of new virulence through migration, mutation, recombination of existing virulences and their selection is more frequent in rust and powdery mildew fungi. Therefore, breeding for resistance to these diseases has always been more dynamic. Physiological races are known to occur for most bunts and smuts; however, evolution and selection of new races is less frequent. Because most bunts and smuts can be easily controlled by chemical seed treatment, little effort is currently placed on resistance breeding. Changes in pathogen races are even less frequent for diseases caused by facultative parasites, possibly because of the lack of significant advantage in survival of the new race over the old races during the off-season in stubble.

Availability, diversity and type of resistance

Diversity for resistance to all important wheat pathogens occurs either within the germplasm or in related species or genera. Resistance is broadly categorized into two groups: race-specific type and race-nonspecific type. A majority of the genes catalogued for resistance to rusts, powdery mildew, smuts and bunts and several other facultative parasites appears to be race-specific and follow the gene-for-gene concept (Flor, 1956). Race-nonspecific type of resistance is also reported to occur for rusts, mildews and various other facultative parasites. The race-specific type of resistance is controlled by genes with major to intermediate effects; whereas the race-nonspecific resistance is mostly controlled by genes with minor to intermediate and additive effects. Race-specific resistance to rusts and powdery mildew is short lived, often lasting for an average of about five years when deployed. In contrast, race-specific resistance to bunts and smuts and facultative parasites tends to last for a long time. Information about the type of resistance is therefore useful in choosing the most appropriate breeding and selection methodology.

Screening methodology and selection environment

The probability of identifying resistant parents and resistant progenies is increased by the availability of a reliable screening methodology and an environment favourable for disease development. Depending on the disease and choice of the type of resistance, the methodology may require simple tests in the greenhouse on seedlings or adult plants, replicated field tests, or even the use of resistance-linked protein and DNA markers. Protocols for screening for resistance to most diseases are well established and can be employed in breeding for resistance. Inclusion of check cultivars for resistance and susceptibility is important to assess the disease pressure and degree of resistance. Choice of field sites with reliable environmental conditions is crucial for progress when selection is to be carried out in field conditions. The wheat breeding programme in Mexico shuttles the segregating populations between sites in Mexico, or sites between Mexico and hot-spot locations outside Mexico, to select multiple disease resistances.

APPROACH TOWARDS BREEDING FOR DISEASE RESISTANCE

The International Maize and Wheat Improvement Center’s (CIMMYT) spring wheat breeding programme has carefully defined its target production environments into six mega-environments (Rajaram et al., 1995). A mega-environment (ME) is defined as a broad not necessarily contiguous area, occurring in more than one country and frequently trans-continental, which is characterized by similar biotic and abiotic stresses, cropping-system requirements, consumer preferences and, for convenience, by volume of production. For a more extensive description of mega-environments, see chapter "CIMMYT international wheat breeding". Germplasm generated for a given mega-environment is useful throughout it, accommodating major stresses, but perhaps not all the significant secondary stresses. The six mega-environments for spring wheats include:

• ME1: Irrigated, Temperate. Leaf and stripe rusts are currently the major diseases. Stem rust, which historically caused havoc, is under control since the cultivation of semidwarf wheats. Karnal bunt and powdery mildew can be a problem in some areas.

• ME2: High Rainfall. Stripe rust, leaf rust, Septorias, Fusariums and barley yellow dwarf (BYD) are the most important diseases. Stem rust and bacteria may also occur in some areas, but are less important. Root rots are considered important, but little progress is achieved in resistance breeding.

• ME3: Acid Soils. Stripe rust is of lesser importance, but stem rust is considered an important disease in this mega-environment. All other diseases of ME2 are of similar importance. Powdery mildew is also common.

• ME4: Low Rainfall. Rusts, especially leaf and stripe, powdery mildew and bunts and smuts are considered important in some years. Years when rainfall is higher, Septorias and Fusariums can be a problem.

• ME5: High Temperature. This mega-environment is subdivided in two MEs: ME5A where leaf rust and Helminthosporiums are major constraints and ME5B where diseases are almost non-existent.

• ME6: High Latitude. Leaf rust and stem rust are the most important diseases, and Helminthosporiums, Septorias and Fusariums are of lesser importance.

Significant progress is evident in incorporating durable resistance to all three rusts and Septoria tritici blotch in CIMMYT’s spring wheat germplasm. Breeding for resistance to scab, spot blotch, Karnal bunt and BYD has been emphasized since the early to mid-1980s, and progress can be noted. The breeding programme at CIMMYT makes all efforts to utilize diverse spring and winter wheats in its crossing programme, which results in diversity for resistance to most diseases in the germplasm. This diversity is identified through the multilocational global testing of advanced breeding lines.

RESISTANCE TO RUSTS

With the discovery of the genetic basis of resistance by Biffen (1905), physiological specialization in rust pathogens by Stakman and Levine (1962) and gene-for-gene interaction by Flor (1956), the utilization of the hypersensitive (race-specific) type of resistance has dominated in wheat improvement. This approach appeared to be very attractive from the crop cleanliness point of view and because it is simple to incorporate into improved germplasm. The phenomenon of the erosion of such genes, or their combinations, led scientists to look for alternative approaches to resistance management. The multilineal approach promoted by Jensen (1952) and Borlaug (1953) emerged out of the frustrations associated with the frequent failures of race-specific genes. Van der Plank (1963) was the first epidemiologist to clearly define the theoretical basis of concepts of resistance. In the late 1960s and 1970s, there was a revival of the concept of general (race-nonspecific) resistance and its application in crop improvement (Caldwell, 1968). This approach was widely used for breeding stem rust resistance in wheat by Borlaug (1972), leaf rust resistance by Caldwell (1968) and yellow rust resistance by Johnson (1988). The wide-scale application of such a concept in breeding for leaf rust resistance, commonly known as slow rusting, has dominated in CIMMYT’s bread wheat improvement for more than 25 years. During this period, numerous terms have been used in the literature to describe various features of resistance. A simple term is often not sufficient, given that resistance can be described based on its epidemiological and genetic characteristics. The terms used in this chapter are defined as following:

