Crop Science Grow Your Career with CSSA
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (81)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Richards, R. A.
Right arrow Articles by van Herwaarden, A. F.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Richards, R. A.
Right arrow Articles by van Herwaarden, A. F.
Agricola
Right arrow Articles by Richards, R. A.
Right arrow Articles by van Herwaarden, A. F.
Related Collections
Right arrow Water Stress
Right arrow Crop Physiology & Metabolism
Right arrow Wheat
Crop Science 42:111-121 (2002)
© 2002 Crop Science Society of America

CROP PHYSIOLOGY & METABOLISM

Breeding Opportunities for Increasing the Efficiency of Water Use and Crop Yield in Temperate Cereals

R. A. Richards*, G. J. Rebetzke, A. G. Condon and A. F. van Herwaarden

CSIRO Plant Industry, P.O. Box 1600, Canberra, ACT, Australia 2601

* Corresponding author (r.richards{at}pi.csiro.au)


   ABSTRACT
TOP
 NOTES
 ABSTRACT
INTRODUCTION
 Trait Identification
 Increased Water Use
 Water-Use Efficiency
 Harvest Index
 Concluding Remarks
 REFERENCES
 
Genetic advances in grain yield under rainfed conditions have been achieved by empirical breeding methods. Progress is slowed, however, by large genotype x season and genotype x location interactions arising from unpredictable rainfall, which is a feature of dry environments. A good understanding of factors limiting and/or regulating yield now provides us with an opportunity to identify and then select for physiological and morphological traits that increase the efficiency of water use and yield under rainfed conditions. The incorporation of these traits into breeders' populations should broaden their genetic base. It also may lead to faster selection methods and selection for the traits may result in correlated gains in yield. Here, we undertake a review of factors that limit yield in rainfed environments and discuss genetic opportunities and genetic progress in overcoming them. The examples given are for wheat (Triticum aestivum L.), but the principles apply to all cereal crops grown in dry environments.

Abbreviations: WUE, water use efficiency • GA, gibberellic acid • SLA, specific leaf area • AFLP, amplified fragment length polymorphism • HI, harvest index • WSC, water soluble carbohydrates


   INTRODUCTION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
Trait Identification
 Increased Water Use
 Water-Use Efficiency
 Harvest Index
 Concluding Remarks
 REFERENCES
 
ACHIEVING GENETIC INCREASES in yields under rainfed conditions has always been a difficult challenge for plant breeders. This is evident from the smaller genetic gains in dry regions compared with those in wetter environments or where irrigation is available. For example, rainfed environments are characterized by unpredictable and highly variable seasonal rainfall and hence highly variable yields. This results in slow genetic advance in breeding programs because genetic variation in yield is masked by large genotype x year and/or genotype x location and genotype x year x location interactions (Calhoun et al., 1994; van Ginkel et al., 1998). A major challenge for breeders in dryland environments is to devise the most effective strategy for maximizing genetic gain. Substantial effort therefore goes into identifying representative sets of environments to evaluate yield of breeding lines (Calhoun et al., 1994), and to devising efficient experimental designs to maximize genetic variation. In turn, as many lines as possible are grown in the smallest number of representative environments and the fewest replications.

It is not surprising that genetic increases in yield potential made by selection in predictable irrigated environments has also resulted in broadly-adapted crops that are often well-suited to both favorable and low-yielding, rainfed environments (Sayre et al., 1995; Cooper et al., 1996). Genetic variation in traits that contribute to high yield in all environments, such as a high harvest index, is greater in predictable environments and is therefore more likely to be selected under favorable conditions. It is likely that genetic advances made in favorable environments will continue to contribute to yield in less favorable environments provided that selected germplasm is widely evaluated under rainfed conditions.

These empirical procedures have been important for yield increases in dryland environments and will remain so. However, there are a host of specific adaptations that may be uniquely important under rainfed conditions. Some of these adaptations may be targeted to improve water use or water-use efficiency to achieve higher regional yields. Selection for these traits in a breeding program could then result in more accurate targeting of factors limiting yield, thereby increasing the rate of yield improvement. For any crop species, appropriate targets for physiological breeding will develop from an understanding of the factors regulating growth, development, and yield of the species in the prevailing farming system and in the region where the crop is grown.

A physiological approach can complement empirical breeding and may enhance the rate of yield improvement in the following ways. First, it may identify important traits for which there is inadequate genetic variation in breeders' populations. This may result in the identification of new parental lines to generate greater variability in key targeted traits for selection. Second, large seasonal variation in yield and subsequent genotype x environment (G x E) interactions will slow genetic gain for yield. Specific targeting of physiological characters that limit yield and have a high heritability may be more effective than direct selection for yield. Third, morphological or physiological traits can sometimes be measured out-of-season, or in controlled conditions, so that several generations can be grown each year with selection. Also, they may be readily amenable to the use of marker-assisted selection. Fourth, selection for physiological traits, particularly in early generations, may be more cost effective. Yield trials are expensive to conduct and if the population can be culled in earlier generations by effective physiological criteria, then this will allow more high-yielding, adapted entries to be tested, greater replication in yield trials to increase selection precision and/or more emphasis on traits such as grain quality. Finally, provided traits are relatively independent of each other, they can be pyramided together to maximize genetic gain in yield.

