TY - JOUR
AU1 - Centritto, Mauro
AU2 - Lauteri, Marco
AU3 - Monteverdi, Maria Cristina
AU4 - Serraj, Rachid
AB - Abstract Genotypic variations in leaf gas exchange and yield were analysed in five upland-adapted and three lowland rice cultivars subjected to a differential soil moisture gradient, varying from well-watered to severely water-stressed conditions. A reduction in the amount of water applied resulted in a significant decrease in leaf gas exchange and, subsequently, in above-ground dry mass and grain yield, that varied among genotypes and distance from the line source. The comparison between the variable J and the Δ values in recently synthesized sugars methods, yielded congruent estimations of mesophyll conductance (gm), confirming the reliability of these two techniques. Our data demonstrate that gm is a major determinant of photosynthesis (A), because rice genotypes with inherently higher gm were capable of keeping higher A in stressed conditions. Furthermore, A, gs, and gm of water-stressed genotypes rapidly recovered to the well-watered values upon the relief of water stress, indicating that drought did not cause any lasting metabolic limitation to photosynthesis. The comparisons between the A/Ci and corresponding A/Cc curves, measured in the genotypes that showed intrinsically higher and lower instantaneous A, confirmed this finding. Moreover, the effect of drought stress on grain yield was correlated with the effects on both A and total diffusional limitations to photosynthesis. Overall, these data indicate that genotypes which showed higher photosynthesis and conductances were also generally more productive across the entire soil moisture gradient. The analysis of Δ revealed a substantial variation of water use efficiency among the genotypes, both on the long-term (leaf pellet analysis) and short-term scale (leaf soluble sugars analysis). Carbon isotope discrimination, drought stress, mesophyll conductance, Oryza sativa, photosynthesis, rice, water-use efficiency Introduction Drought stress is the largest constraint to rice production, affecting 19 million ha of upland rice and over 14 million ha of rainfed lowland rice (Pandey et al., 2007). Current and future climate change is expected to increase the threat of water shortage (Intergovernmental Panel on Climate Change, 2007, http://www.ipcc.ch/), further worsening the water crisis in the rainfed lands (Wassmann et al., 2009a). As a result, the identification and selection of rice germplasm combining enhanced water use efficiency (WUE) and drought resistance, that is, the capacity to produce high yield under water-limited environments, have become priority target traits for genetic improvement (Wassmann et al., 2009b). A better understanding of the mechanisms by which water deficits affect leaf gas exchange, photosynthesis, and plant growth is necessary to improve the drought phenotyping methods and the physiological dissection of potential drought-resistant genotypes in the field (Serraj et al., 2008). At the whole plant level, water deficit is an important environmental constraint influencing all the physiological processes involved in plant growth and development (Sadras and Milroy, 1996). These changes are part of a cascade of responses to drought primarily affecting tissue expansion and gas exchange mechanisms (Serraj et al., 1999). Previous studies have shown that drought significantly affects leaf net CO2 assimilation rate (A), stomatal conductance (gs), and transpiration rate (E) (Chaves, 1991; Lawlor and Cornic, 2002). Stomatal regulation in response to soil drying has also been found to be triggered by root–shoot chemical and/or hydraulic signalling (Tardieu and Davies, 1993) as a key adaptation strategy to avoid tissue dehydration under drought. Most of the carbon stored in the mature rice grains originates from CO2 assimilation during the grain-filling period, with the flag leaf as the most photosynthetically active (Yoshida, 1981; Murchie et al., 1999), factors that lower the photosynthesis rate of the flag leaf during this period could potentially limit grain yield (Dingkuhn et al., 1989). Genotypic variation in rice has long been reported for WUE, measured either as intrinsic WUE, i.e. the ratio of net photosynthesis rate to stomatal conductance (A/gs) or by carbon isotope discrimination (Δ) (Samejima, 1985; Dingkuhn et al., 1991). Dingkuhn et al. (1991) found that Δ was highly correlated with in situ measurements of A/gs. Varietal differences in stomatal response to decreasing leaf water potential have also been reported (Dingkuhn et al., 1989). Comparing WUE values of tropical japonica with those of indica cultivars based on leaf gas exchange rates (A/E) under irrigated conditions, Peng et al. (1998) found that indica cultivars had generally higher E than tropical japonica lines, and the A/E ratio was 25–30% higher for the tropical japonica than for indica. Moreover, lower Δ values in the tropical japonica compared to indica confirmed the observed differences in A/E. Impa et al. (2005) confirmed the relationship between gravimetrically determined WUE and Δ among rice germplasm, indicating that WUE is genetically variable in rice and hence can be exploited through breeding. More recently, Zhao et al. (2008) tried to identify QTLs conditioning A, gs, Ci (intercellular CO2 concentration), and E to facilitate marker-assisted selection for the genetic improvement of rice yield under drought conditions. They found that QTLs affecting these physiological traits tended to cluster in the same genomic regions, which would suggest a common genetic basis for A, gs, Ci, and E, and that there was a high genomic correspondence among QTLs affecting these physiological traits and grain yield under drought-stress conditions. However, there was a low or no correlation between leaf gas exchange parameters and yield (Zhao et al., 2008). Similarly, a number of reports have recently claimed improvements of drought resistance through genetic engineering, including several studies showing an effect of improved photosynthetic capacity and/or WUE under stress conditions (see Hervé and Serraj, 2009, for a review). An expression profiling experiment of rice cultivars differing in their drought resistance levels has recently reported that drought stress strongly down-regulated photosynthesis genes in both sensitive and tolerant cultivars (Degenkolbe et al., 2009). The down-regulation of photosynthetic genes in the tolerant cultivars was interpreted as an adaptive response to prevent photodamage during times of reduced CO2 availability in the mesophyll when stomata are closed due to water shortage. Overall, the relationships among photosynthesis, stomatal conductance, transpiration efficiency, Δ, and yield under water deficit conditions are not yet well understood, and there is still confusion between photosynthetic rates and photosynthetic capacity. In addition, the importance of total diffusive limitations (gt which is related to gs and gm, mesophyll conductance to CO2) to A and, in turn, Cc (chloroplastic CO2 concentration) is still neglected in most drought-phenotyping and genetic-engineering studies. Plants possess several mechanisms to regulate water use with water availability. Regulation of diffusive conductances is known to affect WUE by modulating both transpiration and photosynthesis rates. Regulation of photosynthetic rates also strongly affects WUE. Carbon isotope composition (δ13C) recorded in tissues of C3 plants has been demonstrated to be a useful tool for assessing WUE (for a review see Farquhar et al., 1989; Brugnoli and Farquhar, 2000). The fractionation of carbon isotopes during photosynthesis depends on biochemical and physical phenomena, mainly associated with CO2 diffusion and carboxylation reactions. In fact, the depletion of the heavy isotope 13C in plant tissues, with respect to its abundance in the atmospheric CO2, is directly related to the ratio of Cc to atmospheric CO2 concentration (Ca); this ratio represents the equilibrium between the availability and the requirement of CO2 at the leaf level, that is the set point for gas exchange activity (Ehleringer, 1993). Because Cc/Ca is negatively related to WUE, a mechanistic negative relationship between Δ and long-term WUE is generally observed in C3 plants (Farquhar et al., 1982). According to this theory, carbon isotope discrimination analysis allows an assimilation weighted estimation of both Cc/Ca and WUE integrated over different time scales, depending on which tissues or metabolites are analysed. While the analysis of samples representative of the entire dry mass furnishes an evaluation of WUE integrated over the whole plant life, information from a few hours to days is associated with the isotopic analysis of newly fixed carbon in metabolites such as leaf soluble sugars or starch (Brugnoli et al., 1988). Mesophyll resistance integrates an array of physical (Harley et al., 1992; Loreto et al., 1994; Warren, 2006) and likely biochemical (Gillon and Yakir, 2000; Uehlein et al., 2003; Flexas et al., 2008; Miyazawa et al., 2008) internal resistances that affect the transfer capacity of CO2 between the leaf internal airspace and the site of carboxylation in the chloroplast. Mesophyll conductance is a fundamental property of leaves that influence photosynthetic capacity (Centritto et al., 2003; Flexas et al., 2008; Warren, 2008). Consequently, changes in gm influence the response of photosynthesis to drought. A recent literature review (Flexas et al., 2008) shows that gm generally decreases under water stress. Decreased gm in response to water stress was also found by Scartazza et al. (1998) in rice plants grown in pots. However, no data are available for rice subjected to water stress in field conditions, and it is unclear whether upland- and lowland-adapted genotypes can present a different response of gm to water stress. The objectives of this study were (i) to identify the possible causes of the different sensitivity of both photosynthesis and WUE to different levels and intensities of water deficit in rice varieties, and (ii) to analyse the relationship between intrinsic transpiration efficiency (A/gs) and long-term WUE (as assessed by Δ analysis). This