• Race-specific resistances are easily detected with specific pathotypes or races of the pathogen and are controlled by genes having major effects. In wheat-rust pathosystems, these resistances are recognized by characteristic low-infection types. Numerous genes are now known and have been catalogued by McIntosh et al. (1995). Most of these genes can be detected in seedling evaluations using specific pathotypes. However, detection of a few others requires testing at post-seedling growth stages. Major genes are implicitly vulnerable to pathogen plasticity, and their longevity can range from rapid vulnerability to relative (and often deceiving) durability. It is likely that most specific resistances, whether based on a single major gene or a combination of major genes, will sooner or later succumb to new adaptive pathotypes if careful deployment is not practised.

• Race-nonspecific resistances operate against all pathotypes or races of a pathogen. The genetic nature of this type of rust resistance is usually complex and based on the additive interaction of a few or several genes having minor to intermediate effects.

• Slow rusting and partial resistances are almost synonymous terms. As defined by Caldwell (1968), slow rusting is a type of resistance where disease progresses at a retarded rate, resulting in intermediate to low disease levels against all pathotypes of a pathogen. Partial resistance, as defined by Parlevliet (1975) referring to leaf rust resistance in barley, is a form of incomplete resistance characterized by a reduced rate of epidemic development despite a high- or susceptible-in-fection type. The components that cause slow rusting of a cultivar are longer latent period, low receptivity or infection frequency, as well as smaller uredial size and reduced duration and quantity of spore production. All these components can affect disease progress in the field.

• Durable resistance, as defined by Johnson (1988), is that which has remained effective in a cultivar during its widespread cultivation for a long sequence of generations or period of time in an environment favourable to a disease or pest.

Attaining durable resistance

Since wheat cultivars derived from CIM-MYT germplasm are grown in a large area and exposed to a variety of pathogens under conditions that may favour disease development, the CIMMYT strategy has been to utilize germplasm sources that are as diverse as possible for rust resistance. The flow of germplasm to and from the bread wheat improvement programme is continuous, and the scientists are in constant contact with national programme colleagues to ensure this exchange. Although multilocational testing is not a perfect system for identifying diverse resistance sources, evidence accumulated by CIMMYT over many years indicates that it has greatly facilitated the confirmation of the existence of genetic diversity in CIMMYT’s germplasm. Lines showing stable disease performance across locations are especially useful for understanding the genetic basis of resistance. Genetic studies have suggested that wheat genotypes that are resistant to a given rust disease in many locations, as indicated by low average coefficients of infection, often contain multiple major or minor genes for resistance.

TABLE 8.1
Genetic diversity in 280 advanced lines of the 24^th International Bread Wheat Screening Nursery classified into average coefficients of infection (ACI) for three rusts in international multilocation testing

Table 8.1 shows the phenotypic diversity for resistance to leaf rust (caused by Puccinia triticina), stripe rust (caused by P. striiformis f. sp. tritici) and stem rust (caused by P. graminis f. sp. tritici) of 280 advanced bread wheat lines included in the 24^th International Bread Wheat Screening Nursery (IBWSN). Marked differences among phenotypes suggest the existence of different groups of varieties where response to rust could be under different genetic control. Significant achievement is evident for leaf and stem rust resistance because approximately 60 percent of the entries have average coefficient of infection (ACI) values of less than 5 (Table 8.1). How-ever, much more progress can be expected in stripe rust resistance, as only approximately 10 percent of the entries had such ACI values. This has been due to the absence until 1995 in Mexico of virulence for the Yr9 gene located in the 1B/1R translocation present in wheat lines Veery and Bobwhite and their numerous derivatives. Therefore, several selections that were resistant in Mexico until 1995 did not show enough resistance at locations where Yr9 virulence was present. This does not mean that the presence of 1B/1R is associated with genetic vulnerability; in fact, some 1B/1R lines (including certain selections of Attila, Catbird, Kauz, Lira and Picus) have remained resistant even in the presence of Yr9 virulence, indicating the presence of additional resistance genes. Data from such hot spot locations could serve as an early warning of the breakdown of resistance.

Multilocational testing of regionally important advanced breeding materials in the given epidemiological zone is also encouraged so that performance may be judged against the available variation in the pathogen in diverse environments. Such nurseries could also include a few susceptible and slow rusting check cultivars.

Durability of resistance

Genetic diversity and durability are the two most important features of the resistance sought by CIMMYT for the global wheat improvement programme. Because proof of durability comes only after resistance is deployed over a large area, genetic diversity serves as insurance against vulnerability. Historical performance of resistances could help identify durable resistance sources. Genetic analysis to understand the genetic basis of such resistance could aid the directed transfer of resistance as well as the search for additional genes that could contribute to new durable resistance gene combinations. Durable resistances to all three rusts in CIM-MYT-derived spring wheats are known and are discussed below.

Genes for durable resistance to stem rust

Stem rust resistance gene Sr2, in addition to other unknown minor genes derived from cultivar Hope, provided the durable resistance foundation in Rockefeller-Mexican spring wheat germplasm led by Dr N.E. Borlaug. Cultivar Yaqui 50, released in Mexico during 1950, and other Sr2-carrying wheats released since then have stabilized the stem rust situation in Mexico. In recent studies, change in stem rust pathotypes in Mexico has not been not observed. Released in 1960 in the Indian subcontinent and subsequently grown on millions of hectares, the cultivar Sonalika has also remained resistant. When present alone, the Sr2 gene confers slow rusting, which often may not be adequate under heavy disease pressure. However, it provides adequate resistance in combination with other major or minor genes. The Sr2 gene can be identified by its genetic linkage with the pseudo-black chaff or brown necrosis phenotype seen on the glumes and below the nodes. Unfortunately, not much is known about the other genes and their interactions in the Sr2 complex.

