Cuvillier Verlag Göttingen Setegn Gebeyehu Physiological Response to Drought Stress of Common Bean (Phaseolus vulgaris L.) Genotypes Differing in Drought Resistance Institut für Pflanzenernährung Justus-Liebig-Universität Giessen Aus dem Institut für Pflanzenernährung der Justus-Liebig-Universität Giessen Prof. Dr. S. Schubert Physiological Response to Drought Stress of Common Bean (Phaseolus vulgaris L.) Genotypes Differing in Drought Resistance Dissertation zur Erlangung des Grades eines Doktors der Agrarwissenschaften beim Fachbereich 09 Agrarwissenschaften , Ökotrophologie und Umweltmanagement der Justus-Liebig-Universität Giessen vorgelegt von Setegn Gebeyehu aus Guduru (Oromia), Äthiopien Gießen 2006 Bibliografische Information Der Deutschen Bibliothek Die Deutsche Bibliothek verzeichnet diese Publikation in der Deutschen Nationalbibliografie; detaillierte bibliografische Daten sind im Internet über http://dnb.ddb.de abrufbar. 1. Aufl. - Göttingen : Cuvillier, 2006 © CUVILLIER VERLAG, Göttingen 2006 Nonnenstieg 8, 37075 Göttingen Telefon: 0551-54724-0 Telefax: 0551-54724-21 www.cuvillier.de Alle Rechte vorbehalten. Ohne ausdrückliche Genehmigung des Verlages ist es nicht gestattet, das Buch oder Teile daraus auf fotomechanischem Weg (Fotokopie, Mikrokopie) zu vervielfältigen. 1. Auflage, 2006 Gedruckt auf säurefreiem Papier Tag der Disputation: 16.10.2006 Mitglieder der Prüfungskommission: Vorsitzender: Prof. Dr. S. Hoy 1. Gutachter: Prof. Dr. S. Schubert 2. Gutachter: Prof. Dr. Dr. h.c. W. Friedt Prüfer: Prof. Dr. B. Honermeier Prüfer: Prof. Dr. K.H. Mühling Gedruckt mit Unterstützung des Deutschen Akademischen Austauschdienstes (DAAD) Zugl.: Giessen, Univ., Diss., 2006 ISBN 10: 3-86727-038-4 ISBN 13: 978-3-86727-038-0 ISBN 10: 3-86727-038-4 ISBN 13: 978-3-86727-038-0 I TABLE OF CONTENTS ACKNOWLEDGEMENTS .....................................................................................................III LIST OF FIGURES..................................................................................................................IV LIST OF TABLES ................................................................................................................. VII ABBREVIATIONS...............................................................................................................VIII 1. INTRODUCTION..................................................................................................................1 1.1. Mechanisms and traits related to drought resistance in common bean...........................2 1.1.1. Growth, yield and morphological adaptations .........................................................2 1.1.2. Water-use and water-use efficiency (WUE) ............................................................4 1.1.3. Leaf-water relations and gas-exchange ....................................................................4 1.2. Assimilate metabolism in source and sink organs under drought stress.........................6 1.3. Protein changes in response to drought stress.................................................................8 1.4. Underlying hypotheses and objectives of the study........................................................9 2. MATERIALS AND METHODS.........................................................................................11 2.1. Genotypes......................................................................................................................11 2.2. Plant cultivation.............................................................................................................11 2.3. Experimental procedure ................................................................................................12 2.4. Sample collection and parameters measured ................................................................13 2.4.1. Biomass and seed yield ..........................................................................................13 2.4.2. Plant water-use and leaf-water relations ................................................................14 2.4.3. Growth analysis ......................................................................................................15 2.4.4. Photosynthetic parameters......................................................................................16 2.4.5. Chemical analysis...................................................................................................17 2.5. Determination of the numbers and sizes of cotyledonary cells and amyloplasts..........22 2.6. Proteomic analysis.........................................................................................................24 2.6.1. Protein preparation .................................................................................................24 2.6.2. 2D gel electrophoresis ............................................................................................24 2.6.3. Staining and computer analysis ..............................................................................25 2.7. Data analysis .................................................................................................................26 3. RESULTS ............................................................................................................................27 II 3.1. Effects on seed yield and yield components .................................................................27 3.2. Effects of drought stress on growth and biomass production .......................................30 3.3. Effects on vegetative and reproductive growth rates ....................................................34 3.4. Effects on plant-water relations.....................................................................................37 3.4.1. Relative water content ............................................................................................37 3.4.2. Leaf water potential................................................................................................38 3.5. Effects on water-use and water-use efficiency (WUE).................................................40 3.6. Effects on sink leaf ABA concentration........................................................................43 3.7. Effects on leaf gas-exchange.........................................................................................45 3.8. Effects of drought stress on assimilate metabolism in the source and sink organs.......48 3.8.1. Assimilate synthesis and availability in leaves ......................................................48 3.8.2. Assimilate import and availability in sink organs ..................................................53 3.8.3. Assimilation of storage products in seeds ..............................................................60 3.8.4. Effect of drought stress on seed sink capacity .......................................................61 3.9. Leaf protein changes under drought stress....................................................................62 4. DISCUSSION ......................................................................................................................64 4.1. Effect of drought stress on seed yield ...........................................................................64 4.2. Source and sink limitations and their relationships to yield under drought stress ........66 4.2.1. Assimilate synthesis, availability and supply – source strength ............................66 4.2.1.1. Photosynthesis...............................................................................................66 4.2.1.2. Assimilate availability and supply at source level ........................................69 4.2.2. Carbohydrate import and utilization - sink strength...............................................73 4.2.3. Storage carbohydrates ............................................................................................75 4.3. Growth, biomass accumulation and partitioning ..........................................................77 4.5. Proteome changes in bean leaves ..................................................................................84 5. CONCLUSIONS..................................................................................................................85 6. SUMMARY .........................................................................................................................87 7. ZUSAMENFASSUNG ........................................................................................................90 8. REFERENCES.....................................................................................................................93 9. APPENDIX........................................................................................................................115 III ACKNOWLEDGEMENTS I am indebted to my supervisor, Prof. Dr. Sven Schubert for his friendly guidance, constructive criticisms, kindness and manifold support during all stages of my study. My sincere thank goes to Dr. Heike Wiese for her valuable suggestions and help during the course of my work. I am grateful to Prof. Dr. Dr. Wolfgang Friedt for reading the thesis and for his constructive comments. Many thanks are also extended to Prof. Dr. Diedrich Steffens for all his supports while doing the experiments and writing up this dissertation I highly acknowledge the German Academic Exchange Service (DAAD) for providing me with an adequate financial support during the study. I would also like to thank my employer, the Ethiopian Institute of Agricultural Research (EIAR), for granting me permission to pursue this study in Germany. I express my deep sense of gratitude to the International Center for Tropical Agriculture (CIAT) and to Drs Steve Beebe and Henry Terán, in particular, for supplying seeds of the inbred lines used in this study. I also express my deepest appreciation to the Institute of Plant Ecology of the University of Giessen for allowing me to use their facilities for gas-exchange measurements. Many thanks are also extended to the Institute of Plant Physiology and Biophysics of the University of Würzburg for undertaking the ABA analysis. This piece of work would not have been realized without the genuine help I got from the staff of the Institute of Plant Nutrition based at the experimental station and in the laboratories. I am very much grateful to all those individuals who helped me directly or indirectly during my endeavor of study at the Institute. Finally, I wish to express my special appreciation to my wife, Haimanot, for her love, patience and understanding and taking care of our daughters, Meti and Raji. IV LIST OF FIGURES Figure 1. The effect of drought stress imposed at early pod-filling stage on pod set (total number of pods counted at 5 d stress) and number of productive pods (pods possessing at least one seed at the end of 20 d stress) per plant of two common bean genotypes.......... 29 Figure 2. Biomass ratio (drought tissue / control tissue) of leaves, stems, pods, reproductive organs (pod wall + seeds), above-ground biomass weight (AGBW) (leaves + stems + reproductive materials) and seed yield of two common bean genotypes at different durations of stress................................................................................................ 32 Figure 3. The effect of drought stress initiated at pod-filling stage on reproductive to vegetative biomass ratio of two common bean genotypes. ............................................... 33 Figure 4. The effect of drought stress imposed at vegetative stage on leaf area and specific leaf weight of two common bean genotypes. ....................................................... 34 Figure 5. The effect of drought stress imposed at vegetative phase on the absolute growth rate of two common bean genotypes. ................................................................................ 