information has important implications for developing drought-screening tools, and to provide new insights into the biological effects of drought stress at both the cellular and whole-plant levels. A third objective of this study was to investigate whether gm is involved in limiting photosynthesis of rice genotypes exposed to different drought-stress intensities. Various methods have been developed to assess gm (Harley et al., 1992; Warren, 2006). However, each technique relies on assumptions and on a number of parameters to be measured which affect both precision of the estimations and the feasibility of observations in certain experimental conditions (Harley et al., 1992; Warren, 2006; Rodeghiero et al., 2007). Thus, the fourth objective of this study was to investigate the relationship between two different methods to estimate gm, namely the variable J method (Loreto et al., 1992, 1994) and carbon isotope discrimination in recently synthesized sugars (Lauteri et al., 1997), in order to determine any possible differential effect of the methods applied on the estimation of gm and, consequently, to increase the reliability of this physiological parameter. Materials and methods Plant material and growth conditions Experiments were designed to characterize the physiological and yield responses of eight contrasting rice genotypes: five upland-adapted cultivars (Apo, IR71525-19-1-1, IR55419-04, Moroberekan, and Vandana) and three lowland cultivars (IR64, DK98, and PSBRc80). Experiments were conducted at the IRRI upland experimental farm in Los Bañòs (14o11′ N 121o15′ E, 21 m above sea level), Philippines. The soil is an Andaqueptic Haplaquol with pH of 6.5. The experiments were sown in the dry season of 2006 that extends from January to May. While rainfall is minimal during this period, temperatures and VPD increase significantly. Plants were directly sown in dry soil in rows 25 cm apart at a rate of a rate of 80 kg ha−1. The plants were thinned to a uniform stand of approximately 10 cm between hills within rows. Each variety was planted in six replicates of individual plots of four rows 10 m long. Nitrogen fertilizer was applied in two splits for a total application of 90 kg ha−1. Basal P and K were applied at 30 kg ha−1. Weeds were controlled by a combination of chemical and manual methods. Insects (particularly stem borers) were controlled chemically. Prior to and after the line source treatment period, the entire experiment was irrigated by a conventional sprinkler system (12.2×12.2 m) to maintain the soil water content constantly close to field capacity. Irrigation by overhead sprinkler was done with an interval of 2–3 d depending on soil moisture and evapotranspiration, starting from sowing for 3 h per irrigation event. Line source treatment was initiated 50 d afters sowing (das), which corresponded to the period between panicle initiation and anthesis stages and was applied until the beginning of physiological maturity. Flowering, i.e. days to 50% flowering occurred around 53 das in Vandana, 65 das in IR55419-04 and DK98, 66 das in IR71525-19-1-1, 67 das in Moroberekan, 70 das for Apo, 74 das for IR64, and 79 das for PSBRc80. The line source sprinkler system, previously described (Cruz and O'Toole, 1984) made it possible to generate a differential gradient of soil moisture varying from well-watered to extremely dry. The system consists of a single line of sprinklers which are 4.5 m apart and located at the centre of the plot and perpendicular to the planting rows. The sprinklers produced linearly decreasing amounts of applied water with distance from the line (Fig. 1a). The sprinklers were precalibrated to estimate the amount of water delivered at a given pressure per unit of time, and uniform pressure of 28 psi was maintained for each sprinkler. To minimize field variation and to ensure uniform distribution of irrigation water, the sprinklers were operated at low wind speeds often at predawn. Catch cans were placed at 2 m intervals across the line source gradient at canopy height to measure the actual water applied (Fig. 1a). Fig. 1. View largeDownload slide Amount of water applied as a function of the distance from a line source sprinkler system (r2=0.992, P <0.001) (a), effect of different amounts of water applied on (b) grain yield of Apo (A, r2=0.762, P <0.05), DK98 (D, r2=0.909, P <0.01), IR55419-04 (I5, r2=0.937, P <0.01), IR64 (I6, r2=0.968, P <0.01), IR71525-19-1-1 (I7, r2=0.676, P <0.09), PSBRc80 (P, r2=0.863, P <0.05), and Vandana (V, r2=0.698, P <0.08); (c) above-ground dry mass of Apo (r2=0.001), DK98 (r2=0.980, P <0.01), IR55419-04 (r2=0.911, P <0.01), IR64 (r2=0.711, P <0.07), IR71525-19-1-1 (r2=0.645, P <0.1), PSBRc80 (r2=0.682, P <0.09), Vandana (r2=0.681, P <0.09) and Moroberekan (M, r2=0.491); and (d) harvest index (HI) of Apo (r2=0.841, P <0.05), DK98 (r2=0.900, P <0.05), IR55419-04 (r2=0.880, P <0.05), IR64 (r2=0.911, P <0.01), IR71525-19-1-1 (r2=0.626, P <0.1), PSBRc80 (r2=0.771, P <0.05), and Vandana (r2=0.800, P <0.1). Fig. 1. View largeDownload slide Amount of water applied as a function of the distance from a line source sprinkler system (r2=0.992, P <0.001) (a), effect of different amounts of water applied on (b) grain yield of Apo (A, r2=0.762, P <0.05), DK98 (D, r2=0.909, P <0.01), IR55419-04 (I5, r2=0.937, P <0.01), IR64 (I6, r2=0.968, P <0.01), IR71525-19-1-1 (I7, r2=0.676, P <0.09), PSBRc80 (P, r2=0.863, P <0.05), and Vandana (V, r2=0.698, P <0.08); (c) above-ground dry mass of Apo (r2=0.001), DK98 (r2=0.980, P <0.01), IR55419-04 (r2=0.911, P <0.01), IR64 (r2=0.711, P <0.07), IR71525-19-1-1 (r2=0.645, P <0.1), PSBRc80 (r2=0.682, P <0.09), Vandana (r2=0.681, P <0.09) and Moroberekan (M, r2=0.491); and (d) harvest index (HI) of Apo (r2=0.841, P <0.05), DK98 (r2=0.900, P <0.05), IR55419-04 (r2=0.880, P <0.05), IR64 (r2=0.911, P <0.01), IR71525-19-1-1 (r2=0.626, P <0.1), PSBRc80 (r2=0.771, P <0.05), and Vandana (r2=0.800, P <0.1). Leaf gas exchange measurements Leaf gas exchange and fluorescence were simultaneously measured with a LI-6400-40 leaf chamber fluorometer (Li-Cor Inc., Nebraska, USA) on flag leaves enclosed into a temperature, light, and humidity-controlled cuvette, on plants grown near the line (well-watered treatment), at 7 m (intermediate stress) and 9 m distance (stress) from the sprinkler. The measurements were made in situ between 11.00 h and 15.00 h, in saturating PPFD (1400 μmol m−2 s−1), with relative humidity ranging between 45–55%, and a leaf temperature of 30 °C. Instantaneous measurements of steady-state A, gs, Ci, and (i.e. the quantum yield of PSII in the light) were made on 6–14 plants per treatment. Measurements of Rd (dark respiration) were also made at ambient CO2 concentration in the dark on the same leaves. Intrinsic WUE was calculated as the A to gs ratio. A/Ci curves were performed over a range of CO2 concentrations between ∼40 and 2500 μmol mol−1 on 3–5 plants per treatment. To remove the effect of stomatal limitation on A (and to estimate photosynthetic capacity at high gs), leaves were first preconditioned at low [CO2] (50 μmol mol−1) for up to 1 h to force stomatal opening as described by Centritto et al. (2003). Then [CO2] was progressively increased up to 2500 μmol mol−1. Values for the photosynthetic parameters Amax (the net CO2 assimilation rate under conditions of PPFD and CO2 saturation), Vcmax (RuBP-saturated rate of Rubisco), and PPFD-saturated Jmax (electron transport rate) were estimated by fitting the mechanistic model of CO2 assimilation proposed by Farquhar et al. (1980) to individual A/Ci response data using the method developed by de Pury and Farquhar (1997). The fitting model was run using the in vivo Rubisco kinetics parameters (i.e. Ko, Kc, and their activation energy) measured by Bernacchi et al. (2001), with the exception of Γ* which was calculated by using the Rubisco specific factor estimated for an annual herb by Galmés et al. (2005). Fitting the model involved optimizing the parameter values by adjusting them so as to minimize the sums of residuals between the observed and modelled assimilation values over a range of Ci. Carbon isotope discrimination The same flag leaves previously used for in situ gas exchange measurements were collected in the evening, oven-dried at 70 °C to constant weight, and stored for subsequent soluble sugar extraction, according to Brugnoli et al. (1988). Leaves were ground and boiled for 2 min in water. After centrifugation (6 min at 5000 g) the supernatant was sequentially mixed with cationic (Dowex-50) and anionic (Dowex-1) exchange resins. Soluble sugar samples were freeze-dried and then used for mass-spectrometer determinations. Additional IRMS analyses were performed on the pellet samples remaining after the soluble sugars extraction. Samples of about 1 mg were used for δ13C determination. Sample δ13C was calculated according to the definitions described by Farquhar et al. (1989). Measurements were performed using a continuous-flow triple-collector isotope ratio mass spectrometer (Isoprime, GV, Manchester, UK). Solid samples were quantitatively combusted in an elemental analyser (Model NA 1500, Carlo Erba, Milan, Italy) and CO2 was transferred in helium flow to the mass spectrometer. Carbon isotope discrimination (Δ) of solid samples was calculated according to Farquhar et al. (1989). Carbon isotope composition of air CO2 (δair) was assumed –8.0‰. Estimation of mesophyll conductance Mesophyll conductance to CO2 diffusion, the inverse of the total resistance encountered by CO2 across the leaf mesophyll, was calculated by using both the variable J method and the carbon isotope discrimination method in recently synthesized sugars. The calculation of gm using the variable J method is based on the simultaneous measurements of gas-exchange and fluorescence parameters as described in detail by Harley et al. (1992) and Loreto et al. (1992). The variable J method is quite sensitive to the estimation of Γ* (CO2 compensation point to photorespiration) and Rd. While the latter was measured, Γ* used in the gas exchange algorithm was calculated by using the Rubisco specific factor estimated for an annual herb by Galmés et al. (2005). Then, total conductance (gt) was calculated as: