Field Crops Research 105 (2008) 1–14 www.elsevier.com/locate/fcr Review Drought tolerance improvement in crop plants: An integrated view from breeding to genomics Luigi Cattivelli a,b,*, Fulvia Rizza a, Franz-W. Badeck c, Elisabetta Mazzucotelli b, Anna M. Mastrangelo b, Enrico Francia a,1, Caterina Mare` a, Alessandro Tondelli a, A. Michele Stanca a a CRA – Genomic Research Centre, Via S. Protaso 302, 29017 Fiorenzuola d’Arda (PC), Italy b CRA – Cereal Research Centre, S.S. 16 km 675, 71100 Foggia, Italy c Potsdam Institute for Climate Impact Research, PF 60 12 03, 14412 Potsdam, Germany Received 2 November 2006; received in revised form 12 July 2007; accepted 20 July 2007

Abstract Drought is the most significant environmental stress in agriculture worldwide and improving yield under drought is a major goal of plant breeding. A review of breeding progress pointed out that selection for high yield in stress-free conditions has, to a certain extent, indirectly improved yield in many water-limiting conditions. Further progress requires the introduction of traits that reduce the gap between yield potential and actual yield in drought-prone environments. To achieve this three main approaches can now be exploited: (i) plant physiology has provided new insights and developed new tools to understand the complex network of drought-related traits, (ii) molecular genetics has discovered many QTLs affecting yield under drought or the expression of drought tolerance-related traits, (iii) molecular biology has provided genes useful either as candidate sequences to dissect QTLs or for a transgenic approach. The extent of information that breeders have now offers them new tools for breeding, such as markers for QTLs and single genes for plant transformation. Breeders are thus asked to blend together all knowledge on the traits sustaining yield under drought and to accumulate the most effective QTLs and/or transgenes into elite genotypes without detrimental effects on yield potential. This strategy will lead to new cultivars with high yield potential and high yield stability, that in turn will result in superior performance in dry environments. # 2007 Elsevier B.V. All rights reserved.

Keywords: Drought tolerance; Marker assisted selection; QTL; Stress index; Drought tolerance genes; Plant transformation

Contents

1. Introduction ...... 2 2. Breeding progress in favourable versus drought-prone environments ...... 2 3. Drought tolerance assessment ...... 3 4. Physiological bases for yield under drought ...... 4 5. Molecular markers to dissect drought tolerance-related traits...... 6 6. Genes and metabolites conferring drought tolerance...... 7 7. Field phenotyping ...... 9 8. Future directions ...... 10 Acknowledgements ...... 11 References ...... 11

* Corresponding author at: CRA – Genomic Research Centre, Via S. Protaso 302, 29017 Fiorenzuola d’Arda (PC), Italy. Tel.: +39 0523983758; fax: +39 0523983750. E-mail addresses: [email protected], [email protected] (L. Cattivelli). 1 Present address: University of Modena and Reggio Emilia, Department of Agricultural Sciences, Via J.F. Kennedy 17, I-42100 Reggio Emilia, Italy.

0378-4290/$ – see front matter # 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.fcr.2007.07.004 2 L. Cattivelli et al. / Field Crops Research 105 (2008) 1–14

1. Introduction legumes over the last Century in all countries, including those characterized by vast drought-prone regions. Drought is by far the most important environmental stress in Information on the genetic gain obtained by comparing agriculture and many efforts have been made to improve crop cultivars with different year of release is not absolute due to the productivity under water-limiting conditions. While natural influence of the environment (mainly water availability) on the selection has favoured mechanisms for adaptation and survival, performance of the varieties. When grain yield values of wheat breeding activity has directed selection towards increasing the and barley were estimated in environments with different levels economic yield of cultivated species. More than 80 years of of water availability, yield progress attributed to genetic breeding activities have led to some yield increase in drought improvement was, in absolute terms, higher in the environ- environments for many crop plants. Meanwhile, fundamental ments characterized by a low level of water stress (Slafer et al., research has provided significant gains in the understanding of 1994). However, when yield increases were expressed as a the physiological and molecular responses of plants to water percentage, no differences were recorded for environments deficits, but there is still a large gap between yields in optimal with different degrees of water stress (Araus et al., 2002). This and stress conditions. Minimizing the ‘yield gap’ and suggests that some of the traits selected to improve yield also increasing yield stability under different stress conditions are led to yield increases in dry environments. of strategic importance in guaranteeing food for the future. A number of physiological studies have identified some The evolution of crops since their domestication has been traits for which presence/expression is associated with plant driven by the selection of desired traits recognized at the adaptability to drought-prone environments. Among them, phenotypic level. Nevertheless, direct selection for grain yield traits, such as small plant size, reduced leaf area, early maturity under water-stressed conditions has been hampered by low and prolonged stomatal closure lead to a reduced total seasonal heritability, polygenic control, epistasis, significant genotype- evapotranspiration, and to a reduced yield potential (Fischer by-environment (G Â E) interaction and quantitative trait loci and Wood, 1979; Karamanos and Papatheohari, 1999). (QTLs)-by-environment (QTL Â E) interaction (Piepho, Depending on the stress conditions (timing and intensity) of 2000). The complexity of drought tolerance mechanisms the target environments, some adaptive traits can be considered explains the slow progress in yield improvement in drought- for yield improvement under drought if they enable plants to prone environments. In recent years, crop physiology and cope with a stress event that tends to occur every year at the genomics have led to new insights in drought tolerance same growth stage. For instance, a good level of earliness is an providing breeders with new knowledge and tools for plant effective breeding strategy for enhancing yield stability in improvement (Tuberosa and Salvi, 2006). This article aims to Mediterranean environments where wheat and barley are provide an overview of the breeding progress in drought exposed to terminal drought stress. In this condition shortening tolerance, and highlight future perspectives in plant breeding crop duration, a typical escape strategy, can be useful in that could result from the integration of the recent advances in synchronizing the crop cycle with the most favourable physiology and genomics. environmental conditions. It is known, however, that extreme earliness leads to yield penalty, with earliness not being 2. Breeding progress in favourable versus drought- correlated to grain yield in Mediterranean environments or prone environments fertile conditions (Cattivelli et al., 1994). Late heading and flowering, followed by a short grain-filling period can be Increases in yield potential achieved by plant breeding associated with higher yield when drought stress is experienced during the last Century have been well documented for early in the season, during the vegetative phase (van Ginkel numerous crops. Frequently, genetic gain has been studied by et al., 1998). A more general ‘‘xerophytic’’ breeding strategy to comparing in the same field trial the yield of cultivars improve plant survival through the limitation of evapotran- characterized by different years of release. For most crops a spiration can be applied in extremely harsh environments. linear relation between yield and year of release was found, the Nevertheless, every breeding strategy for drought-prone slope of which gives an estimate of the genetic improvement. environments also has to consider that the timing and intensity Comparison of the different cultivars then enabled identifica- of the stress events vary significantly from year to year and tion of the main morpho-physiological traits modified during plants designed to cope with a specific type of drought may selection in association with yield improvement. For instance, under-perform when the stress conditions are different or studies carried out on wheat and barley genotypes commonly absent. In a typical Mediterranean environment, years with grown in the last Century showed that the increase in grain yield ample water availability during the main cereal growing season was directly correlated to an increase of the harvest index from alternate with years in which terminal drought occurs as well as about 30 up to 55% (reviewed by Slafer et al., 1994; Cattivelli years with early drought during vegetative growth and et al., 1994), while total biomass accumulation remained almost flowering. unchanged over the years due to a concomitant decrease in In mild to moderate drought conditions characterized by a plant height, a process associated with the introgression in the wheat/barley grain yield between 2 and 5 mg haÀ1, selection elite germplasm of a few key genes affecting plant height, for high yield potential has frequently led to some yield mainly Rht (Slafer et al., 1994). In general, a genetic gain from improvements under drought conditions (Araus et al., 2002). In 10 to 50 kg haÀ1 yearÀ1 has been recorded for cereals and these cases the breeders have selected plants characterized by L. Cattivelli et al. / Field Crops Research 105 (2008) 1–14 3 high yield potential and high yield stability, with the latter being despite the greater capacity of the latter to retain water attributed to a minimal G Â E interaction. This implies that measured in terms of water potential and stomatal conductance traits maximizing productivity normally expressed in the (Frederick et al., 1990, 1991). absence of stress, can still sustain a significant yield Many reports suggest that, during the last Century, yield in improvement under mild to moderate drought (Slafer et al., water-limited environments was mainly determined by the 2005; Tambussi et al., 2005). An example is the success of inherent yield potential. This can be explained by considering wheat and rice varieties bred at CIMMYT and IRRI where that the main targets of selection (high harvest index in wheat selection under stress-free environments identified genotypes and barley, stay green in maize and sorghum, resistance to pests with high yield in a wide range of conditions including regions and diseases, nitrogen use efficiency) are equally beneficial with a low yield potential (Trethowan et al., 2002). under dry and wet conditions and, often, the best performances The rationale of this breeding strategy is also supported by for these traits were overriding the differences in drought several retrospective studies where the yield of large sets of adaptability. Nevertheless, examples where selection for stress cultivars was evaluated in parallel fields under different water resistance was effective in improving yield in limited regimes, thus enabling a direct comparison of the performance environments are also