This paper explores opportunities to improve yields in water-limited environments by genetically manipulating physiological or morphological traits that are believed to limit yield. The focus is on temperate cereals, in particular, wheat (Tritcum aestivum L.) and the traits being selected and the examples given are from a breeding program being conducted at CSIRO Plant Industry for wheats specifically adapted to different low-yielding rainfed regions of Australia.


   Trait Identification
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Trait Identification
Increased Water Use
 Water-Use Efficiency
 Harvest Index
 Concluding Remarks
 REFERENCES
 
A physiological breeding approach is risky but, if successful, promises substantial rewards. Success depends on the identification of selectable traits that limit yields. The framework proposed by Passioura (1977) has proved very useful for identifying the important components for yield of grain crops growing in dry environments. This framework is based on grain yield alone, and not on drought protection or on survival under drought which were popular concepts in the past but have largely been unsuccessful. Passioura proposed that when water is limiting then grain yield is a function of (i) the amount of water used by the crop, (ii) how efficiently the crop uses this water for biomass growth (i.e., the water-use efficiency as above-ground biomass/water use), and (iii) the harvest index, (i.e. the proportion of grain yield to above-ground biomass). Since these three components are likely to be largely independent of each other, then an improvement in any one of them should result in an increase in yield.

The regional environment must also be considered in conjunction with this identity. Traits proposed as yield-limiting may only be so in specific environments. Indeed, because of the seasonal variability in rainfall in dry environments, a particular trait may not even be important in every season in a particular region. There are exceptions to this, however, as some traits are universally important in water-limited environments. Examples of these would be successful emergence and therefore good crop establishment, and a high water-use efficiency. An appropriate flowering time has been the single most important factor to maximize yield and adaptation in dry environments (Richards, 1991). Further genetic modification of flowering time in different regions is still possible as management practices are constantly changing and these are providing new cropping opportunities. For example, new machinery, herbicides or tillage practices may enable earlier sowing and the best adapted cultivars to these conditions may require different sensitivities to photoperiod or vernalisation than for currently grown cultivars.


   Increased Water Use
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Trait Identification
 Increased Water Use
Water-Use Efficiency
 Harvest Index
 Concluding Remarks
 REFERENCES
 
A deep root system is synonomous with more water uptake from the soil and better performance under drought. It may be, however, that the root systems of cultivars grown in a given region are already adequate and further improvement may not be required. Information on whether current cultivars extract all the available soil water is required to establish this. If soil water remains after harvest then a genetic improvement in rooting depth and/or distribution may be required. This trait is, of course, difficult to measure. The simplest way to increase rooting depth and root distribution of crops is to increase the duration of the vegetative period (i.e. the period up to anthesis). This may be achieved by sowing earlier or later flowering genotypes if this is feasible. Increased early vigor may also result in both faster growth of roots enabling them to exploit deeper soil layers and production of more adventitious roots in the top soil. The latter may be important for using water and nutrients before evaporative losses dry the top soil. Appropriate ways to select for greater vigor are discussed later. Greater osmotic adjustment may also result in more root growth and an ability to extract additional soil water (Morgan and Condon, 1986). However, selection for osmotic adjustment is not easy at the present time, although a novel method to select in the haploid stage in wheat has recently been demonstrated (Morgan, 1999).

Water uptake by roots may not only be limited by the genetic potential for growing deeper roots but also by soil factors that limit root growth. For example, the presence of root pathogens, mineral deficiencies (Zn, P, for example) or toxicities (salt, pH, boron, for example), or soil physical barriers such as poor structure or high bulk density, can all drastically reduce water use, nutrient uptake and yield. Overcoming these factors may require genetic solutions although the appropriate use of rotations and management practices can also be very effective in improving soil structure and controlling root diseases (Angus and van Herwaarden, 2001).


   Water-Use Efficiency
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Trait Identification
 Increased Water Use
 Water-Use Efficiency
Harvest Index
 Concluding Remarks
 REFERENCES
 
The term water-use efficiency (WUE) is generally used to express the ratio of total dry matter production to evapotranspiration. It can be expressed in the following terms:

[1]
where TE is the transpiration efficiency (above ground dry weight/transpiration), ES is the water lost by evaporation from the soil surface, and T is water lost through transpiration by the crop (Richards, 1991). This expression shows that crop WUE can be increased either by increasing TE or decreasing the magnitude of ES. The relative importance of each of these components of WUE varies according to rainfall distribution. Thus, if a crop is reliant on water stored in the soil, and rainfall during the growing season is low, then increasing TE provides the greatest opportunity to increase WUE. However, if a crop is generally reliant on in-season rainfall then decreasing ES may result in the largest gains. In Mediterranean environments, a typical value for ES from well-managed wheat crops is 40% of evapotranspiration and often substantially more in late-sown or poorly managed crops (French and Schultz, 1984).