gt=gsgm/(gs+gm). The actual [CO2] at the chloroplast site (Cc) was then calculated from the mesophyll conductance value as shown elsewhere (Loreto et al., 1994), and this allowed the A/Cc curves to be estimated from the corresponding A/Ci curves. Cc was then used as a basis for the recalculation of Vcmax and Jmax by fitting the mechanistic model of CO2 assimilation proposed by Farquhar et al. (1980) to individual A/Cc response data as described previously. The method described by Lauteri et al. (1997) was used to calculate gm by using Δ of recently synthesized sugars. Discrepancies eventually existing between Δ observed on leaf soluble carbohydrates (Δobs) and Δ expected on the basis of gas-exchange measurements (Δexp), allow an estimation of gm according to the approach of Evans et al. (1986) as modified by Lloyd et al. (1992):  (1)where b is the discrimination associated with carboxylation reactions and was taken as 27.5‰, bs is the fractionation occurring when CO2 enters the solutions (1.1‰ at 25 °C), a1 is the fractionation during diffusion in water (0.7‰), f is the fractionation associated with photorespiration, and pa is the CO2 partial pressure in free air. Two gm calculations were performed and compared, since two values were taken into account for the fractionation factor f: 7‰, according to Rooney (1988) in Lloyd et al. (1992) and 0, according to von Caemmerer and Evans (1991) and Scartazza et al. (1998). Plant water status measurements and harvesting Leaf water potential (LWP) measurements were made directly in the field, on the youngest mature leaf of the main tiller on each plant. The leaf was sampled and transported quickly to a dark, moist bag for measurement with a Scholander type pressure chamber (Soil Moisture Equipment Corp., Santa Barbara, CA). A leaf from the second tiller on the same plant was harvested for relative water content (RWC) measurement. The leaves were directly placed in a preweighed plastic centrifuge tube and placed on ice in a cooler. The tubes were taken to the laboratory and weighed immediately to determine the initial fresh weight. The tubes were then filled with distilled water, recapped, and placed in a dark 4 oC room overnight. The next morning, leaves were blotted with paper towels and were weighed immediately to determine the fully turgid weight. Leaves were then dried at 70 oC to constant weight for dry mass determination. Plants were harvested at physiological maturity by sampling 1 m of the two central rows of each subplot across the line source gradient. Each plot was subdivided into five subplots along the water gradient (Fig. 1a). Data collected included biomass, plant height, anthesis date, tiller number, straw production, grain yield, and harvest index (HI). Statistical analysis Analyses of variance were conducted using SAS to generate least squares means for each entry. Leaf gas exchange data were tested using a simple factorial ANOVA (three-way maximum interactions), and where, appropriate, the treatment means of leaf properties and gas exchange parameters were compared using Tukey's post-hoc test. Results and discussion Growth responses to soil moisture gradient The use of the line source sprinkler irrigation system (Cruz and O'Toole, 1984) in this study resulted in a differential gradient of soil moisture varying from well-watered to extremely dry. The sprinklers produced linearly decreasing amounts of water with distance from the line source (Fig. 1a). The average total water applied nearest the line was about 390 mm, and the farthest from the line was about 80 mm. The reduction of the amount of water applied resulted in a significant decrease in grain yields (Fig. 1b) and above-ground dry mass (Fig. 1c) that varied among genotypes and distance from the line source. Except for Moroberekan, which failed to set and produce grains possibly because of disease incidence and extreme sensitivity to heat, the upland-adapted genotypes produced grain yields that were comparable with those reported in the literature for upland rice in similar dry-season field trials (Lafitte et al., 2002; Atlin et al., 2006; Botwright Acuňa et al., 2008). However, even under the highest level of irrigation, the grain yield of these upland rice cultivars and breeding lines was much lower than can be expected from aerobic-adapted rice cultivars (Atlin et al., 2006; Bouman et al., 2006). The upland-adapted, short-duration genotype Vandana produced the highest grain yield across the entire soil moisture gradient, followed by the upland genotypes IR71525-19-1-1, Apo, and IR55419-04. The three lowland-adapted cultivars (IR64, DK98, and PSBRc80) had generally lower yields across all water gradients, and their yield under drought was much more severely affected than the five upland-adapted cultivars. The average yield decrease in the driest subplot compared with the wettest varied from 50% in Vandana to over 90% in the three lowland cultivars. These results confirm previous reports on the response of rice genotypes to soil moisture gradients and stress during the reproductive stage (Aragon and De Datta, 1982; Cruz and O'Toole, 1984). In general, above-ground biomass was also inversely correlated with distance from the line, but the effect of water deficit on biomass was much lower than the effect on yield. This was mainly due to the fact that the line source irrigation treatment was initiated during the flowering stage, after vegetative plant growth had already been achieved under well-watered conditions. The upland-adapted cultivars Apo and Moroberekan had the highest and the lowest production of above-ground biomass across the entire soil moisture gradient, respectively, but overall there was no significant difference between upland- and lowland-adapted genotypes (Fig. 1c). The impact of timing and severity of drought stress during crop development of rice on yield components are well established. Drought during the flowering stage causes spikelet sterility, and terminal drought mostly affects grain filling (Ekanayake et al., 1989; Lilley and Fukai, 1994; Kamoshita et al., 2004). Accordingly, as the line-source water deficit treatments were initiated around the flowering stage, the analysis of correlation among yield components and growth parameters showed that the most important parameters affecting grain yield in this experiment were related to grain setting and spikelet sterility (data not shown), which substantially affected the harvest index (HI) (Fig. 1d). In general, HI decreased with distance from the line. It is noteworthy that the early-maturing genotype Vandana showed the highest HI across the entire soil moisture gradient, since, relatively, its early development allowed it to escape the effects of water deficits on grain formation. Photosynthesis limitations in response to drought In general, there was no systematic difference in photosynthesis (Fig. 2a), gs (Fig. 2b), and gm (Fig. 2c) between upland- and lowland-adapted genotypes in both well-watered and water-stressed conditions, although the upland-adapted cultivars IR71525-19-1-1 and Moroberekan showed significantly higher values. There were also no significant, lasting differences in A, gs, and gm between control plants and plants grown at a 7 m distance from the sprinkler (intermediate stress), despite these plants received on average 60% water supplied to the control. On average A, gs, and gm were reduced by about 11%, 17%, and 20%, respectively, in the intermediate stress plants. By contrast, all these gas exchange parameters were similarly, significantly inhibited in the water-stressed plants which received, on average, 40% of the water supplied to the control. In water-stressed plants, A, gs, and gm were reduced by about 62%, 61%, and 76%, respectively. A, gs, and gm were consistently more inhibited (relative to controls) in Apo, IR64, and PSBRc80, whereas in IR71525-19-1-1 and Moroberekan these gas exchange parameters remained consistently higher than in all other water-stressed genotypes. There was no consistent differences in intrinsic transpiration efficiency (A/gs) among genotypes and in response to water deficit (Fig. 2d). Fig. 2. View largeDownload slide Measurements of (a) photosynthesis (A), (b) stomatal conductance (gs), (c) mesophyll conductance (gm), and (d) intrinsic transpiration efficiency (A/gs) in the flag leaves of eight different rice genotypes (i.e. Apo=Apo, DK=DK98, IR5=IR55419–04, IR6=IR64, IR7=IR71525-19-1-1, Mor=Moroberekan, PSB=PSBRc80, Van=Vandana). The measurements were made on the flag leaf in saturating PPFD (1400 μmol m−2 s−1), with relative humidity ranging between 45–55%, and leaf temperature of 30 °C. Data are means of 6–14 plants per treatment ±1 SEM; the letters a–g indicate significant differences at P <0.05. The letters C, I, and S indicate the distance of the plants from the sprinkler: C, near distance (well–watered), I, intermediate distance (intermediate stress), S, far distance (stress); whereas SI indicates plants at the far end from the sprinkler measured after water stress relief caused by a night-time rainfall event followed by an early morning irrigation. Fig. 2. View largeDownload slide Measurements of (a) photosynthesis (A), (b) stomatal conductance (gs), (c) mesophyll conductance (gm), and (d) intrinsic transpiration efficiency (A/gs) in the flag leaves of eight different rice genotypes (i.e. Apo=Apo, DK=DK98, IR5=IR55419–04, IR6=IR64, IR7=IR71525-19-1-1, Mor=Moroberekan, PSB=PSBRc80, Van=Vandana). The measurements were made on the flag leaf in saturating PPFD (1400 μmol m−2 s−1), with relative humidity ranging between 45–55%, and leaf temperature of 30 °C. Data are means of 6–14 plants per treatment ±1 SEM; the letters a–g indicate significant differences at P <0.05. The letters C, I, and S indicate the distance of the plants from the sprinkler: C, near distance (well–watered), I, intermediate distance (intermediate stress), S, far distance (stress); whereas SI indicates plants at the far end from the sprinkler measured after water stress relief caused by a night-time rainfall event followed by an early morning irrigation. Stomatal and mesophyll resistances progressively reduce the CO2 concentration reaching the chloroplasts. There was a