HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Figures Only Full Text (PDF) Free Alert me when this article is cited Alert me if a correction is posted Services Related articles in Crop Science Similar articles in this journal Similar articles in ISI Web of Science Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (12) Citing Articles via Google Scholar Google Scholar Articles by Lafitte, H. R. Articles by Courtois, B. Search for Related Content PubMed Articles by Lafitte, H. R. Articles by Courtois, B. Agricola Articles by Lafitte, H. R. Articles by Courtois, B. Related Collections Water Stress Rice Plant and Environment Interactions

CROP BREEDING, GENETICS & CYTOLOGY

Interpreting Cultivar x Environment Interactions for Yield in Upland Rice

Assigning Value to Drought-Adaptive Traits

^a IRRI, DAPO 7777, Metro Manila, The Philippines
^b CIRAD-Biotrop TA40/03 Avenue Agropolis 34398 Montpellier Cdex 5, France

* Corresponding author (R.Lafitte{at}cgiar.org)

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
Many morphological and physiological traits have been suggested as potential mechanisms of cultivar differences in rice (Oryza sativa L.) drought tolerance, but few data are available to link those traits to variation in grain production in water-limited field environments. We evaluated 45 rice cultivars in managed stress environments to relate cultivar x environment (C x E) interaction for yield to specific putative drought-adaptive mechanisms. Cultivars were sown in aerobic (nonflooded) fields across three seasons, under a range of irrigation systems, to generate nine contrasting environments. Data were collected on yield, plant height, maturity, leaf area, relative water content (RWC), epidermal conductance, root pressure, canopy temperature (CT), and chlorophyll content. Mean grain yield ranged from 0.6 to 2.0 Mg ha^-1 across environments. Correlations between grain yields measured in different environments ranged from -0.14 to 0.86. Pattern analysis revealed different cultivar responses in environments with continuous stress or stress during grain filling compared with environments with ample water or environments with adequate drip irrigation. Traits related to C x E interaction scores for yield included anthesis date, leaf percentage fresh weight (%FW), root pressure, leaf area, and rooting depth. Early maturity was found to be advantageous under drought, even when stress was applied at specific developmental stages for each cultivar. Separate pattern analyses for yield components confirmed that cultivar groups that interacted differently with environments for spikelet fertility and thousand-grain weight (TGW) also differed in specific drought-related traits.

Abbreviations: C x E, cultivar x environment • CT, canopy temperature • Da, drip irrigation with maximum stress at flowering • Dg, drip irrigation with stress during grain filling • Di, drip irrigation with no drought treatment • DS, dry season • Dv, drip irrigation with stress in vegetative stage • Fc, furrow irrigated with continuous moisture stress • Fi, furrow irrigated with no drought treatment • %FW, percentage water in fully turgid leaves • k', derived leaf conductance • LAI, leaf area index • PAR, photosynthetically active radiation • PCA, principal component axis • RWC, relative water content • Sc, sprinkler irrigated with continuous moisture stress • Sv, sprinkler irrigation with stress at vegetative stage • TGW, thousand-grain weight • WS, wet season

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
PLANTS HAVE EVOLVED MANY STRATEGIES to survive and reproduce using scarce and variable water resources. First farmers, and now plant scientists, have attempted to incorporate those strategies that enhance productivity into crop species. The best integrated measure of success in this effort is improved grain yield across a range of stress environments. There is frequently an advantage, however, to focusing on the drought-adaptive strategy itself through the use of physiological or morphological selection criteria. These are referred to as secondary traits. Many secondary traits are easier to measure than yield across representative stress environments, particularly if the trait is expressed constitutively, or if it can be measured on seedlings, or if it can be identified using genetic markers. The challenge to selection for secondary traits as an avenue to crop improvement for water-limited environments is that the trait itself must be strongly correlated with yield in the target environment. Unfortunately, many traits do not pass this test. The difficulties associated with trait-based selection have been well documented (Ludlow and Muchow, 1990; Blum, 1996; Richards, 1996).

