Introduction
In India, available fodder for stock is estimated to be 4050% below requirements, and this scenario is gradually worsening due to the concomitant decrease in grass coverage and increase in livestock population (Indian Council of Agricultural Research 2009). Global climate change in the last decade has been correlated with changes in the productivity of forage grasses and is likely to have a detrimental effect on the overall grass coverage in the long term (Abberton et al. 2008). A huge proportion of land in the country is classified as wasteland due to the problems of soil salinity, alkalinity and waterlogging. The selection of grass germplasm for salinity tolerance is critical for more efficient utilization of these degraded lands by establishing stress-tolerant grasses in non-arable marginal areas (Ashraf 2006). Species that are relatively salt-tolerant show greater endurance and adaptability among the native species (Squires 2015). Therefore there is an urgent need to: identify salt-tolerant traits in wild forage grasses; evaluate their potential for enhancing the productivity of grasslands in their native habitats; and utilize them for the rejuvenation of grasslands and croplands with reduced or lost productivity.
Abiotic stresses, in particular water and salinity stress, play a major role in disrupting the growth and development of grasses including cereals (Tester and Bacic 2005). Salinity limits plant growth and productivity through the toxic effects of Na+ and Cl- ions, which leads to ionic imbalances, osmotic and oxidative stress (Munns and Tester 2008). Native grasses, however, show variable degrees of NaCl tolerance, especially those belonging to the subfamilies Panicoideae and Chloridoideae (Bromham and Bennett 2014; Roy and Chakraborty 2014). Salinity tolerance is a complex trait, governed by several physiological and biochemical parameters and these parameters greatly influence the normal growth and development of plants (Zhu 2000). Salt tolerance of any individual species is demonstrated as the ability to maintain an optimal physiological and biochemical equilibrium under NaCl treatment (Sairam and Tyagi 2004). Ashraf and Harris (2004) suggested different biomarkers as indicators of salinity tolerance, including soluble sugars, proteins, amino acids, ammonium compounds, polyamines, polyols, antioxidants and ATPases.
In the present study however, 6 biochemical markers, viz. relative water content (RWC), proline and soluble sugar concentrations, membrane lipid peroxidation (malondialdehyde, MDA), electrolyte leakage (EL) and H2O2 concentration were selected for use in screening for salinity tolerance of the selected grasses. Increase in leaf RWC in the halophyte Atriplex nummularia with increasing salinity indicated an efficient mechanism to adjust cell cytosol osmotically (Araújo et al. 2006). Accumulation of osmolytes like proline, soluble sugars and glycine betaine and elevated levels of antioxidative enzymes play a vital role in conferring salt tolerance in grasses (Roy and Chakraborty 2014). Accumulation of glycine betaine in Cynodon and Spartina, proline in Paspalum and myo-inositol in Porteresia has been found to confer salinity tolerance (Wyn Jones and Storey 1981; Marcum and Murdoch 1994; Sengupta et al. 2008). Accumulation of proline, fructans and soluble carbohydrates was also correlated with salinity tolerance in salt-tolerant cultivars of wheat (Kafi et al. 2003). MDA concentration has been proposed as an indicator of oxidative damage and a lesser accumulation of the same in root tissues was employed for screening the salttolerant genotypes of Cenchrus ciliaris (Castelli et al. 2009). Electrolyte leakage as an indicator of cell membrane stability of durum wheat cultivars under osmotic stress was demonstrated, with level of electrolyte leakage being inversely related to degree of salt tolerance of cultivars (Bajji et al. 2002).
In addition to the characterization of 12 forage grasses that are widely grazed by and fed to livestock in the eastern parts of the Terai-Duar grasslands by observing the changes in 6 biomarkers for salinity tolerance, the objective of our study was to evaluate the salt-tolerance potential of those grasses by using a rapid screening technique where the inherent tolerance of saline conditions was assessed as a precursor to selective propagation in varied environmentally challenged wastelands.
Materials and Methods
Study area and plant materials
Twelve native grasses were collected from the different regions of the eastern part of the Terai-Duar grasslands (88.22-89.66° E, 26.45-26.86° N; Figure 1). These grasses are widely grazed by livestock and harvested by local people for feeding to domestic animals, viz. Arundo donax L. of the subfamily Arundinoideae; Axonopus compressus (Sw.) P. Beauv., Capillipedium assimile (Steud.) A. Camus, Chrysopogon aciculatus (Retz.) Trin., Digitaria ciliaris (Retz.) Koeler, Arundinella bengalensis (Spreng.) Druce, Imperata cylindrica (L.) Raeusch., Oplismenus burmanni (Retz.) P. Beauv., Setaria pumila (Poir.) Roem. & Schult. and Thysanolaena latifolia (Roxb. ex Hornem.) Honda of the subfamily Panicoideae; and Cynodon dactylon (L.) Pers. and Eragrostis amabilis (L.) Wight & Arn. of the subfamily Chloridoideae. In the subsequent text only the generic names are used.