significant correlation, a typical hyperbolic relationship, between photosynthesis and stomatal conductance after pooling together controls and water-stressed leaves (Fig. 3a), showing that A was limited by declining gs (Galmés et al., 2007; Thompson et al., 2007), although a concomitant gs adjustment to the drought-induced inhibition of photosynthetic rate cannot be ruled out (Wong et al., 1979). A similar hyperbolic relationship was also found between A and gm (Fig. 3b). Moreover, because gm was linearly related to gs (Fig. 3c), a slightly better correlation between A and total CO2 diffusional conductances was found when taking into account both stomatal and mesophyll conductances (Fig. 3d). In general, this is also a clear indication that the chloroplast CO2 concentration, as set by the combination of stomatal and mesophyll resistances to CO2 diffusion, is the main limitation of photosynthesis for well-watered rice genotypes as well as for genotypes sensitive to, and strongly affected by water deficit, as found for plants with different growth form and leaf structure and across different biomes (Loreto et al., 2003; Niinemets et al., 2005; Shi et al., 2006; Galmés et al., 2007; Flexas et al., 2008). Fig. 3. View largeDownload slide Relationships between (a) photosynthesis (A) and stomatal conductance (gs) (r2=0.918, P <0.001), (b) A and mesophyll conductance (gm) (r2=0.890, P <0.001), (c) gm and gs (r2=0.826, P <0.001), and (d) between A and total conductance (gt) (r2=0.923, P <0.001). The letters C, I, and S indicate the distance of the plants from the sprinkler, and SI indicates plants at the far end from the sprinkler measured after having being irrigated in the early morning, as shown in Fig. 2. In figures a, b and d a rectangular hyperbolic function was fitted. Fig. 3. View largeDownload slide Relationships between (a) photosynthesis (A) and stomatal conductance (gs) (r2=0.918, P <0.001), (b) A and mesophyll conductance (gm) (r2=0.890, P <0.001), (c) gm and gs (r2=0.826, P <0.001), and (d) between A and total conductance (gt) (r2=0.923, P <0.001). The letters C, I, and S indicate the distance of the plants from the sprinkler, and SI indicates plants at the far end from the sprinkler measured after having being irrigated in the early morning, as shown in Fig. 2. In figures a, b and d a rectangular hyperbolic function was fitted. The comparison of photosynthesis and conductance responses to water stress indicated that the inhibition of photosynthesis could be fully explained by that of the diffusional conductances. This suggests that there were probably no lasting metabolic limitations to photosynthesis in response to the different levels and intensities of drought imposed in the genotypes tested, and that the stress effects may be reversed if the conductance to CO2 diffusion is restored (Cornic, 2000; Lawlor and Cornic, 2002; Centritto et al., 2003; Brilli et al., 2007). Accordingly, when gas exchanges were remeasured after an night-time rainfall event followed by an early morning irrigation, A of water-stressed genotypes recovered to well-watered rates (Fig. 2a), and this was mirrored by a parallel increase in both gs (Fig. 2b) and gm (Fig. 2c) that also recovered to well-watered values in all genotypes. To our knowledge, this is one of the very few published studies that show that a reduction in mesophyll conductance brought about by environmental stress can be reversed in less than 24 h and that this, in turn, may influence photosynthesis (see Flexas et al., 2008; Warren, 2008, for reviews). To assess further whether drought stress affected mesophyll metabolism, the in vivo multi-enzyme kinetic properties of photosynthesis of the upland-adapted cultivar IR71525-19-1-1, which had intrinsically higher photosynthesis, and the lowland-adapted genotype IR64, which showed intrinsically lower photosynthesis (Fig. 2) were compared. However, after preconditioning leaves at low [CO2] (50 μmol mol−1) to force stomatal opening (Centritto et al., 2003), no significant differences were found in the photosynthetic capacity between IR71525-19-1-1 and IR64 in the well-watered (Fig. 4a) and in the different water-stressed conditions (Fig. 4c, e). The A/Ci curves were not statistically different in both the initial slope (which is related to the activity of the chloroplastic soluble proteins, i.e. Vcmax) and in the saturating portion of the curve at high [CO2] (which is related to both the activity of the thylakoid proteins, i.e. Jmax, and Amax) in control plants and in plants grown at the intermediate distance from the sprinkler (Table 1), indicating that moderate water stress did not affect photosynthetic capacity in either IR71525-19-1-1 or IR64. On the contrary, the shape of the A/Ci curves were dramatically affected by severe water deficit (Fig. 4e), as Amax, but especially Vcmax and J were much lower in water-stressed plants than in control plants (Table 1). This is generally interpreted as an indication that the photosynthetic capacity of water-stressed plants was strongly impaired by biochemical damage. Table 1. Best-fit estimates of Amax (μmol m−2 s−1, maximum photosynthetic rate at saturating PPFD per unit leaf area), and of photosynthetic parameters obtained by fitting the Farquhar et al. (1980) model of leaf photosynthesis to the individual A/Ci and A/Cc response curves shown in Fig. 4   IR 64   IR55419-04     C  I  S  C  I  S  Amax  37.95±2.72 b  37.09±2.23 b  21.23±3.08 a  38.30±1.15 b  38.22±2.11 b  22.57±3.00 a  Jmax (Ci)  175.79±16.59 b  165.29±13.46 b  99.37±16.18 a  173.15±7.17 b  176.82±12.04 b  106.40±15.53 a  Jmax (Cc)  190.07±21.31  178.45±12.34  n.a.  184.42±6.72  187.87±13.91  n.a.  Vcmax(Ci)  108.91±9.26 c  92.14±8.08 bc  38.12±8.66 aA  96.18±13.27 bc  89.21±2.17 bA  39.16±2.16 aA  Vcmax(Cc)  126.61±16.67 b  107.85±8.52 b  83.89±9.39 aB  103.63±9.70 b  115.54±6.96 bB  84.38±5.27 aB  Jmax/Vcmax(Ci)  1.81±0.14 a  2.16±0.23 a  2.80±0.29 b  1.88±0.19 a  1.99±0.18 a  2.67±0.17 b  Jmax/Vcmax(Cc)  1.67±0.09  2.03±0.37  n.a.  1.81±0.14  1.62±0.11  n.a.    IR 64   IR55419-04     C  I  S  C  I  S  Amax  37.95±2.72 b  37.09±2.23 b  21.23±3.08 a  38.30±1.15 b  38.22±2.11 b  22.57±3.00 a  Jmax (Ci)  175.79±16.59 b  165.29±13.46 b  99.37±16.18 a  173.15±7.17 b  176.82±12.04 b  106.40±15.53 a  Jmax (Cc)  190.07±21.31  178.45±12.34  n.a.  184.42±6.72  187.87±13.91  n.a.  Vcmax(Ci)  108.91±9.26 c  92.14±8.08 bc  38.12±8.66 aA  96.18±13.27 bc  89.21±2.17 bA  39.16±2.16 aA  Vcmax(Cc)  126.61±16.67 b  107.85±8.52 b  83.89±9.39 aB  103.63±9.70 b  115.54±6.96 bB  84.38±5.27 aB  Jmax/Vcmax(Ci)  1.81±0.14 a  2.16±0.23 a  2.80±0.29 b  1.88±0.19 a  1.99±0.18 a  2.67±0.17 b  Jmax/Vcmax(Cc)  1.67±0.09  2.03±0.37  n.a.  1.81±0.14  1.62±0.11  n.a.  Jmax (μmol m−2 s−1, PPFD-saturated potential rate of electron transport per unit leaf area under), Vcmax (μmol m−2 s−1, photosynthetic Rubisco capacity per unit leaf area), and Jmax to Vcmax ratio. The letters C, I, and S indicate the distance of the plants from the sprinkler as shown in Fig. 2. Data are means of three plants per treatment ±1 SEM; the letters a, b, and c indicate significant differences at P <0.05 in the same line, whereas the letters A and B indicate significant differences at P <0.05 between Vcmax values estimated from A/Ci (Vcmax(Ci)) and A/Cc (Vcmax(Cc)) curves for the same irrigation treatment. View Large Fig. 4. View largeDownload slide A/Ci curves (a, c, e), and A/Cc curves (b, d, f) in plants grown at the near distance from the sprinkler (a, b), at the intermediate distance from the sprinkler (c, d), and at the far distance from the sprinkler (e, f). Fig. 4. View largeDownload slide A/Ci curves (a, c, e), and A/Cc curves (b, d, f) in plants grown at the near distance from the sprinkler (a, b), at the intermediate distance from the sprinkler (c, d), and at the far distance from the sprinkler (e, f). However, the parameterization of the A/Ci curves relies on the wrong assumption that gm is infinite (i.e. Ci=Cc). Consequently, gm measurements were used to derive A/Cc curves from the corresponding A/Ci curves, and it was then possible to compare the estimates of Vcmax and Jmax based on Ci and Cc (Table 1). The analyses of the A/Cc curves confirmed that there were no differences in photosynthetic capacity between IR71525-19-1-1 and IR64 in the well-watered (Fig. 4b) as well as in the water-stressed treatments (Figs 4d, 5f), and between well-watered (Fig. 4b) and moderately water-stressed plants (Fig. 4d). In these treatments, however, the parameters derived from A/Cc curves were slightly but consistently higher than Vcmax and Jmax based on Ci, although the Jmax to Vcmax ratio remain unaffected (Table 1). However, when A of severely water-stressed plants was re-plotted as function of Cc (Fig. 4f), the initial slope of the A/Cc curves became much steeper than that of A/Ci curves, and Vcmax resulted in a 2-fold increase (Table 1). Moreover, the A/Cc curves of water-stressed plants were not saturated by Ca values of about 2200 μmol mol−1. This is further evidence that A was limited by diffusional limitations more than by biochemical ones, which brings more support to the general consensus that diffusion limitations are the predominant factors controlling photosynthesis under most water-stress situations (Cornic, 2000; Lawlor and Cornic, 2002; Chaves et al., 2003; Centritto et al., 2005; Galmés et al., 2007). In addition, the present data also provide further evidence that the photosynthetic parameterizations are improved by using A/Cc instead of A/Ci curves (Flexas et al., 2007). Fig. 5. View largeDownload slide Relationships between (a) leaf water potential (Ψ) and photosynthesis (A) (r2=0.753, P <0.001), (b) leaf water potential (Ψ) and total conductance (gt) (r2=0.725, P <0.001), (c) RWC (relative water content) and A (r2=0.814, P <0.001), (d) RWC and gt (r2=0.672, P <0.001), (e) between A and grain yield for Vandana (V, - - - -) (r2=0.100, P <0.001) and for the pooled data of Apo, DK98, IR55419-04, IR64, IR71525-19-1-1, and PSBRc80 (____) (r2=0.756, P <0.001), and (f) gt and grain yield for Vandana (r2=0.916, P <0.001) and for the pooled data of Apo, DK98, IR55419-04, IR64, IR71525-19-1-1, and PSBRc80 (r2=0.727, P <0.001). The letters C, I, and