Knott (1988) has shown that adequate levels of multigenic resistance to stem rust can be selected by accumulating approximately five minor genes. In his studies, the genes were different from Sr2. It is likely that similar genes are present in CIMMYT germplasm, but more research is needed to document their existence.

Genes for durable resistance to leaf rust

The South American cultivar Frontana is considered to be one of the best sources of durable resistance to leaf rust (Roelfs, 1988). The Rockefeller-Mexican Program first used the variety in the 1950s. Later derivatives, such as Penjamo 62, Torim 73, Kalyan/Bluebird, etc., showed slow rusting characteristics possibly derived from Frontana. Genetic analysis of Frontana and various CIMMYT wheats possessing excellent partial resistance to leaf rust worldwide has indicated that such adult plant resistance is based on the additive interaction of Lr34 and two or three additional slow rusting genes (Singh and Rajaram, 1992). In Mexico, leaf rust severity on most cultivars can be related to the number of slow rusting genes they carry (Table 8.2). When susceptible cultivars display 100 percent leaf rust severity, cultivars with only Lr34 display approximately 40 percent severity; cultivars with Lr34 and one or two additional minor genes display 10 to 15 percent severity; and cultivars with Lr34 and two or three additional genes display 1 to 5 percent severity. Leaf rust could further increase to unacceptable levels on cultivars carrying only Lr34 or Lr34 and one or two additional genes. However, cultivars with Lr34 and two or three additional genes show a stable response in environments tested so far, with final leaf rust ratings lower than 10 percent. Some cultivars carrying Lr34 and two or three additional genes are listed in Table 8.2. The presence of Lr34 can be indicated by the presence of leaf tip necrosis in adult plants, which is closely linked with it (Singh, 1992a).

TABLE 8.2
Some seedling susceptible bread wheats that carry good adult plant resistance to leaf rust

^aLeaf rust response has two components: percent severity based on the modified Cobb scale (Peterson et al., 1948) and reaction based on Roelfs et al., (1992). The reactions are: MSS = moderately susceptible to susceptible, i.e. medium - to largesized uredia without chlorosis or necrosis; S susceptible, i.e., large uredia without chlorosis or necrosis; N = necrotic, i.e. necrotic leaves following hight leaf rust severity.

Slow rusting can be characterized in greenhouse experiments by evaluating the latent period, uredial number (or infection frequency, or receptivity), uredial size, inoculum production, etc. under quantitative inoculation. Characterization of 27 bread wheats of CIMMYT origin by Singh et al. (1991) indicated that this was phenotypically diverse for all components measured (Table 8.3). The area under disease progress curve of these wheat lines in the field ranged from 1 to 50 percent of the very susceptible check cultivar Morocco (Table 8.3). Singh et al. (1991) also reported the likelihood of pleiotropic genetic control of the components of slow rusting because of highly significant positive or negative phenotypic correlation among the latent period, uredial number and uredial size. If it is assumed that same gene controls various components of slow rusting, then it can be hypothesized that perhaps only a few genes with additive effects could retard disease progress to a rate that final disease level remains to an acceptable low level.

TABLE 8.3
Range of variability for components of slow rusting resistance to leaf rust observed in 27 bread wheats, given as percent of susceptible check variety Morocco

The heterogeneous Mexican wheat cultivar Jupateco 73 was reselected for the presence and absence of Lr34 by Singh (1992a). These isogenic Jupateco 73R (Lr34 present) and Jupateco 73S (Lr34 absent) genotypes and those of Thatcher developed by Dyck (1987) have yielded useful information on the nature of slow rusting resistance. The studies, using the Jupateco pair, have shown that Lr34 affects all three components of slow rusting, i.e. it increases latent period and decreases uredial number and uredial size (Table 8.4). The effect was more pronounced in post-seedling growth stages, although measurable differences also occurred in the seedling stage (Table 8.4). Temperature can also influence the expression of resistance conferred by the gene Lr34 (Singh and Gupta, 1992). Comparison of grain yields of Jupateco isolines in leaf rust protected (by fungicide) and non-protected plots indicated that though leaf rust could significantly reduce grain yield by approximately 15 percent in the presence of Lr34, the reductions in the absence of Lr34 were substantially higher and ranged between 42.5 to 84 percent depending on planting date and year (Singh and Huerta-Espino, 1997).

Rubiales and Niks (1995) studied the infection process and indicated that slow rusting resistance due to Lr34 was based on a reduced rate of haustorium formation in the early stages of infection, in association with no or relatively little plant cell necrosis. Electron microscopic studies of Alvarez-Zamorano (1995) on Jupateco 73 isolines have shown an accumulation of unknown electrodense substances in the cells of Lr34 line near the site where haustorial mother cells try to dissolve the cell wall of mesophyll cells for the formation of haustoria. It would appear that the accumulation (cell wall apposition) causes a thickening of the cell wall, which reduces the establishment of the haustorial tube. If haustoria are formed, the slow mycelial growth may be due to a restricted movement of fungus from one cell to another by a similar phenomenon. Alvarez-Zamorano (1995) also observed structural change in the Lr34 line leading to invagination, or contraction of the cell wall, which may delay the completion of the infection process. These observations indicate a different mechanism for Lr34-based slow rusting than hypersensitivity, which is associated with race-specific genes. Because pathogen isolates can vary for aggressiveness (Lehman and Shaner, 1996), it may be difficult to differentiate pathogenic variation for increased capability to overcome slow rusting resistance of this type from aggressiveness.

