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 Similar articles in this journal Similar articles in Web of Science Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Web of Science (14) Citing Articles via Google Scholar Google Scholar Articles by Collaku, A. Articles by Harrison, S. A. Search for Related Content PubMed Articles by Collaku, A. Articles by Harrison, S. A. Agricola Articles by Collaku, A. Articles by Harrison, S. A. Related Collections Water Stress Crop Physiology & Metabolism Wheat

CROP ECOLOGY, PRODUCTION & MANAGEMENT

Losses in Wheat Due to Waterlogging

^a Pennington BRC, Louisiana State Univ., 6400 Perkins Rd., Baton Rouge, LA 70808
^b Dep. of Agronomy, Louisiana Agric. Exp. Stn.(LAES), Louisiana State Univ. Agricultural Center, Baton Rouge, LA 70803

* Corresponding author (collaka{at}pbrc.edu)

	ABSTRACT

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION REFERENCES

 
Waterlogging stress is one of the limiting factors influencing wheat (Triticum aestivum L.) production, especially in the lower Mississippi valley. A rain-shelter experiment was designed to evaluate the trend response of nine wheat genotypes to four levels of waterlogging treatment: 0, 10, 20, and 30 d of flooding. Genotypes planted in polyvinilchloride (PVC) containers 25 cm long by 10 cm in diameter were waterlogged in plastic tanks under controlled rain conditions. Results indicated significant linear responses for kernels per head and tillers per plant, significant linear and quadratic responses for yield and chlorophyll content, and significant linear and cubic responses for plant height. The linear trend was the most important component, explaining from 92 to 99% of the variability due to waterlogging. Linear prediction equations were obtained to describe the relationship between different traits and waterlogging stress. Losses in yield and yield components were evaluated in a field experiment with 15 genotypes under control and waterlogging treatment. Average yield losses of 44% were mainly caused by a decrease in tiller number and kernels per head. Under waterlogging treatment, tiller number and kernels per head were reduced by 41 and 20%, respectively. Screening of wheat genotypes revealed the potential for waterlogging tolerance in breeding material and identified tolerant cultivars useful for waterlogged environments. ‘Terral LA 422’, ‘Shelby’, and ‘Pioneer 2691’ were the most adapted genotypes for waterlogging treatments. Because of a significant interaction with waterlogging treatment, some of the high-yielding genotypes under non-flooded conditions such as ‘Coker 9663’ and ‘FFR 502W’ showed low tolerance to waterlogging. The results provided information on the methods quantifying losses from waterlogging and identified selection criteria for waterlogging tolerance in wheat.

	INTRODUCTION

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION REFERENCES

 
WHEAT ALONG THE GULF COAST and many other regions is frequently subjected to waterlogging because of heavy rainfall, level topography, and/or inadequate soil drainage. Boyer (1982) estimated that 12% of U.S. agricultural soils are affected by waterlogging. In Louisiana, waterlogging decreases wheat yield, especially during the first months of growth (Musgrave, 1994). Waterlogging can reduce grain yield of winter wheat by about 20 to 50% (Belford, 1981; Cannell et al., 1984; Musgrave and Ding, 1998).

Oxygen deficiency caused by waterlogging reduces shoot and root growth, dry matter accumulation, and final yield. Waterlogging can effect several physiological processes, such as absorption of water (Drew, 1991), root and shoot hormone relations (Huang et al., 1994), and decrease the uptake and transport of ions through roots causing nutrient deficiencies (Trought and Drew, 1980a, b; Hodgson et al., 1989; Huang et al., 1995). Negative effects associated with waterlogging are nitrogen deficiency by stimulating denitrification and leaching (Hodgson et al., 1989; Huang et al., 1994) and accumulation of toxic substances (Ponamperuma, 1984; Huang et al., 1994). In cereals, waterlogging reduces leaf elongation, kernel number, and final yield (Luxomore et al., 1973; Gardner and Flood 1993; Musgrave and Ding, 1998).