S indicate the distance of the plants from the sprinkler as shown in Fig. 2. Fig. 5. View largeDownload slide Relationships between (a) leaf water potential (Ψ) and photosynthesis (A) (r2=0.753, P <0.001), (b) leaf water potential (Ψ) and total conductance (gt) (r2=0.725, P <0.001), (c) RWC (relative water content) and A (r2=0.814, P <0.001), (d) RWC and gt (r2=0.672, P <0.001), (e) between A and grain yield for Vandana (V, - - - -) (r2=0.100, P <0.001) and for the pooled data of Apo, DK98, IR55419-04, IR64, IR71525-19-1-1, and PSBRc80 (____) (r2=0.756, P <0.001), and (f) gt and grain yield for Vandana (r2=0.916, P <0.001) and for the pooled data of Apo, DK98, IR55419-04, IR64, IR71525-19-1-1, and PSBRc80 (r2=0.727, P <0.001). The letters C, I, and S indicate the distance of the plants from the sprinkler as shown in Fig. 2. The decline of both A and total leaf conductance (gt) was correlated to leaf water status, as expressed by leaf Ψ (Fig. 5a, b) and RWC (Fig. 5c, d) (Kramer and Boyer, 1995). Although the effect of drought stress on grain yield seems to be correlated with the effect on A (Fig. 5e), it is not clear whether there would be any causality relationship or if these responses would be just a consequences of an effect of drought on plant water status. Boyer and Westgate (2004) pointed out that inhibition of photosynthesis in response to water stress during plant reproduction phases causes a decreased carbon flux to reproductive organs triggering ovary abortion or pollen sterility (as also found in our study; data not shown), which in turn lead to decreased grain yields. However, they intriguingly hypothesized that the inhibitory effect of low plant water potentials could be mimicked by stomatal limitations to photosynthesis triggered by root–shoot chemical signalling (Tardieu and Davies, 1993). Indeed, as photosynthesis was significantly correlated to gt (Fig. 3d), a linear relationships between grain yield and total diffusional limitations to photosynthesis was also found in our study (Fig. 5f). Overall these data indicate that genotypes which showed higher photosynthesis and conductances were also the more productive (e.g. IR71525-19-1-1), whereas those which had lower photosynthesis and conductances were less productive (e.g. IR64, and PSBRc80) across the entire soil moisture gradient. However, there were two noteworthy exceptions. Moroberekan which showed, together with IR71525-19-1-1, the highest rate of photosynthesis, failed to produce grain yield, whereas Vandana which had intermediate values of photosynthesis was the most productive genotype. In addition, this short-duration genotype showed different relationships between yield and both A and gt with respect to the other upland- and lowland-adapted genotypes, for it had higher grain yield values per unit of both photosynthesis and total leaf conductance. An alternative explanation could be that water deficit affects yield mainly through the effects on the plant water status regulation, with photosynthesis being just a consequence of plant water status effects on leaf gas exchange. It has been recently argued that dehydration avoidance is more relevant as a strategy for relieving drought and maintaining rice yield (Serraj et al., 2008). The conclusion emerging from long-term multi-location drought studies was that rice is mostly a drought avoider, with the genotypes that produce higher grain yield under drought being those that are able to maintain better plant water status around flowering and grain setting (Serraj et al., 2008). Mesophyll conductance: comparison between the variable J method and carbon isotope discrimination in recently synthesized sugars Harley et al. (1992) demonstrated that the variable J method is fairly sensitive to errors in the estimation of the rate of mitochondrial respiration in the light (Rl) and Γ*. Rl can be estimated by either the Laisk or the Kok methods. However, these methods do not consider that respiratory CO2 is partially reassimilated by photosynthesis before it can escape the leaf (Pinelli and Loreto, 2003). These authors demonstrated that, taking into account both components, i.e. CO2 evolution and the fraction of CO2 reassimilated by photosynthesis, Rl becomes similar to Rd. Thus, because of the difficulties of obtaining reliable Rl in the field, Rd was taken as a proxy for Rl. Moreover, in this study, we could not estimate Γ*, and this could have been a potential source of error, although this parameter is remarkably conservative. Furthermore, the variable J method is substantially sensitive to measurement errors especially in small-sized clamp-on cuvettes like the LI-6400-40 leaf chamber fluorometer used in this study (Flexas et al., 2007; Rodeghiero et al., 2007). Although the diffusion leaks through the chamber foam gaskets were taken into account and corrected according to the manufacturer's suggestions (Li-Cor Inc., 2004), it was not possible to rule out the occurrence of other measurement errors (Rodeghiero et al., 2007). Finally, cuticular conductance which may have affected gm in severely stressed plants could not be estimated. Because all the aforementioned problems could have affected the accuracy of gm measurements, this parameter was also estimated by using the carbon isotope discrimination in recently synthesized sugars method described by Brugnoli et al. (1988) and further developed by Lauteri et al. (1997), in order to determine whether the two methods could yield similar results and, consequently, have a double check on the estimation of gm. Scartazza et al. (1998) applied the same method to measure gm in upland rice grown in pots under different watering regimes. They were able to monitor a marked decrease of gm along the ontogeny of both well-watered and water-stressed rice. The two estimations of gm obtained in this work are co-ordinated along a robust relationship (Fig. 6). However, it cannot be ignored that gm estimation by the Δ method is affected by uncertainty in the choice of the appropriate f value (see equation 1). Using a value as high as 7‰ (Rooney, 1988; Lloyd et al., 1992; Lauteri et al., 1997), equation 1 yields gm values almost double (data not shown) with respect to those calculated by the variable J method. Results shown in Fig. 6 have been obtained by using f=0, as reported by von Caemmerer and Evans (1991), and by Scartazza et al. (1998) in work on upland rice. It is interesting to note that, using such a value of f, the relationship between gm estimates improves (data not shown), and the slope of the variable J method versus Δ method regression shifts from 0.5 to 0.7. Fig. 6. View largeDownload slide Linear relationship between mesophyll conductance (gm) calculated using the variable J method and gm calculated with the carbon isotope discrimination in recently synthesized sugars method. The letters C, I, and S indicate the distance of the plants from the sprinkler as shown in Fig. 2. Fig. 6. View largeDownload slide Linear relationship between mesophyll conductance (gm) calculated using the variable J method and gm calculated with the carbon isotope discrimination in recently synthesized sugars method. The letters C, I, and S indicate the distance of the plants from the sprinkler as shown in Fig. 2. The higher values found with the Δ observations are possibly ascribed to the larger time integration of this technique. Whilst gas exchange measurements usually refer to a few minutes, Δ on leaf soluble sugars is related to a time span of a few days. In such circumstances, the isotopic signature of soluble sugars would be affected by the mild conditions (especially in terms of VPD and light intensity) of the early morning and the afternoon, when A, gs and, possibly, gm should be maximized. This could, at least partially, explain the discrepancies observed by comparing the two methodologies. However, gm estimated by using the carbon isotope discrimination in recently synthesized sugars method showed the same substantial variations across the rice genotypes and water treatments as well as similar correlations with A and gs (data not shown) as gm calculated with the variable J method. A strong co-ordination between stomatal and mesophyll conductances, estimated from recently synthesized sugars, was also found in drought-adapted and wet-adapted genotypes of Castanea sativa from different rainfall areas (Lauteri et al., 1997). This was confirmed by the present results, and the congruent estimations of gm obtained by applying both the ‘variable J’ and the ‘Δ in recently synthesized sugars’ methods is a further demonstration that gm is an important factor in the interpretation of photosynthesis data. Finally, these results suggest the possibility of accurate estimates of the fractionation factor associated to photorespiration (f), by combining the ‘on-line Δ’ method (Evans et al., 1986; Lloyd et al., 1992) with the variable J method, that is two methods integrating the same time span of plant response. Genotypic variations of diffusive characteristics may positively affect leaf gas exchange performances, allowing a relatively high intrinsic WUE (Fig. 2d) through a less effective separation between the site of transpiration (stomata) and that of carboxylation (chloroplasts). Lauteri et al. (1997) found that drought-adapted Castanea sativa genotypes had much higher gm and lower WUE (both intrinsic and long-term WUE as evaluated by Δ analysis) than wet-adapted genotypes. They also found a positive relationship between leaf soluble sugar Δ and instantaneous WUE (A/E). Such an unexpected relationship was especially addressed to leaf to air vapour pressure difference (VPD) at the crown level among the different genotypes. However, any positive effect of enhanced gm on WUE of drought-adapted genotypes cannot be ruled out as discussed further in the following section. Long-term and short-term WUE estimation by means of Δ analysis on flag leaf matter (pellet and soluble sugars) An average Δ value of 21.4‰ was observed on the pooled data obtained by pellet analysis (Table 2). The water treatments did not affect carbon isotope discrimination of structural carbon of flag leaves, values ranging from 21.3‰ to 21.5‰. By contrast, highly significant differences were