Opportunities for increasing TE and decreasing ES through both management and breeding are large. Changes in management are widely recognized with fallowing and mulching common practices to increase total water use. An altered sowing date can also improve WUE (Gomez-Macpherson and Richards, 1995), although this may require the breeding of new varieties that are delayed for flowering time.

If it is possible, the simplest way to increase WUE and yield is to sow temperate crops as early as possible before early winter. This improves WUE since the crop may have developed a full canopy by early winter which results in a high TE. This effect on TE will be discussed later. Sowing early also reduces ES as warm soil temperatures result in fast early growth and hence there are fewer rainfall events when the soil is not shaded. The importance of earlier sowing is evident in the higher yields achieved in Western Australia with wheat cultivars that have phenologies more suited to sowing earlier than those previously available (Anderson, 1992). It is important to note that it was not just the introduction of wheats with a changed phenology that enabled earlier sowing and greater yields but also the new agronomic opportunities brought about by new herbicides, rotation crops, larger machinery, and the recognized benefits of minimal cultivation practices.

Substantial new opportunities remain. For example, in most of eastern Australia, spring wheats are planted in late autumn/early winter (May/June). Sowing winter wheat in autumn, however, should result in substantial increases in WUE for the reasons outlined above. Autumn planting in eastern Australia results in biomass and WUE about double that of standard early winter sowings of the same or related spring wheat varieties (Fig. 1) . Surprisingly, the grain yield of earlier sown crops is almost unchanged (Penrose, 1993) and this has been attributed to the low harvest index of early-sown crops (Gomez-Macpherson and Richards, 1995). It is proposed that one way to overcome a low harvest index is to develop shorter winter wheats with fewer tillers to increase soluble carbohydrate reserves and reduce the investment into structural carbohydrates (Richards et al., 1997.) Altering the dry matter partitioning to more soluble carbohydrates may improve harvest index and thereby yield, if these carbohydrates are later mobilized to the grain.



View larger version (13K):
[in this window]
[in a new window]
  		Fig. 1. Relationship between sowing date and water-use efficiency of wheat. Data are from near-isogenic populations differing in flowering time and grown at different locations and years in eastern Australia. Water use efficiency (WUE) was calculated as the total above-ground biomass per unit of rainfall between April and October. Adapted from Gomez-Macpherson and Richards (1995).

 
Reducing Soil Evaporation
Crops are often grown in environments where rainfall is frequent between sowing and stem elongation and therefore where ES, as a proportion of T, is large. This is characteristic of Mediterranean environments around the world. Any increase in early seedling vigor should reduce evaporative losses from the soil surface and also increase the competitiveness of crops resulting in fewer weeds and less herbicide use. The potential for yield gains is large. First, because often half of the growing-season rainfall may be lost through evaporation from the soil and second, because semi-dwarf wheats, which are widely grown, have inherently low vigor compared with taller wheats (Richards, 1992). Greater vigor is not likely to be as important in other environments, although if it results in more growth when vapor pressure deficit is low (de Wit, 1958), then this will also result in a higher TE. Also, if crop duration is very short then vigorous genotypes are likely to yield more grain and biomass than less vigorous types.

Management practices may be important for improving crop vigor. For example, sowing seed at a higher sowing rate, with more nitrogen, or a reduced depth if the seed bed permits, may all improve crop vigor. Also, wheats with large grains result in larger, more vigorous plants than those with smaller grains (Richards and Lukacs, 2001). However, if sowing rate were based on weight per unit area, then there might be little advantage in sowing large grains. Substantial opportunities exist to increase vigor genetically. The first requirement to increasing crop vigor is to ensure that seedling establishment is as high and as fast as possible. The second requirement is that there must be the genetic potential for rapid early growth.

Improved Seedling Establishment. The problem of poor seedling emergence has become particularly important for wheat in recent decades because of the global adoption of gibberellic acid (GA)-insensitive semidwarf cultivars. The reduced height of semidwarf wheat cultivars is principally due to the presence of GA-insensitive alleles, Rht-Blb (Rht1) and Rht-Dlb (Rht2). Although the presence of these alleles does result in a desirable plant height to maximize yield, they also reduce cell size in most above-ground organs (Keyes, et al., 1989; Miralles et al., 1998). In turn, coleoptile length and leaf area are smaller during the early vegetative stage compared with the standard height wheats that are sensitive to GA. Short coleoptiles often result in poor emergence and therefore poor crop establishment, particularly if seeds of semi-dwarf cultivars are deep sown to provide moisture, or are sown into stubble (Schillinger et al., 1998). Better emergence is achieved by sowing wheats with the genetic potential for long coleoptiles. Although increased coleoptile length can be achieved by selection within GA-insensitive semi-dwarf wheat populations (Beharev et al., 1998), substantially greater progress can be made with parents whose seedlings are sensitive to gibberellic acid; although short stature also needs to be selected (Rebetzke et al., 1999). A number of parental sources are available (Gale and Yousseffian, 1985). Some of these sources were widely used before the introduction of the GA-insensitive wheats and others have been derived through mutagenesis. There are also minor genes that are important to further modify plant height. Figure 2 shows that GA-sensitive genes are available which reduce plant height to levels equivalent to that of plants containing GA-insensitive Rht genes yet produce coleoptiles which are up to 100% longer than coleoptiles found in GA-insensitive genetic backgrounds (Rebetzke et al., 1999; Rebetzke and Richards, 2000). It appears that plant height and coleoptile length are unrelated in GA-sensitive backgrounds (Whan, 1976), enabling the simultaneous selection for both characteristics in a segregating population. Our studies (Rebetzke and Richards, 2000) show that semidwarf lines with GA-sensitive dwarfing genes have the same desirable partitioning characteristics of the current semidwarf wheats but may also produce greater biomass if soil conditions at sowing are unfavorable (Rebetzke and Richards, 1999). Furthermore, wheats with long coleoptiles also tend to have larger early leaves and more rapid rates of emergence which together contribute to greater early growth and thereby WUE (Richards, 1992).