In the absence of direct and reliable selection markers for waterlogging tolerance, a successful wheat breeding program requires the identification of physiological and morphological traits that are influenced the most by waterlogging and the existence of useful genetic variability controlling their expression. Some evidence of genotypic differences in tolerance to waterlogging exists in wheat. Davies and Hillman (1988) demonstrated variation in vegetative growth and yield under continuous flooding of various wheat species. Van Ginkel et al. (1991) identified 14 waterlogging-tolerant spring wheat lines. Using a 5-wk waterlogging treatment, Sayre et al. (1994) identified six tolerant genotypes in terms of number of tillers, leaf chlorosis, senescence, fertility, grain yield, and kernel weight. Other varietal studies for waterlogging stress in wheat have been reported (Thompson et al., 1992; Musgrave and Ding, 1998).

Losses caused by waterlogging need to be measured to determine the importance of traits used as waterlogging tolerance indicators. The objectives of this study were to determine the form of response to different levels of waterlogging of several quantitative traits of wheat, including yield, in a rain shelter experiment, to estimate losses from waterlogging in a field experiment, and to evaluate tolerance to waterlogging stress of wheat genotypes common to the Gulf Coast region. Both experiments provided means for quantifying losses from waterlogging and identifying traits to be used as selection criteria in breeding for waterlogging tolerance in wheat.

	MATERIALS AND METHODS

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION REFERENCES

 
Shelter Experiment
This experiment was conducted in a rain shelter to avoid the influence of rainfall on waterlogged treatments. Four periods of waterlogging were applied as treatments: 0 (control), 10, 20, and 30 d. Genotypes included were ‘Pioneer 2643’, Pioneer 2691, ‘LA 87167’, ‘Savannah’, Terral LA 422, Coker 9663, ‘Florida 304’, ‘FFR502W’, and ‘Jaypee’.

Seeds from each genotype were planted in a PVC (SC-10) cone-tainer (Stuewe & Sons, Inc., Corvallis, OR) (25 cm long by 10-cm diam) containing 0.55 kg of Commerce silt loam soil (fine-silty, mixed, nonacid, thermic, Aeric Fluvaquent) taken from the LAES Central Station at Ben Hur Research Farm of the Louisiana State University Agricultural Center at Baton Rouge. An equivalent amount of 90 kg N ha^-1 urea was added to the soil of each cone-tainer. The cone-tainers were held inside a plastic tank (140 cm long by 50 cm wide by 30 cm high). After germination, seedlings were thinned to one plant per cone-tainer. The plants were grown in the greenhouse for 4 wk and then transferred to a rain shelter, under controlled rainfall, but otherwise under similar field weather conditions. Plants were allowed to acclimate 1 wk before starting the waterlogging treatment. Depending on the genotype, waterlogging treatment started at 3- to 4- leaf stage, and was accomplished by raising the level of water in the tanks to the surface of the PVC cone-tainers. Each period of waterlogging was followed by a 2-d period of completely drained tanks and application of the equivalent of 30 kg ha^-1 N as urea. The experimental design was a split-plot design with three replications. Each waterlogging level was randomly assigned to three of the 12 tanks according to a completely randomized design, and wheat genotypes were randomly assigned to one cone-tainer per tank each according to a randomized complete block design (RCBD).

During 1997-1998 and 1998-1999, measurements for each plant were taken for chlorophyll content, tillers per plant, plant height, kernels per spike, and yield. Measurements on chlorophyll content were taken four times, starting before applying waterlogging and at 10-d intervals. Three to four readings were taken for each measurement, and the mean was used for the data analysis. A SPAD meter (Model 502, Minolta Corp., Ramsey, NJ) was used to measure chlorophyll content.