shown by comparing the mean values of the different genotypes (Table 2). In particular, DK98 and PSBRc80 showed the highest discrimination (21.9‰) whilst Moroberekan displayed the lowest Δ value (20.9‰). Statistically, the eight genotypes were divided into three clusters, characterized respectively by high Δ (DK98 and PSBRc80), intermediate Δ (IR55419-04, IR64, IR71525-19-1-1, and Vandana) and low Δ (Apo and Moroberekan) on flag leaf pellets. No significant ‘genotype×watering regime’ interaction was found on Δ of flag leaf pellets. It is interesting to note that the lowland genotypes belong to the ‘high Δ’ type, with the exception of IR64 which displayed an intermediate Δ value. The upland cultivars displayed either intermediate (IR55419-04, IR71525-19-1-1, Vandana) or low (Apo, Moroberekan) Δ values (Table 2). These findings are generally in agreement with those of Peng et al. (1998) and Dingkuhn et al. (1991), confirming the existence of genotypic variation of Δ in rice and showing that improved tropical japonica lines (upland-adapted) exhibited higher A/E and lower Δ than the indica cultivars (lowland adapted) under irrigated conditions. Table 2. Mean carbon isotope discrimination (Δ, ‰) values ±standard errors in leaf pellets of eight rice genotypes under three watering regimes Genotype  Well–watered  Intermediate  Drought  Mean  IR55419-04  21.5±0.1  21.4±0.4  21.2±0.1  21.4 B  IR 64  21.5±0.1  21.5±0.2  21.3±0.1  21.5 B  IR71525-19-1-1  21.5±0.2  21.4±0.1  21.2±0.1  21.4 B  DK98  21.6±0.1  22.6±0.1  21.6±0.1  21.9 A  PSBRc80  21.9±0.2  22.0±0.1  21.8±0.1  21.9 A  Apo  20.9±0.1  21.0±0.1  21.1±0.1  21.0 C  Moroberekan  20.9±0.1  21.1±0.1  20.8±0.2  20.9 C  Vandana  21.5±0.1  21.4±0.5  21.6±0.2  21.5 B  Mean  21.4A  21.5A  21.3A    Genotype  Well–watered  Intermediate  Drought  Mean  IR55419-04  21.5±0.1  21.4±0.4  21.2±0.1  21.4 B  IR 64  21.5±0.1  21.5±0.2  21.3±0.1  21.5 B  IR71525-19-1-1  21.5±0.2  21.4±0.1  21.2±0.1  21.4 B  DK98  21.6±0.1  22.6±0.1  21.6±0.1  21.9 A  PSBRc80  21.9±0.2  22.0±0.1  21.8±0.1  21.9 A  Apo  20.9±0.1  21.0±0.1  21.1±0.1  21.0 C  Moroberekan  20.9±0.1  21.1±0.1  20.8±0.2  20.9 C  Vandana  21.5±0.1  21.4±0.5  21.6±0.2  21.5 B  Mean  21.4A  21.5A  21.3A    Differences among genotype mean values are indicated by different capital letters (ANOVA, P <0.01, post-hoc test). View Large The general average value of Δ on leaf soluble sugars was as high as 18.1‰, indicating a 3.3‰ enrichment in 13C of soluble carbohydrates relatively to flag leaf pellets. Although ANOVA did not show any significant ‘genotype×watering regime’ interaction on Δ of leaf soluble sugars, significant and highly significant effects were found due to the genotypes and to the watering regimes, respectively (Table 3). Limitation in water availability caused a drop of Δ on leaf soluble sugars from 18.4‰ to 17.7‰. A higher span of Δ on leaf soluble sugars was observed among the eight genotypes, values ranking from 17.4‰ to 18.8‰ (Table 3). The highest Δ values on leaf soluble sugars were observed for IR71525-19-1-1 (18.8‰) and Moroberekan (18.5‰). PSBRc80, IR64, and IR55419-04 showed intermediate values: 18.3, 18.1, and 18.0‰, respectively. The lowest values were displayed by Vandana (17.4‰), DK98 (17.5‰), and Apo (17.6‰). It is noteworthy that DK98 and Moroberekan reversed their ranking compared with that based on the Δ in leaf pellets. However, the other genotypes did not show contrasting results when comparing Δ of leaf soluble sugars with that of the leaf pellets. Table 3. Mean carbon isotope discrimination (Δ, ‰) values ±standard errors in leaf soluble sugars of eight rice genotypes under three watering regimes Genotype  Well–watered  Intermediate  Drought  Mean  IR55419-04  17.9±0.4  18.7±2.1  17.3±0.4  18.0 abc  IR 64  18.4±0.4  18.2±0.8  17.6±0.4  18.1 abc  IR71525-19-1-1  19.2±0.2  19.1±0.6  18.2±0.5  18.8 a  DK98  18.1±0.3  17.0±0.4  17.5±0.3  17.5 bc  PSBRc80  19.1±0.4  17.3±0.8  18.5±0.4  18.3 abc  Apo  18.4±0.4  17.3±0.4  17.1±0.5  17.6 bc  Moroberekan  18.7±0.3  18.8±0.4  18.2±0.4  18.5 ab  Vandana  17.9±0.4  17.2±1.4  17.2±0.4  17.4 c  Mean  18.4 A  18.0 AB  17.7 B    Genotype  Well–watered  Intermediate  Drought  Mean  IR55419-04  17.9±0.4  18.7±2.1  17.3±0.4  18.0 abc  IR 64  18.4±0.4  18.2±0.8  17.6±0.4  18.1 abc  IR71525-19-1-1  19.2±0.2  19.1±0.6  18.2±0.5  18.8 a  DK98  18.1±0.3  17.0±0.4  17.5±0.3  17.5 bc  PSBRc80  19.1±0.4  17.3±0.8  18.5±0.4  18.3 abc  Apo  18.4±0.4  17.3±0.4  17.1±0.5  17.6 bc  Moroberekan  18.7±0.3  18.8±0.4  18.2±0.4  18.5 ab  Vandana  17.9±0.4  17.2±1.4  17.2±0.4  17.4 c  Mean  18.4 A  18.0 AB  17.7 B    Differences among genotype mean values are indicated by different lower-case letters (ANOVA, P <0.05, post-hoc test). Differences among mean values of watering regimes are indicated by different capital letters (ANOVA, P <0.01, post-hoc test). View Large Carbon isotope discrimination analysis is considered, at present, to be one of the most versatile methodologies for studying WUE either in natural or experimental environments, as it avoids destructive samplings and environmental artefacts (e.g. lysimeter). This methodology can apply different time-scale integrations, depending on the fraction of plant carbon being analysed (see Brugnoli and Farquhar, 2000, for a review). In the present experiment, Δ analysis was applied on both structural (flag leaf pellet) and fast turnover (flag leaf soluble sugars) carbon. In such a way, the inferred information on WUE contains a dual time integration meaning: the long-term scale referred to the time span involved in the assimilation of carbon used during flag leaf development and expansion; the short-term scale referred to the condition of photosynthetic assimilation, including a few days before the leaf sampling (Brugnoli et al., 1988; Scartazza et al., 1998). However, it should be taken into consideration that a finite gm could affect the well-known negative relationship between Δ and WUE. Variations in Δ as high as 1.0‰ and 1.4‰, respectively, for pellets and leaf soluble sugars, highlight relevant differences in both long-term and short-term WUE. PSBRc80 (lowland cultivar) showed the lowest long-term WUE, the highest being displayed by the upland cultivar Moroberekan. However, whilst PSBRc80 maintained a rather low short-term WUE, both DK98 and Moroberekan showed an inversion of their physiological performance, with relatively high and low short-term WUE, respectively. It is interesting to note that these latter contrasting genotypes also show the highest and lowest switch in Δ from the long-term to the short-term analysis: 4.4‰ variation for DK98 and 2.4‰ for Moroberekan. This evidence strongly indicates a differential sensitivity to the experimental conditions and a widely diversified phenotypic plasticity of physiological traits related to Δ and WUE performances across rice genotypes; this also confirms previous observations in rice by Impa et al. (2005). Carbon isotope discrimination analysis in the long term did not reveal any effect of the watering treatments. By contrast, the decrease in Δ of leaf soluble sugars indicated a clear effect of water shortage in terms of an increased WUE, as expected by the theory (Farquhar et al., 1982). These contrasting evidences clearly suggest that the structural carbon of the flag leaves was assimilated under more favourable environmental conditions than those affecting the synthesis of soluble sugars in the same leaves. The generally higher Δ values observed in the leaf pellets might also partially reflect the presence of 13C-depleted compounds which are absent in the purified carbohydrates. Metabolic compounds like lipids, resins, and terpenoids would be more depleted in 13C than carbohydrates because of isotope fractionation along the fatty acid biosynthetic pathway. However, the limited amount of these compounds within the leaf pellet makes their effective contribution to the wide discrepancy (3.3‰ on average) observed on Δ of pellets, with respect to that of leaf soluble sugars, unlikely. A similar variation of Δ (3.7‰ on average) comparing structural leaf carbon and stem soluble sugars was previously observed on upland rice grown in pots under different watering regimes (Scartazza et al., 1998). In the present work, a highly significant relationship (r=–0.57, P < 0.01, data not shown) was observed between Δ in soluble sugars and A/gs (i.e. intrinsic WUE). Such a finding corroborates the conventional Δ analysis (i.e. isotope discrimination analysis of structural carbon) used here to assess WUE variability across genotypes and watering treatments. However, it must be stated that only a positive tendency, although not a significant relationship, was found between Δ in soluble sugars and Ci/Ca (data not shown). Finally, relatively high mesophyll conductance may have a positive effect on WUE (Lauteri et al., 1997; Flexas et al., 2008). In fact, a negative but weak relationship between Δ in soluble sugars and the A to gm ratio was observed (r = 0.39, P < 0.05, data not shown). In addition, a highly significant, positive, relationship (r = 0.60, P < 0.01) was also found between pooled observations of grain yield and Δ in leaf soluble sugars (data not shown). This is in keeping with the direct relationships found between A, gt, and grain yield (Fig. 5e, f). It is known that carbohydrates synthesized by flag leaves have a crucial role in grain filling and yield (Yoshida, 1981; Murchie et al., 1999). Indeed, the Δ analysis confirms that, under drought, rice genotypes increase their WUE and this is reflected in a general decrease of Δ (Table 3). Decreased Δ values indicate lowered Cc/Ca ratios and, consequently, that at any given VPD variations in WUE result from co-ordinated trade-offs between photosynthetic CO2 uptake and total leaf conductance to CO2. Any effect of gm on WUE must be considered as a result of the contribution of gm itself to this balance (Flexas et al., 2008). Dedicated experiments, however, are needed in order to focus the influence of gm on Δ, WUE, and yield. Conclusions In order to improve the precision of drought-phenotyping methodologies in rice, it is increasingly important to quantify how physiological and agronomic traits respond to soil water