35 Figure 6. The effect of drought stress imposed at pod-filling stage on absolute growth rate (AGR) and relative growth rate (RGR) of reproductive structures (pod walls + aborted pods + seeds) in two common bean genotypes..................................................... 36 Figure 7. Leaf relative water content (RWC) of two common bean genotypes under drought stress and non-stress growth conditions during vegetative and reproductive growth phases..................................................................................................................... 37 Figure 8. The relationship of stomatal conductance (gs) and leaf relative water content (RWC) with net photosynthetic rate (A) of two common bean genotypes grown under drought stress (imposed at reproductive phase) and non-stress growth conditions. ......... 38 Figure 9. Pod water concentration of two common bean genotypes grown under drought stress and non-stress growth conditions............................................................................. 39 Figure 10. Water consumption and water-use efficiency of two common bean genotypes under 10 d drought stress and non-stress growth conditions during the vegetative phase.. ................................................................................................................................ 40 V Figure 11. Instantaneous water-use efficiency (IWUE) of two common bean genotypes grown under drought stress imposed at reproductive stage and non-stress growth conditions. .......................................................................................................................... 43 Figure 12. The relationship between instantaneous water-use efficiency (IWUE) and gas- exchange parameters (A and gs) of two common bean genotypes grown under drought stress (imposed at reproductive phase) and non-stress growth conditions. ....................... 44 Figure 13. The effect of drought stress imposed during the vegetative phase on sink leaf abscisic acid (ABA) concentrations of two common bean genotypes .............................. 44 Figure 14. The effect of drought stress imposed during the vegetative phase on net photosynthetic rate and stomatal conductance of two common bean genotypes differing in drought resistance. ............................................................................................................. 45 Figure 15. The effect of drought stress imposed at early pod-filling stage on net photosynthetic rate and stomatal conductance of two common bean genotypes differing in drought resistance. ............................................................................................................. 46 Figure 16. The effect of drought stress imposed at early pod-filling stage on the ratio of leaf intercellular to ambient CO2 concentration (Ci/Ca) of two common bean genotypes differing in drought resistance. .......................................................................................... 47 Figure 17. The effect of drought stress imposed at pod-filling stage on the ratio of net photosynthetic rate to leaf intercellular CO2 concentration (A/Ci) of two common bean genotypes differing in drought resistance.......................................................................... 48 Figure 18. The effect of drought stress imposed at early pod-filling stage on dark respiration of two common bean genotypes differing in drought resistance..................... 49 Figure 19. Leaf sucrose, hexose sugars and total sugar concentrations of two common bean genotypes under drought stress and non-stress growth conditions during the vegetative phase. ................................................................................................................ 50 Figure 20. The effect of drought stress imposed at early pod-filling stage on leaf sucrose hexose sugars and total sugars concentrations of two common bean genotypes. ............. 51 Figure 21. The effect of drought stress imposed at pod-filling stage on leaf -amino-N concentrations of two common bean genotypes. ............................................................... 52 VI Figure 22. The effect of drought stress imposed at early pod-filling stage on leaf starch and total non-structural carbohydrate (TNC) concentrations of two common bean genotypes. Vertical bars show S.E. of four replications.................................................. 53 Figure 23. The effect of drought stress imposed at early pod-filling stage on the ratio of leaf sucrose to starch concentrations of two common bean genotypes. ............................ 54 Figure 24. The effect of drought stress imposed at early pod-filling stage on stem sucrose concentrations of two common bean genotypes. ............................................................... 54 Figure 25. The effect of drought stress imposed at early pod-filling stage on productive pod sucrose, hexose sugars and total sugars concentration of two common bean genotypes. .......................................................................................................................... 56 Figure 26. Productive pods (Pr-P) and aborted pods (Ab-P) sucrose concentrations of two common bean genotypes grown under drought stress initiated at early pod-filling stage and non-stress growth conditions. ..................................................................................... 57 Figure 27. The effect of drought stress imposed at pod-filling stage on pod -amino-N concentrations of two common bean genotypes. ............................................................... 58 Figure 28. The effect of drought stress imposed at early pod-filling stage on seed sucrose concentrations of two common bean genotypes. ............................................................... 58 Figure 29. The effect of drought stress imposed at pod-filling stage on seed -amino-N concentrations of two common bean genotypes. ............................................................... 59 Figure 30. The effect of drought stress imposed at early pod-filling stage on seed starch concentrations of two common bean genotypes. ............................................................... 60 Figure 31. The effect of drought stress imposed at pod-filling stage on seed protein concentrations of two common bean genotypes. ............................................................... 62 Figure 32. Coomassie-stained 2D gel of total proteins extracted from mature leaves of drought-stressed cv. Brown Speckled. The proteins were separated by two-dimensional isoelectric focusing (IEF)/ SDS-polyacrylamide gel electrophoresis (SDS-PAGE)......... 63 VII LIST OF TABLES Table 1. Some important features of the genotypes used ................................................. 12 Table 2. The effect of drought stress imposed at early flowering stage on seed yield and harvest index of six common bean genotypes. .................................................................. 27 Table 3. Seed yield and yield components of two common bean genotypes 20 d after the commencement of drought stress at pod-filling stage. ..................................................... 28 Table 4. The effect of drought stress imposed at vegetative and pod-filling stages on above-ground fresh and dry biomass weights of two common bean genotypes. ............. 31 Table 5. The effect of drought stress initiated during the vegetative growth phase on relative growth rate (RGR), net assimilation rate (NAR), leaf area ratio (LAR), specific leaf area (SLA) and leaf weight ratio (LWR) of two common bean genotypes................ 35 Table 6. The effect of drought stress imposed at early flowering stage on leaf water potential ( ), osmotic potential ( s) and turgor pressure ( p) of six common bean genotypes. .......................................................................................................................... 39 Table 7. Quantity of water consumed from emergency to maturity and share of the reproductive phase in six common bean genotypes grown under drought stress (initiated at flowering stage) and non-stress growth conditions. ...................................................... 41 Table 8. Water-use efficiency based on above-ground biomass yield (WUEBY) and seed yield (WUESY) of six common bean genotypes grown under drought stress (imposed at flowering stage) and non-stress growth conditions. .......................................................... 42 Table 9. The effect of drought stress imposed at early pod-filling stage on the numbers and volumes of cotyledonary cells and amyloplasts of two common bean genotypes. .... 61 Table 10. The effect of a 10 d drought stress initiated during the vegetative phase on quantitative and qualitative changes in leaf proteins of common bean (cv. Brown Speckled)............................................................................................................................ 62 Table 11. Analysis of source-sink relationships under drought stress (imposed during the reproductive phase) relative to control conditions for the genotypes SEA 15 and Brown Speckled (percent reductions of various parameters). ....................................................... 70 VIII ABBREVIATIONS Leaf water potential p Turgor potential (pressure) s Osmotic potential A Net photosynthetic rate ABA Abscisic acid AGR Absolute growth rate BrSp cv. Brown Speckled Ca Ambient CO2 concentration Ci Leaf intercellular CO2 concentration d day DW Dry weight gs Stomatal conductance to water vapour IWUE Instantaneous water-use efficiency LAR Leaf area ratio LWR Leaf weight ratio NAR Net assimilation rate RGR Relative growth rate RWC Leaf relative water content S.D. Standard deviation S.E. Standard error SEA 15 CIAT inbred line SLA Specific leaf area SLW Specific leaf weight WHC Water-holding capacity WUE Water-use efficiency WUEBY Total biomass yield based water-use efficiency WUESY Seed yield-based water-use efficiency 1 1. INTRODUCTION As much as 60% of common bean (Phaseolus vulgaris L.) production in the developing world occurs under conditions of significant drought stress (Graham and Ranalli, 1997). Consequently, the average global yield of beans remains low (<900 kg ha -1 ) (Singh, 2001; Thung and Rao, 1999). To date, progress in improving common bean cultivars for dry environments of the tropics has been achieved by yield testing of large collections over several locations and years. Such empirical approaches are, however, slow, laborious, and expensive because of the need to assess the yield of large numbers of lines across several locations and years, and the substantial variation from the effects of environment, error, and genotype-environment interactions (Blum, 1988). Success in developing drought- resistant common bean cultivars has further been limited due to the irregularity of available moisture, lacks of screening techniques and practical selection criteria other than yield (Ramirez-Vallejo and Kelly, 1998; Acosta-Gallegos and Adams, 1991). In the above context, there is a strong argument that an indirect (or analytical) approach, based on the understanding of crops at morphological, physiological and molecular levels may help to target the key traits that are currently limiting yield (Araus et al., 2002; Bidinger and Witcombe, 1989; Turner, 1986). The identification of main physiological processes determining yield by comparing genotypes differing in drought tolerance has been proposed as the most reliable and soundest approach to identify the potential secondary traits (Araus et al., 2002; Jat et al., 1991; Bohnert and Jensen, 1996). Comparing physiological bases of the differences in yielding capacity among genotypes released during different periods (retrospective studies) may also serve as a complementary approach (Araus et al., 2002). In fact, examples of the successful use of indirect selection criteria (physiological traits) in breeding for better yields under dry conditions for important crop plants including common bean are rarely found (Ober et al., 2005; Slafer et al., 1994; White and Singh, 1991). Nevertheless, few cases such as selection for low carbon-isotope discrimination ( 13 C) (Passioura, 2002), increased 2 osmotic adjustment (Chimenti et al., 2002; Morgan, 2000), and introgressing QTLs associated with deeper rooting into a high-yielding cultivar (Babu et al., 2003; Shen et al., 2001) have proven the merit of the approach. By the same token, understanding the key adaptive morphological, physiological and biochemical traits/mechanisms linked to growth and yield of common bean under drought stress may contribute to concerted efforts presently under way to develop drought-resistant cultivars. 