Experimental design and NaCl treatment
A rapid screening protocol was implemented for the differentiation of salt-tolerance potential of the forage grasses. The grasses were collected from their natural habitats and placed in small flasks containing 0.1X Hoagland solution with their roots intact, before being transferred to the plant growth chamber in the laboratory of the Department of Botany, University of North Bengal, Siliguri. Before NaCl treatment, the roots were gently washed with sterile dH2O to remove any mud and then again transferred to conical flasks containing 0.1X Hoagland solution. The plants were then allowed to acclimatize for 48 hours in the growth chamber, with a standard temperature of 20-25 °C, RH 65-70% and 16 h photoperiod. Following acclimatization, 2 groups of plants were grown in NaCl treatments of 100 and 200 mM for 9 days, while the third group remained as control and the effects of NaCl on the plants in terms of several biomarkers after 3, 6 and 9 days of treatment were analyzed.
Three individual samplings from 3 different locations (Figure 1) were completed for each grass and the results were expressed as mean ± SD for all parameters analyzed. For grasses with broad leaves like Thysanolaena and Arundo, 3 plants were taken per sampling site, whereas for grasses with small narrow leaves, 5-6 plants were taken per sampling site.
Salt sensitivity index (SSI)
The youngest healthy fully expanded leaves from the plants were briefly washed in deionized water and 1 cm diameter leaf discs were finely cut and floated in a 5 ml solution of NaCl (100 and 200 mM) for 96 hours. Leaf discs floated in sterile dH2O served as the experimental control for the bioassay (Fan et al. 1997). The effects of salt treatment on leaf discs were assessed by observing the phenotypic changes and the extent of NaCl effect in terms of SSI, which was quantified by estimating the chlorophyll concentration in NaCl-treated and control sets. Briefly, the leaf discs were crushed in 80% acetone and the absorbance was recorded in a UV-VIS spectrophotometer at 645 and 663 nm and the chlorophyll concentration was calculated using Arnon's formulae (Arnon 1949). SSI values were then calculated at 100 and 200 mM NaCl as the percent decrease in chlorophyll concentration of the NaCl treatment in comparison with the untreated leaf discs using the following formula:
Biochemical markers for assessment of NaCl tolerance
For an alternative screening of grasses for their salt-tolerant attributes, 6 different biochemical parameters were chosen, viz. relative water content (RWC), proline and soluble sugar concentrations, membrane lipid peroxidation (malondialdehyde, MDA), electrolyte leakage (EL) and H2O2 concentration. For these experiments, the first 3 fully expanded leaves from the top of each grass subjected to the various growth solutions were collected.
Relative water content. RWC was measured following the protocol of Barr and Weatherley (1962). Briefly, fresh leaf samples from control and different treatment sets were weighed to obtain fresh weight (FW). The samples were then immediately hydrated to full turgidity for 4 h, dried of surface moisture and weighed to obtain fully turgid weight (TW). Samples were then oven-dried at 80 °C for 24 h and weighed to determine dry weight (DW). RWC was calculated by the following equation:
RWC (%) = [(FW - DW) / (TW - DW)] x 100
Proline. Extraction and estimation of proline were done by the method of Bates et al. (1973). Leaf tissue was homogenized in 3% sulfosalicylic acid. Ninhydrin reagent was used for the estimation of proline in the extract, which was separated in a separating funnel using toluene, prior to recording the absorbance at 520 nm.
Total sugar. Soluble sugar in leaves was extracted in 95% ethanol following the method of Harborne (1973). Anthrone reagent was used to estimate total sugar following the method of Plummer (1978). Briefly, 4 ml of anthrone reagent was added to 1 ml test solution and kept over boiling water bath for 10 min, after which the absorbance was taken at 620 nm. Total sugar was finally calculated using a standard curve of D-glucose.