In contrast to most other cereal crops, rice originated from a semi-aquatic ancestor, and has limited tolerance to water deficit. A number of traits have been suggested as candidates for the improvement of rice drought tolerance (Fukai and Cooper, 1995; Nguyen et al., 1997) The rice cultivars best adapted to dry conditions are upland types, which are thought to have been selected by generations of farmers beginning some 4000 yr ago (Chang, 1976). Specific traits expected to improve water uptake (e.g., deep rooting; O'Toole and Bland, 1987) or reduce unproductive water loss (e.g., greater epiculticular wax; O'Toole et al., 1979) are generally superior in upland cultivars. Other traits that might mitigate the impact of water deficit on survival, such as osmotic adjustment, tend to be superior in lowland cultivars (Lilley and Ludlow, 1996). Rice cultivars can be differentiated not only on the basis of adaptation but also on the basis of subspecies or isozyme group (Glazmann, 1987). Isozyme Group 6, the japonica subspecies, tends to be characterized by limited tillering, thick roots, and thick leaves. Isozyme Group 1 (indica subspecies) has greater tillering, thin leaves, thinner roots, and greater capacity for osmotic adjustment. Cultivars from isozyme Group 2 (Aus types) have plant type similar to Group 1, except that they have fewer tillers and deeper roots (Lafitte et al., 2001). Within Groups 1 and 6, there are both early- and late-maturing cultivars; Group 2 cultivars tend to be early maturing.

The existence of this strong structure of subspecies and cultivation types in rice complicates the identification of drought-adaptive traits, because specific sets of traits tend to occur together in a given cultivar. Each cultivar has its own adaptation and yield potential that may or may not be causally related to traits under study. One technique that has been used to remove the influence of yield potential on estimation of the adaptive value of a trait is a drought response index (Bidinger et al., 1987). As an alternative approach, in this study, we have used pattern analysis to focus on C x E interaction across a range of environments differing in water availability. Associations between cultivar scores and genotypic covariates have been used to identify traits associated with superior yields in contrasting environments for wheat (Yan and Hunt, 2001). When the cultivar scores are from analysis of the interaction term alone, they should indicate the adaptive values of individual traits in different environments, largely independent of yield potential. This approach is used here to assign value to simple traits that can be measured in the context of a rice cultivar improvement program.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
Nine environments were included (Table 1) , all located at the IRRI Experiment Station (14°11' N, 121°15' E, 21 m above sea level). The soil is an Andaqueptic Haplaquol with pH of 6.5. Plot size in these experiments ranged from 0.9 by 3 m to 2.5 by 5m. Except for the wet season (WS) experiment, which was sown in June, all experiments were sown in the dry season (DS) that extends from January to May. During this period, temperatures and vapor pressure deficit increase. All crops were sown in dry soil at a rate of 80 kg ha^-1, except for the furrow-irrigated experiments (Table 1). Nitrogen fertilizer was applied in 2 or 3 splits for a total application of 90 to 120 kg ha^-1. Basal P and K were applied at 30 kg ha^-1. Chemical and manual weed control was used. Irrigation was provided by either drip, furrow, or sprinkler irrigation, with frequency as shown in Table 1. For well-irrigated treatments, water was applied to maintain soil water potential above -0.04 MPa (monitored using tensiometers at 15- and 30-cm depths), which is the level at which more sensitive rice cultivars begin to exhibit a reduction in relative transpiration (Bois et al., 1984). In the drip-irrigated treatments, the stress was applied at the same developmental stage for each cultivar by using individual plot shut-off valves.