1.1. Mechanisms and traits related to drought resistance in common bean 1.1.1. Growth, yield and morphological adaptations Past research works on adaptation of common beans have demonstrated that compared with shoot traits, root characteristics are of primary importance in determining drought response and differences in yield under low moisture stress (Norman et al., 1995; White and Castillo, 1989). Under drought stress, deeply penetrating and dense roots correlate with leaf gas-exchange (stomatal conductance control) in P. vulgaris (White et al., 1990) and P. acutifolius (Mohamed et al., 2002). At shoot level, beans respond to drought stress by leaf movement (Pastenes et al., 2005; Ehleringer et al., 1991), leaf flagging and shedding (Acosta-Gallegos, 1988; Adams et al., 1985). Loss of leaf area, which could result from reduced size of younger leaves and inhibition of the expansion of developing foliage, is also considered an adaptation mechanism to drought (Acosta-Gallegos, 1988). Early phenology coupled with rapid ground cover and dry matter production in legumes allows greater post-flower water-use leading to greater partitioning of dry matter into seeds (Siddique et al., 2001). Cultivars that show greater phenological adjustment exhibit higher seed yields under drought conditions (Acosta-Gallegos and White, 1995). Slower growth has been suggested as an adaptive feature for plant survival under stress, because it allows plants to divert assimilates and energy, otherwise used for shoot growth, into protective molecules to fight stress (Zhu, 2002) and/or to maintain root growth, improving water acquisition (Chaves et al., 2003). In most drought studies, a single harvest date has been used to correlate growth with the physiological effects of stress. The 3 results from such studies can be misleading when comparing different genotypes or drought treatments because the initial size of the plant can influence the size or rate of growth at harvest (Hunt, 1990). The relative growth rate (RGR) takes this factor into account by dividing the absolute growth rate by the initial weight of the plant. This gives a relative basis on which to compare growth rates of plants. The use of formal growth analysis, therefore, has value in discriminating alternative mechanisms of drought stress at the whole plant level. Shoot biomass accumulation is considered an important trait to attain high seed yield in grain legumes (Saxena et al., 1990). Significant differences have been observed for shoot biomass accumulation among dry bean cultivars grown under moderate to severe drought stress conditions (Rosales-Serna et al., 2002; Ramirez-Vallejo and Kelly, 1998; Acosta- Gallegos and Adams, 1991). Strong positive correlations have often been reported between total plant biomass and seed yield under drought stress and non-stress conditions (Shenkut and Brick 2003; Ramirez-Vallejo and Kelly, 1998). Because plant biomass has moderate to high heritability and exhibits low genotype environment interactions, it has been suggested that the trait could be used as an indirect selection criterion to improve and stabilize seed yield for low moisture areas (Shenkut and Brick, 2003). According to Chaves et al. (2002), in addition to dry matter accumulation, the ability of genotypes to partition stored vegetative biomass to reproductive organs to a large extent determines sink establishment and economic yield under drought stress. In general, drought causes considerable reduction in seed yield of common bean although the ranges of reductions are highly variable due to differences in the timing and intensity of the stress imposed and the genotypes used (Frahm et al., 2004; Shenkut and Brick 2003; Ramirez-Vallejo and Kelly, 1998; Foster et al., 1995; Halterlein, 1983). Seed yield- based genotypic differences for drought resistance have been reported for common bean (Terán and Singh, 2002; Abebe et al., 1998). Bean seed yield reduction due to drought stress are attributed to adverse effects of the stress on individual yield components 4 (number of pods per plant, number of seeds per pod, seed weight and harvest index). The relative importance of individual components as determinants of seed yield varies from experiment to experiment (Shenkut and Brick, 2003; Boutraa and Sanders, 2001; Ramirez-Vallejo and Kelly, 1998; Singh, 1995). 1.1.2. Water-use and water-use efficiency (WUE) Under moisture-limiting environments, productivity in crop plants may be increased by improving water-use efficiency (WUE) (Ehleringer et al., 1993). To achieve this goal, it is important to identify the factors underlying variations in the WUE since they can either positively or negatively be correlated with productivity, depending on the main processes determining changes in WUE (Udayakumar et al., 1998). Carbon isotope discrimination ( 13 C), specific leaf weight (SLW), and canopy temperature have been proposed as potential surrogate tools for selecting genotypes with higher WUE in several legumes (Saranga et al., 1998; Menendez and Hall, 1995; Johnson and Tieszen, 1994; Ismail and Hall, 1993; Gutschick and Currier, 1992; Hattendorf et al., 1990; Farquhar and Richards, 1984). In cereals, traits such as deeper root systems, early vigor, osmoregulation, smaller photosynthetic surfaces and small erect upper canopy leaves may help crops either to use more water or enhance WUE when subjected to drought stress (Araus et al., 2002). Genotypic variation for WUE has been demonstrated in common beans using carbon isotope discrimination ( 13 C) technique (Ehleringer et al., 1990). Also, positive associations between 13 C and bean seed yield have been reported (Ehleringer et al., 1990; White et al., 1990). Nevertheless, key physiological traits that offer a potential to improve WUE in common bean are not thoroughly studied. 1.1.3. Leaf-water relations and gas-exchange Leaf water potential ( ) and its two components, osmotic potential ( s) and turgor potential ( p) are useful as selection criteria for improving drought tolerance in crop plants. Leaf water potential evaluates the water stress intensity sensed by leaves (Hsiao, 5 1973) and is recognized as an index for whole plant water status (Pantuwan et al., 2004; Turner, 1982). It is considered as a reliable parameter for quantifying plant water stress response (Siddique et al., 2000). In general, the maintenance of high determined by the interaction of numerous plant mechanisms at both shoot and root levels is considered to be associated with dehydration avoidance mechanisms (Levitt, 1980). Maintenance of leaf turgor in the face of decreasing soil moisture has been emphasized as an important adaptational trait that contributes to drought tolerance (Hsiao et al., 1976). Jongdee et al. (2002), Pantuwan et al. (2002) and Sibounheuang et al. (2001) found that genotypes with high had less reproductive sterility and produced higher yield than genotypes with lower under drought stress conditions. Other reports suggest that plant metabolic processes are in fact more sensitive to turgor and cell volume than absolute water potential (Jones and Corlett, 1992). Among the physiological mechanisms that act to maintain leaf turgor pressure under lower leaf water potential, decreased osmotic potential resulting either from a decrease in osmotic water fraction or from an osmotic adjustment (net accumulation of solutes in the symplast) has been pointed out (Jones and Turner, 1980). A satisfactory basis for relating cellular water status to metabolism is relative water content (RWC), an easily measured, robust indicator of water status for comparison of tissues and species, which ‘normalizes’ water content by expressing it relative to the fully turgid (hydrated) state (Lawlor and Cornic, 2002). Sinclair and Ludlow (1985) proposed that leaf relative water content (RWC) is a better indicator of water status than was water potential ( ). RWC is a measure of relative change in cell volume; is the resultant of cell turgor ( p) and osmotic potential ( s), and thus depends both on solute concentration and cell wall rigidity and does not relate directly to cell volume (Kramer and Boyer, 1995; Lawlor, 1995; Kaiser, 1987). RWC as an integrative indicator of internal plant water status under drought conditions has successfully been used to identify drought-resistant 6 cultivars of barley (Hordeum vulgare) (Martin et al., 1989) and common bean (Costa França et al., 2000). Photosynthesis is the main process responsible for dry matter accumulation and consequently affects plant development and growth, which are strongly affected by the environment (McCree, 1986). In common bean, drought stress at its initial phase limits photosynthesis due mainly to stomatal closure (Miyashita et al., 2005, Amede et al., 2003b). However, as the stress progresses over a longer period, non-stomatal inhibition of photosynthesis may become more important (Lawlor and Cornic, 2002; Medrano et al., 2002). Increasing evidence suggests that down-regulation of different photosynthetic processes under drought stress depends more on CO2 availability in the mesophyll (i.e. stomatal closure) rather than or RWC (Medrano et al., 2002). Stomatal control is one of the main mechanisms for adapting to water stress in common bean (Laffray and Louguet, 1990). In crops such as beans, stomata often close in response to drought before any change in and/or RWC is detectable (Miyashita et al., 2004; Socías et al., 1997). Information on a common pattern of photosynthetic response to drought for common bean is currently meagre. 1.2. Assimilate metabolism in source and sink organs under drought stress Drought stress decreases photosynthetic rate thereby disrupting carbohydrate metabolism in leaves (Pelleschi et al., 1997; Kim et al., 2000). As a consequence, the amount of assimilates available for export to the sink organs may be reduced leading to an increased rate of reproductive abortion. In drought-stressed maize (Zea mays L.) (Schussler and Westgate, 1991, 1995) and wheat (Wardlaw, 2002), smaller/loss of kernel set was correlated with the extent of loss in photosynthesis and the photosynthate influx into kernels. As sucrose is the principal form of photosynthate for long-distance transport to sink organs, its concentration in leaves represents the current availability of assimilate for reproductive development (Westgate and Thomson Grant, 1989). Leaf sucrose concentration is determined by several factors including the rate of photosynthesis, the 7 partitioning of photosynthetic carbon between starch and sucrose, the rate of sucrose hydrolysis, and the rate of sucrose export (Huber, 1989; Egli et al., 1980). Any effect of drought on these processes would modify leaf sucrose concentration. In sucrose- transporting plants, the sucrose status of a tissue plays a crucial role in the regulation of metabolism, and sucrose export from mature leaves is related to sucrose synthesis (Geiger and Fondy, 1991). In pigeon pea (Cajanus cajan), leaf starch and sucrose concentrations decreased rapidly and became close to zero, while the concentrations of glucose and fructose significantly increased in response to drought stress (Keller and Ludlow, 1993). Similar results have been observed in several plant species under drought conditions (Lawlor and Cornic, 2002). Overall, it is suggested that the starch and sucrose pools in plant leaves are depleted under drought conditions; in the meantime, the resulting high concentrations of hexose may be involved in a feedback regulation of photosynthesis (Chaves et al., 2002). Consequently, the total amount of sucrose for export is significantly decreased. Drought stress can also affect carbohydrate metabolism in plant reproductive organs (Liu et al., 2004). It has been often observed that sucrose concentrations in reproductive structures of drought-stressed plants, i.e., in maize ovaries and rice (Oryza sativa L.) anthers, generally are higher or at least similar to those of the well-watered controls (Setter et al., 2001; Zinselmeier et al., 1995; Sheoran and Saini, 1996). The results imply that rather than sucrose concentration per se, the capacity for sucrose utilization may be affected by drought stress. In drought-stressed maize, accumulation of sucrose in young ovaries coincided with a cessation of ovary growth, an accumulation of sucrose, and a decrease in the concentration of hexose (Zinselmeier et al., 1999). These results suggest that drought-induced changes in carbohydrate status and metabolism in crop reproductive structures during the early stage of development are crucial for successful fruit set. In addition to photosynthate supply, loss of pod set caused by drought stress during the critical, abortion-sensitive phase of soybean pod development was associated with a 8 decrease in water potential and with higher ABA accumulation in the reproductive structure (Liu et al., 2004, 2003). 