CHAPTER 22RECENT ADVANCES IN BREEDING WHEAT FOR DROUGHT AND SALT STRESSESRANA MUNNS AND R.A. RICHARDS CSIRO Plant Industry, GPO Box 1600, Canberra, ACT 2601, AustraliaAbstract:Substantial advances have been made in breeding wheat for dry environments that will also improve performance in saline environments. These genetic gains have been made by conventional breeding. Further gains in productivity will come from the addition of traits that increase the efficiency of water use in dry soils, and control the uptake of salts from saline soils. Conventional breeding methods will continue to be important to provide farmers with higher yielding varieties in dry or saline soils which are resistant to current diseases and which have the grain quality demanded by competitive markets. Trait-based breeding approaches, which often utilize molecular markers to improve selection efficiency, are starting to deliver new and significant gains. To target the most important traits, it is important to know how they will influence yield. Is it through more water use, more efficient use of water or a higher harvest index? For example, the trait of early vigourmay be an advantage in some years but in others may lead to the exhaustion of soil water and a low yield. The challenge for breeders will be to efficiently integrate trait-based and molecular methods to increase yield in dry and saline environmentsKeywords:Drought tolerance, salinity tolerance, trait-based breeding, marker-assisted selection, yield, wheat1.INTRODUCTIONWheat (Triticum aestivum L.) is the most widely grown crop in the world. It is considered a moderately salt-tolerant and drought-tolerant crop and, with barley, it is the preferred cereal in most arid and semi-arid agricultural regions. Wheat is hexaploid and allopolyploid (AABBDD), and is used primarily for baking bread. Wheat has extraordinary adaptation, growing successfully in cool moist regions from Finland to southern Chile to tropical regions such as parts of Asia and Africa. Much of the worlds wheat is produced under irrigation; however, the area sown to rainfed wheat is very substantial and is expected to grow further as water for irrigation declines globally. The rainfed wheat crops are often in semi-arid environments which range from the centre of origin of wheat in the Fertile Crescent, where crops565and Salt Tolerant Crops, 565585. 2007 Springer.M.A. Jenks et al. (eds.), Advances in Molecular Breeding Toward Drought566MUNNS AND RICHARDSrely on current rainfall, to dry regions of north eastern Australia and South Africa, where little rain falls during the growing season and crops rely mostly on stored soil water. The grain yield of wheat is a function of rainfall (or evapotranspiration) and several possible relationships between yield and rainfall (or evapotranspiration) are shown in Figure 1. The broken line represents average yields typically found on-farm in semi-arid environments (rainfed). The thin line represents the yield potential of current well-adapted cultivars. An average farm yield that falls below the potential can be due to management practices (eg, late sowing, insufficient fertilizer), hostile soils (eg, salinity, acidity) or biotic factors (soil-borne or foliar diseases, insect pests). The thick line represents a new yield potential that may be achievable through breeding for dry environments using conventional breeding methods. A new yield potential will come from increases in water productivity; ie, greater or more efficient use of the water resource, than from improvements in drought tolerance or drought resistance. Figure 1 shows that the minimum rainfall required to produce harvested grain is about 100 mm. In other words, if there is no water stored in the soil, then a minimum of 100 mm of rain falling during the growing season is necessary to compensate for the evaporative loss. The slope of the line in Figure 1 is the transpiration efficiency, and is close to 20 kg ha1mm1. Thus, if 100 mm of water were stored in the soil at sowing and there was no further precipitation or evaporative loss, this could result in a wheat yield of 2 t ha1. Barley typically yields more than wheat when crops are reliant on rainfall during the growing season (Lopz-Casteeda and RichardsFigure 1. Relationship between grain yield and rainfall or evapotranspiration. Adapted from French and Schultz (1984)RECENT ADVANCES IN BREEDING WHEAT5671994a), whereas barley achieves a yield equal to or less than wheat when crops are reliant on stored soil moisture. This difference between wheat and barley is attributed to variation in their early leaf area growth which in turn regulates crop water use and not to any inherent variation in their drought resistance. Breeding targets that should improve the grain yield of wheat (thick line in Figure 1) are also evident from this figure. Breeding to increase crop water use will increase yield. This may be achieved, for example, by a more effective root system