known (Ba¨nziger et al., 1999; Morgan, of the same cultivars in drought and non-drought (usually 2000). irrigated) conditions. When 89 barley genotypes representing a Further progress will depend on the introduction in high sample of the germplasm grown in Europe were evaluated in yielding genotypes of traits able to improve drought tolerance Southern Italy in rainfed and irrigated conditions, eight without detrimental effects on yield potential, thus reducing the genotypes with high yield potential and minimal G Â E gap between yield potential and yield in drought-prone interaction were identified. They ranked among the best in both environments. This goal can be achieved via the identification rainfed and irrigated treatments and, although considerably of drought tolerance-related traits and the subsequent reduced in the absence of supplementary irrigation, their yield manipulation of the corresponding genes using marker assisted was superior under all tested conditions (Rizza et al., 2004). Old selection (MAS) and/or gene transformation. varieties were characterized by low yield in rainfed conditions and by a minimal ability to improve yield when water became 3. Drought tolerance assessment available. On the other hand, modern cultivars showed a higher yield in rainfed conditions and strong yield increases in A crucial aspect in all studies dedicated to drought tolerance response to irrigation. A highly significant correlation is the assessment of the degree of drought tolerance of different (r = 0.73***) between yield in rainfed and irrigated conditions genotypes. In many studies the identification of tolerant and was found (Rizza et al., 2004), suggesting that, in a typical susceptible cultivars is based on few physiological measures Mediterranean environment, selection based on the absolute related to drought response. The difficulty in identifying a performance of the genotypes across environments is more physiological parameter as a reliable indicator of yield in dry successful than selecting for the minimum yield decrease under conditions has suggested that yield performance over a range of stress with respect to favourable conditions. Notably, among environments should be used as the main indicator for drought the genotypes with superior yielding capacity there was a large tolerance (Voltas et al., 2005). Furthermore, a number of predominance of varieties selected in regions usually not regression techniques of yield against environmental indices as affected by drought stress (e.g. Northern Italy, Sweden, France, independent variables were developed to evaluate genotype The Netherlands) demonstrating that selection under favour- adaptability. Several indices were proposed to describe yield able environments also has positive effects when plants are performance of a given genotype under stress and non-stress grown in stressed environments. conditions or in comparison with the average yield or the yield A corresponding experiment including 46 sugar beet of a superior genotype (Finlay and Wilkinson, 1963; Eberhart genotypes representing different genetic backgrounds grown and Russell, 1966; Fischer and Maurer, 1978; Soika et al., 1981; in drought and irrigated conditions led to similar results (Ober Lin and Binn, 1988; Yadav and Bhatnagar, 2001). Some authors et al., 2004). Sugar beet genotypes with high yielding capacity have expressed the yielding capacity with regard to an when irrigated also tended to perform well under drought environment-related physiological trait, such as canopy (correlation coefficient r = 0.64***), while the genotypes with temperature or water potential. To estimate the water stress minimum yield loss under stress did not belong to the group of experienced by crop plants Idso et al. (1981) suggested a ‘‘crop high yielding genotypes in either irrigated or drought water stress index’’ (CWSI) derived from the increase in conditions. average canopy temperature in relation to that of a well-watered Comparison between old and recent maize hybrids showed reference plot and evaluated by infrared thermometry. In a that selection for high yield potential is intimately linked to study on Triticum aestivum and Vicia faba water potential index selection for stress resistance; with tolerance to weed (WPI) was suggested as a measure of the total water stress interference, low soil nitrogen and low soil moisture among experienced by a crop in a given environment for a specific time the key factors that sustained the genetic gain in maize interval (Karamanos and Papatheohari, 1999). Other (Tollenar and Wu, 1999; Tollenaar and Lee, 2002). In soybean, approaches aim to quantify the degree of drought based on modern cultivars characterized by high yield under irrigated specific environmental factors (such as weather, soil water conditions, also outperformed older cultivars in rainfed trials, availability, etc.) in order to provide a measure that is 4 L. Cattivelli et al. / Field Crops Research 105 (2008) 1–14 independent of the G Â E interaction. Araus et al. (2003) found 4. Physiological bases for yield under drought that yield was well correlated with water input under different water stress conditions. Motzo et al. (2001) proposed a seasonal The physiologically relevant integrators of drought effects water stress index based on soil-plant-atmosphere interaction, are the water content and the water potential of plant tissues where stress was quantified as 1-(fraction of transpirable soil (Jones, 2007). They in turn depend on the relative fluxes of water). Rizza et al. (2004) proposed an integrated ‘‘water stress water through the plant within the soil-plant-atmosphere index’’ (WSI), based on a simple soil water balance and the continuum. Thus, apart from the resistances and water storage integrated reduction of plant transpiration relative to potential capacities of the plant, it is the gradient of water vapour transpiration. They proposed that yield potential and adapt- pressure from leaf to air, and the soil water content and potential ability of cultivars be related to water stress by means of the that impose conditions of drought on the plant. Once a drop in intercept and slope of a linear regression of yield versus WSI. water potential develops, responses of a wide range of By providing yield analysis as a function of an environ- physiological processes are induced. Some of these responses mental index these different approaches enabled comparison of are directly triggered by the changing water status of the tissues genotype performances under different degrees of water while others are brought about by plant hormones that are limitation. With studies quantifying yield potential, yield signalling changes in water status (Chaves et al., 2003). stability, and susceptibility/resistance to stress, an improvement Physiological traits relevant for the responses to water deficits of yield in drought-prone environments can now be attempted. and/or modified by water deficits span a wide range of vital An ideal genotype would combine the highest yield (highest processes (Table 1). As a consequence, it can be expected that intercept) with the lowest sensitivity to water stress (lowest there is no single response pattern that is highly correlated with slope); nevertheless very often a high yield performance under yield under all drought environments. wet and dry conditions of superior genotypes was found to be The different crop developmental stages show different associated with a high sensitivity to water stress (high intercept sensitivity to drought stress. In wheat most of the floret and high slope) (Pantuwan et al., 2002; Ober et al., 2004; Rizza primordia that reach the fertile floret stage become grains after et al., 2004; Pidgeon et al., 2006). anthesis. The number of fertile florets or grains per m2, the most

Table 1 Physiological traits relevant for response to drought conditions Plant traits Effects relevant for yield Modulation under stress References Stomatal conductance/leaf More/less rapid water consumption. Stomatal resistance increases Jones (1999), Lawlor and temperature Leaf temperature reflects the evaporation and under stress Cornic (2002) hence is a function of stomatal conductance Photosynthetic capacity Modulation of concentration of Calvin cycle Reduction under stress Lawlor and Cornic (2002) enzymes and elements of the light reactions Timing of phenological phases Early/late flowering. Maturity and growth Wheat and barley advanced flowering, Slafer et al. (2005), duration, synchrony of silk emergence and rice delayed, maize asynchrony Richards (2006) anthesis, reduced grain number Anthesis-silking interval ASI is negatively associated with yield Drought stress at flowering causes a Bolanos and Edmeades (1993), (ASI) in maize in drought conditions delay in silk emergence relative Edmeades et al. (2000) to anthesis Starch availability during A reduced starch availability leads to Inhibition of photosynthetic activity Boyer and Westgate (2004) ovary/embryo development abortion, reduced grain number reduces starch availability Partitioning and stem Lower/higher remobilization of reserves Compensation of reduced current leaf Blum (1988), reserve utilization from stems for grain-filling, effecting photosynthesis by increased Slafer et al. (2005) kernel weight remobilization Stay green Delayed senescence Rajcan and Tollenaar (1999) Single plant leaf area Plant size and related productivity Reduced under stress (wilting, Walter and Shurr (2005) senescence, abscission) Rooting depth Higher/lower tapping of soil Reduced total mass but increased Hoad et al. (2001), water resources root/shoot ratio, growth into wet soil Sharp et al. (2004) layers, regrowth on stress release Cuticular resistance and Higher or lower water loss, modification Kerstiens (1996) surface roughness of boundary layer and reflectance