1.3. Protein changes in response to drought stress In addition to the physiological and biochemical responses of plants to water stress, the information on the molecular mechanisms of drought stress adaptation could be useful for the genetic improvement of drought-resistant crops/genotypes. Proteomics are a recent addition to the molecular tools used to analyze drought-affected plants (Salekdeh et al., 2002), and have been applied to the study of drought response of barley (Neslihan-Ozturk et al. 2002), maritime pine (Costa et al., 1998), maize (Riccardi et al., 1998) and wild watermelon (Kawasaki et al., 2000). Two-dimensional gel electrophoresis (2DE) is known to be a powerful method to resolve qualitative variations (positional shifts, present and absent) and quantitative variations (increase or decrease) of proteins and to follow the modification of gene expression under various conditions (Damerval et al., 1986). Water deficit induces the expression of proteins that are directly or indirectly related to the stress and some functions have been assigned to some of the sequenced proteins. Among the stress-induced proteins identified are those implicated in the biosynthesis of osmolytes (Bohnert et al., 1995; Ishitani et al., 1995), in the uptake and compartmentation of ions (Lisse et al., 1996; Niu et al., 1995), in hydroxyl-radical scavenging (Ingram and Bartels, 1996; Bohnert et al., 1995; Smirnoff and Cumbes, 1989) and in protein turnover (Kiyosue et al., 1994; Koizumi et al., 1993). Some induced proteins are expressed in order to protect the cellular machinery. These protective proteins include different classes of late embryogenesis abundant (LEA) proteins such as dehydrins (Neslihan-Ozturk et al., 2002; Colmenero-Flores et al., 1997; Lisse et al., 1996). There is a strong circumstantial evidence for the involvement of LEA proteins in the plant adaptation to water deficit through their protective role in maintaining specific cellular structures or ameliorate the effects of drought stress (Lisse et al., 1996). Proteins that show significant down- regulation under drought stress were observed for photosynthesis-related function 9 (Neslihan-Ozturk et al., 2002). Water deficit may also induce the expression of proteins, which are not specifically related to the stress but rather to reactions against cell damage, and those whose functions are not directly related to the stress (reviewed by Riccardi et al., 1998). 1.4. Underlying hypotheses and objectives of the study Past studies have shown that common bean genotypes selected for specific adaptations to drought conditions produce significantly higher seed yield compared with landraces and standard cultivars grown under similar drought conditions (Téran and Singh, 2002). Profound differences have also been reported among old and modern cultivars of other crops in terms of water-use and water-use efficiency when subjected to drought stress (Koç et al., 2003; Siddique et al., 1990). In agreement with these findings, we hypothesized that common bean genotypes selected for specific adaptation to drought stress exhibit significant variation from those developed for wider agro-ecological adaptations in terms of drought resistance and water-use efficiency. Differential responses in growth, yield and biomass partitioning under drought stress of the genotypes may account for such differences. The differences in drought resistance (determined based on grain yield) among drought- resistant and susceptible genotypes are often related to the ability to partition biomass stored in vegetative biomass to reproductive organs and the subsequent capacity to establish new sink under drought stress conditions (Koç et al., 2003; Siddique et al., 1990). In line with this, drought stress, when initiated during the reproductive phase, may differentially affect the sink strength (i.e. capacity to establish new sink) of common bean genotypes differing in drought resistance. We supposed that genotypic differences in sink strength are due to the differential effect of drought stress on assimilate synthesis and availability at source level and/or availability of assimilates for metabolism in the sink organs of the genotypes. In accordance with the observations of Schulze (1986) and Kubiske and Abrams (1993) plants of a drought-resistant bean genotype may maintain 10 higher rates of photosynthesis and stomatal conductance than plants of a drought- susceptible bean genotype when subjected to drought stress at different growth stages of the crop. The disparity in gas-exchange rate between the contrasting genotypes may lead to different rates of assimilate synthesis and availability for export to sink organs. Drought stress may also affect the accumulation of seed storage products by limiting the seed sink capacities (i.e. reduces the number and volume of storage organelles). Drought stress induces changes in proteins, which play a pivotal role in the adaptive response of plants to stress (Riccardi et al., 1998; Bray, 1997; Ingram and Bartels, 1996). Accordingly, relative to non-stressed growth conditions drought stress initiated during the vegetative phase may induce quantitative and qualitative changes in proteins of mature bean leaves. The objective of this study was to test the hypotheses that I) differences exist in biomass accumulation, yield and water-use efficiency among common bean cultivars developed for wider agro-ecological adaptation and inbred lines selected for specific adaptation to drought situations when subjected to drought stress; II) a drought-resistant genotype has a higher sink strength than a susceptible genotype and the difference between the genotypes is related to the ability to maintain assimilate synthesis and availability of assimilates for metabolism in the reproductive sink organs under drought stress; III) drought stress induces higher accumulation of ABA in sink leaves of a drought- susceptible genotype than in the leaves of a resistant genotype; and IV) relative to non- stressed plants, drought stress alters the protein patterns in a mature bean leaf. . 11 2. MATERIALS AND METHODS 2.1. Genotypes Three adapted cultivars and three inbred lines of common bean (Phaseolus vulgaris, L.) varying in seed characteristics and growth habits were initially screened to assess seed yield-based drought resistance and water-use efficiency of the genotypes (Table 1). The adapted (old) cultivars were chosen among varieties developed by the national bean research program of Ethiopia for wider adaptations to different agro-ecological conditions of the country. A recent yield-testing study carried out across years and locations representing major bean-growing regions of the country demonstrated that the cultivar Mexican 142 was relatively stable across the range of environments used, whilst the other two cultivars (Roba 1 and Brown Speckled) were more adapted to marginal environments (Mekbib, 2003). The inbred lines were obtained from the bean research program of CIAT. The drought-resistance degrees of these materials have been demonstrated in earlier field studies (CIAT, 2002). A drought-tolerant (SEA 15) and a drought-susceptible genotype (Brown Speckled) selected from the screening experiment were used in subsequent experiments carried out thereafter. 2.2. Plant cultivation Seeds of the tested genotypes were grown in either Mitscherlich or Ahr pots filled with 6 or 13 kg of Kleinlindener soil, respectively. At the time of planting, the soil was fertilized with Blaukorn (12.0% N, 5.2% P, 14.1% K, 1.2% Mg and 6.0% S). Eight seeds per pot were initially sown and later thinned to three (for Mitscherlich pots) or four (for Ahr pots) plants when the first trifoliate leaves were unfolded. Plants were raised in a vegetation hall. The daily minimum and maximum temperatures (mean S.D.) during the growth periods of 2003, 2004 and 2005 were (12.2 2.6; 27.3 4.7), (12.6 3.2; 26.2 5.1) and (11.2 3.2; 23.8 4.8) C, respectively. The respective daily average temperatures during same period were 22.2 3.7, 21.3 4.4 and 19.3 4.1. 12 Table 1. Some important features of the genotypes used Genotype Source/ Background Seed size Growth Habit † Special features Mexican 142 AC, Ethiopia Small III Popular export-type cultivar Roba 1 AC, Ethiopia Medium II Popular food-type cultivar Brown Speckled AC, Ethiopia Large II Less popular food-type cultivar SEA 15 IL, CIAT Medium II Combines Middle American races with Mesoamerica and Durango races SEA 23 IL, CIAT Medium II Combines Middle American race Mesoamerica BAT 881 IL, CIAT Medium I Combines Middle American and Andean races Note: AC = adapted cultivar; IL = inbred line † Bean Growth Type I = determinate bush; Type II = indeterminate bush; Type III = indeterminate prostrate (Singh, 1982) The pots were weighed daily and watered to restore the appropriate moisture by adding a calculated amount of water. Daily additions of water (equivalent to the amount of water lost) to each pot were recorded to calculate the total water consumed (kg plant -1 ) by the genotypes under contrasting soil moisture regimes. In order to minimize the variation, which may arise due to differences in the original fresh weights of the genotypes, the amount of water applied in both watering regimes was corrected for the fresh weight per plant determined shortly before the initiation of drought stress. 2.3. Experimental procedure The descriptions, vegetative and reproductive phase experiment may be used as required in the forthcoming sections in order to facilitate communication throughout the manuscript. They refer to set(s) of experiments carried out with drought stress initiated at either the vegetative or reproductive growth stage of the crop. 13 Before the commencement of drought stress treatment at either growth stage of the crop, plants were grown under optimal soil moisture conditions. Drought stress was imposed by withholding the amount of water applied in order to keep the moisture level at about 30% of the maximum water-holding capacity (WHC) of the soil. For control treatments, the soil moisture was maintained at 70% of the maximum WHC until the plants were harvested. The exposure of plants to the indicated intensity of stress for the vegetative phase experiment began at growth stage V6, when plants had six trifoliate leaves. The plants attained this stage 30 d after planting. For experiments in which drought stress was imposed at reproductive phase, the stress treatment began at early pod-filling stage, when plants had at least one pod that had grown to its maximum length. Only for the initial genotype-screening experiment, drought stress was initiated at 100% bloom stage, when the plants had at least one open flower. In all experiments, the treatments were replicated four times and the pots were regularly randomized. 