View larger version (14K):
[in this window]
[in a new window]
  		Fig. 2. Variation in coleoptile length and plant height of GA-insensitive semidwarf cultivars, GA-sensitive standard height cultivars, and GA-sensitive semidwarf lines with long coleoptiles from 3 populations. Coleoptile length of lines was determined at 11 and 15°C whereas plant height of lines was measured at maturity in a single favorable environment following a late autumn sowing.

 
Coleoptile length of cereals is easily measured and large numbers can be screened rapidly. Heritability, particularly on a line-mean basis, is high owing to small genotype x temperature interactions (Rebetzke et al., 1999) thereby enabling substantial genetic progress. To achieve better establishment and high yields, longer coleoptiles must be combined with semidwarf stature. Thus, long coleoptile plants or lines must be further tested to select for the desired plant height. The availability of molecular markers for some GA-sensitive dwarfing genes in wheat can further enhance selection efficiency (Korzun et al., 1998; Rebetzke and Richards, 2000).

Seedling Vigor. After ensuring a high proportion of seedlings emerge, it is then desirable to have a fast growing canopy. There is little important genetic variation for this characteristic among semidwarf wheat cultivars currently grown (Rebetzke and Richards, 1999), and since almost all are GA-insensitive, there is little scope for genetic improvement because of the association between small cell size and GA-insensitivity. Characterization of seedling vigor differences between wheat and barley (Hordeum vulgare L.) has been illuminating and very important to understanding factors contributing to high vigor in cereals. Barley achieves almost double the early leaf area of wheat primarily because of its earlier emergence (around 1 d earlier), larger embryo and greater specific leaf area (SLA) (López-Castañeda et al., 1996). Also, barley produces large coleoptile tillers which emerge before the primary tillers, and it has been shown that these also contribute to increased vigor of wheat (Liang and Richards, 1994). Exhaustive screening of international wheat collections and a broadly-based composite cross population has revealed several important sources of genetic variation for early plant vigor (Richards and Lukacs, 2001). As in barley, the greater early vigor of these wheats arises through their larger embryo, high specific leaf area and large coleoptile tillers. Pyramiding these characteristics together has produced progeny with greater vigor than the original wheat parents and a leaf area and dry weight approximately double that of most CIMMYT-derived wheats. These lines are now being used as parents in our breeding program for improved establishment and vigor to improve WUE and grain yield. Our experiments show that heritability for early leaf area is small (Rebetzke and Richards, 1999). However, leaf breadth averaged over the first two leaves has a much higher narrow-sense heritability as well as a strong genetic correlation with leaf area (Fig. 3) . The combination of a higher heritability and strong genetic correlation suggests that selection for early leaf area development using leaf breadth as a surrogate for leaf area is more effective than selection for total leaf area itself. Furthermore, the procedure is very fast and can be conducted out of season allowing several generations of selection and progeny-testing each year.



View larger version (14K):
[in this window]
[in a new window]
  		Fig. 3. Relationship between leaf breadth of seedling leaves and leaf area for random families derived from a convergent cross population differing in vigor. Lines were grown in 10 m2 field plots at Condobolin together with semidwarf check varieties Janz, Hartog, and Amery. Note that the Australian cultivars are amongst the lines with the lowest vigor. (Data adapted from Rebetzke and Richards, 1999.)

 
Field observations on wheats containing GA-insensitive dwarfing genes indicate they produce smaller leaf areas and plant dry weights early in the season compared with wheats which are GA-sensitive (Richards, 1992). Furthermore, we have found that the potential of larger embryos and greater SLA to increase seedling vigor is reduced when selected in GA-insensitive Rht genetic backgrounds because of smaller cell sizes associated with Rht-Blb and Rht-Dlb. We believe that the expression of seedling vigor using larger embryos and greater SLA would be enhanced in genetic backgrounds containing GA-sensitive dwarfing genes. We are combining the high vigor characteristics of large embryo, greater SLA, and a high frequency of coleoptile tillers into GA-sensitive, semidwarf genetic backgrounds.