Each experiment was analyzed individually using the SAS GLM procedure (SAS Institute, 1992). Yield means were obtained from all nine environments, means of spikelet number and fertility were from six environments, and TGW was measured in eight environments. Pattern analysis was conducted using IRRISTAT software (IRRI, 2000) with separate analyses for yield, spikelet number per panicle, spikelet fertility, and TGW. The mean polish procedure was used for data standardization. In this procedure, analysis is on the residuals after fitting row and column main effects by least squares, so the principal component axes reflect the interaction terms, similar to AMMI analysis (McLaren, 1996). Cultivar and environment classification was by incremental least squares. The number of cultivar and environment groups was selected to retain roughly the same proportion of the variation captured as was captured in the significant principal component axes. On the basis of these criteria, the 45 cultivars were divided into eight or nine groups, and the nine environments were separated into five or six groups. The ordination process resulted in three principal component axes; primarily the first two axes are reported here. Correlations between principal component scores for cultivars and other traits were calculated using the SAS CORR procedure. Because of the large number of comparisons, only correlation coefficients with P < 0.005 were considered significant. This generally corresponded to an association that explained more than one-fourth of the observed variation (r² > 0.25). Cultivar groups were used as treatment factors in an analysis of variance of secondary traits (SAS procedure GLM) to test the hypothesis that groups defined by C x E interaction scores also differed in other traits.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
Grain Yield
In the furrow-irrigated continuous stress (Fc) and sprinkler-irrigated continuous stress (Sc) environments, the irrigation supplied was not adequate to fully meet evaporative demand, and the crop was under mild continuous stress for much of the season. In the other experiments, the irrigation applied was at least 1.6 times the evaporative demand, except during specific stress periods as indicated in Table 1. Environments, cultivars, and interaction between environments and cultivars accounted for 36, 30, and 34% of the total sum of squares. The largest yields were recorded in the WS environment (Table 1). The main effect of cultivar was large, and some cultivars attained much greater mean yields than others (e.g., IR55419, N22, IRAT212, Aus257; Table 2). Correlations of cultivar yield among environments varied widely (Table 3) . Wet season yields were well correlated with yields in DS trials having ample moisture from preflowering to maturity via furrow (Fi) or sprinkler irrigation (Sv). Yields in the continuous stress environments (Fc, Sc) were well correlated, and were also associated with yields in the drip-irrigated environments with vegetative stage (Dv) or grain filling stage (Dg) stress.

View larger version (21K): [in this window] [in a new window]   Fig. 1. Biplot of the first two ordination scores for standardized grain yield for diverse rice cultivars in nine managed environments. The environment vectors are labeled as in Table 1. The codes for cultivars are from Table 2. PCA = principal component axis.

As an alternative to correlation analysis between cultivar scores and plant traits, cultivar group membership identified in the pattern analysis for grain yield was used to conduct an analysis of variance for the measured traits (Table 5) . In some cases, the pattern of traits was related to maturity. For example, later groups tended to have greater delay of flowering caused by stress, lower CT in a continuous stress environment, and greater leaf rolling. Other traits that differed among cultivar groups, but were not consistently related to maturity, included root pressure in Di, %FW, leaf chlorophyll measured during grainfilling stress, and epidermal conductance. Group Y7, which interacted positively with continuous stress environments, had minimum anthesis delay in Dv, low %FW in Fc, little leaf rolling, high CT, and long roots (Table 5). Groups that interacted positively with wetter environments (Y2, Y9) flowered later, and had large flowering delay with vegetative stress, low chlorophyll, rolled leaves, low CT, low root pressure, and short root length. The group with moderate yields and minimum interaction with environment (Y3) had high root pressure, a large change in %FW from Fc to WS, highest leaf chlorophyll after stress, little rolling, and low epidermal conductance.

Yield Components
Data on spikelets per panicle were available for six environments. The continuous stress environments had the fewest spikelets per panicle (Table 1). Consistently high spikelet numbers were found in WAB181 (40), Moroberekan (22), and IAC25 (28), all japonica cultivars. In the pattern analysis of C x E interactions for spikelets per panicle, the PCA1 accounted for 35% of the variation, and, in contrast to the analysis for grain yield, it separated Dg from environments with stress during the vegetative stage (Sc, Fc, Dv; Fig. 2) . Cultivars that interacted positively with the grain-filling stress environment were upland japonica types, including Dinorado (24), IAC25 (29), and IR63371 (47). The second principal component axis accounted for 26% of the variation, and it separated the Di and WS environments from those with stress. The second principal component axis was negatively correlated with root pressure measured in WS (Table 4).