Field Experiment
To evaluate losses from waterlogging under field conditions, 15 wheat genotypes were grown under no waterlogging (control) and 5 wk of continuous flooding, starting at the 3- to 4-leaf stage. The experimental design was a split-plot with two replications in the first year, and three replications in the second and third year of the study. Waterlogging treatments were assigned to the main plots according to a RCBD and genotypes were assigned to the subplots according to a RCBD. The 15 wheat genotypes represent soft red winter wheat genotypes that are important along the Gulf Coast, are parental components in the LAES wheat breeding program, and have shown good performance in wheat trials (Harrison et al., 1997). Plots consisted of six rows 1 m long and 20 cm apart. Soil type was a Commerce silt loam (fine-silty, mixed, nonacid, thermic, Aeric Fluvaquent). Pre-plant fertilizer at a rate of 240 kg ha^-1 of N-P-K and Glean (clorsulfurun) herbicide at 200 L ha^-1 were applied. Levees constructed for each main plot were 30 to 40 cm in height. Waterlogging was accomplished by pumping water from a nearby water service and flooding the plots assigned to the waterlogging treatment. The soil was kept saturated with water above the field capacity by continuous flooding, usually every day to create an oxygen deficience environment. A top dressing of 90 kg-ha^-1 N as ammonium nitrate was applied immediately after the termination of waterlogging treatment.

Soil oxygen content was measured with gas samplers constructed of porous bronze cups attached to sampling taps. Gas samplers were buried in the soil between rows at a depth of 10 cm in every main plot. Measurements on redox potential were taken weekly during the waterlogging period. Gas samples of 20 mL were withdrawn and oxygen concentration was measured as described by Musgrave and Ding (1998).

Data on chlorophyll content, plant height, and kernels per spike were taken on three plants per plot, and the means per plot were used for data analysis. Tiller number was measured before harvesting by counting the number of spikes in 20 cm of an interior row in each plot, for the third year of the study only. Kernel weight was determined from a 100-seed sample from each plot. Yield was measured by hand harvesting and threshing two interior rows per plot. This experiment was conducted for three years: 1997-1998, 1998-1999, and 1999-2000, at LAES Central Station of Louisiana State University Agricultural Center at Baton Rouge.

Statistical Analysis
Data on chlorophyll content for the shelter experiment were analyzed as repeated measures design in a split-plot arrangement, with measurement as a third factor not randomly assigned. The univariate model for chlorophyll data was:

where µ is the overall mean effect, w_i is the effect of the ith waterlogging treatment (i = 1 to i = 4), g_j is the effect for the jth genotype (j = 1 to j = 9), m_l is the effect of time of measurement (l = 1 to l = 4), (wg)_ij is the interaction of the ith waterlogging level with the jth genotype, (wm)_il is the interaction effect of the ith treatment with the lth measurement, and e_ijl is the random error among measurements across the plants. For hypothesis testing, e_ijl is assumed to be normally distributed with mean zero, variance ², and a covariance structure that satisfies the sphericity requirement of the univariate approach to repeated measures (Huynh and Feldt, 1970). Genotypes were considered fixed for all the statistical analysis performed. A univariate model was used to study the trend response of chlorophyll content to waterlogging treatments for the linear, quadratic, and cubic component. Data were unbalanced, and the analyses were performed in PROC MIXED of SAS, ver. 6.12 (SAS Institute Inc., 1996).

Data for yield, tillers per plant, plant height, and kernels per spike were analyzed according to a split-plot design (Steel and Torrie, 1981; Hinkelmann and Kempthorne, 1994). A trend analysis based on orthogonal polynomials was performed to determine the form of response of traits to waterlogging.

Data of field experiment were analyzed as a split-plot design according to:

where y_i is the effect of the ith year, r(y)_l(i) is the effect of the lth replication within the ith year, w_j is the effect of the jth level of watterlogging; [wr(y)]_jl(i) is the interaction effect of jth waterlogging level with the lth replication within a given year, representing the error for waterlogging effect; g_k is the effect of the kth genotype, (yw)_ij is the year by waterelogging interaction effect; (yg)_ik is the year by genotype interaction effect; (wg)_jk is the genotype by waterlogging interaction effect; (ywg)_ijk is the interaction effect of year by waterlogging by genotype; and e_ijklm is the random error for genotypes and interactions of first and second order that involve genotype, waterlogging and year effects. Data for this experiment were balanced. A fixed model was assumed for waterlogging, genotypes, and years. The analyses were performed in PROC GLM, SAS, ver. 6.12 (SAS Inc., 1996).