TABLE 8.4
Comparison of the three components of slow rusting resistance to leaf rust in seedling and flag leaves of the near-isogenic Lr34 Jupateco 73 reselections tested at 15°C

Slow rusting resistance to leaf rust is common in spring wheat germplasm. Studies have shown that at least 10 to 12 slow rusting genes are involved in the adult plant resistance of CIMMYT wheats. Lines such as Amadina (Table 8.2), where Lr34 is absent but whose level of resistance is high, have also been identified. Thus durable resistance is feasible even in the absence of Lr34, as in the case of Pavon 76 (Table 8.2) where a new gene Lr46 for slow rusting in chromosome 1B has been identified. Genes other than Lr34 and Lr46 are not yet named, though work is currently underway at CIMMYT to identify their chromosomal locations so that they may be catalogued. CIMMYT is also searching for molecular markers that may facilitate quick detection of the slow rusting genes.

Genes for durable resistance to stripe rust

In recent studies, Singh (1992b) and McIntosh (1992) have indicated that the moderate level of durable adult plant resistance to stripe rust of the CIMMYT-derived US wheat cultivar Anza and winter wheats such as Bezostaja is controlled in part by the Yr18 gene. This gene is completely linked with the Lr34 gene. The level of resistance it confers is usually not adequate when present alone. However, combinations of Yr18 and two to four additional slow rusting genes result in adequate resistance levels in most environments (Singh and Rajaram, 1994). Cultivars carrying such Yr18 complexes are listed in Table 8.5. Genes Lr34 and Yr18 occur frequently in germplasm developed at CIMMYT and in various countries. Using the Jupateco 73 near-isogenic reselections, studies at CIMMYT have shown that the gene Yr18 also increases latent period and decreases infection frequency and length of infection lesions (stripes) to stripe rust in greenhouse experiments (Table 8.6). The conclusion again was that these components were under pleiotropic genetic control. Comparison between stripe rust protected and non-protected treatments showed that stripe rust infection caused grain yield losses of 31 to 52 percent in Yr18-carrying Jupateco 73R and 74 to 94 percent Yr18-lacking Jupateco 73S (Ma and Singh, 1996). This shows that slow rusting resistance based on Yr18 protected grain yield in the range of 36 to 58 percent depending on the year and sowing date. This level of protection was not considered sufficient in the environment of Toluca, Mexico, where the experiments were conducted.

TABLE 8.5
Some seedling susceptible bread wheats that carry good adult plant resistance to stripe rust

^aStripe rust response has two components: percent severity based on the modified Cobb scale (Peterson et al., 1948) and reaction based on Roelfs et al. (1992). The reactions are: M = moderately resistant to moderately susceptible, i.e. sporulating stripes with necrosis and chlorosis; S = susceptible, i.e. sporulating stripes without chlorosis or necrosis.

TABLE.8.6
Comparison of the three components of slow rusting resistance to stripe rust in flag leaves of the near-isogenic Yr18 Jupateco 73 reselections tested at 15°C

FIGURE 8.1
Distribution of 308 wheat varieties that displayed susceptible seedling reactions to Puccinia striiformis race 14E14, for their area under the disease progress curve (AUDPC),Toluca, Mexico, 1994.

The slow rusting type of resistance is quite common in spring wheats, as illustrated in Figure 8.1 for 308 non-B/1R cultivars showing seedling susceptibility to Mexican pathotype 14E14. These cultivars have diverse origins. As evident in Figure 8.1, most cultivars have only intermediate levels of resistance. The genetic study on a few selected cultivars indicated the presence of genes different from Yr18. The authors believe that these slow rusting genes can be pyramided to achieve adequate resistance levels. The durability of such slow rusting genes is not known; however, when combinations are deployed, the longevity of the resistance is expected to be high.

Because stripe rust can develop systemically, it is different from the other two rusts where every new pustule develops from a new infection. The epidemiology of stripe rust is also different from that of the other two rusts. Johnson (1988) presented examples of adult plant resistance genes that are race-specific in nature. It is difficult to distinguish such resistance from the resistance conferred by genes of a race-nonspecific nature based on the adult plant infection type. Low disease severity to stripe rust is most often associated with at least some reduction in infection type. However, it has been observed that in the case of potentially durable slow rusting resistance, the first uredia to appear are moderately susceptible to susceptible. Subsequent growth of the fungal mycelium causes some chlorosis and necrosis; therefore, the final infection type is usually rated as moderately resistant to moderately susceptible. Durability of such resistance can be expected if the cultivar’s low disease severity is due to the additive interaction of several (three to five) partially effective genes.

Resistance based on the additive interactions of slow rusting genes

It is often believed that selecting for resistance based on additive minor genes is difficult. However, at CIMMYT certain steps aimed at enhancing the accumulation of such genes are being taken. These steps are:

• Selecting parents that lack effective major genes and have moderate to good levels of slow rusting resistance to the local rust patho-types. Such parents are easily identified by testing them at the seedling stage in the green-house and as adult plants in the field with the same pathotype. The parents of interest should show susceptibility at the seedling stage and slow rusting in the field. Known cultivars with durable resistance are also included.

• Maintaining genetic diversity. Parents having different sets of additive genes based on available information are used in crossing. If such information is not available, parents of diverse origins or diverse pedigrees are selected for crosses.

• Establishing high disease pressure in the breeding nursery with chosen rust patho-types. Spreader rows are planted at optimum distance and artificially inoculated to ensure homogeneous disease spread of desired rust pathotypes in the plot. Susceptible and slow rusting checks are included to assess disease pressure.

• Selecting plants with low to moderate terminal disease severity in F₂ and F₃, and from F₄ onwards selecting plants or lines with low terminal severity. Because adequate resistance levels require the presence of three to five additive genes, the level of homozygosity from the F₄ generation onwards is usually sufficient to identify plants or lines that combine adequate resistance with good agronomic features. Moreover, selecting plants with low terminal disease severity under high disease pressure means that more additive genes may be present in those plants.