2.4. Sample collection and parameters measured 2.4.1. Biomass and seed yield Plants were harvested at the end of 5 and 10 d stress (for the vegetative phase experiment) and 5, 10 and 20 d after imposing drought stress for the reproductive growth phase experiments. Above-ground fresh weight was determined by adding up the various plant parts (leaves, stems, pods and seeds) harvested separately. Similarly, above-ground dry weight was obtained by adding up various plant parts dried at 80 o C for 48 h. Biomass partitioning ability of the genotypes was evaluated by computing the ratio of reproductive structures (pods/pod walls + seeds) to vegetative biomass (leaf + stem dry weight). At harvest, pods were categorized into two groups as productive pods (Pr-P) and aborted pods (Ab-P). The classification of pods was based on length attained and whether or not they bore seeds at the time of harvesting. Productive-pods (Pr-P) were defined as pods that were longer than 5 cm (for the harvest made at 5 d stress) and bore at least one seed 14 per pod (for the harvests made at 10 and 20 d stress). During the course of pod growth and development, it was observed that the underdeveloped pods (whether dropped off the plant or loosely hanging to the reproductive branches) had less than 5 cm length. These pods were considered as aborted pods (Ab-P). Also, pods that grew to a length of more than 5 cm but did not possess typical and healthy seeds (usually found at 10 and 20 d stress) were regarded as aborted pods. Seed yield (g plant -1 ) was calculated as a product of the yield components (number of productive pods per plant, number of seeds per pod and seed weight). Hundred seed weight (HSW, g) was determined on 100 seeds randomly sampled from all plants harvested per pot. Harvest index (HI) was calculated as the proportion of seed weight to the above-ground dry weight (stem + leaves + pod + seed) at harvest. 2.4.2. Plant water-use and leaf-water relations Water-use efficiency (WUE, mg g -1 ) of the bean plants at vegetative phase was calculated according to the following formula: WUE = (w2 – w1) / T, where w1 and w2 are the total dry weights at the end of 5 and 10 d stress, respectively, and T is the total amount of water used for transpiration between the first and the second harvest. Seed yield-based water-use efficiency (WUESY) was estimated as the ratio of seed yield to the amount of water consumed from emergence to physiological maturity of the genotypes. Instantaneous water-use efficiency (IWUE, µmol mol -1 ) was calculated as the ratio of net photosynthetic rate (A) to stomatal conductance (gs) determined during the reproductive phase. Leaf growth and water relation parameters were determined on young expanding trifoliate leaves. During the vegetative phase, these leaves were located at the 7 th and 8 th (5 d stress) and at the 8 th and 9 th (10 d stress) main stem nodes of Brown Speckled and SEA 15, respectively. The central leaflets of the selected trifoliate leaves were cut and fresh weight (FW) taken immediately. The weighed leaves were then placed in a petri-dish containing 15 wet filter paper and kept in the dark. After 24 h, the turgid weight (TW) was obtained. For the dry weight (DW), the leaflets were oven-dried for 48 h at 80°C. The second leaflet from the same trifoliate leaf used for fresh weight determination was cut and the leaf area (LA) was measured using a leaf area meter AM200 (ADC BioScientific Ltd., UK). Leaf relative water content (RWC, %) and specific leaf weight (SLW, g m -2 ) were calculated as follows: RWC = [(FW-DW) / (TW-DW)] x 100 SLW = DW / LA where FW, DW, TW and LA are the fresh weight, dry weight, turgid weight and leaf area, respectively. Leaf water potential and its components were determined at developmental stage R5 (plants had at least one pod with fully developed seeds) for bean plants subjected to drought stress at flowering stage. Water potential ( ) was measured with the central leaflet of the youngest expanding leaf using the pressure probe method (Scholander et al., 1965). The second leaflets were cut simultaneously, put in a 5 ml syringe, frozen in liquid nitrogen and kept in a cool box until transfer to a deep freezer where they were kept at –80 o C. The solute osmolality was measured using Osmomat 030 (Gonotec, Berlin) in duplicate from the press sap of the frozen leaves after pressing mechanically at room temperature. Readings were converted to pressure units by using the van´t Hoff equation ( = -cRT), where is the osmotic pressure, c is the osmolality (mosmol kg -1 ), R is the gas constant and T the temperature (K). Turgor potential ( p) was calculated as the difference between osmotic and water potentials. 2.4.3. Growth analysis To investigate the effect of drought stress on plant growth of two distinct common bean genotypes, the relative growth rate (RGR, g g -1 d -1 ), net assimilation rate (NAR, g m -2 d -1 ), leaf area ratio (LAR, m 2 g -1 ), specific leaf area (SLA, m 2 g -1 ), and leaf weight ratio (LWR, g g -1 ) were calculated according to Beadle (1993). For the growth analysis, shoot (leaf + 16 stem) dry weight and estimated total leaf area of bean plants subjected to drought stress for 5 (t1) and 10 days (t2) during the vegetative phase were used. Total leaf area per plant was estimated by measuring maximum length (mL) and width (mW) of leaves and multiplying these inputs (mL x mW) by a correction factor of 0.6 derived from the actual leaf area determined with a leaf area meter AM200 (ADC BioScientific Ltd., UK). The estimations were considered accurate because the differences in correction factor between the two genotypes and the leaf age were very small, so that comparisons between the genotypes and the watering regimes were not significantly biased. RGR and its components were calculated between sampling dates as follows: RGR = (lnW2 – lnW1) / t2 – t1 NAR = [(W2 - W1) / (A2 - A1)] [(ln A2 / A1) / (t2 – t1)] LAR = [(A2 - A1) / (W2-W1)] [(ln W2 / W1) / (ln A2 / A2)] SLA = [(A2 - A1) / (WL2-WL1)] [(ln WL2 / WL1) / (ln A2 / A2)] LWR = [(WL2 - WL1) / (W2-W1)] [(ln W2 / W1) / (ln WL2 / WL1)] where W is the total dry weight, t is the time, A is the total leaf area, WL is the total dry weight of leaves, and 1 and 2 are 5 and 10 d stress periods, respectively. Absolute growth rate (AGR, g d -1 ) and relative growth rate (RGR, g g -1 d -1 ) of reproductive structures (pods + seeds) were also calculated to study whether growth rates of the genotypes were similarly affected when subjected to drought stress during different growth phases. Both growth rate parameters were computed using dry weights of reproductive structures obtained from the harvests made at the end of 5 and 10 d stress. 2.4.4. Photosynthetic parameters Gas-exchange characteristics, net photosynthetic rate (A), stomatal conductance (gs), and intercellular CO2 concentration (Ci) were measured on the central leaflets of fully- matured upper canopy leaves of both stressed and non-stressed treatments using a portable photosynthesis system (Li-COR LI-6200, Li-Cor, Inc., Lincoln, NE) assembled with an infra-red gas analyzer (Li-COR LI-6250) and data logger. Measurements were 17 made on 5 th and 10 th d of the stress imposition during the vegetative phase experiment. Five measurements were made during the reproductive phase beginning on day two of the stress imposition and continued on alternate days until 10 th d Leaf gas-exchange measurements were initiated (usually between 09.30 and 13.30 h) at ambient relative humidity and temperature, when CO2 concentration in the 0.25 L leaf chamber approached ambient concentration. When the photosynthetically active radiation (PAR) was below 800 µmol m 2 s -1 , leaflets were illuminated by a light source to maintain a saturating irradiation of up to 1200 µmol m -2 s -1 PAR. Photosynthetically active radiation (PAR), leaf temperature (Tl), and air temperature (Ta) were recorded simultaneously. For dark respiration analyses, the chamber was covered with black plastic sheath for 2 min so that the leaf was in complete darkness (a PAR value close to zero was displayed on the LI-6200 Console’s monitor). Measurements began immediately after an increase in CO2 concentration in the leaf chamber was detected. 2.4.5. Chemical analysis For sugar and starch analyses, leaf, stem, pod and seed samples were obtained from the harvests made at 5 and 10 d stress (for the vegetative phase) and at 5, 10 and 20 d stress (for the reproductive phase). On the other hand, plant materials (leaf, stem, pod and seed) for -amino N and crude protein analyses were collected from intact (growing) plants that were subjected to drought stress for periods of 7, 14 and 21 days. The various plant parts were dried at 80 o C for 48 h and finely ground materials were used for the chemical analyses. Sugars: Three-hundred mg ground plant material was weighed into a 50 ml volumetric flask and 30 ml of double-demineralized water was added. The material was then extracted by incubating in a shaking water bath at 60 o C for 30 min. The flask was quickly cooled on ice, and filled up to the mark with double-demineralized water followed by filtration with (blue-band) filter paper (Faltenfilter 595 1/2 , Scheicher and Schüll Co., 18 Dassel, Germany). Sugars (sucrose, glucose and fructose) were determined by using enzymatic test kits and absorbances of the solutions were read at 340 nm. Principles of the determination of sucrose, D-glucose and D-fructose using Enzymatic BioAnalysis: The D-glucose concentration is determined before and after the enzymatic hydrolysis of sucrose; D-fructose is determined subsequent to the determination of D-glucose: Determination of D-glucose before inversion: At pH 7.6, the enzyme hexokinase (HK) catalyzes the phosphorylation of D-glucose by adenosine-5´-triphosphate (ATP) with the simultaneous formation of adenosine-5´- diphosphate (ADP) (1). (1) D-Glucose + ATP HK G-6-P + ADP In the presence of glucose-6-phosphate dehydrogenase (G6P-DH), the D-glucose-6- phosphate (G-6-P) formed is specifically oxidized by nicotinamide-adenine dinucleotide phosphate (NADP + ) to D-gluconate-6-phosphate with the formation of reduced nicotinamide-adenine dinucleotide phosphate (NADPH) (2). (2) G-6-P + NADP + D-gluconate-6-phosphate + NADPH + H + The NADPH formed in this reaction is stoichiometric to the amount of D-glucose and is measured by means of its light absorbance at 340 nm. Determination of D-fructose: Hexokinase also catalyzes the phosphorylation of D-fructose to D-fructose-6-phosphate (F-6-P) with the aid of ATP (3). (3) D-Fructose + ATP HK F-6-P + ADP On the completion of the reaction (3) F-6-P is converted by phosphoglucose isomerase (PGI) to G-6-P (4) F-6-P PGI G-6-P G-6-P reacts again with NADP + with the formation of D-gluconate-6-phosphate and NADPH (2). The amount of NADPH formed now is stoichiometric to the amount of D- fructose. Enzymatic inversion: G6P-DH 19 At pH 4.6, sucrose is hydrolyzed by the enzyme -fructosidase (invertase) to D-glucose and D-fructose (5). (5) Sucrose + H2O -fructosidase D-glucose + D-fructose The determination of D-glucose after inversion (total D-glucose) was carried out according to the principle outlined above. The sucrose concentration is calculated from the difference of the D-glucose concentration before and after enzymatic inversion. Amino acids: Free amino acid concentrations were determined by quantifying -amino N using the ninhydrin method. Ground dry materials (100 mg) of both leaves and pods were extracted with 20 ml