	RESULTS AND DISCUSSION

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION REFERENCES

 
Trend Response to Waterlogging
The average chlorophyll content of genotypes decreased with time for both years of study (Table 1). Significant reduction in chlorophyll content from waterlogging has been previously reported (Huang et al., 1994). Decrease in chlorophyll content as a result of waterlogging reduces photosynthesis in wheat with a significant effect on yield (Drew, 1991). The univariate corrections using Greenhouse-Geisser and Huynh-Feldt gave the same probability values, suggesting that the univariate model was a good approach for the repeated measures data (Moser et al., 1990). Duration of waterlogging from 0 to 30 d had a significant effect on the average chlorophyll content of genotypes. There was a significant linear reduction in chlorophyll content with longer period of waterlogging for both years of study (Table 1). The quadratic component was significant for the first year, and the cubic component was significant for the second year of study, but their P values were smaller than P values for the linear trend.

A significant linear response to waterlogging was observed for kernels per spike and tillers per plant, a significant linear and quadratic response was observed for grain yield per plant, and a significant linear and cubic response was observed for height (Table 2). The linear trend was the most important component, explaining from 92 to 99% of the variability caused by waterlogging. This result was confirmed by regression analysis, which showed that linear was the main trend describing the response to waterlogging of grain yield per plant, kernels per spike, tillers per plant, and plant height (Table 3). Both parameters of linear regression were significantly different from zero for all the traits. Values of regression coefficients significantly different from zero were occasionally observed in the quadratic and cubic model. Nevertheless, these models explained only a small part of the variability, which ranged from 0.5 to 8% (Table 2).

View larger version (20K): [in this window] [in a new window]   Fig. 1. Relationship of wheat genotypes to four levels of waterlogging for (a) yield, (b) kernels per head, (c) tillers per plant, and (d) height in a rain-shelter experiment.

Evaluating Losses from Waterlogging
A field experiment under waterlogging treatment was used to evaluate losses in yield and yield components. Redox potential data for the three years of this experiment ranged from 290 to 337 mV for the flooded plots, showing a high degree of waterlogging stress, as compared to the control, which ranged from 548 to 617 mV (Table 4).

Screening of Wheat Genotypes for Waterlogging Tolerance
Genotypes exhibited a differential yield response to waterlogging stress. Although waterlogging significantly reduced yield for most of the genotypes, their respective response was very different, ranging from 34% for Pioneer 2691 to 60% for FFR 502W (Table 6). Mean yield of Pioneer 2691, Terral LA 422, Shelby, and ‘Pioneer 2684’ were significantly higher than other genotypes under waterlogging treatment. The four genotypes were ranked in the highest-yielding group under waterlogging stress (group ab, Table 6). High yield response of Terral LA 422 was mainly because of the ability to maintain a high number of kernels per spike and a high number of tillers. Under waterlogging, this cultivar was ranked among genotypes with the highest number of kernels per spike and tillers per plant (group ab, Table 6). Similarly, high yield of Shelby was due to high kernels per spike and high kernel weight. This cultivar had the highest mean of 35 kernels per spike (Table 6) and for kernel weight was classified in the ab group with a mean of 3.55 g 100^-1 kernels (Table 6). Genotypes ‘Roberts’ and Jaypee had the smallest tiller number reduction from waterlogging. They were ranked in the highest group for this trait under waterlogging conditions but their yield performance was low as a result of a poor response to waterlogginig for other yield components.

A few genotypes such as ‘AR-584A-3-2’, Roberts, and Pioneer 2643 had little yield loss from waterlogging, ranging from 15 to 27%, but their yield performance was low. These three genotypes were classified in the low yielding group, in both waterlogged and control conditions (Table 6).

Yield of Coker 9663, FFR 502W, and ‘Mason’ were depressed most drastically by waterlogging (Table 6). In control conditions, they were ranked among the highest yield genotypes (group ab). While under waterlogging treatment, they were classified in the low yield groups (bcd). Their yield reduction from waterlogging ranged from 3.3 Mg ha^-1 (57%) for Coker 9663 to 2.8 Mg ha^-1 (51%) for Mason (Table 6). Although under normal conditions, these cultivars belong to the group of high yield wheat genotypes, their economic importance is greatly reduced under waterlogging stress.

