Heat tolerance in food legumes as evaluated by cell membrane thermostability and chlorophyll fluorescence techniques


Key words: chickpea, chlorophyll fluorescence, groundnut, heat tolerance, membrane stability, photosystem, pigeonpea, screening methods, soya bean, thermostability -----


The genotypic variation for heat tolerance in chickpea, groundnut, pigeonpea, and soya bean was evaluated by testing membrane stability and photosystem (PS II) function in leaves at high temperatures. The legumes could be ranked from heat-tolerant to sensitive in the order: groundnut, soya bean, pigeonpea and chickpea. The damage to cell membranes (as reflected by an increased leakage of electrolytes) and PS II (as reflected by a decrease in the ratio of variable to maximum fluorescence) was less, and recovery from heat stress was faster in groundnut than in other crops. Prior exposure of plants to 35oC for 24 h led to a reduced leakage of electrolytes at high temperatures in all crops but the differences among legumes were consistent. Substantial genotypic variation for heat tolerance was found in all legumes. Membrane injury was negatively associated with specific leaf weight in groundnut (r= -0.69**) and soya bean (r= -0.56**) but not in the pulses. Electrolyte leakage and fluorescence ratio were negatively correlated in all legumes. The potential use of electrolyte leakage and fluorescence tests as screening procedures for breeding heat-tolerant legumes is discussed. -----

Abbreviations: RI -- relative injury; Fo -- initial fluorescence; Fm -- maximum fluorescence; Fv -- variable fluorescence; PS II -- photosystem II; PAR --photosynthetically active radiation -----


High temperatures (>30oC) limit growth and adaptation of legumes in many countries (Brown, 1960; Ketring, 1984; Wery et al., 1994). Heat stress is, therefore, a major cause for the unstable and low seed yields (~1 t ha-1) that are far below the potential yields of 7-8 t ha-1 in groundnut (Arachis hypogaea L.) and soya bean (Glycine max L. Merr.), and 4 t ha-1 in chickpea (Cicer arietinum L.) and pigeonpea (Cajanus cajan L. Millsp.). Further, frequency of hot weather may increase in future due to global warming (Schneider, 1989). Improvement of heat tolerance can contribute to sustainability and provides a means of extending legume cultivation to previously unsuitable regions and seasons. For instance, development of heat-tolerant pigeonpea enables its sowing early in summer, thereby allowing timely sowing of wheat and leading to high yields in pigeonpea-wheat rotation in northwestern India (Table 1).

Limited efforts have been made in breeding heat-tolerant legumes, perhaps because yield losses due to heat are more subtle than those due to disease or insect infestations (Summerfield et al., 1984). The two most serious barriers are the lack of (a) information on the range in genetic diversity for heat tolerance, and (b) screening techniques (Wery et al., 1994). Screening for seed yield under heat stress is one method but it needs full-season field data and is always not convenient or efficient, especially in mesic locations due to the variation in the timing and intensity of stress. Further, a negative feature as regards the selection for yield, is its low heritability in hot and dry environments (Blum, 1988). Therefore, the development of screening methods based on traits involved in heat reactivity is strongly desirable.

A screening method is effective if it can show distinct differences in injury to a tissue or process. Plant responses to heat stress are diverse, and include cessation of cytoplasmic streaming (Alexandrov, 1964), protein denaturation (Bernstam, 1978), changes in lipid composition (Suss & Yordanov, 1986), reduction in membrane stability (Shen & Li, 1982) and efficiency of photosynthesis (Bar-Tsur et al., 1985), etc. The relative importance of each can vary with species however. Membrane dysfunction is a physiological process disturbed most by heat stress (Levitt, 1980; Quinn, 1989). It results in an increased permeability and leakage of electrolytes, which, in turn, reduces photosynthetic or mitochondrial activity, and the ability of plasmalemma to retain solutes and water (Lin et al., 1985). The electrolyte leakage test was used to examine variation for heat tolerance in common bean (Schaff et al., 1987) and soya bean (Sapra & Anaele, 1991) but the relative tolerance of legumes under uniform growing conditions has not been assessed.

Electron transfer from photosystem II (PS II) is also heat-sensitive. Measurements of chlorophyll fluorescence can give quantitative assessment of inhibition or damage to electron transfer (Baker et al., 1989). The technique is rapid, sensitive, non-destructive, relatively cheap, and able to detect injury even before visible symptoms appear (Wilson & Greaves, 1990). Hall (1992) considered that fluorescence properties are more suitable for screening than photosynthetic rates as the former reflect both time-integrated damage to PS II and time-integrated effects on photosynthetic rates. Chlorophyll fluorescence has been employed to screen for heat tolerance in potato (Smillie & Hetherington, 1990) and wheat (Moffat et al., 1990) but the usefulness of the technique has not been examined in legumes.