• Maintaining leaf tip necrosis or mild pseudo-black chaff phenotypes. Because leaf tip necrosis is linked with durable resistance genes Lr34 and Yr18, and pseudo-black chaff is linked with Sr2, these traits are useful mor-phological markers.

• Conducting multilocational testing. As discussed earlier, multilocational testing of useful advanced lines can indicate the effectiveness and stability of resistance across environments. Based on the results, new lines are identified for future crossing.

• Genetic analyses of selected lines. To confirm the presence of resistance based on additive genes, important lines are genetically analysed.

The authors believe that the chances of selecting resistance based on complex additive genes would increase greatly by following these steps. Some of these complexes may turn out to be durable when used commercially.

RESISTANCE TO KARNAL BUNT

Search for resistance to Karnal bunt (caused by Tilletia indica Mitra) at CIMMYT initiated during the early 1980s. Some cultivars of bread wheat were reported to be resistant in India. Extensive search for resistance at CIM-MYT has identified four principal sources of resistance: Indian, Chinese, Brazilian and the synthetic wheats produced at CIMMYT. Resistance appears to be based on a few partially dominant or partially recessive genes and is additive (Fuentes-Dávila et al., 1995). Synthetic wheats have derived their resistance from both the Triticum turgidum as well as Aegilops tauschii (syn. T. tauschii) parents, and some synthetics have shown immunity to the pathogen in repeated tests. It is also interesting to note that Karnal bunt does not occur in China or Brazil, but germplasm developed for resistance to scab carried Karnal bunt resistance. Several high-yielding advanced lines are now available that carry good levels of resistance to Karnal bunt.

RESISTANCE TO SEPTORIA TRITICI BLOTCH

Breeding for resistance to Septoria tritici blotch (caused by Septoria tritici) at CIMMYT initiated in early 1970. The susceptibility of the early semidwarf cultivars involved in the green revolution became evident when they were grown in North Africa. Steady progress has been made since then. Currently, several high-yielding semidwarf wheats with good resistance are available. Resistance in these wheats is derived from Argentine, Brazilian, Russian, West European and Chinese sources. The main problem encountered in early breeding work was to break the association of resistance with lateness and tallness present in the above sources. Two high-rainfall sites, Toluca (Mexico State) and Patzcuaro (Michoacan State), are used in Mexico for S. tritici resistance breeding. Some high-yielding, semidwarf and resistant lines are Milan (resistance derived from French source), Corydon (Brazilian source), Catbird (Chinese source), Bobwhite (Russian source), etc. Efforts are being made to combine these resistances. Genetic studies conducted on CIMMYT wheats indicate that between five to eight genes are operating in resistance to Septoria tritici blotch depending on the source population (Briceno, 1992; Jlibene et al., 1992; Matus-Tejos, 1993). Two to three genes are generally needed to confer an acceptable level of resistance, and genes have predominantly additive effects. The selection methodology, therefore, is similar to that described for combining minor, additive genes for leaf and stripe rust resistances.

Some synthetic wheats (T. turgidum/Ae. tauschii) developed at CIMMYT have shown excellent resistance, which appears to be leading towards immunity to the disease. These sources offer new genetic diversity of resistance originating from durum wheat and/or Ae. tauschii. Attempts are being made to transfer this resistance to semidwarf wheats and combine the resistance present in these synthetic wheats with other sources currently present in semidwarf wheats.

RESISTANCE TO SCAB

Scab, caused by Fusarium spp., is a major production constraint in the humid and semi-humid wheat areas of subtropical countries. Several species of the genus Fusarium are known to cause scab (Wiese, 1977). Fusarium graminearum Schwabe (perfect stage Gibberella zeae [Schw.] Petch) predominates in wheat-growing areas of China and North and South America (Luzzardi, 1985; Schroeder and Christensen, 1963; Zhong and Miller, 1988). CIMMYT has been involved in breeding for resistance to this disease since 1985. Sources of scab resistance have been divided into three groups according to their geographic origin: China and Japan, Argentina and Brazil, and Eastern Europe (Liu and Wang, 1991; Snijders, 1990).

Progress in breeding for scab resistance has been built largely due to the recent expansion of collaboration between Chinese and CIMMYT scientists. The dual objective was to introduce high-yield genes from new CIMMYT materials into resistant Chinese materials for those areas in China where direct CIMMYT material had not adapted well. Most foreign germplasm introductions into China had occurred in the 1950s and 1960s, and more recent materials had not been used in the Chinese breeding programmes. In addition, the aim was to utilize Chinese scab resistance in CIMMYT’s global breeding programme. In the past ten years, Chinese researchers have shared with CIMMYT more than 500 cultivars and advanced lines, many of them carrying high levels of resistance to scab. More than 10 000 materials were distributed from CIMMYT, either in the form of regular International Nurseries or materials selected by Chinese scientists themselves in CIMMYT’s fields in Mexico. The Chinese cultivars that best combined with CIMMYT materials to transmit scab resistance are Sumai#3, Ning 7840, Shanghai#5, Yangmai#6, Suzhoe#6, Wuhan#3 and Chuanmai 18. The following new germplasm for Chinese agriculture has evolved from the joint shuttle breeding effort:

• Sichuan Province: Chuanmai 25 (SW2089; Genaro 81 cross); SW5193, SW89-1862 (Veery cross); SW89-5422 (Alondra cross); SW90-1648 (Seri 82 cross); Catbird (Bagula cross made at CIMMYT).

• Jiangsu Province: Ningmai 7 (Shanghai 4-23B-0Y); Ningmai 8, Ning 9338 (Yangmai 158/Kauz); Ning 9341 (Yang87-158/Fasan); Ning 8675/Catbird; Ning 9350 (Catbird); Ning 9415 (Shanghai 7//Parula/Veery#6).