Waterlogging tolerance is closely related to the way genotypes interact with waterlogging treatment. In this study, the interaction of genotypes with waterlogging was significant for yield, and for some of the yield components such as tiller number, kernels per spike and kernel weight (Table 5). In terms of tolerance to waterlogging, the most tolerant genotypes are those with the smallest magnitude of interaction with waterlogging treatment. A larger yield reduction of Coker 9663 or FFR 502W was caused by their higher interaction with waterlogging treatment as compared to that of Terral LA 422, or Pioneer 2691 (Fig. 2) . Genotypes such as Terral LA 422 which interacts with waterlogging treatment but maintains a high yield performance, would be a reliable choice under waterlogging stressful environments. Genotypes such as Coker 9663 or FFR 502W may be a good choice for normal growing conditions but their placement in environments with expected waterlogging stress would not be prudent. Genotypes with the smallest interaction with waterlogging treatment such as ‘AR-584A’, ‘GA 87139’, and Pioneer 2643 are waterlogging tolerant. For their low yield performance under control and waterlogging conditions, these genotypes represent little interest as production varieties, but, for their tolerance, such genotypes may be valuable in breeding programs that include waterlogging tolerance.

View larger version (20K): [in this window] [in a new window]   Fig. 2. Yield response to waterlogging of Pioneer 2691, Terral LA 422, Coker 9663, AR-584A, and FFR 502W.

	ACKNOWLEDGMENTS

 
We thank Drs. G.O. Myers, B. Venuto and P. Bell for their helpful comments and suggestions on the manuscript.

	NOTES

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION REFERENCES

 
Approved for publication by the Director of the LAES as manuscript no. 01-09-220.

Received for publication March 20, 2001.

	REFERENCES

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION REFERENCES

 