deficit. The major components limiting photosynthesis are the CO2 diffusional factors, especially CO2 diffusion through stomata and mesophyll, and the biochemical components related to triosephosphate formation. A change in any of these components will alter the final photosynthesis rate. While there have been a number of studies that investigated the individual effects of drought on photosynthetic limitations and WUE/yield, few studies have investigated the integrative response of photosynthesis components and WUE/yield to increasing drought stress. All these factors were integrated in the present study with the aim of phenotyping the responses of eight contrasting rice genotypes in the field to water stress, which was applied from the flowering stage until the beginning of physiological maturity by mean of a line source sprinkler system. Our data demonstrate that gm is a major determinant of photosynthesis, for rice genotypes with inherently higher mesophyll conductance and, in general, with higher total diffusive conductances, were capable of keeping higher assimilation rates in stressed conditions. Moreover, the effect of drought stress on grain yield was correlated with the effects on both A and total diffusional limitations to photosynthesis. Overall these data indicate that genotypes which showed higher photosynthesis and conductances were also more productive, whereas those which had lower photosynthesis and conductances were less productive across the entire soil moisture gradient. Two noteworthy exceptions included the variety Moroberekan, which had the highest A but failed to produce grain, and Vandana, which had intermediate values of A but was the most productive genotype. These two cases highlight the importance of considering photosynthesis as a dynamic process in relation to plant water status, escape and dehydration avoidance mechanisms affecting yield under drought. Despite the fact that more work is needed to address the hypothesis of carbon limitation of grain yield induced by a decrease of photosynthesis under drought, and to distinguish the photosynthetic causes from the stomatal closure causes of differences in WUE, the present data provide direct evidence of a correlation between photosynthesis and diffusional conductances and yield, and may provide the mechanistic basis for designing future drought-phenotyping protocols. The authors would like to extend their sincere thanks to Rolly Torres and Glenn Dimayuga for their technical assistance in the field at IRRI, and Luciano Spaccino and Matilde Tamantini for their help with the analyses of carbon isotope discrimination. The authors are also grateful to Professor G Cornic and Dr Xinyou Yin for their constructive and helpful comments, and the Rockefeller Foundation Project ‘Physiological Basis of Tolerance in Rice’ for providing funding. Finally, we thank two anonymous reviewers for their constructive and helpful comments on the manuscript. References Aragon EL,  De Datta SK.  Drought response of rice at different nitrogen levels using line source sprinkler system,  Irrigation Science ,  1982, vol.  3 (pg.  63- 73) Google Scholar CrossRef Search ADS   Atlin G,  Lafitte HR,  Tao D,  Laza RC,  Amante M,  Courtois B.  Developing rice cultivars for high-fertility upland systems in the Asian tropics,  Field Crops Research ,  2006, vol.  97 (pg.  43- 52) Google Scholar CrossRef Search ADS   Bernacchi CJ,  Singsaas EL,  Pimentel C,  Portis ARJr,  Long SP.  Improved temperature response functions for models of Rubisco-limited photosynthesis,  Plant, Cell and Environment ,  2001, vol.  24 (pg.  253- 259) Google Scholar CrossRef Search ADS   Botwright Acuňa TL,  Lafitte HR,  Wade LJ.  Genotype×environment interactions for grain yield of upland rice backcross lines in diverse hydrological environments,  Field Crops Research ,  2008, vol.  108 (pg.  117- 125) Google Scholar CrossRef Search ADS   Bouman BAM,  Xiaoguang Y,  Huaqui W,  Zhimin W,  Junfang Z,  Bin C.  Performance of aerobic rice varieties under irrigated conditions in north China,  Field Crops Research ,  2006, vol.  97 (pg.  53- 65) Google Scholar CrossRef Search ADS   Boyer JS,  Westgate ME.  Grain yields with limited water,  Journal of Experimental Botany ,  2004, vol.  55 (pg.  2385- 2394) Google Scholar CrossRef Search ADS PubMed  Brilli F,  Barta C,  Fortunati A,  Lerdau M,  Loreto F,  Centritto M.  Response of isoprene emission and carbon metabolism to drought in white poplar (Populus alba) saplings,  New Phytologist ,  2007, vol.  175 (pg.  244- 254) Google Scholar CrossRef Search ADS PubMed  Brugnoli E,  Farquhar GD.  Leegood RC,  Sharkey TD,  von Caemmerer S.  Photosynthetic fractionation of carbon isotopes,  Photosynthesis: physiology and metabolism, Advances in photosynthesis ,  2000, vol.  Vol. 9  The Netherlands Kluwer Academic Publishers(pg.  399- 434) Brugnoli E,  Hubick KT,  von Caemmerer S,  Wong SC,  Farquhar GD.  Correlation between the carbon isotope discrimination in leaf starch and sugars of C3 plants and the ratio of intercellular and atmospheric partial pressures of carbon dioxide,  Plant Physiology ,  1988, vol.  88 (pg.  1418- 1424) Google Scholar CrossRef Search ADS PubMed  Centritto M,  Loreto F,  Chartzoulakis K.  The use of low [CO2] to estimate diffusional and non-diffusional limitations of photosynthetic capacity of salt-stressed olive saplings,  Plant, Cell and Environment ,  2003, vol.  26 (pg.  585- 594) Google Scholar CrossRef Search ADS   Centritto M,  Wahbi S,  Serraj R,  Chaves MM.  Effects of partial rootzone drying (PRD) on adult olive tree (Olea europaea) in field conditions under arid climate. II. Photosynthetic responses,  Agriculture Ecosystems and Environment ,  2005, vol.  106 (pg.  303- 311) Google Scholar CrossRef Search ADS   Chaves MM.  Effects of water deficits on carbon assimilation,  Journal of Experimental Botany ,  1991, vol.  42 (pg.  1- 16) Google Scholar CrossRef Search ADS   Chaves MM,  Maroco JP,  Pereira JS.  Understanding plant responses to drought: from genes to the whole plant,  Functional Plant Biology ,  2003, vol.  30 (pg.  239- 264) Google Scholar CrossRef Search ADS   Cornic G.  Drought stress inhibits photosynthesis by decreasing stomatal aperture: not by affecting ATP synthesis,  Trends in Plant Science ,  2000, vol.  5 (pg.  187- 188) Google Scholar CrossRef Search ADS   Cruz RT,  O'Toole JC.  Dryland rice response to an irrigation gradient at flowering stage,  Agronomy Journal ,  1984, vol.  76 (pg.  178- 183) Google Scholar CrossRef Search ADS   Degenkolbe T,  Phuc TD,  Zuther E,  Repsilber D,  Walther D,  Hincha DK,  Köhl KI.  Expression profiling of rice cultivars differing in their tolerance to long-term drought stress,  Plant Molecular Biology ,  2009, vol.  69 (pg.  133- 153) Google Scholar CrossRef Search ADS PubMed  Dingkuhn M,  De Datta SK,  Dorffling K,  Javellana C.  Varietal differences in leaf water potential, leaf net CO2 assimilation, conductivity, and water use efficiency in upland rice,  Australian Journal of Agricultural Research ,  1989, vol.  40 (pg.  1183- 1192) Google Scholar CrossRef Search ADS   Dingkuhn M,  Farquhar GD,  De Datta SK,  O'Toole JC.  Discrimination of 13C among upland rices having different water use efficiencies,  Australian Journal of Agricultural Research ,  1991, vol.  42 (pg.  1123- 1131) Google Scholar CrossRef Search ADS   Ehleringer JR.  Ehleringer JR,  Hall AE,  Farquhar GD.  Carbon and water relations in desert plants: an isotopic perspective,  Stable isotopes and plant carbon–water relations ,  1993 San Diego Academic Press(pg.  155- 172) Ekanayake I,  De Datta S,  Steponkus P.  Spikelet sterility and flowering response of rice to water stress at anthesis,  Annals of Botany ,  1989, vol.  63 (pg.  257- 264) Google Scholar CrossRef Search ADS   Evans JR,  Sharkey TD,  Berry JA,  Farquhar GD.  Carbon isotope discrimination measured concurrently with gas exchange to investigate CO2 diffusion in leaves of higher plants,  Australian Journal of Plant Physiology ,  1986, vol.  13 (pg.  281- 292) Google Scholar CrossRef Search ADS   Farquhar GD,  Ehleringer JR,  Hubick KT.  Carbon isotope discrimination and photosynthesis,  Annual Review of Plant Physiology and Plant Molecular Biology ,  1989, vol.  40 (pg.  503- 537) Google Scholar CrossRef Search ADS   Farquhar GD,  O'Leary MH,  Berry JA.  On the relationship between carbon isotope discrimination and the intercellular carbon dioxide concentration in leaves,  Australian Journal of Plant Physiology ,  1982, vol.  9 (pg.  121- 137) Google Scholar CrossRef Search ADS   Farquhar GD,  von Caemmerer S,  Berry JA.  A biochemical model of photosynthetic CO2 assimilation in leaves of C3 species,  Planta ,  1980, vol.  149 (pg.  78- 90) Google Scholar CrossRef Search ADS PubMed  Flexas J,  Diaz-Espejo A,  Berry JA,  Cifre J,  Galmés J,  Kaldenhoff R,  Medrano H,  Ribas-Carbó M.  Analysis of leakage in IRGA's leaf chambers of open gas exchange systems: quantification and its effects in photosynthesis parameterization,  Journal of Experimental Botany ,  2007, vol.  58 (pg.  1533- 1543) Google Scholar CrossRef Search ADS PubMed  Flexas J,  Ribas-Carbó M,  Diaz-Espejo A,  Galmés J,  Medrano H.  Mesophyll conductance to CO2: current knowledge and future prospects,  Plant, Cell and Environment ,  2008, vol.  31 (pg.  602- 621) Google Scholar CrossRef Search ADS   Galmés J,  Flexas J,  Keys AJ,  Cifre J,  Mitchell RAC,  Madgwick PJ,  Haslam RP,  Medrano H,  Parry MA.  Rubisco specificity factor tends to be larger in plant species from drier habitats and in species with persistent leaves,  Plant, Cell and Environment ,  2005, vol.  28 (pg.  571- 579) Google Scholar CrossRef Search ADS   Galmés J,  Medrano H,  Flexas J.  Photosynthetic limitations in response to water stress and recovery in Mediterranean plants with different growth forms,  New Phytologist ,  2007, vol.  175 (pg.  81- 93) Google Scholar CrossRef Search ADS PubMed  Gillon JS,  Yakir D.  Internal conductance to CO2 diffusion and C18OO discrimination in C3 leaves,  Plant Physiology ,  2000, vol.  123 (pg.  201- 213) Google Scholar CrossRef Search ADS PubMed  Harley PC,  Loreto F,  Di Marco G,  Sharkey TD.  