View larger version (19K): [in this window] [in a new window]   Fig. 2. Biplot of the first two ordination scores for standardized number of spikelets per panicle (SPP) for diverse rice cultivars in six managed environments. The environment vectors are labeled as in Table 1. The codes for cultivars are from Table 2. PCA = principal component axis. Group S1, comprising Cultivars 7, 17, 25, 36, 44, and 48, clusters at the origin and has been omitted.

View larger version (20K): [in this window] [in a new window]   Fig. 3. Biplot of the first two ordination scores for standardized fraction of fertile spikelets for diverse rice cultivars in six managed environments. The environment vectors are labeled as in Table 1. The codes for cultivars are from Table 2. PCA = principal component axis.

View larger version (21K): [in this window] [in a new window]   Fig. 4. Biplot of the first two ordination scores for standardized thousand-grain weight (TGW) for diverse rice cultivars in eight managed environments. The environment vectors are labeled as in Table 1. The codes for cultivars are from Table 2. PCA = principal component axis.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

It is clear that anthesis date, unlike some other traits discussed here, is never neutral; it confers either an advantage or a disadvantage in different water stress scenarios. This is unfortunate because timing of stress in rainfed environments is uncertain, and cultivars can only be tailored to the environment in a general sense. For yield in this set of environments, maturity was a major factor that influenced how cultivars performed in wet environments compared with environments with moist soil but a confined wetting front (Di or Dv), as indicated by the strong correlation between PCA1 score and maturity. In contrast, the PCA that separated wetter environments from those with stress at flowering or continuous stress (PCA2) was not related to maturity. With this type of stress, positive interactions were associated with other adaptive mechanisms, such as greater root length, greater sap production under stress, reduced flowering delay, and specific patterns of change in leaf fresh weight across environments. These results allow the development of testable hypotheses about mechanisms that confer tolerance in different types of stress environments.

Root Pressure
Sap exudation from an excised tiller has been used as an indication of root activity (Hirasawa et al., 1992). Root pressure may also play an important role in refilling cavitated xylem vessels. This trait exhibits large variation in the field, and it is difficult to consistently demonstrate varietal differences. Nonetheless, interaction groups differed significantly for root pressure. Cultivars that interacted positively with wet environments consistently produced little sap (Tables 4, 5). This may indicate a cost to high root pressure in more favorable environments, or it may be an artifact of the greater yield potential but poor stability of full-season indica cultivars (also noted in IRRI, 1989), which coincidentally produced little sap. Tentative QTLs for root pressure have been identified in a rice mapping population (Lafitte and Courtois, 2000). It remains to be seen whether sap per tiller or sap per square meter is the most meaningful expression of root pressure.

Leaf Water Content and Specific Leaf Area
The %FW reflects leaf density, which is a function of internal leaf structure (Garnier and Laurent, 1994). Low %FW is generally associated with a large structural component (small cells or more supporting tissue) or a large volume of intercellular air spaces in the mesophyll (Niinemets, 1999). A relationship with %FW was seen in interactions for yield, spikelet fertility, and TGW. The ranking of %FW was not, however, consistent across the four environments where it was measured (Di, Dv, WS, and Fc). Groups Y4 and Y7 had low %FW in both Fc and WS, and %FW in drip-irrigated environments was similar to that in WS for those groups. The groups that interacted positively with only wet environments, Groups Y2 and Y9, had a unique pattern of %FW in that they had low %FW in drip environments but higher than average values in Fc and WS. The change in %FW for a cultivar across environments may be predictive of yield response to stress. Leaf growth in cultivars with a strong reduction in %FW in drip environments relative to WS may be sensitive to soil impedance (Young et al., 1997). If this is due to root signals, the same signals could influence reproductive development as well. Leaf growth in other cultivars may be more directly dependent on turgor, which has less of an effect on the hydraulically protected panicle. Tentative QTLs have been identified for %FW in rice, and one was found to cosegregate with a QTL for the decrease in spikelet fertility with water deficit (Lafitte and Courtois, 2000).