• Heilongjiang Province: Longmai 19, Ke92-779 (Roller cross).

Presently, throughout China, 5 to 7 million ha are cultivated to new varieties carrying CIMMYT germplasm in their pedigree, which represents about 25 percent of the total Chinese wheat area.

The Chinese sources are probably the best resistances currently available and must be combined with other sources of resistance. Genetic analysis results indicate that a few additive genes confer resistance in Chinese and Brazilian wheats, and genes present in Chinese sources are different from those in Brazilian sources (Singh et al., 1995; van Ginkel et al., 1996). Some synthetic wheats have recently been identified whose moderate resistance must be derived from Ae. tauschii (syn. T. tauschii) as the T. turgidum parents used in the generation of synthetics are highly susceptible (Gilchrist et al., 1997). These new sources should add newer genetic diversity, which is crucial to enhance the resistance level currently present in hexaploid wheats. Because genes for scab resistance are additive, a careful crossing and selection scheme should allow combinations of several genes leading to high levels of resistance and reduced accumulation of Fusarium toxins in the grains.

RESISTANCE TO SPOT BLOTCH AND TAN SPOT

The first crosses to incorporate spot blotch (caused by Bipolaris sorokiniana) resistance into CIMMYT wheats were made about 20 years ago. These crosses involved moderately resistant cultivars, such as BH1146 from Brazil. However, the level of resistance in progenies was inadequate when tests were carried out at Poza Rica, Mexico, CIMMYT’s ME5 testing site. In the mid-1980s, wheat genotypes carrying resistance to scab and obtained from the Yangtze River Valley of China, showed varying levels of spot blotch resistance when tested at Poza Rica. These Chinese lines included Suzhoe 1 to 10, Wuhan 1 to 3, Shanghai 1 to 8 and certain Ningmai and Yangmai lines. About the same time, the wide crossing programme at CIMMYT produced resistant lines, which contain Thinopyrum curvifolium in their pedigree (Villareal et al., 1995). Some of these lines and their derivatives are showing good resistance and appear to be promising in Bangladesh, low-land Bolivia and Nepal. Resistance in these wheats, such as Sabuf, Chyria 1 and Cugap, appear to be controlled by two to three genes (Velazquez-Cruz, 1994) whereas Longmai 10 and Yangmai 6 may carry polygenic resistance with high narrow sense heritabilities (Sharma et al., 1997). A few synthetic wheats developed at CIMMYT also carry resistance derived from the Ae. tauschii accessions. A key problem with selection for spot blotch resistance is the negative correlation of disease severity with heading date and plant height (Duveiller and Gilchrist, 1994). Therefore, care must be taken if short types with early maturity are required. Current strategy followed at CIMMYT is to combine resistances from these diverse sources. Identification of some highly resistant lines from such crosses indicate that resistance is additive.

Tan spot (caused by Drechslera tritici-repentis) resistance is not widely dispersed in CIMMYT germplasm but moderate resistance is known to occur (Rees and Platz, 1992). Some newer CIMMYT lines, such as Milan, Attila, Corydon and Tinamou, and some Chinese wheats and their derivatives, such as Luan, are also reported to carry high to moderate resistance (Diaz de Ackermann and Kohli, 1998). Tan spot is increasing in areas where reduced tillage practices are being combined with stubble retention. CIMMYT has an ongoing project to search new and better sources of resistance to tan spot for these areas.

TOLERANCE AND RESISTANCE TO BYD

Tolerance to barley yellow dwarf virus in cultivar Anza and several other CIMMYT wheats is due to the gene Bdv1 (Singh et al., 1993), which is linked to durable leaf and stripe rust resistance genes Lr34 and Yr18 (Singh, 1993). Presence of this gene does not reduce virus titre but does cause slow yellowing of plants. Gene Bdv1 is widespread in CIMMYT wheats because Lr34 and Yr18 occur in a large number of CIMMYT wheats. CIMMYT’s highland field location at Toluca, Mexico, has an endemic presence of BYD. Some CIMMYT lines, such as Milan, show a much higher level of tolerance/resistance than Anza and are likely to carry genes in addition to Bdv1. Wheat lines developed through an Australian-Chinese collaboration and carrying a chromosome 7DL translocation from Th. intermedium have shown true resistance to BYD (lower virus titre). However, these sources show high symptoms in Mexico despite the low titre. CIMMYT’s programme is attempting to combine this resistance with other tolerance genes and high yield potential.

RESISTANCE TO BACTERIAL LEAF STREAK

Selection for resistance to bacterial leaf streak (caused by Xanthomonas translucens pv. undulosa) has been carried out for many years at CIMMYT because the disease occurs at the highland locations of Toluca and El Batan in Mexico. Genetic analysis indicates that three genes confer moderate resistance to CIMMYT wheats and are widely dispersed in the germplasm through commonly used parents Pavon 76 and Mochis 88 (Duveiller et al., 1993).

Resistance to all facultative parasites described above involves genes with minor to intermediate and additive effects. Therefore, the selection scheme is similar to that described for combining minor, additive gene based resistance to rusts. Considerable progress is evident for breeding resistance to most of these important diseases. Future methodologies for resistance breeding should emphasize networking involving National Agricultural Research Systems, CIMMYT and Advanced Research Institutes. Molecular markers for resistance genes are expected to enhance the efficiency of selection in the future, and novel resistance from transformation should bring new dimensions to resistance breeding.

REFERENCES

Alvarez-Zamorano, R. 1995. Patogenesis de Puccinia recondita Rob. ex Desm. f. sp. tritici y la resistencia en trigo. Ph.D. thesis. Montecillos, Mexico, Colegio Postgraduados. 76 pp.