Photosynthetic pathway C3/C4/CAM, higher WUE and greater heat Cushman (2001) tolerance of C4 and CAM Osmotic adjustment Accumulation of solutes: ions, sugars, Slow response to water potential Serraj and Sinclair (2002) poly-sugars, amino acids, glycinebetaine Membrane composition Increased membrane stability and changes Regulation in response to water Tyerman et al. (2002) in aquaporine function potential changes Antioxidative defense Protection against active oxygen species Acclimation of defence systems Reddy et al. (2004) Accumulation of Involved in the protection of cellular Accumulated under stress Ramanjulu and Bartels (2002), stress-related proteins structure and protein activities Cattivelli et al. (2002) L. Cattivelli et al. / Field Crops Research 105 (2008) 1–14 5 relevant component in ensuring high yield in drought having high rates of leaf conductance when soil moisture was conditions (Slafer and Whitechurch, 2001), is determined favourable, and markedly reduced leaf conductance when soil during stem elongation, a few weeks before anthesis. An moisture was limiting. Old cultivars, on the other hand, were extended duration of the stem elongation phase without a characterized by a ‘‘conservative strategy’’ with lower leaf change in the timing of anthesis has been proposed as a conductance even at high soil moisture. An effect of physiological determinant of an increase in the number of developmental plasticity affecting yield response under drought grains per m2 without altering the amount of water used by the stress was shown by the fact that modern cultivars used less crop (Slafer et al., 2005). In maize a close synchrony between water in the pre-anthesis period and had more water available in pollen shed at anthesis and silk emergence is required for high the post-anthesis period (Siddique et al., 1990). kernel set and a negative relationship exists between final Stay green is an important trait in several crops (e.g. maize, kernel number and the extent of anthesis-silking interval (ASI). rice, sorghum). Stay green plants are characterized by a post- Drought stress at flowering causes a delay in silk emergence flowering drought resistance phenotype that gives plants relative to anthesis (Bolanos and Edmeades, 1993). Meiosis, resistance to premature senescence, stalk rot, and lodging anthesis, and male and female fertility are all extremely when subjected to drought during grain-filling. Stay green has susceptible to drought stress and their failure directly affects the been extensively used in plant breeding to improve yield kernel number, thus leading to a significant yield penalty. In potential and yield stability in all environments, including maize ovary abortion is greatest when drought is experienced a drought-prone areas (Campos et al., 2004; Tollenar and Wu, few days before pollination (Boyer and Westgate, 2004). Water 1999). The improvement of grain-filling capacity under deficit around pollination increases the frequency of kernel terminal drought could also be achieved by increasing the abortion in maize due to a lack of photosynthate and to the mobilization of the vegetative reserves from stems to ears disruption of carbohydrate metabolism in ovaries (Zinselmeier (Blum, 1988). Selection for stem reserve contribution to grain- et al., 1995). In the absence of stress, the sucrose, converted by filling was extensively carried out in cereals (Blum, 1988), invertase into hexoses, is the source for starch deposition in the however, this mechanism can only be useful in environments ovaries and ovary development. Water deficit inhibits photo- where terminal drought is the main recurrent problem. synthesis and, therefore, the photosynthetic flux decreases, a Osmotic adjustment (OA) is a key mechanism enabling condition associated with a loss in invertase activity and a plants under drought to maintain water absorption and cell depletion of starch and sugars in the developing maize ovaries turgor pressure, thus contributing to sustained higher photo- and embryos. When drought treated maize plants were fed with synthetic rate and expansion growth. A general analysis of OA sucrose to ensure the photosynthate flux even when photo- in wheat under several drought stress conditions has shown that synthesis was inhibited, ovary abortion was largely prevented osmoregulation can be an effective selection criterion for suggesting that sugar starvation is the main reason sustaining drought tolerance, and that it has a role in reducing drought- ovary abortion (Zinselmeier et al., 1999; Boyer and Westgate, dependent yield loss especially when water deficit occurs 2004). Carbohydrate metabolism is also a key factor in ensuring during the reproductive growth stage (Morgan, 1983; Moi- male fertility; in wheat it has been shown that the disruption of nuddin et al., 2005). Nevertheless, a number of contrasting starch deposition in pollen grains due to losses of invertase reports on the role of OA have been published. For instance, the activity, leads to increased male sterility (Dorion et al., 1996). recent study of Turner et al. (2007) concluded that differences In rice, low water potentials around the time of anthesis may in OA were not associated with yield benefits in a population of lead to a failure of anther dehiscence which leads to male chickpea advanced breeding lines developed from a cross sterility (Saini and Westgate, 2000). In maize, water deficit later between cultivars with high and low OA. A comparative in development tends to reduce kernel size rather than number, analysis of many studies dedicated to OA has suggested that OA and size is mainly determined by photosynthetic activity and cannot be considered equally useful in all crops and/or drought available photosynthetic reserves (Boyer and Westgate, 2004; conditions, but that a general positive association between yield Saini and Westgate, 2000). and OA can be found under severe water stress where yields Several studies have addressed yield under drought stress as tend to be low (Serraj and Sinclair, 2002). a function of single physiological traits in attempts to A further example of a successful breeding program for dry understand which metabolic processes and/or morpho-physio- environments, based on a physiological trait, was reported by logical traits are crucial in ensuring high yield performance Rebetzke et al. (2002). They used carbon isotope discrimination under a wide range of environments. Fischer et al. (1998) (D) as a surrogate for water use efficiency to select wheat lines studied representative semi-dwarf spring wheat cultivars for with high water use efficiency in drought-prone environments. changes in stomatal conductance associated with selection During photosynthesis plants discriminate against the heavy 13 progress. They found that stomatal conductance and maximal isotope of carbon ( C) and, as a result, in several C3 species, D rates of photosynthesis were positively correlated with is positively correlated with the ratio of internal leaf CO2 increased yields of advanced cultivars, while leaf temperatures concentration to ambient CO2 concentration (Ci/Ca) and were negatively correlated. Looking at the relationships negatively associated with transpiration efficiency. Thus, a between conductance and leaf water potential in modern high Ci/Ca leads to a higher D and a lower transpiration wheat cultivars, Siddique et al. (1990) suggested that wheat efficiency (Farquhar and Richards, 1984). In wheat, when varieties may be ‘‘opportunistic’’ in relation to available water, employing divergent selection, the resulting low D lines had 6 L. Cattivelli et al. / Field Crops Research 105 (2008) 1–14 increased aboveground biomass and kernel weight. Yield was goal for present and future research. Looking for the increased by about 2% under mild stress conditions and up to coincidence of loci for specific traits and loci for yield under about 10% under the driest conditions (Rebetzke et al., 2002). drought stress and in stress-free environments, it is possible to test more precisely whether a specific trait is of significance in 5. Molecular markers to dissect drought tolerance- improving drought tolerance and yield potential. For example related traits in rice, QTLs for plant yield under drought were coincident with QTLs for root traits and OA (Babu et al., 2003). Several Molecular markers can be used to explore germplasm major loci for yield under different environmental regimes were through segregation and association mapping to identify useful mapped along with QTLs for late senescence of the flag leaf in alleles in both cultivated varieties and wild relatives. Although winter wheat (Verma et al., 2004). Similarly, Lanceras et al. association mapping is intrinsically more powerful than (2004) found that favourable alleles for yield components were ‘classical’ genetic linkage mapping because it scrutinizes the located in a region of rice chromosome 1 where QTLs for many results of thousands of generations of recombination and drought-related traits (root dry weight, relative water content, selection (Syva¨nen, 2005), most of the data available up to date leaf rolling and leaf drying) were previously identified (Zhang on drought tolerance are based on segregation mapping and et al., 2001). These results may suggest that selection for QTL analysis. Many efforts have been dedicated to under- drought tolerance could become more efficient thanks to the standing the genetic basis of physiological traits conferring availability of handle markers tightly linked to loci for stress advantages in dry environments, while less attention has been related traits. Marker diagnostics of individual QTLs represents given to understanding the high yield stability in dry and wet an important surrogate for physiological trait measurements, environments. and may ultimately improve breeding efficiency through MAS. Drought tolerance is a typical quantitative trait, however However, we also need to consider that favourable alleles for single genes, such as those controlling flowering time, plant drought tolerance and for yield components detected in QTL height, ear type and OA, may have important roles in adaptation studies could be contributed by the two contrasting parents to drought-prone environments (Forster et al., 2004). For (Lanceras et al., 2004). Therefore, it is possible that selection instance, a single gene influencing OA in wheat was mapped on for drought tolerance using these regions as targets may incur a the short arm of chromosome 7A (or gene—Morgan and Tan, penalty on grain yield by decreasing the yield components. 1996), and a breeding program for the or gene was shown to One of the possible disadvantages of molecular markers is increase yield under reduced water supply conditions (Morgan, that the genetic linkage between a specific random DNA marker 2000). and a target locus allele, established by QTL studies, can be During the last 10 years, the application of QTL analysis broken by genetic recombination, although a QTL/gene can be provided unprecedented opportunities to identify chromosome tagged by two flanking markers to reduce the recombination regions controlling variations in almost all the physiological, risk. Furthermore an accurate QTL mapping usually resulting morphological and developmental changes observed during in a small QTL interval, is also a pre-requisite to improve MAS- plant growth in water-limiting conditions. Particular attention QTL efficiency. These intrinsic difficulties together with the has been paid to: (i) genetic variation of the OA (Teulat et al., polygenic nature of drought tolerance and the interaction with 1998; Robin et al., 2003), (ii) genetic bases of phenological the environment, makes MAS for drought tolerant QTLs traits—e.g. stay green phenotype (Sanchez et al., 2002; Jiang extremely difficult (Francia et al., 2005) due to the number of et al., 2004; Verma et al., 2004); (iii) the ability of the roots to genes involved and the interactions among them (epistasis). exploit deep soil moisture to meet evapotranspirational demand The fact that numerous genes are involved in the expression of (Johnson et al., 2000; Nguyen et al., 2004); (iv) the limitation of polygenic traits means that the individual genes generally have water-use by reduction of leaf area and shortening of growth small effects on the plant phenotype. This implies that several period (Anyia and Herzog, 2004); (v) isotope discrimination regions (i.e. QTLs) must be manipulated at the same time in (Martin et al., 1989; Juenger et al., 2005; Saranga et al., 2004); order to obtain a significant impact, and that the effect of (vi) the limitation of non-stomatal water loss from leaves—e.g. individual regions is not readily identifiable. Replicated field through the cuticle (Lafitte and Courtois, 2002) and (vii) the tests are needed in order to accurately characterize the effects of response of leaf elongation rate to soil moisture and evaporative QTLs and to evaluate their stability across environments. demand (Reymond et al., 2003). Although significant QTL effects can be detected across