Specific leaf area depends on both leaf density and leaf thickness. These two components can vary independently. In this study, specific leaf area varied across isozyme groups, and it was weakly correlated with PCA2 for yield, but it was not identified as a trait that differed across interaction groups for yield or yield components.

Leaf Water Status and Canopy Temperature
Relative water content under stress was not significantly correlated with PCAs for yield interactions, but was correlated for spikelet fertility. In addition, Group Y3, which interacted positively with the vegetative stress environment for yield, had high RWC after 3 wk of vegetative-stage stress. The relationship between performance in other stress environments and RWC measured during the stress period was not clear.

Rice leaves roll readily under water deficit, and rolling has been used as an indication of a cultivar's ability to continue to maintain a favorable water status under stress. Leaf rolling is easily scored, highly heritable, and has a clear genetic basis (Price and Courtois, 1999). It also differs strongly among isozyme groups (Table 8) , and isozyme Group 6 cultivars roll at higher leaf water potentials than isozyme Group 1 cultivars (Dingkuhn et al., 1989). In this study, cultivars that interacted positively with stress environments were characterized by minimum leaf rolling. High rolling scores were found for cultivars that interacted positively with the wetter environments.

Canopy temperature is an indicator of crop transpiration, with higher temperatures indicating stomatal closure. We found that cultivars having positive interactions with stress environments tended to have high CT measured in a continuous stress environment. This result differs from the conclusion made on the basis of main effects alone, where lines with low canopy temperatures were found to produce greater yields under stress (Garrity and O'Toole, 1995). In some cases, CT can reflect the ability of a line to access soil moisture, but in continuous stress environments it may indicate the capacity to limit water loss from leaves. In this case, high CT may indicate that the plant more effectively controlled water loss through stomatal closure and limited nonstomatal water loss. Alternatively, this result may simply indicate that other processes, such as the ability to maintain the grain sink under stress, were more important than continued transpiration in this set of environments.

Epidermal Conductance
Rice has a very high epidermal conductance and this has been suggested as an important trait to improve for water-limited environments (Nguyen et al., 1997). The mean epidermal conductance measured in this set of cultivars was 0.32 mm s^-1 in Fc and 0.25 mm s^-1 in WS, indicating significant nonstomatal water loss. Values reported for rice leaves where the stomata were closed by placing them in a pure CO₂ atmosphere ranged from 0.13 to 0.25 mm s^-1 (O'Toole et al., 1979). The values calculated here from drying leaves may include a residual loss through incompletely closed stomata. Epidermal conductance was found to differ among interaction groups, and high conductance was noted for two cultivars with positive interactions for spikelet fertility in the continuous stress environment and for groups with positive C x E interactions for yield in Di and Dv (Table 5).

Root Traits
Thick roots and deep rooting have been emphasized as important adaptations to stress in rice (Nguyen et al., 1997; Table 5). Our results indicate that large maximum root length measured in a greenhouse study was associated with positive interactions with severe stress environments. This effect was primarily seen for yield but not for individual yield components. In some cases, cultivars that behaved quite similarly in terms of interaction with environments differed significantly for root length. Maximum root length is strongly negatively associated with tiller number in rice (Yoshida et al., 1982), and the influence on yield may have been through this component. Root thickness was not found to be important in this study.

Effects of Environment and Yield Component Compensation
The WS environment and the Di environment were often more contrasting than other pairs of environments, even though their mean yields were similar. The Di environment resulted in substantial changes in ranking of cultivars relative to the WS, as indicated by the low correlation between the two environments (Table 2). Di was characterized by ample soil moisture within the zone around the emitter, but very hard dry soil outside that zone, and also by comparatively high vapor pressure deficit. The cultivars that could exploit this environment were mostly early cultivars from eastern India. The interaction between cultivar and environment for spikelet fertility was similar in Di and continuous stress environments, indicating that Di did experience stress. These results are consistent with a role of root-sourced signals on fertilization or early embryo growth. Spikelet fertility in rice is reported to be sensitive to nonhydraulic root signals (Kobata et al., 1994), and genetic differences in root-sourced abscisic acid have been reported (Asch et al., 1995). The cultivars with similar maturity that could or could not exploit the Di environment would constitute a useful set of contrasts for studies on root signaling or VPD sensitivity in rice. This environment is different than the continuous stress environments and also than the furrow-irrigated environment. Further data from multilocation trials are required to identify which managed environment is most predictive of the actual rainfed target environments where upland rice is grown.