Biffen, R.H. 1905. Mendel’s laws of inheritance and wheat breeding. J. Agr. Sci., 1: 4-48.

Borlaug, N.E. 1953. New approach to the breeding of wheat varieties resistant to Puccinia graminis tritici. Phytopathology, 43: 467.

Borlaug, N.E. 1972. A cereal breeder and exforester’s evaluation of the progress and problems involved in breeding rust resistant forest trees. In "Moderator’s Summary". Biology of Rust Resistance in Forest Trees. Proc. NATO-IUFRO Advanced Study Institute, 17-24 Aug. 1969, p. 615-642. USDA Forest Service Misc. Publication 1221.

Briceno, F.G. 1992. Inheritance of resistance to Septoria leaf blotch in selected spring bread wheat genotypes. M.Sc. thesis. Corvallis, Oregon, USA, Oregon State University. 86 pp.

Caldwell, R.M. 1968. Breeding for general and/or specific plant disease resistance. In Proc. 3rd Int. Wheat Genetics Symp., p. 263-272. Canberra, Australia.

Diaz de Ackermann, M. & Kohli, M.M. 1998. Research on Pyrenophora triticirepentis tan spot of wheat in Uruguay. In E. Duveiller, H.J. Dubin, J. Reeves & A. McNab, eds. Helminthosporium Blights of Wheat: Spot Blotch and Tan Spot. Proc. Int. Workshop, El Batan, Mexico, 9-14 Feb. 1997, p. 134-141. Mexico, DF, CIMMYT.

Duveiller, E. & Gilchrist, L. 1994. Production constraints due to Bipolaris sorokiniana in wheat: current situation and future prospects. In D.A. Saunders & G. Hettel, eds. Wheat in Heat-Stressed Environments: Irrigated, Dry Areas and Rice-Wheat Farming Systems. Proc. Int. Conf., p. 343-352. Mexico, DF, CIM-MYT.

Duveiller, E., van Ginkel, M. & Thijssen, M. 1993. Genetic analysis of resistance to bacterial leaf streak caused by Xan-thomonas campestris pv. undulosa in bread wheat. Euphytica, 66: 35-43.

Dyck, P.L. 1987. The association of a gene for leaf rust resistance with the chromosome 7D suppressor of stem rust resistance in common wheat. Genome, 29: 467-469.

Flor, H.H. 1956. The complementary genetic systems in flax and flax rust. Adv. Genet., 8: 29-54.

Fuentes-Davila, G., Rajaram, S. & Singh, G. 1995. Inheritance of resistance to Karnal bunt (Tilletia indica Mitra) in bread wheat (Triticum aestivum L.). Plant Breed., 114: 250-252.

Gilchrist, L., Rajaram, S., Mujeeb-Kazi, A., van Ginkel, M., Vivar, H. & Pfeiffer, W. 1997. Fusarium scab screening program at CIMMYT. In H.J. Dubin, L. Gilchrist, J. Reeves & A. McNab, eds. Fusarium Head Scab: Global Status and Future Prospects. Proc. Workshop, El Batan, Mexico, 13-17 Oct. 1996. Mexico, DF, CIMMYT.

Jensen, N.F. 1952. Intervarietal diversification in oat breeding. Agron. J., 44: 30-34.

Jlibene, M., Gustafson, J.P. & Rajaram, S. 1992. A field disease evaluation method for selecting wheat resistant to Mycosphaerella graminicola. Plant Breed., 108: 26-32.

Johnson, R. 1988. Durable resistance to yellow (stripe) rust in wheat and its implications in plant breeding. In N.W. Simmonds & S. Rajaram, eds. Breeding strategies for resistance to the rusts of wheat, p. 63-75. Mexico, DF, CIM-MYT.

Knott, D.R. 1988. Using polygenic resistance to breed for stem rust resistance in wheat. In N.W. Simmonds & S. Rajaram, eds. Breeding strategies for resistance to the rusts of wheat, p 39-47. Mexico, DF, CIMMYT.

Lehman, J.S. & Shaner, G. 1996. Genetic variation in latent period among isolates of Puccinia recondita f. sp. tritici on partially resistant wheat cultivars. Phytopathology, 86: 633-641.

Liu, Z.Z. & Wang, Z.Y. 1991. Improved scab resistance in China: sources of resistance and problems. In D.A. Saunders, ed. Wheat for the Nontraditional, Warm Areas. Proc. Int. Conf., Foz do Iguacu, Brazil, 29 Jul.-3 Aug. 1990, p. 179-188. Bangkok, CIMMYT.

Luzzardi, G.C. 1985. Wheat breeding for scab resistance. In Wheat breeding for more tropical environments, p. 159-168. Mexico, DF, CIMMYT.

Ma, H. & Singh, R.P. 1996. Contribution of adult plant resistance gene Yr18 in protecting wheat from yellow rust. Plant Dis., 80: 66-69.

Matus-Tejos, I.A. 1993. Genetica de la resistencia a Septoria tritici en trigos harineros. M.Sc. thesis. Montecillos, Mexico, Colegio Postgraduados. 82 pp.

McIntosh, R.A. 1992. Close genetic linkage of genes conferring adult-plant resistance to leaf rust and stripe rust in wheat. Plant Pathol., 41: 523-527.

McIntosh, R.A. 1998. Breeding wheat for resistance to biotic stresses. In H.J. Braun et al., eds. Wheat prospects for global improvement, p. 71-86. Dordrecht, Netherlands, Kluwer Academic Press.

McIntosh, R.A., Hart, G.E. & Gale, M.D. 1995. Catalogue of gene symbols for wheat. In Z.S. Li & Z.Y. Xin, eds. Proc. 8th Int. Wheat Genetics Symp., Beijing, 20-25 Jul. 1993, p. 1333-1500. Beijing, China Agricultural Scientech.