a range Comparative analysis of QTL results clearly shows that of environments, the evaluation of the QTL  E interactions chromosomal regions determining variation in agronomic and remains a major constraint on the efficiency of MAS (Beavis physiological drought-related traits cover a large proportion of and Keim, 1996). the whole genome. For many crop plants information on The contribution of genomics-assisted breeding to the drought-related QTL findings have been collected in open development of drought-resistant cultivars has so far been source databases, such as GRAMENE (http://www.grame- marginal and only few significant examples of MAS for traits ne.org/) or GRAINGENES (http://wheat.pw.usda.gov/GG2/). associated with drought tolerance have been reported. Due to The association between variation in drought-related the high cost of molecular marker technology and the relative quantitative traits and, ultimately, the effects of these traits low efficiency in MAS for QTL, particularly for traits with a on yield in drought and favourable environments is the main strong G  E interaction and epistatic effects, the cost L. Cattivelli et al. / Field Crops Research 105 (2008) 1–14 7 effectiveness of QTL-MAS versus standard breeding strategies tag the different alleles of the drought-related genes (QTN— in commercial programs still needs to be accurately evaluated quantitative trait nucleotide). (Barker et al., 2005). Ribaut and Ragot (2007) used a marker In most QTL studies the work has not been extended beyond assisted backcross in maize to introgress the favourable alleles their detection for a given trait under drought. The development at five QTLs explaining about 38% of the total phenotypic of consensus QTL maps generated from a number of crosses is variance for the interval between anthers and silks extrusion an important step towards the identification of regions (ASI—Bolanos and Edmeades, 1996), a trait negatively commonly associated with drought tolerance. A major associated with yield in drought conditions. Grain yield of challenge remaining is to confirm that QTLs discovered in a the best maize hybrids selected with molecular markers for four given mapping population will improve drought tolerance when generations was, on average, 50% higher than control hybrids introduced into high yielding elite genotypes. This is under severe water stress conditions. It is worth noting that no particularly difficult when the traits are governed by ‘‘context yield penalty was observed under well-watered conditions. dependent’’ gene effects (i.e. interaction with other genes and/ Similarly, when a pearl millet major QTL for grain yield under or environment). In these cases the value of the QTL alleles can terminal drought stress was transferred into a drought sensitive differ depending on the genetic structure of the current genotype the genotypes carrying the introgression at the target germplasm set in the breeding program (Wade, 2002). Under QTL showed a consistent grain yield advantage (Serraj et al., these conditions, the value of a given QTL allele can change 2005). In rice, MAS was used to transfer several QTLs for deep during selection due to changes in the background effects at any roots from the japonica upland cultivar ‘‘Azucena’’, adapted to given time in the breeding process. As a consequence, when the rainfed conditions, to the lowland indica variety ‘‘IR64’’. MAS background effects are important, the stacking of desirable selected lines showed a greater root mass and higher yield in alleles by MAS becomes inadequate because the initial target drought-stressed trials (Courtois et al., 2003). In sorghum, combination of alleles may no longer be the best target, or even molecular markers were exploited to develop near isogenic a relevant target, for increasing trait performance in subsequent lines each containing one of four stay green QTLs previously breeding cycles. The ‘‘Mapping As-You-Go’’ strategy (Podlich identified (Harris et al., 2007). Favourable alleles in each of the et al., 2004) involves repeated re-estimation and validation of four loci contributed to the lower rate of leaf senescence under the QTL effects throughout the breeding process to ensure that post-anthesis water deficit. they remain relevant throughout. This method results in An important step towards the application of molecular substantial increases in MAS efficiency compared with markers in breeding for drought tolerance is the cloning of standard approaches based on the evaluation of the QTL DNA sequences underlying QTLs. To date most plant QTLs effects only at the beginning of the breeding program, have been cloned by the positional cloning approach, although particularly when epistasis and/or genotype  environment alternative strategies based on candidate genes and linkage interactions play a significant role. disequilibrium may represent an interesting shortcut to QTL cloning (Salvi and Tuberosa, 2005). A candidate gene for 6. Genes and metabolites conferring drought tolerance drought tolerance usually refers to a sequence for which the expression profile or protein function can be associated with the New chances to further improve yield and/or yield stability stress response/adaptation process, and the position on the under limiting conditions come from the last 10 years’ progress genome co-maps with a QTL conferring drought tolerance. in the identification of the genetic determinants of the Candidate genes can be selected from literature data, by physiological responses related to stress tolerance. Adaptation mapping of known stress responsive genes (Tondelli et al., of plants to drought and to the consequent cellular dehydration 2006), or using bioinformatic analysis of all genes present in induces an active plant molecular response. This response QTL-underlined genomic regions. significantly improves the tolerance to negative constraints and So far no QTL for drought tolerance has been cloned in crop it is to a great extent under transcriptional control. Many stress- species, although a recent study in Arabidopsis has led to the related genes have been isolated and characterized in the last cloning of the ERECTA gene, a sequence beyond a QTL for two decades in a variety of crop species (Cattivelli et al., 2002; transpiration efficiency (Masle et al., 2005). In plants with large Ramanjulu and Bartels, 2002), however, the complexity of the genomes, the generation of molecular-linkage maps based on whole molecular response to drought in crop plants has only candidate genes (molecular-function maps) is one way to recently been revealed by large transcriptome analyses (Hazen identify the genetic determinants of QTLs—i.e. functional et al., 2005; Buchanan et al., 2005; Kollipara et al., 2002). markers (Causse et al., 2004; Chen et al., 2001), in spite of the Molecular analysis in Arabidopsis has sketched the complex time-consuming fine mapping. This candidate gene strategy network constituting cell communication during drought shows promise to bridge the gap between quantitative genetic response. From model plants, genetic information is being and molecular genetic approaches to study complex traits and, moved to crops exploiting genome synteny, taking advantages for example, has been applied to find genes potentially involved of conserved molecular pathways, including those controlling in barley and rice drought tolerance (Zheng et al., 2003; stress tolerance. Following this approach the regulatory Nguyen et al., 2004; Diab et al., 2004; Tondelli et al., 2006). components of the drought response are being searched and The identification of the QTL corresponding genes will also identified in crop plants (Li et al., 2005; Shen et al., 2003; Kizis provide the best markers for MAS, those designed to directly and Pages, 2002; Mare` et al., 2004). 8 L. Cattivelli et al. / Field Crops Research 105 (2008) 1–14

Transgenic plants have been developed either to up-regulate (OsCDPK7) also resulted in enhanced levels of stress- the general stress response or to reproduce specific metabolic or responsive genes that contribute to improved salt and drought physiological processes previously shown to be related to tolerance (Saijo et al., 2000). drought tolerance by classical physiological studies (Table 2). Notably, not all transcription factors involved in drought Transcription factors as well as components of the signal signal transduction are suitable targets for biotechnology. transduction pathways that coordinate expression of down- DREB2 plays a major role in drought-related gene expression, stream regulons are thought to be optimal targets for nevertheless, when overexpressed in transgenic plants, DREB2 engineering of complex traits, such as stress tolerance. caused only a weak induction of the downstream genes and did Successful examples are transgenic crops engineered with not result in enhanced stress tolerance, probably because of lack genes encoding the DREBs/CBFs transcription factors of post-translational modifications (Sakuma et al., 2006). On (tomato—Hsieh et al., 2002; rice—Dubouzet et al., 2003; Ito the contrary, CBF3/DREB1A in transgenic rice increased et al., 2006 and wheat—Pellegrineschi et al., 2004). The drought and salinity tolerance without affecting growth transgenic plants showed increased stress tolerance as well as undesirably (Oh et al., 2005). Overexpression of transcription the over induction of downstream stress related genes and/or factors may also activate additional non-stress-related genes higher levels of soluble sugars and proline. A recent report has that adversely affect the normal agronomic characteristics of a shown that rice plants overexpressing the SNAC1 (stress- crop, producing deleterious effects on the phenotype and thus responsive NAC1) transcription factor showed improved yield (Wang et al., 2003). Common detrimental effects due to drought tolerance and yield potential under field conditions. constitutive expression of regulative factors are growth The leaves of SNAC1-overexpressing plants lost water more retardation, reduced fruit and seed numbers and fresh weight slowly showing an increased stomatal closure and ABA of transgenic plants under normal conditions. Alternative sensitivity (Hu et al., 2006). Ectopic expression of a stress strategies based on stress-inducible promoters cause minimal induced rice gene encoding a calcium-dependent protein kinase negative effects under normal growth conditions and enhance