Cultivars that interacted positively with the continuous stress environments interacted negatively with the higher yield potential environments, except where the stress coincided with anthesis. Conclusions about interactions with the Da environment are also complicated because that environment had the weakest interaction with genotype. This is consistent with the observation that rice is extremely sensitive to stress at this stage (Garrity and O'Toole, 1994), with limited genetic variation among cultivars. The stress environments other than Sv interacted similarly with cultivars for grain yield.

Environment groupings differed for yield relative to yield components, indicating yield component compensation. Similar patterns of interaction for the number of spikelets per panicle were found in Fc, Sc, and Dv. This is consistent with a reduction in the number of potential grains when stress occurs before flowering, but Sv was not included in this group. For spikelet fertility, the grain-filling stress treatment interacted differently with cultivar than did other drip-irrigated environments, suggesting that superior grain set with terminal stress may rely on different mechanisms than other stress environments. While the Dg treatment was designed to influence only TGW, it also had a significant effect on spikelet fertility. Dv interaction with cultivars was different than in Dg for spikelets per panicle, but the two environments interacted similarly with cultivars for TGW. The TGW of Da was expected to recover when water was reapplied, but it is possible that the endosperm cell number was largely determined by that time or that photosynthetic capacity was damaged, and TGW was unable to recover.

Some interaction groupings identified in the pattern analysis were consistent with adaptation or isozyme groupings, but others were not. The set we tested was a mixture of traditional and improved cultivars, and some of the latter included between-group hybridizations. Although all could be assigned to an isozyme group based on five loci, many improved cultivars had incorporated alleles from other groups, which could result in a blurred structure. Isozyme Group 2 had consistently positive interactions with the Di, Dv, Dg, and continuous stress environments. Some early-maturing Group 1 cultivars also responded in this way. Other factors that result in differential yield of isozyme groups in the WS vs. the DS are lodging and disease. Group 2 cultivars generally respond poorly to inputs, and lodging was noted in several plots. There were not, however, apparent correlations between performance and lodging scores or observations of disease.

	CONCLUSIONS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
Cultivars that tended to interact positively with stress environments had root traits associated with stress avoidance (long roots, high root pressure with stress) and leaf traits that were not stress-avoiding (little rolling and high CT). Positive interaction between cultivars and wetter environments was associated with traits that have been considered drought avoidance strategies, such as leaf rolling and anthesis delay, but these pessimistic strategies resulted in negative interactions with stress environments. Some traits seem to confer an advantage in stress environments without a related cost in wet environments. Among the traits measured, maturity and, to a lesser extent, %FW patterns and root pressure, appear to provide information on the ability of a cultivar to tolerate stress. New germplasm resources, particularly near-isogenic lines and mutants, will be needed to further dissect the value of specific traits independent of differences in maturity and genetic background.

	ACKNOWLEDGMENTS

 
We thank L. Holongbayan, N. Turingan, and N. Trillana for technical assistance, and G. McLaren for helpful comments on the manuscript. We are grateful to three anonymous reviewers for their suggestions.

Received for publication June 21, 2001.