Greater Transpiration Efficiency
The other component of water-use efficiency, that is TE, the ratio of dry matter to transpiration, also is amenable to improvement. There are numerous ways to increase TE of wheat. Perhaps the simplest way is to ensure that the period of maximum biomass accumulation occurs during the cooler periods of the growing season. It was noted earlier that sowing temperate crops before the onset of winter results in a substantially greater WUE than crops sown later. Much of this increase is due to the relationship between climate and TE. There is a direct link between saturation vapor pressure deficit of the air (evaporative demand) and TE. Figure 4 shows the TE of wheat sown at different times in a mild climate (similar to northern California). It is clear that less water is required for growth when it is cool. A greater TE can therefore be achieved by a change in planting time so that full canopy closure occurs before the onset of the coolest period. This may require wheats with a different phenology. Since this is such a simple way to improve TE, opportunities to alter management practices to sow earlier, including dry seeding, should be explored where practical. An improvement in early crop vigor outlined earlier should also result in a higher leaf area index allowing more light interception and growth by the crop when it is cool to increase TE. Thus, greater crop vigor should result in an increased WUE through both a reduction in ES and an increase in TE.



View larger version (14K):
[in this window]
[in a new window]
  		Fig. 4. Relationship between transpiration efficiency of wheat at Wagga Wagga and pan evaporation (USWB class A). Unpublished data of Warren and Lill, cited in Fischer (1979). Data adapted to northern hemisphere seasons.

 
There are many ways in which TE can be improved genetically. For example, increased surface reflectance to lower the surface temperature of photosynthetic tissue will increase TE. Selection for glaucousness and/or pubescence are two ways to achieve this in cereals (Richards et al., 1986). Smaller photosynthetic surfaces are more effective at dissipating heat than larger surfaces if it is dry and hot after flowering. Selection for awns that maintain photosynthetic activity and for a small erect upper canopy of leaves may therefore be effective. Some transpiration occurs at night through incompletely closed stomata and through the cuticle. This is unlikely to be large in most environments, but it could result in a nightly loss of up to 0.5 mm of water if the vapor pressure deficit is high (Rawson and Clarke, 1988). Cuticular loss during the day may also be important. Substantial genetic variation for cuticular transpiration has been found in wheat and a relatively simple procedure for selection has been established (Clarke and McCaig, 1982).

Carbon Isotope Discrimination. A very promising method for improving TE is the measurement of carbon isotope discrimination of plant material. Discrimination against 13CO2 in favor of 12CO2 during CO2 diffusion into the sub-stomatal cavities and during photosynthesis in C3 plants is closely related to the TE integrated over the life of the plant material sampled (Farquhar and Richards, 1984; Condon et al., 1992). We are now using this technique to breed wheats with a higher TE.

Carbon isotope discrimination ({Delta}) is a measure of the ratio of the stable isotopes of carbon (13C/12C) in plant dry matter relative to the value of the 13C/12C ratio in the air that plants use in photosynthesis. The extent of discrimination against 13C varies substantially among genotypes (Condon and Richards, 1992). We have established that there are several properties of {Delta} that make it appealing as a potential breeding tool (Richards and Condon, 1992). Most importantly, {Delta} is a lot easier and faster to measure than TE itself. Measuring {Delta} of plant dry matter sampled from a collection of genotypes provides a relative estimate of variation in leaf-level TE integrated over the time that the dry matter was laid down. We have also established that there is substantial genetic variation for {Delta}, G x E is low so heritability of {Delta} is high, and that {Delta} can be measured on freshly sampled dry matter or on samples that have been stored for extended periods.

A breeding program was initiated to introgress low {Delta} into the variety Hartog, an important commercial variety in the northern wheat belt of Australia. The backcrossing program and the results contrasting high and low {Delta} groups derived from the backcrossing program are reported in Rebetzke et al. (2002). Selection for low {Delta} resulted in an increased grain yield in the Hartog background (Fig. 5) . It also resulted in a greater biomass, harvest index, and kernel weight. The yield advantage was greatest in the drier environments. The line-mean heritability for {Delta} was higher than for grain yield or above-ground biomass . Also, the genetic correlation was significant between {Delta} and yield and {Delta} and biomass . The high heritabilities and strong genetic correlations indicate that indirect selection for low {Delta} should be more efficient on a single-plot basis than selection for grain yield (+122%) or above-ground biomass (+203%). These efficiencies decline to 76% and 100% when calculated on a line-mean basis obtained from replicated field trials over 9 environments. A new wheat variety, selected for low {Delta} and that has higher yield than the recurrent parent, will be released in Australia in 2001.



View larger version (10K):
[in this window]
[in a new window]
  		Fig. 5. Grain yield advantage of near-isogenic Hartog lines selected for low carbon isotope discrimination ({Delta}) compared with lines selected for high {Delta} evaluated across nine environments differing in seasonal rainfall. Grain yield advantage (%) was estimated as: [({zeta}Low {Delta} group - {zeta}High {Delta} group)/{zeta}High {Delta} group] x 100. Yield varied from around 1.5 Mg ha-1 with 250 mm rainfall to 5 Mg ha-1 with 450 mm rainfall.