Table 2 Relevant examples of genes conferring drought tolerance Genes Function Mechanism of action References DREBs/CBFs; ABF3 Stress induced transcription factors Enhanced expression of downstream Oh et al. (2005), stress related genes confers drought/cold/salt Ito et al. (2006) tolerance. Constitutively overexpression can lead to stunting growth SNAC1 Stress induced transcription factor SNAC1 expression reduces water loss Hu et al. (2006) increasing stomatal sensitivity to ABA OsCDPK7 Stress induced Ca-dependent Enhanced expression of stress responsive genes Saijo et al. (2000) protein kinase Farnesyl-transferase (ERA1) Negative-regulator of ABA sensing Down-regulation of farnesyltransferase Wang et al. (2005) enhances the plant’s response to ABA and drought tolerance reducing stomatal conductance Mn-SOD Mn-superoxide dismutase Overexpression improves stress tolerance McKersie et al. (1996) also in field conditions AVP1 Vacuolar H+-pyrophosphatase Overexpression facilitate auxin fluxes Gaxiola et al. (2001), leading to increased root growth Park et al. (2005) HVA1; OsLEA3 Stress induced LEA proteins Over-accumulation of LEA increases Bahieldin et al. (2005), drought tolerance also in field conditions Xiao et al. (2007) ERECTA A putative leucine-rich repeat ERECTA acts as a regulator of transpiration Masle et al. (2005) receptor-like kinase is a major efficiency with effects on stomatal density, contributor to a locus for D on epidermal cell expansion, mesophyll cell Arabidopsis chromosome 2 proliferation and cell–cell contact otsA and otsB Escherichia coli trehalose Increased trehalose accumulation correlates Garg et al. (2002) biosynthetic genes with higher soluble carbohydrate levels, elevated photosynthetic capacity and increased tolerance to photo-oxidative damage P5CS d-Pyrroline-5-carboxylate synthetase Enhanced accumulation of proline Kavi Kishor et al. (1995), leads to increased osmotolerance Zhu et al. (1998) mtlD Mannitol-1-phosphate dehydrogenase Mannitol accumulation leads to Abebe et al. (2003) increased osmotolerance GF14l 14-3-3 protein Lines overexpressing GF14l have a ‘‘stay green’’ Yan et al. (2004) phenotype, improved water stress tolerance and higher photosynthetic rates under water deficit conditions NADP-Me NADP-malic enzyme The overexpression decreased stomatal Laporte et al. (2002) conductance and improves WUE L. Cattivelli et al. / Field Crops Research 105 (2008) 1–14 9 stress tolerance, although it remains to be established if the conductance and transpiration under drought stress conditions. threshold stress under which the promoter is active corresponds Furthermore, transgenic plants were more resistant to water well to stress levels in target environments (Garg et al., 2002; deficit-induced seed abortion during flowering (Wang et al., Wang et al., 2005). 2005). Metabolic engineering for increasing osmolyte contents was A more robust root system enables plants to take up greater successful in several plants subjected to stress (Wang et al., amounts of water during water deficit stress, resulting in a more 2003), although real advantages of such a strategy are always a favourable plant water status and less injury. Although this subject of debate (Serraj and Sinclair, 2002). Given that the consideration has been obvious for many years, only recent target compounds did not achieve levels sufficient to sustain a studies have found the way to increase root size through single role in OA, chaperone-like activity and scavenging of reactive gene transformation. The gene coding for the vacuolar H+- oxygen species were proposed as alternative functions in plant pyrophosphatase (H+-PPase)-AVP1-plays an important role in protection during stress exposure. The first example of root development through the facilitation of auxin fluxes. metabolic engineering for drought tolerance was the over- Overexpression of AVP1 in Arabidopsis and tomato resulted in production of proline in transgenic tobacco (Kavi Kishor et al., more pyrophosphate-driven cation transport into root vacuolar 1995) and rice (Zhu et al., 1998), resulting in an enhanced fractions, an increased root biomass, and an enhanced recovery biomass under stress conditions. Garg et al. (2002) developed of plants from an episode of soil water deficit stress (Gaxiola drought tolerant transgenic rice lines showing tissue- or stress- et al., 2001; Park et al., 2005). inducible accumulation of trehalose, which accounted for higher soluble carbohydrate levels, a higher capacity of 7. Field phenotyping photosynthesis and a concomitant decrease in photo-oxidative damage, and more favourable mineral balance under both stress A comprehensive and careful field evaluation of mapping and non-stress conditions, without negative effects. A populations and transgenic plants is urgently needed in order to significant improvement of wheat tolerance to water deficit provide reliable information on the effectiveness of QTLs, was also achieved by Abebe et al. (2003), through the ectopic candidate genes and transgenes. Due to the multigenic nature of expression of the mannitol-1-phosphate dehydrogenase (mtlD) drought tolerance, the introduction of a single gene or QTL into gene that caused a small increase in the level of mannitol. an elite germplasm may result in a subtle phenotypic effect or An emerging strategy to broaden stress tolerance in plants is yield increase. Capacity for precise phenotyping under reliable to maintain energy homeostasis under stress conditions. In fact, conditions probably represents the most limiting factor for the it is known that the high energy consumption of the plant’s progress of genomic studies on drought tolerance. Often field stress response increases respiration rate with a linked experiments designed to evaluate genetic differences in drought production of reactive oxygen species (Tiwari et al., 2002). tolerance are faced with contrasting requirements. There is a De Block et al. (2005) obtained Brassica napus plants tolerant need for a high precision because the differences may be small to multiple stresses by preventing over activation of mitochon- and subtle, and detailed physiological measurements (i.e. drial respiration and high energy consumption related to stress- evaluation of the photosynthetic activity) are difficult when a inducible poly(ADP-ribosyl)ation activity. large numbers of genotypes are involved; the QTL studies for A ‘‘stay green-like’’ phenotype was obtained when the instance are based on segregating populations with hundreds of Arabidopsis gene GF14 lambda encoding a 14-3-3 protein was progenies in the same trial. To achieve an accurate phenotyping introduced into cotton plants. The transgenic plants showed it is also important to control stress levels and timing. The improved water stress tolerance and a higher photosynthesis intensity of drought stress is often different from year to year rate due to an altered stomatal conductance under water deficit and within fields because of variations in soil composition conditions (Yan et al., 2004). Since an earlier-than-normal which determine the capability of the soil to retain water. To stomatal closure in a crop is considered a positive trait to reduce the signal-to-noise ratio in field based experiments there improve water use efficiency in drought environments (Sinclair is a need to select research plots with low spatial variability in and Muchow, 2001), developing transgenic plants with a soil properties. Application of nutrients and the control of drought-avoidance phenotype represents a possible strategy for weeds/pests should be carried out precisely and uniformly; crop improvement. A phenotype with decreased conductance experimental design should control within-replica variability and higher water use efficiency was obtained in tobacco plants (Edmeades et al., 2004). Variations in rainfall amount and overexpressing a maize NADP-malic enzyme, the primary distribution strongly influence the level and timing of the stress; decarboxylating enzyme in C4 photosynthesis (Laporte et al., the use of rain shelters and supplementary irrigation can help to 2002). The implication of abscissic acid hormone as a control the stress conditions and improve the quality of the molecular signal in drought activated pathways and in the phenotyping work. control of stomatal closure makes ABA synthesis and response In the last decade many differently engineered plants have a possible target for improving drought tolerance. When a been proposed and tested for improved performance under farnesyl-transferase acting as a negative-regulator of ABA drought. Nevertheless, in many reports desiccation and salt sensing was down-regulated in a drought inducible manner in stresses applied are ‘shock’ treatments, while for most crops Brassica napus the transgenic plants showed enhanced ABA drought tends to develop slowly as the soil dries. The evaluation sensitivity, as well as a significant reduction in stomatal of drought tolerance of transgenic plants has often been based 10 L. Cattivelli et al. / Field Crops Research 105 (2008) 1–14 on survival capacity, with very limited analyses of the transgene through the unconscious pyramiding of yield-related traits or effects on yield potential. Furthermore, the assessment of plant loci. Research in the last three decades has opened up three water status with different methods, often with a visual score, main approaches: (i) plant physiology provided new tools to makes the comparison among different reports difficult. A understand the complex network of drought-related traits and comparative evaluation of all transgenic plants available for several drought-related traits useful to improve selection drought tolerance would provide important information for the efficiency have been proposed (see Table 1); (ii) molecular exploitation of transgenic work in breeding programs. genetics has led to the discovery of a large number of loci Few reports have been published on transgenic plants tested affecting yield under drought or the expression of drought for drought tolerance in field trials, although more trials are tolerance-related traits; and (iii) molecular biology has presently carried out by private breeding companies and a first provided genes that are either useful as candidate sequences generation of drought tolerant corn varieties is expected within to dissect QTLs or for transgenic approaches (see Table 2). The the coming years (Campos et al., 2004). Obviously, only field integration of molecular genetics with physiology is leading to trials under real stress conditions allow for conclusive remarks the identification of the most relevant loci controlling drought on stress tolerance and yield performance of a genotype. tolerance and drought-related traits. Routine cloning of the Transgenic wheat plants from six independent transgenic genes underlying the QTLs is still a long way off, but it will, events overexpressing the barley HVA1 gene were tested under ultimately, provide simple markers for an effective MAS. irrigated and rainfed conditions over six seasons (Bahieldin Nevertheless MAS for drought tolerance will not be an easy et al., 2005). Although the effect of the transgene was changing task because dozens of QTLs for drought-related traits have from year to year, the field trials showed that the HVA1 gene has been identified. Selecting which QTLs/traits follow with MAS the potential to confer drought stress protection in field is now crucial. The improvement of drought tolerance should conditions. Results from 3 consecutive years of field trials with not be achieved with a parallel limitation of yield potential. transgenic Brassica napus carrying a down-regulation of a Hence, drought tolerance traits should be tested in both stressed farnesyl-transferase acting as a negative regulator of ABA and non-stressed environments before being introduced in a sensing suggested that with adequate water, transgenic plants MAS breeding program. QTLs for drought-related traits produced the same amount of seed as the parental control, while coincident with QTLs for yield potential should be considered under moderate drought stress conditions at flowering seed as priority targets for MAS. yield of transgenic canola was significantly higher than the The success of any selection process relies on the control (Wang et al., 2005). Encouraging field trial data have availability of superior alleles for the target trait. Most QTLs also been published for alfalfa plants overexpressing the