The objective of this work was to assess variation for heat tolerance among four legumes (chickpea, groundnut, pigeonpea, and soya bean) and among genotypes in each legume by evaluating differences in electrolyte leakage and PS II efficiency at high temperatures. -----

Materials and methods

Plant material. Seeds of chickpea, groundnut, pigeonpea, and soya bean were obtained from the International Crops Research Institute for the Semi-Arid Tropics (ICRISAT), India and the National Institute of Agrobiological Resources, Japan. The selection of genotypes was based on the climatic conditions likely to occur in the regions (including the sites that experience high temperature) where the genotypes were developed, or were successfully grown. Genotypes differing in plant types, maturities and adaptation to various climates were also chosen.

Plant culture. Seeds were sown in vinyl pots filled with vermiculite. Three to five days after emergence, two seedlings were transplanted into a plastic pot containing 3 kg powdered Kaolinitic hyper-thermic Ultisol. The soil was mixed with lime at 5 g kg-1, and NPK fertilizer (14-14-14) at 0.5 g kg-1. Plants were grown in a glasshouse set at 25+1.0oC (day/night).

Cell membrane thermostability analysis. The procedure was similar to that of Martineau et al. (1979). In groundnut, pigeonpea and soya bean, 10 leaf discs cut with a 13 mm punch were used. In chickpea, alternate pinnules were allocated to various treatments while ensuring that all treatments received equal amounts of tissue. Leaf tissues were washed with 3-4 changes of distilled water and placed in a test tube (150x25 mm) containing 2 ml of pre-heated (to the treatment temperature) water. Tubes were covered with plastic wrap and placed in a water bath at the desired temperature for 15 min., while the control tubes were kept at 25oC. After cooling to room temperature, distilled water was added to make up the final volume of 10 ml. Samples were incubated at 10oC for 16 h and conductivity measured with a conductance meter (CM-115, Kyoto Electronics, Japan). The tubes were covered with aluminum foil and autoclaved at 120oC for 15 min. to release all electrolytes. After cooling tubes to 25oC, the contents were mixed and final conductance measured. The injury was determined as follows:

Relative Injury[RI] (%) = {1-[1-(T1/T2)]/[1-(C1/C2)]} * 100

where T and C refer to the conductance in treatment and control tubes, and subscripts 1 and 2 refer to readings before and after autoclaving respectively. The conductance in treatment tubes is a measure of electrolytes leaked from cells due to high temperature and is assumed to be proportional to the degree of injury to membranes. The control gives a measure of leakage due solely to the cutting and incubation of leaf discs. Incubation of chickpea pinnules at 10oC for 16 h led to a high leaching of electrolytes even in the controls. Therefore, while assessing genotypic variation, initial conductivity was measured after shaking pinnules in distilled water for 2 h at room temperature. The remaining procedure was the same as above.

Fully expanded leaves from 30-day old seedlings of chickpea (cv. Annigeri), groundnut (cv. Florunner), pigeonpea (cv. ICPL 85010) and soya bean (cv. Bragg) were first used to measure RI at 35, 40, 45, 50 and 55oC. After identifying the critical range of temperatures, RI was measured at temperatures with a narrow interval of 2oC. The temperature at which 50% injury occurs (heat killing temperature) was determined. The differences in heat killing time (the duration for leakage of 50% of electrolytes at 50oC), and heat acclimation potential (the extent of reduction in RI upon subjecting plants to 35oC for 24 h) were then determined. Genotypes in each species were compared by assessing RI at the chosen temperature for 15 min. Leaves of similar age were sampled during vegetative, flowering and pod filling stages. Simple correlations between RI and specific leaf weight, and solute concentration were calculated. Specific leaf weight (g m-2) was determined after measuring area and dry weight.