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 

• Asch, F., K. Dorffling, and M. Dingkuhn. 1995. Response of rice varieties to soil salinity and air humidity: a possible involvement of root-borne ABA. Plant Soil 177:11–19.
• Bidinger, F.R., V. Mahalakshmi, and P. Rao. 1987. Assessment of drought resistance in pearl millet [Pennisetum americanum (L.) Leeke]. II. Estimation of genotype response to stress. Aust. J. Agric. Res. 38:49–59.
• Blum, A. 1996. Crop responses to drought and the interpretation of adaptation. Plant Growth Regul. 20:135–148.
• Bois, J.F., P. Couchat, and P. Moutonnet. 1984. Water stress response for some cultivars of upland and lowland rice. I. Effects on transpiration. Plant Soil 80:227–236.
• Cape, J.N., and K.E. Percy. 1996. The interpretation of leaf drying curves. Plant Cell Environ. 19:356–361.
• Catsky, J. 1974. Water content. p. 121–156. In J. Jacobs et al. (ed.) Ecological studies 9: Methods of studying plant water relations. Springer-Verlag, New York.
• Chang, T.T. 1976. The rice cultures. Philos. Trans. R. Soc. London, Ser. B. 275:143–157.
• Courtois, B., W. Chaitep, S. Moolsri, K. Prasad, P.K. Sinha, G.Trebiuil, and R. Yadav. 1996. Drought resistance and germplasm improvement: ongoing research of the Upland Rice Consortium. p. 154–175. In C. Piggin et al. (ed.) Upland rice research in partnership. IRRI Discussion Paper Ser. No. 16. IRRI, Manila, The Philippines.
• Dingkuhn, M., S.K. De Datta, K. Dorffling,, and C. Javellana. 1989. Varietal differences in leaf water potential, leaf net CO₂ assimilation, conductivity and water use efficiency in upland rice. Aust. J. Agric. Res. 40:1183–1192.
• Fukai, S., and M. Cooper. 1995. Development of drought-resistant cultivars using morpho-physiological traits in rice. Field Crops Res. 49:67–86.
• Garnier, E., and G. Laurent. 1994. Leaf anatomy, specific leaf mass and water content in congeneric annual and perennial grass species. New Phytol. 128:725–736.
• Garrity, D.P., and J.C. O'Toole. 1994. Screening rice for drought resistance at the reproductive phase. Field Crops Res. 39:99–110.
• Garrity, D.P., and J.C. O'Toole. 1995. Selection for reproductive stage drought avoidance in rice using infrared thermometry. Agron. J. 87:773–779.[Abstract/Free Full Text]
• Glazmann, J.C. 1987. Isozymes and classification of Asian rice varieties. Theor. Appl. Genet. 74:21–30.[ISI]
• Hirasawa, T., M. Tsuchida, and K. Ishihara. 1992. Relationship between resistance to water transport and exudation rate and the effect of the resistance on the midday depression of stomatal aperture in rice plants. Jpn. J. Crop Sci. 61:145–152.
• Ingram, K.T., J.G. Real, M.A. Maguiling, M.A. Obien, and G.C. Loresto. 1990. Comparison of selection indices to screen lowland rice for drought resistance. Euphytica 48:253–260.
• IRRI. 1989. Program report for 1988. IRRI, Los Baños, The Philippines.
• IRRI. 1993. Program report for 1992. IRRI, Los Baños, The Philippines.
• IRRI. 1996. Standard Evaluation System for rice. 4th ed. Int. Network for Genetic Evaluation of Rice, Genetic Resource Center, Los Baños, The Philippines.
• IRRI. 2000. IRRISTAT for Windows. Version 4.02.b. IRRI, Metro Manila, The Philippines.
• Kobata, T., S. Tanaka, and M. Utumi. 1994. Sterility in rice (Oryza Sativa L.) subject to drought during the booting stage occurs not because of lack of assimilate or of water deficit in the shoot but because of dehydration of the root zone. Jpn. J. Crop Sci. 63:510–517.
• Lafitte, H.R., M.C. Champoux, G. McLaren, and J.C. O'Toole. 2001. Rice root morphological traits are related to isozyme group and adaptation. Field Crops Res. 73:1–14.