for drought tolerance, rather than being chosen for their overall superoxide dismutase gene (McKersie et al., 1996), and for rice agronomic value, have been identified in segregating popula- overexpressing the stress responsive NAC1 transcription factor tions derived from parental lines chosen to maximize the (Hu et al., 2006) as well as the stress responsive OsLEA3 gene differences in the target traits. Typically, a segregating (Xiao et al., 2007). The report showed that the higher yield population from a cross between modern and old varieties under drought conditions achieved by transgenic OsLEA3 rice, allows the identification of many QTLs, nevertheless a majority was due to increased spikelet fertility. These phenotypic data as of the positive QTL alleles might derive from the modern well as the fact that the OsLEA3 gene is located within the parental line and therefore are already present in the best interval of QTL controlling spikelet fertility under drought performing cultivars. A chance to find new useful alleles is conditions illustrate the key role of ovary fertility in drought represented by the exploitation of wild germplasm. During the tolerance (Xiao et al., 2007). domestication process, wild plants carrying promising traits Even if the use of biotechnology in agriculture offers great were cultivated, leading to locally adapted landraces. These lost potential benefits to farmers, the effectiveness of a transgenic- many undesirable alleles and useful alleles became enriched in based breeding strategy is nowadays hampered by non- the cultivated gene pool (Tanksley and McCouch, 1997). Many biological constraints, related to the commercialisation of studies have demonstrated the value of alleles originating from transgenic crops, particularly in Europe. To generate commer- non-cultivar germplasm (Tanksley and Nelson, 1996; Tanksley cially useful products scientists must increasingly integrate and McCouch, 1997), showing that centuries of selective social, legal, and economic issues as well as technical issues breeding have thrown away useful alleles in addition to many from the earliest stages of project design. The commercial useless ones. The effect of a given QTL/locus can also be success of these products will depend upon the development of influenced by the background of the genotypes used in the regulatory frameworks that are clearly defined and scientifi- breeding program epistatic interactions among QTLs, for cally based, and upon public acceptance of transgenic plant example, might hamper the development of an efficient MAS products (McElroy, 1999). program. Notably, when four genomic segments carrying QTLs for root length were transferred from the rice cultivar 8. Future directions ‘‘Azucena’’ into ‘‘Kalinga III’’, only one of them significantly increased root length in the new genetic background (Steele When phenotypic selection was the only tool available to et al., 2006). improve yield under drought, the improvements in crop yield Transgenic breeding will also have a role in the future and observed were likely due to an increase in yield potential the possibility of cloning stress-related QTLs will enable the L. Cattivelli et al. / Field Crops Research 105 (2008) 1–14 11 simultaneous engineering of multiple genes governing quanti- from the Italian Ministry of Agriculture (Progetto FRUMISIS), tative traits. However, the scarcity of field trials for drought the Italian Ministry of Science (Progetto AGROGEN) and by tolerant transgenic plants does not allow for final conclusions. EU INCOA3 Project MABDE No. ICA3-2002-10073. New transgenic plants where the gene is introduced into elite genotypes have to be tested under optimal as well as drought conditions to evaluate the impact of the transgenes on yield References potential and stress tolerance. Compared with ten years ago, breeders now have new Abebe, T., Guenzi, A.C., Martin, B., Cushman, J.C., 2003. Tolerance of perspectives for plant improvement for drought. First, they mannitol-accumulating transgenic wheat to water stress and salinity. Plant Physiol. 131, 1748–1755. will soon have new tools, such as markers for QTLs and single Anyia, A.O., Herzog, H., 2004. Water-use efficiency, leaf area and leaf gas genes for plant transformation that will provide an exchange of cowpeas under mid-season drought. Eur. J. Agron. 20, 327– opportunity to move the selection from phenotype to 339. genotype. Second, they will have a new role, to blend Araus, J.L., Slafer, G.A., Reynolds, M.P., Royo, C., 2002. Plant breeding and together all knowledge of the traits sustaining yield under drought in C3 cereals: what should we breed for? Ann. Bot. 89, 925–940. Araus, J.L., Villegas, D., Aparicio, N., Garcı`a del Moral, L.F., El Hani drought and to accumulate randomly dispersed QTLs and/or Rharrabti, Y., Ferrio, J.P., Royo, C., 2003. Environmental factors determin- transgenes into elite genotypes. This so called ‘‘breeding by ing carbon isotope discrimination and yield in durum wheat under Med- design’’ strategy (Peleman and Van der Voort, 2003) will lead iterranean conditions. Crop Sci. 43, 170–180. to new highly yielding cultivars able to improve performance Babu, C.R., Nguyen, B.D., Chamarerk, V., Shanmugasundaram, P., Chezhian, in both high and low yielding environments. A significant P., Juyaprakash, P., Ganesh, S.K., Palchamy, A., Sadasivam, S., Sarkarung, S., Wade, L.J., Nguyen, T.H., 2003. Genetic analysis of drought resistance in example showing how the integration of stress physiology and rice by molecular markers: association between secondary traits and field genomics can lead to an integrated view of plant breeding is performance. Crop Sci. 43, 1457–1469. represented by the studies on transpiration efficiency. During Bahieldin, A., Hesham, H.T., Eissa, H.F., Saleh, O.M., Ramadan, A.M., Ahmed, drought stress plants generally coordinate photosynthesis and I.A., Dyer, W.E., El-Itriby, H.A., Madkour, M.A., 2005. Field evaluation of transpiration, although significant genetic variation in transgenic wheat plants stably expressing the HVA1 gene for drought tolerance. Physiol. Plant 123, 421–427. transpiration efficiency has been identified both between Ba¨nziger, M., Edmeades, G.O., Lafitte, H.R., 1999. Selection for drought and within species (Rebetzke et al., 2001). After the tolerance increases maize yield across a range of nitrogen levels. Crop demonstration that D of plant matter is a reliable and Sci. 39, 1035–1040. sensitive marker negatively related to variation in transpira- Barker, T., Campos, H., Cooper, M., Dolan, D., Edmeades, G., Habben, J., tion efficiency (Martin et al., 1989), several selection Schussler, J., Wright, D., Zinselmeier, C., 2005. Improving drought toler- ance in maize. Plant Breed. Rev. 25, 173–253. programs based on D were carried out (Rebetzke et al., Beavis, W.D., Keim, P., 1996. Identification of quantitative trait loci that are 2002; Juenger et al., 2005) and new wheat cultivars, Drysdale affected by environment. In: Kang, M.S., Gauch, H.G. (Eds.), Genotype- and Rees, with improved water use efficiency have already by-Environment Interaction. CRC Press, Boca Raton, FL, pp. 123–149. been released in Australia. The isolation of a gene, ERECTA, Blum, A., 1988. Improving wheat grain filling under stress by stem reserve that regulates transpiration efficiency in Arabidopsis (Masle mobilisation. Euphytica 100, 77–83. Bolanos, J., Edmeades, G.O., 1993. Eight cycles of selection for drought et al., 2005), and the transcriptional analysis of wheat tolerance in lowland tropical maize.1. Responses in grain-yield, biomass, genotypes with contrasting transpiration efficiency (Xue and radiation utilization. Field Crops Res. 31, 233–252. et al., 2006) are providing the molecular bases of the isotopic Bolanos, J., Edmeades, G.O., 1996. The importance of the anthesis-silking discrimination parameter. Hence, in the near future, an interval in breeding for drought tolerance in tropical maize. Field Crop Res. integrated approach for transpiration efficiency could involve 48, 65–80. Boyer, J.S., Westgate, M.E., 2004. Grain yields with limited water. J. Exp. Bot. the use of the physiological test (analysis of D), the markers 55, 2385–2394. for QTLs controlling D,andtheERECTA gene as well as of Buchanan, C.D., Lim, S., Salzman, R.A., Kagiampakis, I., Morishige, D.T., other genes either to search for allelic variations in the Weers, B.D., Klein, R.R., Pratt, L.H., Cordonnier-Pratt, M.M., Klein, P.E., germplasm or as a tool for plant transformation. Every day, it Mullet, J.E., 2005. Sorghum bicolor’s transcriptome response to dehydra- becomes more evident that successful breeding for stable high tion, high salinity and ABA. Plant Mol. Biol. 58, 699–720. Campos, H., Cooper, M., Habben, J.E., Edmeades, G.O., Schussler, J.R., 2004. yield under drought conditions will only be possible when a Improving drought tolerance in maize: a view from industry. Field Crop true integration of traditional breeding with physiology and Res. 90, 19–34. genomics is achieved. Thus, to face drought stress and to Cattivelli, L., Baldi, P., Crosatti, C., Di Fonzo, N., Faccioli, P., Grossi, M., achieve sufficient grain yield in the future requires a Mastrangelo, A.M., Pecchioni, N., Stanca, A.M., 2002. Chromosome multidisciplinary approach based on plant genomics, phy- regions and stress-related sequences involved in resistance to abiotic stress in Triticeae. Plant Mol. Biol. 48, 649–665. siology and modeling (Yin et al., 2004; Wollenweber et al., Cattivelli, L., Delogu, G., Terzi, V., Stanca, A.M., 1994. Progress in barley 2005). breeding. In: Slafer, G.A. (Ed.), Genetic Improvement of Field Crops. Marcel Dekker, Inc., New York, pp. 95–181. Acknowledgements Causse, M., Duffe, P., Gomez, M.C., Buret, M., Damidaux, R., Zamir, D., Gur, A., Chevalier, C., Lemaire-Chamley, M., Rothan, C., 2004. A genetic map of candidate genes and QTLs involved in tomato fruit size and composition. The authors wish to thank an anonymous referee who J. Exp. Bot. 55, 1671–1685. provided a number of stimulating comments useful for the Chaves, M.M., Maroco, J.P., Pereira, J.S., 2003. Understanding plant responses improvement of this review. This work was supported by grants to drought—from genes to the whole plant. Funct. Plant Biol. 30, 239–264. 12 L. Cattivelli et al. / Field Crops Research 105 (2008) 1–14