Chlorophyll fluorescence analysis. Fluorescence properties of dark-adapted (for 45 min.) leaves were measured using a modulated fluorimeter (Model MK II MFMS/2S, Hansatech Electronics Ltd., UK.). A leaflet was placed in a clip and fiber-optic cables connected to a Bjorkman lamp were used to send actinic and saturating light. Initial (Fo) and maximum (Fm) fluorescence were recorded, and variable fluorescence, Fv, derived by subtracting Fo from Fm. Fv/Fm ratios were then calculated. In each replicate, Fv/Fm ratio was derived as the mean of 10-12 measurements. Initially, the differences in recovery of Fv/Fm were measured in vivo after exposing 45-day old plants of chickpea (cv. Annigeri), groundnut (cv. ICG 1236), pigeonpea (cv. ICPL 84023) and soya bean (cv. Sooty) to 50oC for 4 h in an incubator (Eyelatron, FL1 301NH, Tokyo Rikakikai Co. Ltd., Japan) under 250 Ámol m-2 s-1 PAR. Plants were then shifted to a glasshouse set at 25+1oC (day/night) under natural light. Fluorescence properties were measured at various intervals in the control and treated plants.

Simultaneously, a method to measure fluorescence values using detached leaves was developed. Fully expanded leaves were excised between 9.00 and 9.30 am (to avoid effects due to diurnal variations in tolerance), placed in water, and brought to the laboratory. Leaves of similar age and nodal position(s) were used to ensure uniformity. Leaflets were individually placed in test tubes containing 2 ml distilled water and heated for 5 min. in a water bath (35- 60oC). In another trial, leaflets were held in air by enclosing the base in a wet cotton plug fitted on to the open end of the test tube. Tubes were placed in a thermal gradient chamber (Shimadzu TG 100-AD, Nippon Medical and Chemical Instruments Co. Ltd., Tokyo) and exposed to high temperatures for 5 min. Because results from both tests were similar, the former method was employed due to its simplicity. After identifying temperatures at which genotypic differences were distinct, in vitro measurements were conducted on all genotypes. In pigeonpea, the changes in Fv/Fm ratio with time were monitored in vivo in two genotypes, ICPL 84031 and ICPL 85045, upon transfer to a warm regime (35/30oC day/night).

Data analysis. All data were subjected to standard analysis of variance procedures using IRRISTAT 92 (IRRI, 1992) software. Duncan's multiple range test at the 0.05 probability level was used to separate treatment means. Simple correlations were used to determine the strength of association between two variables. -----


Cell membrane thermostability. There were clear differences among legumes in RI at and above 40oC (Fig. 1a). Injury was low even at 50oC in groundnut while it was high even at 40oC in chickpea. Pigeonpea and soya bean showed similar RI values. In all species, RI seemed to be a sigmoidal function of temperature. Exposing tissues to critical temperatures (38-48oC for chickpea, 46-56oC for pigeonpea and soya bean, and 48-58oC for groundnut) showed distinct differences in temperature that caused 50% RI. Heat killing temperatures in chickpea, pigeonpea, soya bean and groundnut were 44.3, 49.9, 50.8 and 54.0oC respectively (Fig. 1b). Heat killing time was much longer in groundnut (139 min.) than in soya bean (51 min.), pigeonpea (47 min.) or chickpea (41 min.) (Fig. 1c). Exposure of plants to 35oC for 24 h led to a decreased RI in all crops but the differences among species were consistent (Table 2).

Genotypic variation in RI was evaluated at 45oC for chickpea, 50oC for pigeonpea, 51oC for soya bean, and at 54oC for groundnut, with a uniform treatment duration of 15 min. In chickpea, Annigeri, ILC 482, and ICCV 10 were more thermostable than other genotypes. In contrast, K 850, a genotype grown in central and northern India, showed high sensitivity. Injury values decreased with the development of the crop (Table 3). In groundnut, Virginia types (Florunner, Virginia Bunch 67, and Penghu 1) showed less injury than Spanish ones. However, ICG 1236 was an exception. The varieties originating from the southern regions of the former Soviet Union (Chico and Tashkent 112), and Senegal (Senegal 48-15) showed high RI. Injury at different stages was largely similar (Table 4).

Substantial variation in RI was found in pigeonpea at 50oC. ICPLs 85010 and 84031, and OK86-282 had low RI but all wild species showed high RI (Table 5). In soya bean, RI was low in Sooty, Biloxi, Peking, and AGS-2 but it was high in Improved Pelican, Kitamishiro, Tracy, and Hawkeye. The origin of genotypes seemed to have no relation with the degree of tolerance. For example, cold tolerant genotypes from Sweden (Fiskeby V) and Russia (Dobruzhanka 18) showed less RI than the genotypes from the subtropical parts of Thailand (Chieng Mai SKG) and USA (Harosoy) (Table 6). In all crops, genotypic ranking was inconsistent across stages but a few genotypes were superior throughout.