phosphate buffer in a 100 ml poly flask with an end-over-end shaker for 1 h. After filtration of the extract (Faltenfilter 595 1/2 , Scheicher and Schüll Co., Dassel, Germany), 0.4 ml of the sample solution was mixed with 4 ml citrate buffer and 4 ml ninhydrin solution in a reagent glass and incubated for 15 min in a water bath at 100 o C. After cooling down the reagent glass in water for 5 min, the solution was added into a micro cuvette and -amino N concentration was determined by means of a spectrophotometer at a wavelength of 570 nm. A calibration curve was made with L- glutamine, which was prepared in the same way with the sample solution, and data were expressed in mmol -amino-N kg -1 dry weight (DW). The reaction between alpha-amino acid and ninhydrin involved in the development of color is described by the following five mechanistic steps: alpha-amino acid + ninhydrin reduced ninhydrin + alpha-imino acid + H2O alpha-amino acid + H2O alpha-keto acid +NH3 alpha-keto acid + NH3 aldehyde + CO2 Step (1) is an oxidative deamination reaction that removes two hydrogen from the alpha- amino acid to yield an alpha-imino acid. Simultaneously, the original ninhydrin is reduced and loses an oxygen atom with the formation of a water molecule. In Step (2), the NH group in the alpha-imino acid is rapidly hydrolyzed to form an alpha-keto acid with the production of an ammonia molecule. This alpha-keto acid further undergoes 20 decarboxylation reaction of Step (3) under a heated condition to form an aldehyde that has one less carbon atom than the original amino acid. A carbon dioxide molecule is produced here. These first three steps produce the reduced ninhydrin and ammonia that are required for the production of color in the last two Steps (4) and (5). The overall reaction for the above reactions is simply (slightly inaccurately) expressed in Reaction (6) as follows: alpha-amino acid + 2 ninhydrin CO2 + aldehyde + final complex (blue) + 3 H2O In summary, ninhydrin, which is originally yellow, reacts with amino acid and turns deep purple. It is this purple color that is detected in this method. Ninhydrin will react with a free alpha-amino group, NH2-C-COOH. This group is contained in all amino acids, peptides, or proteins. Whereas the decarboxylation reaction will proceed for a free amino acid, it will not happen for peptides and proteins. Thus, theoretically only amino acids will lead to the color development. Starch: Starch determination was performed following enzymatic assay procedure using the starch determination kit from Boehringer (Mannheim, Germany). Homogenized ground seed and leaf samples of 300 mg were weighed into Erlenmeyer flasks, and 20 ml of dimethylsulfoxide and 5 ml HCl (8 mol/l) were added. The sealed flask was then incubated for 30 min at 60 o C in a shaking water bath. The sample solutions were cooled quickly to room temperature and approximately 50 ml water were added. The pH was adjusted to 4 – 5 with sodium hydroxide (5 M) under vigorous shaking. The solution was then transferred to a 100 ml volumetric flask, rinsed with water, filled up to the mark with water and filtered using Faltenfilter 595 1/2 (Scheicher and Schüll Co., Dassel, Germany). Principle of starch determination using the Boehringer method: In the presence of the enzyme amyloglucosidase, starch is hydrolyzed to glucose. The content of glucose is determined with hexokinase and glucose-6-phosphate dehydrogenase. 1. Starch + (n-1)H2O n D-glucose The glucose is phosphorylated to glucose-6-phosphate by ATP in the presence of hexokinase with formation of ADP. 21 2. D-glucose + ATP G-6-P + ADP The glucose-6-phosphate is oxidized by (NADP + ) to gluconate-6-phosphate with formation of (NADPH). 3. G-6-P + NADP + D-gluconate-6-phosphate + NADPH + H + The amount of NADPH formed in the above reaction is stoichiometric to the amount of D-Glucose formed by hydrolysis of starch. NADPH is determined by means of light absorbance by means of spectrophotometer at the absorption maximum of 340 nm. Protein: The nitrogen concentration was determined by means of sulphuric acid digestion in a Büchi K-324 (BÜCHI Labortechnik AG, Switzerland). Ground leaf samples of 500 mg were digested by adding 20 ml of H2SO4 and a Kjeldhal Cu–Se catalytic pill. The digestion process was left to run its course until the samples were clarified. The samples were then diluted to 50 ml with distilled water. In order to determine the nitrogen concentration, 5 ml of ionic strength adjuster was added to 5 ml of measuring solution. Measurements were performed with an ammonia-selective electrode using 0.1 mM of ammonium chloride as a standard. The nitrogen content quantified by this method was multiplied by an approximate factor of 6.25 to estimate the crude protein content of the bean seed samples. ABA: Freeze-dried sink leaf tissue samples of two common bean genotypes were homogenized and extracted in 80% aqueous methanol solution. Extracts were passed through a Sep Pak C18-cartridge. Methanol was removed under reduced pressure and the aqueous residue was partitioned three times against ethyl acetate at pH 3.0. The ethyl acetate of the combined organic fractions was removed under reduced pressure. The newly obtained residue was taken up in TBS-buffer (Tris buffered saline; 150 mmol/L NaCl 1 mmol/L MgCl2 and 50 mmol/L Tris at pH 7.8) and subjected to an immunological ABA assay (ELISA) as described earlier (Mertens et al., 1985). The accuracy of the ELISA has been verified in earlier investigations (Hartung et al., 1994). 22 2.5. Determination of the numbers and sizes of cotyledonary cells and amyloplasts The numbers and sizes of cotyledonary cells and amyloplasts were determined on 10 randomly selected seeds per replication obtained from the last harvest (20 d stress). Seed volume was obtained using Archimedes principle (Wessel-Beaver et al., 1984). Dried bean seed weight was determined immediately before measurement of seed volume. Seeds were softened by soaking in distilled water for one night and then separated into seed coat, cotyledons, and embryonic axes. The cotyledons were cut into small pieces, dried at 104 C for 24 h, and dry weights were determined. The cotyledon samples were then immersed in an enzymatic solution (sorbitol 0.45M; MgCl2 10 mM; KH2PO4 1 mM; MES 20 mM; Macerozyme R-10 1%; pH 5.6) under vacuum conditions for few minutes. The samples were then placed in an oven at 37 C for 72 h and then macerated gently with mortar and pestle. Macerated cells were separated on a 300 µm nylon mesh to obtain a homogeneous 100 ml cotyledonary cell suspension. Parts of the cell suspensions were transferred to a 20 ml tube and vortexed before transferring 2 µl aliquots with a micropipette to the middle of a counting grid on a hemacytometer (Medicihaus, Berlin). A cover slip was applied and moved in a circular motion to evenly distribute the cells. Cells were counted under a microscope (Leitz, Wetzlar, Germany) at 25 magnification, and the counts were taken from the four outer squares of the counting chamber (each 1mm 2 ) for four aliquots for each cell suspension. Counts from the four aliquots were averaged to compute the number of cells per unit volume (cells ml -1 ) in the cell suspension. This average of cells ml -1 was multiplied by the volume of the total suspension to give an estimate of cell number per cotyledon. The estimate of cell number per cotyledon was divided by cotyledonary mass to yield an estimate of cotyledon cells per unit mass according to the following equations: Cellsss = (cell counth/volumeh) (suspension volume) Cells per unit mass = cellss/massss 23 Cotyledon cells per seed = (cotyledon mass) (cellss/massss) Mass per cell = cellss/massss Volume per cell = (mass per cell) (seed volume/seed mass); where cellsss = total number of cells in the cotyledon; cell counth = the average of the cell counts observed in the hemacytometer grid; volumeh = the volume of the grid space used for cell counts within the hemacytometer; suspension volume = total volume of the cotyledon cell suspension; and massss = oven-dry mass of the cotyledon sample. Cotyledon cell number per seed was estimated as the product of cotyledon cells per unit mass and the measured cotyledon mass per seed (as calculated above). Individual cotyledon cell volume was estimated by dividing mass per cell by the seed density previously determined from seed weight and seed volume as described earlier. This assumed that seed density and cotyledon density were similar. Because the seeds were about 90% cotyledon, this appeared to be an acceptable assumption. From same aliquot used for cell counts, 1 ml was removed from the suspension and diluted with an equal volume of an iodine solution (3.3 g/l I2 + 6.7 g/l KI) to stain the starch granules. Stained starch granules (amyloplasts) were counted on a hemacytometer (Medicihaus, Berlin) at 40 magnification. The counts were then multiplied by the number of cotyledonary cells to determine number of amyloplasts per seed. The same solution used for the determination of number of amyloplasts was also used for measuring the size of the granules. Approximately 3 to 5 µl of the solution was transferred to the middle of a slide on a microscope (LEICA DM IRB, LUDL electronics, NY) equipped with a digital camera (CoolsnapCF, Photometrics). Pictures of 15 to 20 randomly selected cells observed under microscope were acquired to the computer and the sizes of 3 to 15 granules per cell with distinct boundaries from the neighboring amyloplasts were measured using Meta Vue Software (Universal Imaging Corporation). The distance for measurement was calibrated at 40 magnification (0.11625 µm / pixel). 24 2.6. Proteomic analysis 2.6.1. Protein preparation Proteins were prepared for isoelectric focusing using a DTT–TCA–acetone precipitation method adopted from Zörb et al. (2004). Plants of the genotype Brown Speckled grown under non-stress and drought stress imposed at vegetative growth stage were used for the analysis. Mature leaf material was disrupted by grinding the tissue under liquid nitrogen in a mortar. Ground powder was stored at -80 °C. Protease activity was inhibited by lowering the temperatures of the cell material (4 °C) and the use of strong denaturants, such as urea and TCA, in the protein sample buffer supplemented by the use of the protease inhibitor Pefablock. 1.6 ml lysis buffer (10% TCA in acetone) was added to ¾- filled ground tissue in a 2 ml Eppendorf tube. After vortexing, samples were incubated for 15 min in an ice-cold ultrasonic bath and incubated at -20 °C for 1 h or overnight before centrifugation (20000 g, 15 min, 4 °C). The precipitant was resuspended in 1 ml 4 °C cold buffer A (50 mM DTT; 2 mM EDTA, in acetone). Samples were incubated for 10 min in an ice-cold ultrasonic bath. This procedure was repeated twice. Pellets were lyophylized under N2. The collected pellets were resuspended in 1 ml protein sample buffer (8 M urea, 2 M thiourea, 0.5% pharmalyte buffer (v/v, pH 3–10); 4% CHAPS; 30 mM DTT; 20 mM Tris–base, pH 8.8; 5 mM Pefablock). For solubilization of proteins, samples were incubated for 2 h at 33 °C and for 15 min in an ice-cold ultrasonic bath. After vortexing, samples were centrifuged (18000 g, 30 min) and the supernatant was subjected to isoelectric focusing (IEF). Protein concentration was determined in 1:50 dilutions of the samples according to the 2D QUANT protein determination kit from Amersham Biosciences. 2.6.2. 