Parlevliet, J.E. 1975. Partial resistance of barley to leaf rust, Puccinia hordei. I. Effect of cultivar and development stage on latent period. Euphytica, 24: 21-27.

Peterson, R.F., Campbell, A.B. & Hannah, A.E. 1948. A diagrammatic scale for estimating rust intensity of leaves and stem of cereals. Can. J. Res. Sect. C, 26: 496-500.

Rajaram, S. & van Ginkel, M. 1996. A guide to the CIMMYT bread wheat section. In Wheat Special Report No. 5. Mexico, DF, CIMMYT.

Rajaram, S., van Ginkel, M. & Fischer, R.A. 1995. CIMMYT’s wheat breeding mega-environments (ME). In Z.S. Li & Z.Y. Xin, eds. Proc. 8th Int. Wheat Genetics Symp., Beijing, 20-25 Jul. 1993, p. 1101-1106. Beijing, China Agricultural Scientech.

Rees, R.G. & Platz, G.J. 1992. Tan spot and its control - some Australian experiments. In L.J. Francl, J.M. Krupinsky & M.P. McMullen, eds. Advances in Tan Spot Research. Proc. 2nd Int. Tan Spot Work-shop, Fargo, ND, USA, 25-26 Jun. 1992, p 1. North Dakota Agric. Exp. Sta.

Roelfs, A.P. 1988. Resistance to leaf rust and stem rust in wheat. In N.W. Simmonds & S. Rajaram, eds. Breeding strategies for resistance to the rusts of wheat, p. 10-22. Mexico, DF, CIMMYT.

Roelfs, A.P., Singh, R.P. & Saari, E.E. 1992. Rust diseases of wheat: concepts and methods of disease management. Mexico, DF, CIMMYT. 81 pp.

Rubiales, D. & Niks, R.E. 1995. Characterisation of Lr34, a major gene conferring nonhypersensitive resistance to wheat leaf rust. Plant Dis., 79: 1208-1212.

Schroeder, H.W. & Christensen, J.J. 1963. Factors affecting resistance of wheat to scab caused by Gibberella zeae. Phytopathology, 53: 831-838.

Sharma, R.C., Dubin, H.J., Devokota, R.N. & Bhatta, M.R. 1997. Heritability estimates of field resistance to spot blotch in four spring wheat crosses. Plant Breed., 116: 64-68.

Singh, R.P. 1992a. Association between gene Lr34 for leaf rust resistance and leaf tip necrosis in wheat. Crop Sci., 32: 874-878.

Singh, R.P. 1992b. Genetic association of leaf rust resistance gene Lr34 with adult plant resistance to stripe rust in bread wheat. Phytopathology, 82: 835-838.

Singh, R.P. 1993. Genetic association of gene Bdv1 for tolerance to barley yellow dwarf virus with genes Lr34 and Yr18 for adult plant resistance to rusts in bread wheat. Plant Dis., 77: 1103-1106.

Singh, R.P. & Gupta, A.K. 1992. Expression of wheat leaf rust resistance gene Lr34 in seedlings and adult plants. Plant Dis., 76: 489-491.

Singh, R.P. & Huerta-Espino, J. 1997. Effect of leaf rust resistance gene Lr34 on grain yield and agronomic traits of spring wheat. Crop Sci., 37: 390-395.

Singh, R.P. & Rajaram, S. 1992. Genetics of adult-plant resistance to leaf rust in ‘Frontana’ and three CIMMYT wheats. Genome, 35: 24-31.

Singh, R.P. & Rajaram, S. 1994. Genetics of adult plant resistance to stripe rust in ten spring bread wheats. Euphytica, 72: 1-7.

Singh, R.P., Payne, T.S. & Rajaram, S. 1991. Characterisation of variability and relationship among components of partial resistance to leaf rust in CIMMYT bread wheats. Theor. Appl. Genet., 82: 674-680.

Singh, R.P., Burnett, P.A., Albarran, M. & Rajaram, S. 1993. Bdv1: a gene for tolerance to barley dwarf virus in bread wheat. Crop Sci., 33: 231-234.

Singh, R.P., Ma, H. & Rajaram, S. 1995. Genetic analysis of resistance to scab in spring wheat cultivar Frontana. Plant Dis., 79: 238-240.

Snijders, C.H.A. 1990. Genetic variation for resistance to Fusarium head blight in bread wheat. Euphytica, 50: 171-179.

Stakman, E.C. & Levine, M.N. 1962. Analytical key for the identification of physiologic races of Puccinia graminis tritici. (Processed) Division of Cereal Crops and Dis., USDA, Minnesota Agric. Exp. Sta. 7 pp.

Van der Plank, J.E. 1963. Plant diseases: epidemics and control. New York, NY, USA, Academic Press.

van Ginkel, M., van der Schaar, W. & Zhuping, Y. 1996. Inheritance of resistance of scab in two wheat cultivars from Brazil and China. Plant Dis., 80: 863-867.

Velazquez-Cruz, C. 1994. Genetica de la resistencia a Bipolaris sorokiniana en trigos harineros. M.Sc. thesis. Montecillos, Mexico, Colegio Postgraduados. 84 pp.

Villareal, R.L., Mujeeb-Kazi, A., Gilchrist, L. & Del Toro, E. 1995. Yield loss to spot blotch in spring bread wheat in warm non-traditional wheat production areas. Plant Dis., 79: 893-897.

Wiese, M.V. 1977. Compendium of wheat diseases. St Paul, MN, USA, APS Press.

Zhong, W.Y. & Miller, J.D. 1988. Screening techniques and sources of resistance to Fusarium head blight. In A.R. Klatt, ed. Wheat Production Constraints in Tropical Environments. Proc. Int. Conf., Chiang Mai, Thailand, 19-23 Jan. 1987, p. 239-250. Mexico, DF, CIMMYT.