• Belford, R.K. 1981. Response of winter wheat to prolonged waterlogging under outdoor conditions. J. Agric. Sci. (Cambridge) 97:557– 568.
• Boyer, J.S. 1982. Plant productivity and environment. Science 218: 443–448.[Abstract/Free Full Text]
• Cannell, R.Q., R.K. Belford, K. Gales, R.J. Thomson, and C.P. Webster. 1984. Effects of waterlogging and drought on winter wheat and winter barley grown on a clay and a sandy soil. I. Crop growth and yield. Plant Soil 80:53–66.
• Davies, M.S., and G.S. Hillman. 1988. Effect of soil flooding on growth and grain yield of tetraploid and hexaploid species of wheat. Cereal Res. Comm. 12:135–141.
• Drew, M.C. 1991. Oxygen deficiency in the root environment and plant mineral nutrition. p. 301–316. In M.B. Jackson et al (ed.) Plant life under oxygen deprivation. Academic publishing, The Hague.
• Gardner, W.K., and R.G. Flood. 1993. Less waterlogging damage with long season wheats. Cereal Res. Comm. 21:337–343.
• Harrison, S.A., L.P. Brown, D.R. Burns, A. Collaku, P.D. Colyer, R.J. Habetz, W.B. Hallmark, D. Lanclos, H.J. Mascagni, S.H. Moore, J. Rabb, M.E. Reeder, J.S. Russin, C. Seale, A. Shadow, M.L. Tarpley, W. Shelton, and P. Vernon. 1997. Small grain performance trials and research reports. Louisiana Agricultural Experiment Station. Mimeo. Series No. 129. Baton Rouge, LA.
• Hinkelmann, K., and O. Kempthorne. 1994. Designs and analysis of experiments. John Wiley and Sons, Inc., New York.
• Hodgson, D.R., G.M. Whitely, and Anna E. Bradnam. 1989. Effects of waterlogging in the spring on soil conditions and the growth and yield of spring barley in three cultivation systems. J. Agric. Sci. (Cambridge) 112:265–276.
• Huang, B., J.W. Johnson, S. Nesmith, and D.C. Bridges. 1994. Growth, physiological and anatomical responses of two wheat genotypes to waterlogging and nutrient supply. J. Exp. Bot. 45:193–202.[Abstract/Free Full Text]
• Huang, B., J.W. Johnson, S. Nesmith, and D.C. Bridges. 1995. Nutrient accumulation and distribution of wheat genotypes in response to waterlogging and nutrient supply. Plant Soil 173:47–54.
• Huynh, H., and L.S. Feldt. 1970. Conditions under which mean square ratios in repeated measurements design have exact F-distribution. J. Am. Stat. Assoc. 65:1582–1589.
• Luxomore, R.J., R.A. Fisher, and L.H. Stolzy. 1973. Flooding and soil temperature effects on wheat during grain filling. Agron. J. 65:361–364.[Abstract/Free Full Text]
• Moser, E.B., A.M. Saxton, and S.R. Pezeshki. 1990. Repeated measures analysis of variance: Application to a tree research. Can. J. For. Res. 20:524–535.
• Musgrave, M.E. 1994. Waterlogging effects on yield and photosynthesis in eight winter wheat cultivars. Crop Sci. 34:1314–1318.[Abstract/Free Full Text]
• Musgrave, M.E., and N. Ding. 1998. Evaluating wheat cultivars for waterlogging tolerance. Crop Sci. 38:90–97.[Abstract/Free Full Text]
• Ponnamperuma, F.N. 1984. Effects of flooding in soils. p. 9–45. In T.T. Kozlowski (ed.) Flooding and plant growth. Academic Press, London.
• SAS Inst., Inc. 1996. SAS/Proc Mix, Appendix I, 491–504; chapter 4, 135–169; chapter 18, 533–610. User's guide. SAS Inst., Inc., Cary, NC.
• Sayre, K.D., M. van Ginkel, S. Rajaram, and I. Monasterio. 1994. Tolerance to water-logging losses in spring bread wheat, effect of time of onset on expression. Colorado State Univ. In Annual Wheat Newsletter 40:165–171.
• Sharma, D.P., and A. Swarup. 1989. Effect of short-term waterlogging on growth, yield and nutrient composition of wheat in alkaline soils. J. Agric. Sci. (Cambridge) 112:191–197.
• Steel, R.G.D., and J.H. Torrie. 1981. Principles and procedures of statistics. McGraw-Hill Int. Book Company, Berkshire, UK.
• Thompson, C.J., T.D. Colmer, E.L.J. Watkin, and H. Greenway. 1992. Tolerance of wheat (Triticum aestivum cvs. Gamenya and Kite) and triticale (Triticoscecale cv. Muir) to watterlogging. New Phytol. 120:335–344.
• Trought, M.C.T., and M.C. Drew. 1980a.. The development of waterlogging damage in young wheat plants in anaerobic solution cultures. J. Exp. Bot. 31:1573–85.[Abstract/Free Full Text]
• Trought, M.C.T., and M.C. Drew. 1980b. The development of waterlogging damage in wheat seedlings (Triticum aestivum L). Plant Soil 54:77–94.
• Van Ginkel, M., S. Rajaram, and M. Thijssen. 1991. Waterlogging in wheat, germoplasm evaluation and methodology development. p. 115–124. In G.T. Douglas and W. Mwangi (ed.) The seventh regional wheat workshop for Eastern, Central and Southern Africa, Nakuru, Kenya. CIMMYT.

This article has been cited by other articles:

				J. A. Christianson, I. W. Wilson, D. J. Llewellyn, and E. S. Dennis The Low-Oxygen-Induced NAC Domain Transcription Factor ANAC102 Affects Viability of Arabidopsis Seeds following Low-Oxygen Treatment Plant Physiology, April 1, 2009; 149(4): 1724 - 1738. [Abstract] [Full Text] [PDF]

				A. Collaku and S. A. Harrison Heritability of Waterlogging Tolerance in Wheat Crop Sci., February 23, 2005; 45(2): 722 - 727. [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 Similar articles in this journal Similar articles in Web of Science Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Web of Science (14) Citing Articles via Google Scholar Google Scholar Articles by Collaku, A. Articles by Harrison, S. A. Search for Related Content PubMed Articles by Collaku, A. Articles by Harrison, S. A. Agricola Articles by Collaku, A. Articles by Harrison, S. A. Related Collections Water Stress Crop Physiology & Metabolism Wheat