 
Although low discrimination is likely to be important in all water-limited environments, it is likely to be more effective in environments where crop growth is most dependent on moisture stored in the soil profile from rain that falls outside the main crop growth phase, such as in the northern Australian wheat belt (Condon et al., 2002). In these environments, transpiration makes up a high proportion of total crop water use and a high TE leads to conservation of soil water up to and after anthesis which sustains the yield determining processes during the terminal drought. It is also notable that when water is not limited, high {Delta} is positively related to genotypic differences in yield (Condon et al., 1987; Condon et al., 2002).

A potential limitation to the use of {Delta} in breeding programs is its cost, and surrogates that are related to {Delta} may help reduce the cost of culling the population for {Delta}. Possible surrogates are: ash content of dry matter (Masle et al., 1992), near-infrared reflectance of bulk tissue (Clark et al., 1995), stomatal conductance (Condon et al., 1987; Rebetzke et al., 2001), chlorophyll content of leaves (Araus et al., 1997), and SLA (Wright et al., 1988). We have also identified AFLP molecular markers that account for much of the among-family variance in {Delta} in breeding populations.


   Harvest Index
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Trait Identification
 Increased Water Use
 Water-Use Efficiency
 Harvest Index
Concluding Remarks
 REFERENCES
 
Morphological and physiological traits discussed so far all contribute to greater yields through increases in total biomass. At maturity, a proportion of this biomass of grain crops is the grain, and a high proportion is desirable to achieve high yields, i.e. a high harvest index (HI). Like water use and WUE, the genetic manipulation of HI in variable rainfed environments is possible but not simple. This is because for crops that experience drought there are two separate determinants of HI and hence two factors that can be genetically manipulated to maximize HI to achieve a high grain yield (Richards, 1991). The first determinant of HI is independent of drought. It is the HI in the absence of drought in a given environment. The second determinant of HI is drought dependent and depends largely on water availability during grain filling, but also on other factors such as pre-anthesis partitioning between structural and soluble carbohydrates.

Drought-Independent Harvest Index
A high HI under optimal conditions is likely to contribute to high yield in all environments, provided there is no sacrifice in biomass. Selection for yield in cereal breeding programs, usually under favorable conditions, has indirectly resulted in higher HI's (Riggs et al., 1981; Austin et al., 1989; Sayre et al., 1997). A high drought-independent HI is a prerequisite to high yield under rainfed conditions as it sets the genetic potential. A high drought-independent HI has been achieved by a greater partitioning of dry matter to reproductive than to non-reproductive organs. Genes contributing to height reduction and to early flowering are simple and effective ways to increase HI. Both traits are highly heritable and both reduce the growth of vegetative organs. In the case of height reduction, there is less competition for assimilates between the growing spike and the elongating stem. This is the main reason for the advantage of semidwarf wheat varieties over standard height varieties and why semidwarfs have been so successful in both irrigated and rainfed environments. It is likely that we are approaching the limit for genetic increases in HI due to height reduction and to earlier flowering. However, other opportunities remain to increase partitioning to reproductive organs such as genetic manipulation of phasic development (Miralles et al., 2000), carpel size (Calderini et al., 1999), sterile tiller reduction (Richards, 1988), and grain set (Fischer et al., 1998).

Drought-Dependent Harvest Index
It is only when the HI of a given genotype in the absence of drought in the target environment is already high that the genetic improvement of the drought-dependent HI becomes important. Drought-dependent HI is often a function of post-anthesis water use. If post-anthesis water use as a proportion of total water use is large then HI will be large (Passioura, 1972; Siddique et al., 1990). Thus, if soil water is finite then conserving soil water before flowering so that it can be used for grain filling should increase HI. The achievement of a high grain yield (and a high HI) will then depend on the balance between growth before and after anthesis. Getting this balance right in unpredictable rainfed environments is difficult. For example, too little growth before anthesis may limit total dry matter production, maximize HI, but leave water behind in the soil. On the other hand, too much growth before anthesis will ensure that total dry matter yield is maximized because all the available soil water is used, but it could result in a low HI and a low grain yield.

Water use is a function of evaporative demand and leaf area. There is little that can be done to alter evaporative demand, although there may be flexibility for changing the phenology of the crop and therefore the timing of crop growth, but there are a number of opportunities to genetically reduce leaf area development. The most likely conditions in which these opportunities may apply is where the crop is growing on stored soil water. A discussion of some of the important traits that may regulate leaf area and thereby water use for increasing the drought-dependent HI in these conditions follows.

Phenology is a major determinant of drought-independent HI as it can determine the amount of pre-anthesis and post-anthesis water use by the crop. For example, a few days earlier flowering may mean an extra 10 to 15 mm of soil water for post-anthesis use (drought escape) and thereby a higher HI and maybe greater yield (Angus and van Herwaarden, 2001). However, if earlier flowering also results in less pre-anthesis growth, yield may not be greater despite the higher HI. Seasonal variation in yields is generally large under rainfed conditions and sowing earlier flowering cultivars may not always be advantageous. With decades of breeding and yield testing, it is likely that anthesis time of successful varieties in a region is close to optimum. Nevertheless, flowering earlier should result in a higher WUE in environments where temperatures increase after anthesis. Combining earlier flowering with greater vigor or frost resistance may also assist in improving HI and yield.