• Lafitte, H.R., and B. Courtois. 2000. Genetic variation in performance under reproductive stage water deficit in a doubled-haploid rice population in upland fields. p. 97–102. In J.M. Ribaut and D. Poland (ed.) Molecular approaches for the genetic improvement of cereals for stable production in water-limited environments. A strategic planning worksh. held at CIMMYT, El Batan, Mexico. 21–25 June 1999. CIMMYT, Mexico City.
• Lilley, J.M., and M.M. Ludlow. 1996. Expression of osmotic adjustment and dehydration tolerance diverse rice lines. Field Crops Res. 48:185–190.
• Ludlow, M.M., and R.C. Muchow. 1990. A critical evaluation of traits for improving crop yields in water-limited environments. Adv. Agron. 43:107–153.
• McLaren, C.G. 1996. Methods of data standardization used in pattern analysis and AMMI models for the analysis of international multi-environment variety trials. p. 225–242. In M. Cooper and G.T.L. Hammer (ed.) Plant adaptation and crop improvement. CAB Int., Cambridge.
• Mitchell, J.H., D. Siamhan, M.H. Wamala, J.B. Risimeri, E. Chinyamakobvu, S.A. Henderson, and S. Fukai. 1998. The use of seedling leaf death score for evaluation of drought resistance of rice. Field Crops Res. 55:129–139.
• Nguyen, H.T., R.C. Babu, and A. Blum. 1997. Breeding for drought resistance in rice: physiology and molecular genetics considerations. Crop Sci. 37:1426–1434.[Abstract/Free Full Text]
• Niinemets, U. 1999. Components of leaf dry mass per area—thickness and density—alter leaf photosynthetic capacity in reverse directions in woody plants. New Phytol. 144:35–47.
• O'Toole, J.C., and W.L. Bland. 1987. Genotypic variation in crop plant root systems. Adv. Agron. 41:91–145.
• O'Toole, J.C., R.T. Cruz, and J.N. Seiber. 1979. Epicuticular wax and cuticular resistance in rice. Physiol. Plant. 47:239–244.
• Price, A., and B. Courtois. 1999. Mapping QTLs associated with drought resistance in rice: Progress, problems and prospects. Plant Growth Regul. 29:123–133.
• Richards, R.A. 1996. Defining selection criteria to improve yield under drought. Plant Growth Regul. 20:157–166.
• SAS Institute. 1992. SAS/STAT user's guide, version 6. 4th ed. Vol. 1 and 2. SAS Inst., Cary, N.C.
• Squire, G.R. 1990. The physiology of tropical crop production. CAB Int., Oxon, UK.
• Yan, W., and L.A. Hunt. 2001. Interpretation of genotype x environment interaction for winter wheat yield in Ontario. Crop Sci. 41:19–25.[Abstract/Free Full Text]
• Yoshida, S., D.P. Bhattacharjee, and G.S. Cabuslay. 1982. Relationship between plant type and root growth in rice. Soil Sci. Plant Nutr. 28:473–482.
• Young, I.M., K. Montagu, J. Conroy, and A.G. Bengough. 1997. Mechanical impedance of root growth directly reduces leaf elongation rates of cereals. New Phytol. 135:613–619.

Related articles in Crop Science:

This issue in Crop science

Crop Science 2002 42: 1393-1395. [Full Text]  

This article has been cited by other articles:

				V. Stiller, H. R. Lafitte, and J. S. Sperry Hydraulic Properties of Rice and the Response of Gas Exchange to Water Stress Plant Physiology, July 1, 2003; 132(3): 1698 - 1706. [Abstract] [Full Text] [PDF]

This Article Abstract Figures Only Full Text (PDF) Free Alert me when this article is cited Alert me if a correction is posted Services Related articles in Crop Science Similar articles in this journal Similar articles in ISI Web of Science Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (12) Citing Articles via Google Scholar Google Scholar Articles by Lafitte, H. R. Articles by Courtois, B. Search for Related Content PubMed Articles by Lafitte, H. R. Articles by Courtois, B. Agricola Articles by Lafitte, H. R. Articles by Courtois, B. Related Collections Water Stress Rice Plant and Environment Interactions