Chen, X., Salamini, F., Gebhardt, C., 2001. A potato molecular-function map Hazen, S.P., Pathan, M.S., Sanchez, A., Baxter, I., Dunn, M., Estes, B., Chang, for carbohydrate metabolism and transport. Theor. Appl. Genet. 102, 284– H.S., Zhu, T., Kreps, J.A., Nguyen, H.T., 2005. Expression profiling of rice 295. segregating for drought tolerance QTLs using a rice genome array. Funct. Courtois, B., Shen, L., Petalcorin, W., Carandang, S., Mauleon, R., Li, Z., 2003. Integr. Genomics 5, 104–116. Locating QTLs controlling constitutive root traits in the rice population IAC Hoad, S.P., Russell, G., Lucas, M.E., Bingham, I.J., 2001. The management of 165  Co39. Euphytica 134, 335–345. wheat, barley and oat root systems. Adv. Agron. 74, 193–246. Cushman, J.C., 2001. Crasulacean acid metabolism. A plastic photosynthetic Hsieh, T.H., Lee, J.T., Charng, Y.Y., Chan, M.T., 2002. Tomato plants ectopi- adaptation to arid environments. Plant Physiol. 127, 1439–1448. cally expressing Arabidopsis CBF1 show enhanced resistance to water De Block, M., Verduyn, C., De Brouwer, D., Cornelissen, M., 2005. Poly(ADP- deficit stress. Plant Physiol. 130, 618–626. ribose) polymerase in plants affects energy homeostasis, cell death and Hu, H., Dai, M., Yao, J., Xiao, B., Li, X., Zhang, Q., Xiong, L., 2006. stress tolerance. Plant J. 41, 95–106. Overexpressing a NAM, ATAF, and CUC (NAC) transcription factor Diab, A.A., Teulat, B., This, D., Ozturk, N.Z., Benscher, D., Sorrells, M.E., enhances drought resistance and salt tolerance in rice. Proc. Natl. Acad. 2004. Identification of drought-inducible genes and differentially expressed Sci. U.S.A. 103, 12987–12992. sequence tags in barley. Theor. Appl. Genet. 109, 1417–1425. Idso, S.B., Reginato, R., Reicosky, D., Hatfield, J., 1981. Determining soil- Dorion, S., Lalonde, S., Saini, H.S., 1996. Induction of male sterility in wheat by induced plant water potential depression in alfalfa by means of infrared meiotic-stage water deficit is preceded by a decline in invertase activity and thermometer. Agron. J. 73, 826–830. changes in carbohydrate metabolism in anthers. Plant Physiol. 111, 137–145. Ito, Y., Katsura, K., Maruyama, K., Taji, T., Kobayashi, M., Seki, M., Shinozaki, Dubouzet, J.G., Sakuma, Y., Ito, Y., Kasuga, M., Dubouzet, E.G., Miura, S., K., Yamaguchi-Shinozaki, K., 2006. Functional analysis of rice DREB1/ Seki, M., Shinozaki, K., Yamaguchi-Shinozaki, K., 2003. OsDREB genes in CBF-type transcription factors involved in cold-responsive gene expression rice, Oryza sativa L., encode transcription activators that function in in transgenic rice. Plant Cell Physiol. 47, 1–13. drought-, high-salt- and cold-responsive gene expression. Plant J. 33, Jiang, G.H., He, Y.Q., Xu, C.G., Li, X.H., Zhang, Q., 2004. The genetic basis of 751–763. stay-green in rice analyzed in a population of doubled haploid lines derived Eberhart, S.A., Russell, W.A., 1966. Stability parameters for comparing from an indica by japonica cross. Theor. Appl. Genet. 108, 688–698. varieties. Crop Sci. 6, 36–40. Johnson, W.C., Jackson, L.E., Ochoa, O., van Wijk, R., Peleman, J., St. Clair, Edmeades, G.O., Bolan˜os, J., Elings, A., Ribaut, J.-M., Ba¨nziger, M., Westgate, D.A., Michelmore, R.W., 2000. A shallow-rooted crop and its wild pro- M.E., 2000. The role and regulation of the anthesis-silking interval in maize. genitor differ at loci determining root architecture and deep soil water In: Westgate, M.E., Boote, K.J. (Eds.), Physiology and Modelling Kernel extraction. Theor. Appl. Genet. 101, 1066–1073. Set in Maize. CSSA Special Publication No. 29. CSSA, Madison, WI, pp. Jones, H.G., 1999. Use of thermography for quantitative studies of spatial and 43–73. temporal variation of stomatal conductance over leaf surfaces. Plant Cell Edmeades, G.O., McMaster, G.S., White, J.W., Campos, H., 2004. Genomics Environ. 22, 1043–1055. and physiologist: bridging the gap between genes and crop response. Field Jones, H.G., 2007. Monitoring plant and soil water status: established and novel Crop Res. 90, 5–18. methods revisited and their relevance to studies of drought tolerance. J. Exp. Farquhar, G.D., Richards, R.A., 1984. Isotopic composition of plant carbon Bot. 58, 119–130. correlates with water-use efficiency of wheat genotypes. Aust. J. Plant Juenger, T.E., Mckay, J.K., Hausmann, N., Keurentjes, J.J.B., Sen, S., Stowe, Physiol. 11, 39–552. K.A., Dawson, T.E., Simms, E.L., Richards, J.H., 2005. Identification and Finlay, K.W., Wilkinson, G.N., 1963. The analysis of adaptation in a plant characterization of QTL underlying whole plant physiology in Arabidopsis breeding programme. Aust. J. Agric. Res. 14, 742–754. thaliana: d13C, stomatal conductance and transpiration efficiency. Plant Fischer, R.A., Maurer, R., 1978. Drought resistance in spring wheat cultivars. I. Cell Environ. 28, 697–708. Grain yield response. Aust. J. Agric. Res. 29, 897–912. Karamanos, A.J., Papatheohari, A.Y.,1999. Assessment of drought resistance of Fischer, R.A., Rees, D., Sayre, K.D., Lu, Z.M., Condon, A.G., Larque Saavedra, crop genotypes by means of the Water Potential Index. Crop Sci. 39, 1792– A., 1998. Wheat yield progress associated with higher stomatal conductance 1797. and photosynthetic rate and cooler canopies. Crop Sci. 38, 1467–1475. Kavi Kishor, P.B., Hong, Z., Miao, G.H., Hu, C.A.A., Verma, D.P.S., 1995. Fischer, R.A., Wood, J.T., 1979. Drought resistance in spring wheat cultivars. Over-expression of d-pyrroline-5-carboxylate synthetase increases proline III. Yield association with morpho-physiological traits. Aust. J. Agric. Res. production and confers osmotolerance in transgenic plants. Plant Physiol. 30, 1001–1020. 25, 1387–1394. Forster, B.P., Ellis, R.P., Moir, J., Talame`, V., Sanguineti, M.C., Tuberosa, R., Kerstiens, G., 1996. Cuticular water permeability and its physiological sig- This, D., Teulat-Merah, B., Ahmed, I., Mariy, S.A.E., Bahri, H., El Ouahabi, nificance. J. Exp. Bot. 47, 1813–1832. M., Zoumarou-Wallis, N., El-Fellah, M., Ben Salem, M., 2004. Genotype Kizis, D., Pages, M., 2002. Maize DRE-binding proteins DBF1 and DBF2 are and phenotype associations with drought tolerance in barley tested in North involved in rab17 regulation through the drought-responsive element in an Africa. Ann. Appl. Biol. 144, 157–168. ABA-dependent pathway. Plant J. 30, 679–689. Francia, E., Tacconi, G., Crosatti, C., Barabaschi, D., Bulgarelli, D., Dall’Aglio, Kollipara, K.P., Saab, I.N., Wych, R.D., Lauer, M.J., Singletary, G.W., 2002. E., Vale`, G., 2005. Marker assisted selection in crop plants. Plant Cell Tissue Expression profiling of reciprocal maize hybrids divergent for cold germi- Organ Cult. 82, 317–342. nation and desiccation tolerance. Plant Physiol. 129, 974–992. Frederick, J.R., Woolley, J.T., Hesketh, J.D., Peters, D.B., 1990. Seed yield and Lafitte, H.R., Courtois, B., 2002. Interpreting cultivar  environment interac- agronomic traits of old and modern soybean cultivars under irrigation and tions for yield in upland rice assigning value to drought-adaptive traits. Crop soil water-deficit. Field Crops Res. 27, 71–82. Sci. 42, 1409–1420. Frederick, J.R., Woolley, J.T., Hesketh, J.D., Peters, D.B., 1991. Water deficit Lanceras, J.C., Pantuwan, G., Jongdee, B., Toojinda, T., 2004. Quantitative trait development in old and new soybean cultivars. Agron. J. 82, 76–81. loci associated with drought tolerance at reproductive stage in rice. Plant Garg, A., Kim, J., Owens, T., Ranwala, A., Choi, Y., Kochian, L., Wu, R., 2002. Physiol. 135, 384–399. Trehalose accumulation in rice plants confers high tolerance levels to Laporte, M.M., Shen, B., Tarczynski, M.C., 2002. Engineering for drought different abiotic stresses. Proc. Natl. Acad. Sci. U.S.A. 99, 15898–15903. avoidance: expression of maize NADP-malic enzyme in tobacco results in Gaxiola, R.A., Li, J., Undurraga, S., Dang, L.M., Allen, G.J., Alper, S.L., Fink, altered stomatal function. J. Exp. Bot. 53, 699–705. G.R., 2001. Drought- and salt-tolerant plants result from overexpression of Lawlor, D.W., Cornic, G., 2002. Photosynthetic carbon assimilation and the AVP1 H+-pump. Proc. Natl. Acad. Sci. U.S.A. 98, 11444–11449. associated metabolism in relation to water deficits in higher plants. Plant Harris, K., Subudhi, P.K., Borrel, A., Jordan, D., Rosenow, D., Nguyen, H., Cell Environ. 25, 275–294. Klein, P., Klein, R., Mullet, J., 2007. Sorghum stay-green QTL individually Li, X.P., Tain, A.G., Luo, G.Z., Gong, Z.Z., Zhang, J.S., Chan, S.Y., 2005. reduce post-flowering drought-induced leaf senescence. J. Exp. Bot. 58, Soybean DRE-binding transcription factors that are responsive to abiotic 327–338. stresses. Theor. Appl. Genet. 110, 1355–1362. L. Cattivelli et al. / Field Crops Research 105 (2008) 1–14 13

Lin, C.S., Binn, M.R., 1988. A superiority measure of cultivar performance for Rebetzke, G.J., Condon, A.G., Richards, R.A., Farquhar, G.D., 2002. Selection cultivar  location data. Can. J. Plant Sci. 68, 193–198. for reduced carbon isotope discrimination increases aerial biomass and Mare`, C., Mazzucotelli, E., Crosatti, C., Francia, E., Stanca, A.M., Cattivelli, L., grain yield of rainfed bread wheat. Crop Sci. 42, 739–745. 2004. Hv-WRKY38: a new transcription factor involved in cold- and Rebetzke, G.J., Condon, A.G., Richards, R.A., Read, J.J., 2001. Phenotypic drought-response in barley. Plant Mol. Biol. 55, 399–416. variation and sampling for leaf conductance in wheat (Triticum aestivum L) Martin, B., Nienhuis, J., King, G., Schaeffer, A., 1989. Restriction fragment breeding populations. Euphytica 121, 335–341. length polymorphism associated with water use efficiency in tomato. Reddy, A.R., Chaitanya, K.V., Vivekanandan, M., 2004. Drought-induced Science 243, 1725–1728. responses of photosynthesis and antioxidant metabolism in higher plants. Masle, J., Gilmore, S.R., Farquhar, G.D., 2005. The ERECTA gene regulates J. Plant Physiol. 161, 1189–1202. plant transpiration efficiency in Arabidopsis. Nature 436, 866–870. Reymond, M., Muller, B., Leopardi, A., Charcosset, A., Tardieu, F., 2003. McElroy, D., 1999. Moving agbiotech downstream. Nat. Biotechnol. 17, 1071– Combining quantitative trait loci analysis and an ecophysiological model to 1074. analyze the genetic variability of the responses of maize leaf growth to McKersie, B.D., Bowley, S.R., Harjanto, E., Leprice, O., 1996. Water-deficit temperature and water deficit. Plant Physiol. 131, 664–675. tolerance and field performance of transgenic alfalfa overexpressing super- Ribaut, J.M., Ragot, M., 2007. Marker-assisted selection to improve drought oxide dismutase. Plant Physiol. 111, 1177–1181. adaptation in maize: the backcross approach, perspectives, limitations, and Moinuddin, A., Fischer, R.A., Sayre, K.D., Reynolds, M.P., 2005. Osmotic alternatives. J. Exp. Bot. 58, 351–360. adjustment in wheat in relation to grain yield under water deficit environ- Richards, R.A., 2006. Physiological traits used in the breeding of new cultivars ments. Agron. J. 97, 1062–1071. for water-scarce environments. Agric. Water Manage. 80, 197–211. Morgan, J.M., 2000. Increases in grain yield of wheat by breeding for an Rizza, F., Badeck, F.W., Cattivelli, L., Li Destri, O., Di Fonzo, N., Stanca, A.M., osmoregulation gene: relationship to water supplies and evaporative 2004. Use of a water stress index to identify barley genotypes adapted to demand. Aust. J. Agric. Res. 51, 971–978. rainfed and irrigated conditions. Crop Sci. 44, 2127–2137. Morgan, J.M., 1983. Osmoregulation as a selection criterion for drought Robin, S., Pathan, M.S., Courtois, B., Lafitte, R., Carandang, S., Lanceras, S., tolerance in wheat. Aust. J. Agric. Res. 34, 607–614. Amante, M., Nguyen, H.T., Li, Z., 2003. Mapping osmotic adjustment in an Morgan, J.M., Tan, M.K., 1996. Chromosomal location of a wheat osmoregula- advanced back-cross inbred population of rice. Theor. Appl. Genet. 107, tion gene using RFLP analysis. Aust. J. Plant Physiol. 23, 803–806. 1288–1296. Motzo, R., Giunta, F., Deidda, M., 2001. Factors affecting the genotype x Saijo, Y., Hata, S., Kyozuka, J., Shimamoto, K., Izui, K., 2000. Over-expression environment interaction in spring triticale grown in a Mediterranean of a single Ca2+-dependent protein kinase confers both cold and salt/ environment. Euphyitica 121, 317–324. drought tolerance on rice plants. Plant J. 23, 319–327. Nguyen, T.T., Klueva, N., Chamareck, V., Aarti, A., Magpantay, G., Millena, Saini, H.S., Westgate, M.E., 2000. Reproductive development in grain crops A.C., Pathan, M.S., Nguyen, H.T., 2004. Saturation mapping of QTL during drought. Adv. Agron. 