2D gel electrophoresis Two-dimensional gel electrophoresis was done following the method described by Zörb et al. (2004). IPG strips (11 cm, pH 3–10, Amersham Biosciences) were placed in the trays and 200 µl of the protein solution (150 µg protein) were applied. Strips were covered with paraffin oil. IEF was carried out in a IPGphor chamber (Amersham 25 Biosciences) applying the following conditions: 10 h rehydration; 100 V, 2 h; 500 V, 1 h; 1000 V, 2 h; 8000 V, 2 h. Temperature was 20 °C and current was 45 µA per strip. After running the first dimension, the strips were placed in equilibration buffer (50 mM Tris– HCl, pH 8.8; 6 M urea; 30% glycerol; 2% (w/v) SDS; bromophenol blue, 0.001% (w/v) containing 1% DTT (w/v)) and carefully shaken for 15 min. Thereafter, the strips were incubated for additional 15 min in equilibration buffer with 4% (w/v) iodoacetamide without DTT under slow agitation. The strips were then rinsed with SDS-PAGE running buffer (25 mM Tris–base; 192 mM glycine; 0.1% (w/v) SDS) for 15 min. The second dimension SDS gels contained 12.5% (v/v) acrylamide. Molecular weight standards in a range from 10 to 220 kDa were obtained from Invitrogen. The marker lane was positioned at the acidic side (pH 3) of the gel. Strips and marker dyes were mounted onto the gel surface and sealed with 1% (w/v) agarose containing 0.001% (w/v) bromophenol blue. The second dimension was run at 20 °C and with a constant current of 45 mA per gel in a Hoefer (20 cm 20 cm) vertical gel electrophoresis chamber. Electrophoresis was stopped when the bromophenol blue left the gel and thereafter the gels were fixed with 50% ethanol and 12% acetic acid. 2.6.3. Staining and computer analysis Coomassie staining was done according to a hot-staining protocol with Coomassie R350 tablets (Westermeier and Naven, 2002). Gels were digitized by scanning on an image scanner (Amersham Biosciences) with 300 dpi and 16 bits per pixel. The Coomassie- stained gel replicates for each of the drought-stressed and non-stressed treatments were fused and subsequent spot quantification was performed using Delta2D software (version 3.3) (Decodon, Greifswald, Germany). Matching of protein/peptide spots was performed manually. The most interesting spots in terms of expression levels (up- or down-regulated by at least the factor of 2 or newly appearing or disappearing) were displayed using the statistical tools option of the software. 26 2.7. Data analysis Data were analyzed using the statistical package MSTAT-C, developed by Michigan State University (MSTAT-C, 1989). Data were subjected to analysis of variance (ANOVA) to determine significant differences among treatments for various parameters. Means of the treatments that exhibited significant differences were separated using the least significant difference (LSD) test. The differences of means between control and drought-stressed treatments were tested for statistical significance using the t-test. Relationships between selected parameters were determined using the Pearson’s simple correlation test. For all analyses, a P-value of less than 0.05 was interpreted as statistically significant. 27 3. RESULTS 3.1. Effects on seed yield and yield components During the initial screening experiment in which six common bean genotypes were subjected to drought at early flowering stage, the stress caused significant reduction in seed yield that ranged from 30% (in BAT 881) to 72% (in Brown Speckled) (Table 2). With only about 33% decrease due to drought relative to the non-stressed treatment, SEA 15 produced the highest seed yield under both growth conditions. Seed yield under drought stress and non-stress conditions were highly and positively correlated (r = 0.96, p<0.01). The old adapted cultivars generally suffered higher yield losses due to drought stress compared with the inbred lines. Severe yield losses encountered by the old adapted cultivars under drought conditions were a consequence of reductions in individual yield components (data not presented). Table 2. The effect of drought stress imposed at early flowering stage on seed yield and harvest index of six common bean genotypes. Data are means S.E. of four replications. Seed yield (g plant -1 ) Harvest index (%) Genotype Control Stress % Reduction Control Stress % Reduction Mex.142 9.0 0.8 5.2 0.6 ** 42 37.8 2.5 27.3 2.9 * 28 Roba 1 13.1 0. 5 6.9 0.7 ** 47 57.6 1.7 45.7 4.3 21 Br.Speckl. 8.4 0.9 2.4 0.5 ** 72 28.4 2.4 13.5 2.2 * 52 SEA 15 20.5 0.6 13.8 1.4 ** 33 64.0 1.2 63.4 4.2 1 SEA 23 17.7 0.9 11.2 1.4 * 37 62.1 0.7 55.4 7.3 11 BAT 881 13.5 0.2 9.5 0.4 ** 30 45.8 0.8 42.8 1.4 6 **, * The difference between drought stressed and control treatments are significant at 1 and 5% levels of probability, respectively, according to t-test. For both Brown Speckled and SEA 15, the degree of seed yield reduction due to drought stress imposed at pod-filling stage was comparable with drought stress initiated at early 28 flowering stage. Drought stress commenced at early pod-filling stage and lasted for 20 d resulted in 53 and 30% seed yield reductions for Brown Speckled and SEA 15, respectively (Table 3). The effect of drought on seed yield was primarily due to the significant reduction in number of seeds per plant (Table 3). The smaller numbers of seeds per plant under stress for Brown Speckled (20 under drought vs. 41 under control) were ascribed to the significant decrease by about 26% in the numbers of pods per plant and ca. 28% reduction in numbers of seeds per pod. For SEA 15, however, the reduction in the number of seeds per plant owing to drought was due mainly to ca. 25% less number of productive pods retained per plant. The seed weight of both genotypes remained relatively stable under the contrasting soil moisture regimes (Table 3). Table 3. Seed yield and yield components of two common bean genotypes 20 d after the commencement of drought stress at pod-filling stage. Data are the means S.E. of four replications. Treatment No. Seeds (pod –1 ) No. Seeds (plant -1 ) 100-seed wt. (g) Seed yield (g plant –1 ) Harvest Index (%) Control 3.18 0.16 b 40.8 2.5 c 21.0 1.7 bc 8.5 0.3 c 23.9 0.9 b Br Sp Stress 2.29 0.20 c 21.4 1.1 d 18.4 0.9 c 3.9 0.2 d 16.9 1.5 c Control 4.10 0.14 a 62.1 2.5 a 24.0 0.9 b 14.9 0.6 a 61.1 2.1 aSEA15 Stress 3.59 0.14 b 40.4 2.2 b 25.7 0.8 a 10.4 0.5 b 58.2 1.1 a Means in the same column having same letters in common are not significantly different according to LSD test at 5% level of probability. As reported earlier, the number of pods per plant destined for final harvest to a large extent determined the differences in yielding levels of the tested genotypes under drought conditions. Five days after the commencement of drought stress, Brown Speckled had higher pod numbers per plant than SEA 15 under both soil moisture supply regimes (Fig. 1). Nevertheless, relative to the initial pod number, the number of productive pods retained per plant at 20 d stress was considerably more reduced for Brown Speckled 29 (32%) than for SEA 15 (49%) (Fig. 1). In fact, the drought-susceptible genotype had a higher rate of pod abortion than the resistant one under control conditions, too. For the susceptible genotype, the absolute number of aborted pods per plant increased significantly due to drought stress, whilst the plants of the resistant genotype maintained a similar number of aborted pods under the contrasting soil moisture supply regimes (data not shown). 0 5 10 15 20 25 30 35 40 All pods Productive pods P o d s (n o p la n t -1 ) BrSp Control BrSp Stress SEA15 Control SEA15 Stress Fig. 1 The effect of drought stress imposed at early pod-filling stage on pod set (total number of pods counted at 5 d stress) and number of productive pods (pods possessing at least one seed at the end of 20 d stress) per plant of two common bean genotypes. Mean-values for each pod category having same letters in common are not significantly different according to LSD test at 5% level of probability. Vertical bars show S.E. of four replications. Numbers above bars are percentages of productive pods (Pr-P) relative to total pod set counted at 5 d stress. † All pods = productive pods + aborted pods. During the initial genotype screening experiment, drought stress reduced the harvest indices of the adapted cultivars (Roba 1, SEA 15 and BAT 881) between 21 and 52% (Table 1). No such effect was found for the inbred lines. When the two selected genotypes a a ab b b c a b 39 32 55 49 † 30 (Brown Speckled and SEA 15) were compared for the same parameter determined 20 d after the imposition of drought stress at pod-filling stage, the reduction owing to drought for Brown Speckled was about 29% relative to the control treatment (Table 3). SEA 15, on the other hand, maintained comparable harvest indices under the contrasting soil moisture regimes (Table 3). In fact, the inherent difference in the harvest indices of the two bean genotypes was substantial. Under both growth conditions, biomass remobilization ability assessed using harvest index values was much lower for Brown Speckled compared with SEA 15. 3.2. Effects of drought stress on growth and biomass production The effects of genotype and soil moisture regimes were highly significant for above- ground fresh and dry weights determined 5 and 10 d after the imposition of drought stress during the vegetative phase. Compared with Brown Speckled, SEA 15 accumulated significantly higher above-ground biomass under both growth conditions (Table 4). Relative to non-stressed treatments, drought stress caused significant reductions in above- ground biomass yield in the range of 36 - 40% for SEA 15 and 16 - 29% for Brown Speckled. Likewise, drought stress imposed at early pod-filling stage of the crop significantly reduced above-ground fresh and dry weights (leaves + stems + pods + seeds) of both genotypes at all durations of stress considered (Table 4). Drought-induced reductions in above-ground fresh weight (30 - 36% for Brown Speckled vs. 21 - 37% for SEA 15) and dry weight (24 - 33% for Brown Speckled vs. 17 - 29% for SEA 15) were comparable between the genotypes. Leaf biomass (with up to 40 and 50% reduction for Brown Speckled and SEA 15, respectively) was the most affected fraction of above-ground dry biomass yield due to drought stress (Fig. 2). On the other hand, stem dry weight was relatively unaffected due to drought stress, in general, and for the drought-resistant cultivar, in particular. Unlike the other components of biomass, the difference in seed yield biomass ratio of drought- 31 stressed to control treatments between SEA 15 (0.70) and Brown Speckled (0.47) was significant (Fig. 2). Table 4. The effect of drought stress imposed at vegetative and pod-filling stages on above-ground fresh and dry biomass weights (g plant -1 ) of two common bean genotypes. Data are the means S.E. of four replication. Vegetative phase Reproductive phase Duration of stress (d) Treatment Fresh wt. Dry wt Fresh wt. Dry wt Control 44.3 2.3 b 6.1 0.2 c 162.1 8.5 a 27.3 0.5 a Br Sp Stress 33.6 0.9 c 5.1 0.2 c 113.9 5.6 c 20.9 0.9 b Control 72.7 5.9 a 11.6 0.8 a 133.3 2.3 b 24.9 1.0 a5 SEA 15 Stress 41.4 0.8 bc 7.4 0.2 b 105.5 8.7 c 20.7 1.4 b Control 67.1 2.6 b 9.9 0.4 b 153.2 2.7 a 28.5 0.6 aBr Sp Stress 44.8 3.3 c 7.0 0.5 c 99.9 3.9 b 20.3 0.8 b Control 93.7 5.1 a 17.8 0.8 a 144.5 8.1 a 30.7 1.8 a10 SEA 15 Stress 54.7 0.5 c 10.7 0.1 b 93.8 0.7 b 21.7 0.6 b Control - - 134.4 3.3 a 35.4 0.7 aBr Sp Stress - - 85.5 4.3 c 23.7 1.3 b Control - - 102.8 7.3 b 24.3 0.6 b20 SEA 15 Stress - - 65.2 3.0 d 17.8 0.5 c Mean values within the same column and for the same duration of stress having similar letters in common are not significantly different according to LSD test at 5% level of probability. Under drought as well as non-stress growth conditions, SEA 15 maintained higher reproductive (pods + seeds) to vegetative (leaves + stems) mass ratio than Brown Speckled (Fig. 3). Soil moisture supply regime, however, did not affect the partitioning of 32 Fig. 2. Biomass ratio (drought tissue / control tissue) of leaves, stems, pods, reproductive organs (pod wall + seeds), above-ground biomass weight (AGBW) (leaves + stems + reproductive materials) and seed yield of two common bean genotypes at different durations of stress. For the calculation of the biomass ratio, g of dry weight was the unit used. Vertical bars show S.E. of four replications. * Difference between the two genotypes is significant at 5% level of probability according to t-test. 0.0 0.2 0.4 0.6 0.8 1.0 1.2 BrSp SEA15 