Chlorophyll fluorescence

Variation among legumes in recovery of Fv/Fm ratio following heat stress (in vivo)

Exposure of plants to 50oC for 4 h led to a sharp decrease in Fv/Fm ratio in all crops except groundnut. The differences were distinct soon after heat stress. Groundnut showed least damage while chickpea showed maximum damage (54% of the control) to PS II. The recovery in groundnut was fast, with complete recovery to the control level occurring within 3 h, while it took nearly 24 h in soya bean and pigeonpea, and about 48 h in chickpea (Fig. 2).

Variation among legumes in fluorescence properties measured on detached leaves (in vitro)

Distinct differences in chlorophyll fluorescence properties were seen upon exposure of leaves to temperatures above 45oC. In chickpea and pigeonpea, Fo rose with an increasing temperature. In contrast, the rate of increase in Fo was low in groundnut (Fig. 3a). Fm declined sharply in all species but the reduction was less in chickpea and pigeonpea especially beyond 50oC, largely due to an increased Fo (Fig. 3b). Fv/Fm ratio declined with an increasing temperature in all species but the magnitude of reduction was more in chickpea and pigeonpea than in soya bean and groundnut (Fig. 3c).

The differences among species were consistent in both in vitro and in vivo tests, although a greater decline in Fv/Fm ratio was observed in detached than in intact leaves. However, as in vitro tests are plant-conserving, and are easier and quicker than in vivo tests, and enable comparison of many genotypes using limited amount of leaf tissue, genotypic variation was evaluated by in vitro tests.

Genotypic variation in each legume

In chickpea, Annigeri, ICCC 37, ICCV 92901 and ILC 482 maintained a higher Fv/Fm ratio than others at 45oC. ICCV 88202 was the most sensitive (Table 3). In groundnut, ICG 1236, Florunner and Virginia Bunch maintained a higher Fv/Fm ratio than others at 55oC (Table 4). Shulamit, a cultivar known for high water use efficiency, was as sensitive to high temperatures as Chico. ICGS 44, a variety recommended for cultivation in the post-rainy season in India, also showed high sensitivity. Among cultivated genotypes of pigeonpea, ICPLs 84031, 84023 and 85010 maintained a higher Fv/Fm ratio than others at 50oC. ICPL 85045 was the most sensitive (52% of the control). All wild species were, in general, more sensitive than the cultivated ones but Atylosia sericea was superior to the other two species (Table 5). In soya bean, Sooty, Tropical, and AGS-2 maintained a higher Fv/Fm ratio than others at 50oC. In contrast, genotypes such as Dobruzhanka 18 and Karikachi showed significant reduction in the fluorescence ratio (Table 6).

The Fv/Fm responses of contrasting pigeonpea genotypes (ICPLs 84031 and 85045) upon transfer to a warm regime (35/30oC day/night) were compared. Genotypic differences in Fv/Fm ratio were not evident during the first 7 days of transfer but they became distinct after 10 days. Fv/Fm ratio in the sensitive genotype, ICPL 85045, decreased gradually to reach about 72% of the control by the 20th day of transfer. The decrease in the tolerant genotype, ICPL 84031, was much less (Fig. 4). Leaves in ICPL 85045 senesced earlier than in ICPL 84031 (data not shown). The flush of new leaves that developed in the warm regime in both genotypes, however, maintained high Fv/Fm ratios (data not shown).

RI was negatively associated with specific leaf weight in groundnut(-0.69**) and soya bean (-0.56**) but not in the pulses. In contrast, RI was not related to solute concentration in all legumes (data not shown). Fv/Fm ratio and RI were negatively correlated in all crops (chickpea, -0.69*; groundnut -0.87***; pigeonpea, -0.69** and soya bean, -0.57**).



The studies have shown that groundnut was the most tolerant, and chickpea the most sensitive, in terms of membrane stability and PS II function at high temperatures. Groundnut is native to South America but it is now grown throughout the tropical, subtropical and warm temperate regions (40oN to 40oS) indicating its wider adaptation. The high sensitivity of chickpea is perhaps not surprising as it is the only cool season legume tested here. Malhotra & Saxena (1993), however, reported that critical temperatures for heat tolerance may be higher in chickpea than in legumes such as lentil, pea and faba bean. Pigeonpea and soya bean seem to have similar levels of leaf heat tolerance.