Theoretical considerations when estimating the mesophyll conductance to CO2 flux by analysis of the response of photosynthesis to CO2,  Plant Physiology ,  1992, vol.  98 (pg.  1429- 1436) Google Scholar CrossRef Search ADS PubMed  Hervé P,  Serraj R.  GM technology and drought: a simple solution for a complex trait?,  African Journal of Biotechnology ,  2009  (in press) IPCC (Intergovernmental Panel on Climate Change) Climate change 2007,  2007  http://www.ipcc.ch/ Impa SM,  Nadaradjan S,  Boominathan P,  Shashidhar G,  Bindumadhava H,  Sheshshayee MS.  Carbon isotope discrimination accurately reflects variability in WUE measured at a whole plant level in rice,  Crop Science ,  2005, vol.  45 (pg.  2517- 2522) Google Scholar CrossRef Search ADS   Kamoshita A,  Rodriguez R,  Yamauchi A,  Wade L.  Genotypic variation in response of rainfed lowland to prolonged drought and rewatering,  Plant Production Science ,  2004, vol.  7 (pg.  406- 420) Google Scholar CrossRef Search ADS   Kramer PJ,  Boyer JS. ,  Water relations of plants and soils ,  1995 San Diego Academic Press Lafitte HR,  Courtois B,  Arraudeau M.  Genetic improvement of rice in aerobic systems: progress from yield to genes,  Field Crops Research ,  2002, vol.  75 (pg.  171- 190) Google Scholar CrossRef Search ADS   Lauteri M,  Scartazza A,  Guido MC,  Brugnoli E.  Genetic variation in photosynthetic capacity, carbon isotope discrimination and mesophyll conductance in provenances of Castanea sativa adapted to different environments,  Functional Ecology ,  1997, vol.  11 (pg.  675- 683) Google Scholar CrossRef Search ADS   Lawlor DW,  Cornic G.  Photosynthetic carbon assimilation and associated metabolism in relation to water deficits in higher plants,  Plant, Cell and Environment ,  2002, vol.  25 (pg.  275- 294) Google Scholar CrossRef Search ADS   Li-Cor Inc,  Using the LI-6400 portable photosynthesis system ,  2004  Version 5. Li–Cor, Inc., Lincoln, NE, USA Lilley J,  Fukai S.  Effect of timing and severity of water deficit on four diverse rice cultivars III. Phenological development, crop growth and yield,  Field Crops Research ,  1994, vol.  37 (pg.  225- 234) Google Scholar CrossRef Search ADS   Lloyd J,  Syvertsen JP,  Kriedemann PE,  Farquhar GD.  Low conductances of CO2 diffusion from stomata to the sites of carboxylation in leaves of woody species,  Plant, Cell and Environment ,  1992, vol.  15 (pg.  873- 899) Google Scholar CrossRef Search ADS   Loreto F,  Centritto M,  Chartzoulakis K.  Photosynthetic limitations in olive cultivars with different sensitivity to salt stress,  Plant, Cell and Environment ,  2003, vol.  26 (pg.  595- 601) Google Scholar CrossRef Search ADS   Loreto F,  Di Marco G,  Tricoli D,  Sharkey TD.  Measurements of mesophyll conductance, photosynthetic electron-transport and alternative electron sinks of field-grown wheat leaves,  Photosynthesis Research ,  1994, vol.  41 (pg.  397- 403) Google Scholar CrossRef Search ADS PubMed  Loreto F,  Harley PC,  Di Marco G,  Sharkey TD.  Estimation of mesophyll conductance to CO2 flux by three different methods,  Plant Physiology ,  1992, vol.  98 (pg.  1437- 1443) Google Scholar CrossRef Search ADS PubMed  Miyazawa S-I,  Yoshimura S,  Shinzaki Y,  Maeshima M,  Miyake C.  Deactivation of aquaporins decreases internal conductance to CO2 diffusion in tobacco leaves grown under long-term drought,  Functional Plant Biology ,  2008, vol.  35 (pg.  553- 564) Google Scholar CrossRef Search ADS   Murchie EH,  Chen Y-Z,  Hubbart S,  Peng S,  Horton P.  Interactions between senescence and leaf orientation determine in situ patterns of photosynthesis and photoinhibition in field-grown rice,  Plant Physiology ,  1999, vol.  119 (pg.  553- 563) Google Scholar CrossRef Search ADS PubMed  Niinemets Ü,  Cescatti A,  Rodeghiero M,  Tosens T.  Leaf internal diffusion conductance limits photosynthesis more strongly in older leaves of Mediterranean evergreen broad-leaved species,  Plant, Cell and Environment ,  2005, vol.  28 (pg.  1552- 1566) Google Scholar CrossRef Search ADS   Pandey S,  Bhandari H,  Ding S,  Prapertchob P,  Sharan R,  Naik D,  Taunk SK,  Sastri A.  Coping with drought in rice farming in Asia: insights from a cross-country comparative study,  Agricultural Economists ,  2007, vol.  37 (pg.  213- 224) Google Scholar CrossRef Search ADS   Peng S,  Laza RC,  Khush GS,  Sanico AL,  Visperas RM,  Garcia FV.  Transpiration efficiencies of indica and improved tropical japonica rice grown under irrigated conditions,  Euphytica ,  1998, vol.  103 (pg.  103- 108) Google Scholar CrossRef Search ADS   Pinelli P,  Loreto F.  CO2)-C12 emission from different metabolic pathways measured in illuminated and darkened C3 and C4 leaves at low, atmospheric and elevated CO2 concentration,  Journal of Experimental Botany ,  2003, vol.  54 (pg.  1761- 1769) Google Scholar CrossRef Search ADS PubMed  de Pury DGG,  Farquhar GD.  Simple scaling of photosynthesis from leaves to canopies without the errors of big-leaf model,  Plant, Cell and Environment ,  1997, vol.  20 (pg.  537- 557) Google Scholar CrossRef Search ADS   Rodeghiero M,  Niinemets Ü,  Cescatti A.  Major diffusion leaks of clamp-on leaf cuvettes still unaccounted: how erroneous are the estimates of Farquhar et al. model parameters?,  Plant, Cell and Environment ,  2007, vol.  30 (pg.  1006- 1022) Google Scholar CrossRef Search ADS   Rooney MA. ,  Short-term carbon isotope discrimination by plants ,  1988 Madison, WI University of Wisconsin  PhD Thesis Sadras VO,  Milroy SP.  Soil-water thresholds for the responses of leaf expansion and gas exchange: a review,  Field Crops Research ,  1996, vol.  2-3 (pg.  253- 266) Google Scholar CrossRef Search ADS   Samejima M.  Intraspecific variations of 13C-discrimination in Oryza sativa L,  Bulletin of the National Institute of Agrobiological Resources ,  1985, vol.  1 (pg.  63- 84) Scartazza A,  Lauteri M,  Guido MC,  Brugnoli E.  Carbon isotope discrimination in leaf and stem sugars, water–use efficiency and mesophyll conductance during different developmental stages in rice subjected to drought,  Australian Journal of Plant Physiology ,  1998, vol.  25 (pg.  489- 498) Google Scholar CrossRef Search ADS   Serraj R,  Allen HL,  Sinclair TR.  Soybean leaf growth and gas exchange response to drought under carbon dioxide enrichment,  Global Change Biology ,  1999, vol.  5 (pg.  283- 292) Google Scholar CrossRef Search ADS   Serraj R,  Dimayuga D,  Gowda V,  Guan Y,  Hong He,  Impa S,  Liu DC,  Mabesa RC,  Sellamuthu R,  Torres R.  Serraj R,  Bennett J,  Hardy B.  Drought resistant rice: physiological framework for an integrated research strategy,  Drought frontiers in rice: crop improvement for increased rainfed production ,  2008 Singapore World Scientific Publishing  Los Baňos, Philippines: International Rice Research Institute, 139–170 Shi Z,  Liu S,  Liu X,  Centritto M.  Altitudinal variation in photosynthetic capacity, diffusional conductance, and δ13C of butterfly bush (Buddleja davidii Franch.) plants growing at high elevations,  Physiologia Plantarum ,  2006, vol.  128 (pg.  722- 731) Google Scholar CrossRef Search ADS   Tardieu F,  Davies WJ.  Integration of hydraulic and chemical signalling in the control of stomatal conductance and water status of droughted plants,  Plant, Cell and Environment ,  1993, vol.  16 (pg.  341- 349) Google Scholar CrossRef Search ADS   Thompson AJ,  Andrews J,  Mulholland BJ, et al.  Overproduction of abscisic acid in tomato increases transpiration efficiency and root hydraulic conductivity and influences leaf expansion,  Plant Physiology ,  2007, vol.  143 (pg.  1905- 1917) Google Scholar CrossRef Search ADS PubMed  Uehlein N,  Lovisolo C,  Siefritz F,  Kaldenhoff R.  The tobacco aquaporin NtAQP1 is a membrane CO2 pore with physiological functions,  Nature ,  2003, vol.  425 (pg.  734- 737) Google Scholar CrossRef Search ADS PubMed  von Caemmerer S,  Evans JR.  Determination of the average partial pressure of CO2 in chloroplasts from leaves of several C3 plants,  Australian Journal of Plant Physiology ,  1991, vol.  18 (pg.  287- 305) Google Scholar CrossRef Search ADS   Warren CR.  Estimating the internal conductance to CO2 movement,  Functional Plant Biology ,  2006, vol.  33 (pg.  431- 442) Google Scholar CrossRef Search ADS   Warren CR.  Stand aside stomata, another actor deserves centre stage: the forgotten role of the internal conductance to CO2 transfer,  Journal of Experimental Botany ,  2008, vol.  59 (pg.  1475- 1487) Google Scholar CrossRef Search ADS PubMed  Wassmann R,  Jagadish SVK,  Heuer S,  Ismail A,  Redona E,  Serraj R,  Singh RK,  Howell G,  Pathak H,  Sumfleth K.  Climate change affecting rice production: the physiological and agronomic basis for possible adaptation strategies,  Advances in Agronomy ,  2009  (in press) Wassmann R,  Jagadish SVK,  Sumfleth K,  Pathak H,  Howell G,  Ismail A,  Serraj R,  Redona E,  Singh RK,  Heuer S.  Regional vulnerability of climate change impacts on Asian rice production and scope for adaptation,  Advances in Agronomy ,  2009  (in press) Wong SC,  Cowan IR,  Farquhar GD.  Stomatal conductance correlates with photosynthetic capacity,  Nature ,  1979, vol.  282 (pg.  424- 426) Google Scholar CrossRef Search ADS   Yoshida S. ,  Fundamentals of rice crop science ,  1981 Los Baňos, Philippines International Rice Research Institute Zhao XQ,  Xu JL,  Zhao M,  Lafitte R,  Zhu LH,  Fu BY,  Gao YM,  Li ZK.  QTLs affecting morph–physiological traits related to drought tolerance detected in overlapping introgression lines of rice (Oryza sativa L.),  Plant Science ,  2008, vol.  174 (pg.  618- 625) Google Scholar CrossRef Search ADS   © The Author [2009]. Published by Oxford University Press [on behalf of the Society for Experimental Biology]. All rights reserved. For Permissions, please e-mail: journals.permissions@oxfordjournals.org
TI - Leaf gas exchange, carbon isotope discrimination, and grain yield in contrasting rice genotypes subjected to water deficits during the reproductive stage
JF - Journal of Experimental Botany
DO - 10.1093/jxb/erp123
DA - 2009-05-14
UR - https://www.deepdyve.com/lp/oxford-university-press/leaf-gas-exchange-carbon-isotope-discrimination-and-grain-yield-in-LKW3UbRJsE
SP - 2325
EP - 2339
VL - 60
IS - 8
DP - DeepDyve
ER -