68, 59–96. regions and identification of putative candidate genes for drought tolerance Sakuma, Y., Maruyama, K., Osakabe, Y., Qin, F., Seki, M., Shinozaki, K., in rice. Mol. Gen. Genomics 272, 35–46. Yamaguchi-Shinozaki, K., 2006. Functional analysis of an Arabidopsis Ober, E.S., Clark, C.J.A., Le Bloa, M., Royal, A., Jaggard, K.W., Pidgeon, J.D., transcription factor, DREB2A, involved in drought-responsive gene expres- 2004. Assessing the genetic resources to improve drought tolerance in sugar sion. Plant Cell 18, 1292–1309. beet: agronomic traits of diverse genotypes under droughted and irrigated Salvi, S., Tuberosa, R., 2005. To clone or not to clone plant QTLs: present and conditions. Field Crop Res. 90, 213–234. future challenges. Trends Plant Sci. 10, 297–304. Oh, S.J., Song, S.I., Kim, Y.S., Jang, H.J., Kim, S.Y., Kim, M., Kim, Y.-K., Kim, Sanchez, A.C., Subudhi, P.K., Rosenow, D.T., Nguyen, H.T., 2002. Mapping N.Y.-K., Nahm, B.H., Kim, J.-K., 2005. Arabidopsis CBF3/DREB1A and QTLs associated with drought resistance in sorghum (Sorghum bicolor L ABF3 in transgenic rice increased tolerance to abiotic stress without Moench). Plant Mol. Biol. 48, 713–726. stunting growth. Plant Physiol. 138, 341–351. Saranga, Y., Jiang, C.-X., Wright, R.J., Yakir, D., Paterson, A.H., 2004. Genetic Pantuwan, G., Fukai, S., Cooper, M., Rajatasereekul, S., O’Toole, J.C., 2002. dissection of cotton physiological responses to arid conditions and their Yield response of rice (Oryza sativa L) genotypes to different types of inter-relationships with productivity. Plant Cell Environ. 27, 263–277. drought under rainfed lowlands. Part1. Grain yield and yield components. Serraj, R., Hash, C.T., Rizvi, S.M., Sharma, A., Yadav, R.S., Bidinger, F.R., Field Crop Res. 73, 153–168. 2005. Recent advances in marker-assisted selection for drought tolerance in Park, S., Li, J., Pittman, J.K., Berkowitz, G.A., Yang, H., Undurraga, S., Morris, pearl millet. Plant. Prod. Sci. 8, 334–337. J., Hirschi, K.D., Gaxiola, R.A., 2005. Up-regulation of a H+-pyropho- Serraj, R., Sinclair, T.R., 2002. Osmolyte accumulation: can it really increase sphatase (H+-PPase) as a strategy to engineer drought-resistant crop plants. crop yield under drought conditions? Plant Cell Environ. 25, 333–341. Proc. Natl. Acad. Sci. U.S.A. 102, 18830–18835. Sharp, R.E., Poroyko, V., Hejlek, L.G., Spollen, W.G., Springer, G.K., Bohnert, Peleman, J.D., Van der Voort, J.R., 2003. Breeding by design. Trends Plant Sci. H.J., Nguyen, T., 2004. Root growth maintenance during water deficits: 8, 330–334. physiology to functional genomics. J. Exp. Bot. 55, 2343–2351. Pellegrineschi, A., Reynolds, M., Pacheco, M., Brito, R.M., Almeraya, R., Shen, Y.G., Zhang, W.K., He, S.J., Zhang, J.S., Liu, Q., Chen, S.Y., 2003. An Yamaguchi-Shinozaki, K., Hoisington, D., 2004. Stress-induced expression EREBP/AP2-type protein in Triticum aestivum was a DRE-binding tran- in wheat of the Arabidopsis thaliana DREB1A gene delays water stress scription factor induced by cold, dehydration and ABA stress. Theor. Appl. symptoms under greenhouse conditions. Genome 47, 493–500. Genet. 106, 923–930. Pidgeon, J.D., Ober, E.S., Qi, A., Clark, C.J.A., Royal, A., Jaggard, K.W., 2006. Siddique, K.H.M., Tennant, D., Perry, M.W., Belford, R.K., 1990. Water use and Using multi-environment sugar beet variety trails to screen for drought water use efficiency of old and modern wheat cultivars in a Mediterranean- tolerance. Field Crop Res. 95, 268–279. type environment. Aust. J. Agric. Res. 41, 431–447. Piepho, H.-P., 2000. A mixture-model approach to mapping quantitative trait Sinclair, T.R., Muchow, R.C., 2001. System analysis of plant traits to increase loci in barley on the basis of multiple environment data. Genetics 156, grain yield on limited water supplies. Agron. J. 93, 263–270. 2043–2050. Slafer, G.A., Araus, J.L., Royo, C., Del Moral, L.F.G., 2005. Promising eco- Podlich, D.W., Winkler, C.R., Cooper, M., 2004. Mapping as you go: an physiological traits for genetic improvement of cereal yields in Mediterra- effective approach for marker-assisted selection of complex traits. Crop nean environments. Ann. Appl. Biol. 146, 61–70. Sci. 44, 1560–1571. Slafer, G.A., Satorre, E.H., Andrade, H., 1994. Increases in grain yield in bread Rajcan, I., Tollenaar, M., 1999. Source-sink ratio and leaf senescence in maize. wheat from breeding and associated physiological changes. In: Slafer, G.A. I. Dry matter accumulation and partitioning during the grain-filling period. (Ed.), Genetic Improvement of Field Crops. Marcel Dekker, Inc., New Field Crop Res. 90, 245–253. York, pp. 1–67. Ramanjulu, S., Bartels, D., 2002. Drought- and desiccation-induced modulation Slafer, G.A., Whitechurch, E.M., 2001. Manipulating wheat development to of gene expression in plants. Plant Cell Environ. 25, 141–151. improve adaptation and to search for alternative opportunities to increase 14 L. Cattivelli et al. / Field Crops Research 105 (2008) 1–14

yield potential. In: Reynolds, M.P., Ortiz-Monasterio, J.I., McNab, A. a yield determinant in winter wheat under optimal and drought-stressed (Eds.), Application of Physiology on Plant Breeding. CYMMIT, Me´xico, environments. Euphytica 135, 255–263. DF, pp. 160–170. Voltas, J., Lopez-Corcoles, H., Borras, G., 2005. Use of biplot analysis and Soika, R.E., Stolzy, L.H., Fischer, R.A., 1981. Seasonal drought response of factorial regression for the investigation of superior genotypes in multi- selected wheat cultivars. Agron. J. 73, 838–845. environment trials. Eur. J. Agron. 22, 309–324. Steele, K.A., Price, A.H., Shashidhar, H.E., Witcombe, J.R., 2006. Marker- Wade, M.J., 2002. A gene’s eye view of epistasis, selection and speciation. J. assisted selection to introgress rice QTLs controlling root traits into an Evol. Biol. 15, 337–346. Indian upland rice variety. Theor. Appl. Genet. 112, 208–221. Walter, A., Shurr, U., 2005. Dynamics of leaf and root growth: endogenous Syva¨nen, A.C., 2005. Toward genome-wide SNP genotyping. Nat. Genet. 37, control versus environmental impact. Ann. Bot. 95, 891–900. S5–S10. Wang, W., Vinocur, B., Altman, A., 2003. Plant response to drought, salinity Tambussi, E.A., Nogues, S., Ferrio, P., Voltas, J., Araus, J.L., 2005. Does higher and extreme temperatures: toward genetic engineering for stress tolerance. yield potential improve barley performance in Mediterranean conditions? A Planta 218, 1–14. case of study. Field Crop Res. 91, 149–160. Wang, Y., Ying, J., Kuzma, M., Chalifoux, M., Sample, A., McArthur, C., Tanksley, S.D., McCouch, S.R., 1997. Seed banks and molecular maps: Uchacz, T., Sarvas, C., Wan, J., Tennis, D.T., McCourt, P., Huang, Y., 2005. unlocking genetic potential from the wild. Science 277, 1063–1066. Molecular tailoring of farnesylation for plant drought tolerance and yield Tanksley, S., Nelson, J., 1996. Advanced backcross QTL analysis: a method protection. Plant J. 43, 413–424. for the simultaneous discovery and transfer of valuable QTLs from Wollenweber, B., Porter, J.R., Lubberstedt, T., 2005. Need for multidisciplinary unadapted germplasm into elite breeding lines. Theor. Appl. Genet. 92, research towards a second green revolution. Curr. Opin. Plant Biol. 8, 337– 191–203. 341. Teulat, B., This, D., Khairallah, M., Borries, C., Ragot, C., Sourdille, P., Leroy, Xiao, B., Huang, Y., Tang, N., Xiong, L., 2007. Over-expression of a LEA gene P., Monneveux, P., Charrier, A., 1998. Several QTLs involved in osmotic in rice improves drought resistance under the field conditions. Theor. Appl. adjustment trait variation in barley (Hordeum vulgare L). Theor. Appl. Genet. 115, 35–46. Genet. 96, 688–698. Xue, G.P., McIntyre, C.L., Chapman, S., Bower, N.I., Way, H., Reverter, A., Tiwari, B.S., Belenghi, B., Levine, A., 2002. Oxidative stress increased Clarke, B., Shorter, R., 2006. Differential gene expression of wheat progeny respiration and generation of reactive oxygen species, resulting in ATP with contrasting levels of transpiration efficiency. Plant Mol. Biol. 61, 863– depletion, opening of mitochondrial permeability transition, and pro- 881. grammed cell death. Proc. Natl. Acad. Sci. U.S.A. 128, 1271–1281. Yadav, O.P., Bhatnagar, S.K., 2001. Evaluation of indices for identification of Tollenaar, M., Lee, E.A., 2002. Yield stability and stress tolerance in maize. pearl millet cultivars adapted to stress and non-stress conditions. Field Crop Field Crop Res. 75, 161–169. Res. 70, 201–208. Tollenar, M., Wu, J., 1999. Yield in temperate maize is attributable to greater Yan, J., He, C., Wang, J., Mao, Z., Holaday, S.A., Allen, R.D., Zhang, H., 2004. stress tolerance. Crop Sci. 39, 1604–1897. Overexpression of the Arabidopsis 14-3-3 protein GF14 lambda in cotton Tondelli, A., Francia, E., Barabaschi, D., Aprile, A., Skinner, J.S., Stockinger, leads to a ‘‘stay-green’’ phenotype and improves stress tolerance under E.J., Stanca, A.M., Pecchioni, N., 2006. Mapping regulatory genes as moderate drought conditions. Plant Cell Physiol. 45, 1007–1014. candidates for cold and drought stress tolerance in barley. Theor. Appl. Yin, X.Y.,Struik, P.C., Kropff, M.J., 2004. Role of crop physiology in predicting Genet. 112, 445–454. gene-to-phenotype relationships. Trends Plant Sci. 9, 426–432. Trethowan, R.M., van Ginkel, M., Rajaram, S., 2002. Progress in breeding Zhang, J., Zheng, H.G., Aarti, A., Pantuwan, G., Nguyen, T.T., Tripathy, J.N., wheat for yield and adaptation in global drought affected environments. Sarial, A.K., Robin, S., Babu, R.C., Nguyen, B.D., Sarkarung, S., Blum, A., Crop Sci. 42, 1441–1446. Nguyen, H.T., 2001. Locating genomic regions associated with components Tuberosa, R., Salvi, S., 2006. Genomics-based approaches to improve drought of drought resistance in rice: comparative mapping within and across tolerance of crops. Trends Plant Sci. 11, 405–412. species. Theor. Appl. Genet. 103, 19–29. Turner, N.C., Abbo, S., Berger, J.D., Chaturvedi, S.K, French, R.J., Ludwig, C., Zheng, B.S., Yang, L., Zhang, W.P., Mao, C.Z., Wu, Y.R., Yi, K.K., Liu, F.Y., Mannur, D.M., Singh, S.J., Yadava, H.S., 2007. Osmotic adjustment in Wu, P., 2003. Mapping QTLs and candidate genes for rice root traits under chickpea (Cicer arietinum L) results in no yield benefit under terminal different water-supply conditions and comparative analysis across three drought. J. Exp. Bot. 58, 187–194. populations. Theor. Appl. Genet. 107, 1505–1515. Tyerman, S.D., Niemietz, C.M., Bramley, H., 2002. Plant aquaporins: multi- Zhu, B.C., Su, J., Chan, M.C., Verma, D.P.S., Fan, Y.L., Wu, R., 1998. Over- functional water and solute channels with expanding roles. Plant Cell expression of a d-pyrroline-5-carboxylate synthetase gene and analysis of Environ. 25, 173–194. tolerance to water-stress and salt-stress in transgenic rice. Plant Sci. 139, van Ginkel, M., Calhoun, D.S., Gebeyehu, G., Miranda, A., Tian-you, C., Pargas 41–48. Lara, R., Trethowan, R.M., Sayre, K., Crossa, J., Rajaram, S., 1998. Plant Zinselmeier, C., Jeong, B.R., Boyer, J.S., 1999. Starch and the control of kernel traits related to yield of wheat in early, late, or continuous drought number in maize at low water potentials. Plant Physiol. 121, 25–35. conditions. Euphytica 100, 109–121. Zinselmeier, C., Westgate, M.E., Schussler, J.R., Jones, R.J., 1995. Low water Verma, V., Foulkes, M.J., Worland, A.J., Sylvester-Bradley, R., Caligari, P.D.S., potential disrupts carbohydrate metabolism in maize (Zea mais L) ovaries. Snape, J.W., 2004. Mapping quantitative trait loci for flag leaf senescence as Plant Physiol. 107, 385–391.