Significant variation in RI values among species and genotypes indicates that the electrolyte leakage test is sensitive to evaluate heat tolerance and that the trait is amenable through breeding. Although high replication is necessary to achieve a small standard error, the test shows promise as a screening method. Preliminary tests showed that a minimum of 8 discs must be used to reduce variability within the genotype. In our study, 10 discs per replicate were used. Evidence from other crops suggests that tolerant genotypes selected by this test perform well and give stable yields in hot environments. For instance, Saadalla et al., (1990) reported that heat tolerant genotypes of wheat, determined on the basis of electrolyte leakage, outyielded sensitive ones by 19% under field conditions. Kuo et al. (1993) showed that vegetable species with low RI were more stable in different growing seasons.

Both morphological and biochemical traits determine cell membrane stability. Electrolyte leakage is affected by differences in leaf structure (MacRae et al., 1986), cell wall composition (Jarvis et al., 1988), the degree of membrane's lipid saturation (Tal & Shannon, 1983), etc. In general, leaves in groundnut are thicker (250-295 Ám) than in soya bean (220-255 Ám), pigeonpea (140-170 Ám) and chickpea (150-190 Ám). Further, the high solute concentration per unit leaf area, and thick, waxy cuticle on the abaxial side of the leaf in groundnut may help in maintaining a better osmotic balance and improved water status at high temperatures. Sutter & Langhans (1982) showed that water loss was higher in cabbage plants without structured epicuticular wax than in plants with epicuticular wax.

Like electrolyte leakage, chlorophyll fluorescence analysis is simple, rapid, reproducible and many samples can be analyzed within a short time. These advantages make the test ideal for large scale screening of populations. The results suggest that the temperatures from 45 to 55oC are suitable for determining genotypic variation. The substantial variation in each legume again indicates the possibility of manipulating this trait through breeding. Smillie & Hetherington (1990) and Wilson & Greaves (1990) suggested that a large number of samples is required for fluorescence analysis to reduce variability. In our studies, we found that 10-12 measurements per replicate, and 4-5 replicates are adequate to detect genotypic variation.

Heat treatment (50oC for 4 h) had only marginal adverse effects on Fv/Fm ratio in groundnut, indicating its thylakoid membrane integrity and stable photochemical efficiency (Krause & Weiss, 1984). In contrast, a drastic reduction in Fv/Fm ratio was seen in the pulse crops. Such inter-specific differences may be due to variations in ability to keep repair mechanisms functional (Smillie et al., 1988). Thermal inactivation of PS II may be related to 3 different effects (Weiss & Berry, 1988). These include (a) heat-induced disturbance of lateral distribution of pigment complexes in thylakoid membranes, (b) damage to oxygen evolving system that interrupts electron donation to PS II as reflected by a decline in Fv, and (c) and an irreversible injury reflected by a sharp rise in Fo. The results confirm the occurrence of the latter two in pigeonpea and chickpea. An increase in Fo presumably reflects the physical dissociation of the PS II reaction center from the light harvesting system (Sundby et al., 1986). In contrast, the decrease in Fo in groundnut and soya bean probably indicates the occurrence of non-radiative energy dissipation (Demmig et al., 1987). Fluorescence properties thus suggest that groundnut possesses two mechanisms to cope with heat stress i.e., rapid recovery (perhaps due to less injury), and operation of photoprotective mechanisms.

The response of Fo seems complex and may differ depending on the material and conditions. Demmig & Bjorkman (1987) reported that photoinhibition related to a decreased Fo recovered fully whereas photoinhibition that leads to an increased Fo was largely irreversible. Somersalo & Krause (1989), however, showed that recovery occurred in spinach even when Fo increased. Our results support the latter findings as the Fv/Fm ratio in pulses recovered to the control within 1-2 days, despite an increase in Fo immediately after heat stress.

In our study, both in vitro and in vivo tests gave largely similar results. Potvin (1985) reported that leaf detachment damaged PS II extensively and questioned the use of such leaves to probe stress responses. Smillie et al. (1987), however, showed that both detached and attached leaves behaved uniformly if a similarity in treatments was strictly maintained. In vitro tests are advantageous in that they are plant-conserving and this feature is important in breeding programmes. For most crops, comparisons of the heat sensitivity must be done with detached leaves as it is not possible to apply a controlled stress to field-grown plants. However, in vitro tests may not be of much use to test recovery mechanisms as there is no further supply of photosynthates or water necessary for recovery from other plant parts.

In conclusion, the present study has shown that wide genetic diversity for leaf heat tolerance exists among legumes and that electrolyte leakage and fluorescence tests are useful to screen for heat tolerance. Studies are in progress to determine if tolerant genotypes selected by these tests perform well at high temperatures. -----



If you have comments or need further information, please email me at ancha

Last revised