0 .0 0 .2 0 .4 0 .6 0 .8 1 .0 1 .2 0 .0 0 .2 0 .4 0 .6 0 .8 1 .0 1 .2 * 5 d stress 10 d stress 20 d stress Leaf Stem Pod Reprod. AGBW Seed walls organs Yield D ro u g h t- to -c o n tr o l b io m as s ra ti o 33 0 1 2 3 4 5 10 20 5 10 20 Duration of stress (d) R ep ro d u ct iv e- to -v eg et at iv e b io m as s ra ti o Control Stress Fig. 3. The effect of drought stress initiated at pod-filling stage on reproductive to vegetative biomass ratio of two common bean genotypes. Vertical bars show S.E. of four replications. biomass between the vegetative and reproductive parts of both genotypes. During the final harvest (20 d stress), the biomass weight of reproductive organs was two and three- fold larger than the vegetative parts for drought-stressed and non-stressed SEA 15 plants, respectively (Fig. 3). On the contrary, the dry weights of reproductive and vegetative structures of Brown Speckled remained fairly proportional under both growth conditions (Fig. 3). Main effects due to genotype and soil moisture regime were significant for leaf area and specific leaf weight (SLW) determined during the vegetative growth phase of the crop. SEA 15 reacted to the stress imposed with an enormous leaf area reduction (by about 65% relative to the control treatment at both sampling times) compared with Brown Speckled, which encountered only ca. 40 % reduction (Fig. 4A). A drought stress period of 10 d during the vegetative phase significantly increased (by ca. 16%) the specific leaf weight (SLW) of SEA 15 relative to the control treatment (Fig. 4B). On the other hand, drought BrSp SEA 15 34 stress did not significantly alter the SLW of Brown Speckled at both sampling times. SLW exhibited significant correlation with leaf dry matter content (LDMC, ratio of leaf fresh weight to dry weight) (r = 0.98, p<0.01) but not with leaf area (r = - 0.27, p>0.05). 0 15 30 45 60 75 90 5 d stress 10 d stress L ea f ar ea ( cm 2 ) BrSp Control BrSp Stress SEA15 Control SEA15 Stress 0 10 20 30 40 50 60 5 d stress 10 d stress S p ec if ic l ea f w ei g h t (g m -2 ) BrSp Control BrSp Stress SEA15 Control SEA15 Stress Fig. 4. The effect of drought stress imposed at vegetative stage on leaf area (A) and specific leaf weight (B) of two common bean genotypes. Means followed by the same letter during the same duration of stress are not significantly different according to LSD test at 5% level of probability. Vertical bars are S.E. of four replications. 3.3. Effects on vegetative and reproductive growth rates Relative to the control treatments, drought stress significantly reduced the vegetative absolute growth rate (AGR) computed on shoot dry weight basis by about 38 and 47% for Brown Speckled and SEA 15, respectively (Fig. 5). Relative growth rates (RGR) of both genotypes were also negatively affected by drought stress (Table 5). Drought stress imposed during vegetative growth phase reduced the net assimilation rate (NAR) of Brown Speckled and SEA 15 by ca. 18 and 28%, respectively (Table 5). Leaf area ratio (LAR) is a composite parameter, determined partly by allocation (leaf weight a c b d a b a b c bc ab a c c b a A B 35 0.0 0.4 0.8 1.2 1.6 Control Stress Control Stress Br.Sp. SEA 15 A b so lu te g ro w th r at e (g d -1 ) Fig. 5. The effect of drought stress imposed at vegetative phase on the absolute growth rate of two common bean genotypes. Vertical bars show S.E. of four replications. Table 5. The effect of drought stress initiated during the vegetative growth phase on relative growth rate (RGR), net assimilation rate (NAR), leaf area ratio (LAR), specific leaf area (SLA) and leaf weight ratio (LWR) of two common bean genotypes. Data are means ( 10 -2 ) S.E. of four replications. Treatment RGR (g g -1 d -1 ) NAR (g m -2 d -1 ) LAR (m 2 g -1 ) SLA (m 2 g -1 ) LWR (g g -1 ) Control 9.46 0.78 a 450.4 49.6 a 2.12 0.09 a 3.05 0.16 a 69.6 0.6 a BrSp Stress 7.17 0.26 b 366.8 11.9 b 1.95 0.04 ab 2.89 1.00 ab 67.7 1.3 a Control 10.56 0.03 a 594.8 21.1 a 1.78 0.06 bc 2.68 0.06 b 66.5 1.8 a SEA15 Stress 7.27 0.33 b 424.4 29.7 b 1.72 0.04 c 2.92 0.07 ab 59.0 0.5 b Means having similar letter within the same column are not significantly different according to LSD test at 5% level of probability. ratio, LWR) and partly by leaf morphology (specific leaf area, SLA). Brown Speckled maintained higher LAR than SEA 15 under drought as well as control conditions (Table 36 5). Whereas drought stress did not have a significant impact on specific leaf area (SLA) of both genotypes, the leaf weight ratio (LWR) component of LAR was significantly reduced for SEA 15 due to drought stress (Table 5). RGR correlated significantly with NAR (R 2 = 0.81, p<0.05) but not with leaf area ratio (LAR) (R 2 = 0.18, p>0.05). Absolute growth rate (AGR) and relative growth rate (RGR) of reproductive structures (dry weights of aborted pods + pod walls + seeds) were also computed to examine whether the bean genotypes maintain similar growth rates when subjected to drought stress at different growth phases. AGR of the reproductive structures due to drought stress was significantly reduced by 49 and 32% for Brown Speckled and SEA 15, respectively (Fig. 6). On the other hand, the decrease in RGR caused by drought was significant only for Brown Speckled. Compared with Brown Speckled, SEA 15 generally maintained higher AGR and RGR of reproductive structures under both soil moisture supply regimes. Fig. 6. The effect of drought stress imposed at pod-filling stage on absolute growth rate (AGR) and relative growth rate (RGR) of reproductive structures (pod walls + aborted pods + seeds) in two common bean genotypes. Vertical bars show S.E. of four replications. 0.00 0.04 0.08 0.12 0.16 0.20 Control Stress Control Stress BrSp SEA15 R G R ( g g -1 d -1 ) 0.0 0.4 0.8 1.2 1.6 2.0 A G R ( g d -1 ) RGR AGR 37 3.4. Effects on plant-water relations 3.4.1. Relative water content The leaf relative water content (RWC) of Brown Speckled was influenced by drought stress at neither the vegetative nor the reproductive growth stages (Fig. 7A, B). On the contrary, SEA 15 maintained significantly lower RWC under drought stress relative to control treatments at 5 d stress during the vegetative phase and during most sampling times of the reproductive phase (Fig. 7A, B). RWC as a key reference parameter of leaf water status, exhibited a positive and significant correlation with net photosynthetic rate (r = 0.54, p < 0.05) and stomatal conductance (r = 0.57, p < 0.01) during the reproductive phase. However, the degree of relationship between RWC and net photosynthetic rate (A) was smaller (R 2 = 0.33, p<0.01) as compared with the relationship between A and stomatal conductance (gs) (R 2 = 0.89, p < 0.01) (Fig. 8A, B ). 0 20 40 60 80 100 5 d stress 10 d stress R W C ( % ) BrSp Control BrSp Stress SEA15 Control SEA15 Stress 60 70 80 90 100 2 4 6 8 2 4 6 8 BrSp SEA15 Duration of stress (d) R W C ( % ) Control Stress Fig. 7. Leaf relative water content (RWC) of two common bean genotypes under drought stress and non-stress growth conditions during vegetative (A) and reproductive (B) growth phases. Means followed by the same letter during the same duration of stress are not significantly different according to LSD test at 5% level of probability. Vertical bars are S.E. of four replications. a ab a b a a ab b A B BrSp SEA 15 38 y = 52.19x + 1.37 R 2 = 0.87** 0 2 4 6 8 10 12 0.00 0.05 0.10 0.15 0.20 gs (mmol m -2 s -1 ) A ( µ m o l m -2 s -1 ) y = 0.33x - 22.61 R 2 = 0.47** 0 2 4 6 8 10 12 70 75 80 85 90 95 RWC (%) A ( µ m o l m -2 s -1 ) Fig. 8. The relationship of stomatal conductance (gs) and leaf relative water content (RWC) with net photosynthetic rate (A) of two common bean genotypes grown under drought stress (imposed at reproductive phase) and non-stress growth conditions. * Significant at 1 % level of probability. A marked difference was found between the two genotypes for the maintenance of pod relative water content under drought stress (Fig. 9). Drought-induced reduction in pod water concentration of Brown Speckled was higher than that of SEA 15 (Fig. 9). Twenty days after drought stress was initiated at early pod-filling stage, pod water concentration dropped from ca. 86% (determined at 5 d stress for both genotypes) to 54 and 74% for Brown Speckled and SEA 15, respectively (Fig. 9). 3.4.2. Leaf water potential The tested genotypes exhibited significant differences for water potential and its components under drought stress imposed at early flowering stage (Table 6). Drought- induced reductions in leaf water potential ( ) determined at early pod-filling stage (ca. 15 d after drought stress was commenced) were significant only for Brown Speckled and BAT 881 (Table 6). Under drought stress, SEA 15 (–1.17 MPa) and Mexican 142 (–1.03 MPa) maintained the highest and lowest leaf water potentials, respectively. Higher solute accumulation due to drought stress enabled the tested genotypes (except Brown Speckled and SEA 15) to maintain significantly lower osmotic potentials ( s), which ranged from 39 –2.05 to –1.86 MPa (Table 6). Except for Mexican 142 and BAT 881, drought stress did not significantly affect the turgor pressures ( p) of the bean genotypes. 40 60 80 100 5 10 20 5 10 20 Duration of stress (d) P o d w at er c o n ce n tr at io n (% ) Control Stress Fig. 9. Pod water concentration of two common bean genotypes grown under drought stress and non-stress growth conditions. *, ** The differences between the drought- stressed and non-stressed treatments are significant at 5 and 1% levels of probability, respectively. Vertical bars show S.E. of four replications. Table 6. The effect of drought stress imposed at early flowering stage on leaf water potential ( ), osmotic potential ( s) and turgor pressure ( p) of six common bean genotypes. (MPa) s (MPa) p (MPa) Genotype Control Stress Control Stress Control Stress Mex.142 -0.70 -0.83 -1.67 -2.05** 0.98 1.22** Roba 1 -0.78 -0.88 -1.67 -1.89** 0.90 1.01 Br.Speckl. -0.76 -0.94* -1.78 -1.93 1.03 0.99 SEA 15 -1.03 -1.17 -1.62 -1.86* 0.59 0.69 SEA 23 -0.85 -0.99 -1.91 -2.12 1.06 1.13 BAT 881 -0.99 -1.19* -1.96 -2.41** 0.98 1.22* *, ** The differences between drought-stressed and non-stressed treatments are significant at 5 and 1% levels of probability, respectively. ** * * * BrSp SEA15 40 3.5. Effects on water-use and water-use efficiency (WUE) Under non-stress growth conditions, SEA 15 consumed ca. 36% more water than Brown Speckled during a ten-day period of the vegetative phase (Fig. 10A). Nonetheless, the amount of water used by the two genotypes was more or less comparable under drought stress (Fig. 10A). The effects of genotype and soil moisture supply regimes were highly significant for water-use efficiency (WUE, mg dry matter produced per g water used) determined at vegetative growth stage of the crop. Drought stress imposed during the same period increased WUE by about 35 and 37% for Brown Speckled and SEA 15, respectively (Fig. 10B). Nevertheless, the increase in WUE owing to drought stress during the vegetative phase was significantly higher for SEA 15 (3.12 mg g -1 ) compared with Brown Speckled (2.45 mg g -1 ). 0.0 0.4 0.8 1.2 1.6 2.0 Control Stress Control Stress BrSp SEA15 W at er c o n su m p ti o n ( k g p la n t -1 ) 0 1 2 3 4 Control Stress Control Stress BrSp SEA15 W U E ( m g D W g -1 H 2 O ) Fig. 10. Water consumption (A) and water-use efficiency (B) of two common bean genotypes under 10 d drought stress and non-stress growth conditions during the vegetative phase. Mean values for each parameter having same letter in common are not significantly different according to LSD test at 5% level of probability. Vertical bars are S.E. of four replications. A B d b c a b c a c 41 The amount of water consumed (from emergence to maturity) by the inbred lines SEA 15 and SEA 23 was significantly less than the old adapted cultivars under drought stress as well as non-stress growth conditions (Table 7). Also, the two inbred lines used more than two-third and one-ha