A reduction in tillering may contribute to a higher HI both in the presence and absence of drought because of the high level of tiller mortality. Temperate cereals typically initiate up to 50% more tillers than can survive, even under favorable conditions (Stapper and Fischer, 1990). Although these unproductive tillers can retranslocate some of their nitrogen and carbon to productive tillers, there is a cost of production of these tillers in the form of soil water used for transpiration, as well as carbon, phosphorous, and other nutrients that form the structural material and respiratory losses. Reduced tillering may also contribute to a higher HI under drought because a smaller leaf area in the period leading up to anthesis may reduce transpiration leaving more water for grain filling. However, this may not be desirable in Mediterranean environments where restricting soil evaporation may give greater benefit.

Narrower xylem vessels in the seminal roots would have a similar effect to restricted tillering (Richards and Passioura, 1981). The seminal roots are responsible for uptake of water in the deeper soil layers and reducing the diameter of xylem vessels in the roots should increase the hydraulic resistance. This should result in a restricted leaf area and slower water use before anthesis, if it is dry, and result in more soil water available for grain filling and a higher yield. In a breeding program to reduce the root xylem vessel diameter of wheat, yield was increased 7% on average, in two different genetic backgrounds (Richards and Passioura, 1989). An important feature of reducing xylem vessel diameter is that it is likely to be advantageous under dry conditions but neutral in favorable conditions, as the nodal roots, which are in the topsoil, will supply the crop with its water requirement if the topsoil is wet. Other ways to reduce pre-anthesis transpiration may be to select for smaller uppermost leaves, including the flag leaf, or for a lower stomatal conductance, and/or a lower night-time leaf conductance.

Substantial genetic variation exists for tillering, xylem vessel diameter, leaf dimensions and stomatal or cuticular water loss. In the case of tillering in wheat there is a major gene on chromosome 1AS that inhibits tillering (Richards, 1988). The penetrance of this gene varies with climate and genetic background and so it provides substantial scope for the regulation of tillering and therefore leaf area. This gene can result in plants with primarily only productive tillers without significantly compromising vegetative growth (Richards, 1988; Duggan, 2000). There is significant genetic variation among wheats for xylem vessel diameter and its measurement is not difficult (Richards and Passioura, 1981). Genotypic variation and selection for stomatal conductance and/or night-time leaf conductance were discussed previously.

The manipulation of pre- and post-anthesis water use is one way to increase the drought-dependent HI, but there are others. Surplus assimilates are stored in stems and leaf sheaths of cereals around the time of anthesis. These are in the form of water soluble carbohydrates (WSC) and can amount to 25% of the total above-ground dry weight at anthesis (van Herwaarden et al., 1998b). Carbohydrates, in the form of sucrose, can then be translocated to the developing kernels during grain filling and, if it is dry, may form up to 50% of the final grain yield (Fettell, 1992). Lack of these reserves, especially in crops of high nitrogen status, has been found to be responsible for haying-off, resulting in a low yield of shrivelled grain (van Herwaarden et al., 1998a). The redistribution of stored WSC can therefore be very important to increase HI and yield. Figure 6 shows the relationship between loss in stem weight between anthesis and physiological maturity of bread wheat (Triticum aestivum L.), durum wheat (Triticum turgidum L.), triticale (x Triticosecale Wittm), barley (Hordeum vulgare L.), and oats (Avena sativa L.) grown side by side in a water-limited environment. Similar results have been found in different years and locations (López-Castañeda, Richards, Fettell, and van Herwaarden, unpublished, 1996). There is substantial genetic variation within temperate cereal species in the storage and remobilization of assimilate (Fig. 6). For wheats with the same anthesis time, WSC can vary among genotypes from 160 g m-2 to 360 g m-2 in large field plots (van Herwaarden, unpublished). Some of this variation has been found to be associated with the tiller inhibitor gene of wheat mentioned earlier. Table 1 shows the increased grain yield, harvest index, and kernel mass of commercially grown wheats, and wheats with reduced tillering, as well as WSCs in the above-ground biomass at anthesis. The lines with reduced tillering are unselected for yield.



View larger version (14K):
[in this window]
[in a new window]
  		Fig. 6. Relationship between apparent remobilization of stem and leaf sheath reserves and grain yield for bread wheat, durum wheat, barley, triticale, and oats. Grain yield of barley and oats has been corrected for husk mass. (Adapted from López-Castañeda and Richards, 1994.)

 

View this table:
[in this window]
[in a new window]
  		Table 1. Mean values of grain yield, yield components, and water soluble carbohydrates of wheat lines with a gene for reduced tillering compared with standard tillering types grown in seven contrasting environments in south-eastern Australia.

 
Effective selection techniques for greater accumulation of WSC have not yet been developed, although novel procedures using senescence agents have been attempted (Blum et al., 1983). An effective way to identify the best parents in a breeding program is to determine the weight loss in stems between anthesis and maturity of genotypes grown in bordered field plots. Alternatively, near-infrared reflectance spectroscopy can be used to increase throughput of WSC determinations. These analyses must be determined on a ground area basis and should provide an estimate of assimilate remobilisation. Some morphological traits may also be effective to increase assimilate storage. For example, the tiller inhibition gene results in thicker stems, and lines with this gene have very high storage capacity. Also, there is variation in the size and anatomy of the internode cavity of the wheat stem which may be important for assimilate storage.

Some dry environments have a high probability of rainfall during the grain filling period. For these, it may be possible to extend the duration of grain filling, and the leaf area duration, resulting in a higher HI. This would also allow time for further translocation of assimilates to the grain as well as extra photosynthesis. Genetic variation to delay leaf senescence and extend the duration of grain filling is well established in maize (Zea mays L.) (Russell, 1991) and sorghum (Sorghum bicolor L.) (Borrell et al., 2000) and similar variation is likely to be present in other species. Selection for leaf rolling may also be effective in delaying senescence. Leaf rolling is an important trait for shedding radiant energy and is likely to result in cooler leaf temperatures, less transpiration, and lower respiratory losses. It may also be important for maximizing photosynthesis and TE by unrolling in the morning when the plant has a high leaf water potential and vapor pressure deficit is low, and rolling when conditions become more unfavorable.


   Concluding Remarks
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Trait Identification
 Increased Water Use
 Water-Use Efficiency
 Harvest Index
 Concluding Remarks
REFERENCES
 
There is no easy route to achieving genetic improvements in yield and water use efficiency in dryland environments. Even the most assured methods such as empirical breeding where plot yield is the unit of selection, are difficult and slow because of the unpredictable variation in temperature and rainfall across locations and years. A physiological approach, where the underlying physiological limitations to yield can be identified, so as to more accurately target the limiting factors, also carries risks. If successful, however, the benefits are likely to be important. A century of wheat breeding has not increased crop biomass, one of the two components of grain yield. But we have identified a number of traits which singly, or pyramided together, are likely to increase biomass and yield. The effective use of these traits in a breeding program is a major opportunity and challenge. There are likely to be both predictable and unpredictable pleiotropic effects of different traits which adds a further degree of complexity. For example, increasing early vigor may increase crop water use leaving less available for grain filling and hence the possibility of a lower harvest index and yield if there is a terminal drought. It may be that increased early vigor has to be combined with earlier flowering to maximize the yield benefits. Furthermore, a greater SLA may be important to increase early vigor but it may reduce TE early in the season through its effect on assimilation. When lines have been selected for specific traits and are being evaluated in breeding programs, it will be important to retain adequate genetic variation for other traits so that any predictable/unpredictable effects may be overcome. Another challenge lies in trait validation because a character may be important in one year but not in the next. The most fruitful relationship will be where physiological breeding complements empirical breeding programs.


   NOTES
TOP
 NOTES
ABSTRACT
 INTRODUCTION
 Trait Identification
 Increased Water Use
 Water-Use Efficiency
 Harvest Index
 Concluding Remarks
 REFERENCES
 
Presented at the 1999 CSSA Symposium on Water Use Efficiency, organized by Div. C-2 char, Dr. Tom Gerik.

Received for publication September 19, 2000.


   REFERENCES
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Trait Identification
 Increased Water Use
 Water-Use Efficiency
 Harvest Index
 Concluding Remarks
 REFERENCES
 




This article has been cited by other articles:


Home page
 Agron. J.Home page
C. Royo, V. Martos, A. Ramdani, D. Villegas, Y. Rharrabti, and L. F. Garcia del Moral
Changes in Yield and Carbon Isotope Discrimination of Italian and Spanish Durum Wheat during the 20th Century
Agron. J., February 26, 2008; 100(2): 352 - 360.
[Abstract] [Full Text] [PDF]

Home page
 Agron. J.Home page
F. Alvaro, J. Isidro, D. Villegas, L. F. Garcia del Moral, and C. Royo
Breeding Effects on Grain Filling, Biomass Partitioning, and Remobilization in Mediterranean Durum Wheat
Agron. J., February 26, 2008; 100(2): 361 - 370.
[Abstract] [Full Text] [PDF]

Home page
 Crop Sci.Home page
R. Tuberosa, S. Salvi, S. Giuliani, M. C. Sanguineti, M. Bellotti, S. Conti, and P. Landi
Genome-wide Approaches to Investigate and Improve Maize Response to Drought
Crop Sci., December 18, 2007; 47(Supplement_3): S-120 - S-141.
[Abstract] [Full Text] [PDF]

Home page
 Crop Sci.Home page
M. P. Reynolds, C. S. Pierre, A. S.I. Saad, M. Vargas, and A. G. Condon
Evaluating Potential Genetic Gains in Wheat Associated with Stress-Adaptive Trait Expression in Elite Genetic Resources under Drought and Heat Stress
Crop Sci., December 18, 2007; 47(Supplement_3): S-172 - S-189.
[Abstract] [Full Text] [PDF]