Membrane lipid peroxidation. Membrane lipid peroxidation was measured in terms of concentration of malondialdehyde (MDA) produced by the thiobarbituric acid (TBA) reaction, following the method of Heath and Packer (1968). Leaves were homogenized in 0.1% (w/v) trichloroacetic acid (TCA) and estimation was done with 0.5% (w/v) TBA in 20% TCA. The absorbance of the reaction mixture was determined at 532 and 600 nm and the MDA content was calculated using an extinction coefficient of 155 mM/cm.
Electrolyte leakage. Electrolyte leakage (EL) was measured as described by Lutts et al. (1996). Leaves were washed thoroughly with deionized water and placed in culture tubes containing 10 ml of deionised water on a rotary shaker for 24 h. Subsequently, the electrical conductivity of the solution (Lt) was determined and the samples were then autoclaved at 120 °C for 20 min and cooled to room temperature before determining the final electrical conductivity (L0). EL was calculated as follows:
Electrolyte leakage (%) = (Lt / L0) x 100
H2O2 concentration. The extraction and estimation of H2O2 were done by the method given by Jana and Choudhuri (1981) with slight modification. Leaf tissue was homogenized in 50 mM phosphate buffer (pH 6.5) and mixed with 0.1% titanium sulphate in 20% (v/v) H2SO4 and centrifuged at 6,000 rpm for 15 min. Absorbance was measured at 410 nm and H2O2 concentration was measured using the extinction coefficient of 0.28 umol/cm.
Hierarchical cluster analysis
For cluster analysis of the grasses for their NaCl tolerance, the data for fold change values of RWC, proline, soluble sugar, MDA, EL and H2O2 after NaCl treatments for 3, 6 and 9 days with respect to the control sets were taken. Hierarchical cluster analysis was performed using the CLUSTER 3.0 program by the uncentered matrix and complete linkage method following the protocol of de Hoon et al. (2004). The resulting tree figure was displayed using the software package, Java Treeview, as described by Chan et al. (2012).
Statistical analysis
All experiments were repeated with sampling from 3 different locations (n = 3) for each species. Species and treatment means were statistically analyzed using Least Significant Difference (P<0.05) for a completely randomized design.
Results
Salt sensitivity index (SSI) of grasses
Chlorophyll concentration in fresh untreated leaves varied from 0.72 mg/g (Capillipedium) to 1.45 mg/g (Oplismenus). SSIs of grasses determined by leaf disc assay and represented in terms of % decrease in chlorophyll concentration in the leaf discs floated in 100 mM and 200 mM NaCl solutions relative to the control sets, i. e. leaf discs kept in sterile dH2O, are shown in Table 1. At 100 mM NaCl, the senescence assay indicated that Setaria, Thysanolaena, Imperata and Cynodon were least affected with SSI values of 0.45-7.36. At the same time, Capillipedium, Axonopus and Arundinella were much more sensitive (SSI values of 24.20-18.37). However, at 200 mM NaCl, Imperata, Digitaria and Cynodon were least affected by salt concentration (SSI values of 6.5915.00). Interestingly, Thysanolaena and Setaria were more affected by 200 mM NaCl, showing marked increases in SSI values (23.38 and 57.98, respectively). Capillipedium showed the highest sensitivity to both 100 and 200 mM NaCl with SSI values of 24.20 and 61.93, respectively. This result was also reciprocated by the phenotypical changes in the leaf discs floated in NaCl solutions, which can be clearly observed in Figure 2.
Values for chlorophyll concentration are mean ± SD (n = 3). Greater values of salt sensitivity index denote greater sensitivity or susceptibility to NaCl, whereas lower values denote lesser sensitivity.
Effect of NaCl on biochemical markers for analysis of salinity tolerance
Relative water content. Leaf RWC values were found to decrease in all grasses with both increase in NaCl concentration and duration of treatment (Table 2). The fold change values of RWC in plants subjected to 100 and 200 mM NaCl in comparison with the control sets revealed the smallest changes in Cynodon and Imperata and the largest changes in Chrysopogon and Digitaria (Figure 3a).
1LSD (P<0.05) Species _ 2.23; Treatment _ 1.12. 2LSD (P<0.05) Species _ 3.41; Treatment _ 1.7. 3LSD (P<0.05) Species _ 5.19; Treatment = 2.59. Values represent mean ± SD, where n = 3.
Proline concentration. Proline concentration in fresh untreated leaves varied from 11.6 ug/g (Chrysopogon) and 12.4 ug/g (Setaria) to 63.1 ug/g (Imperata) and 64.5 Ug/g (Digitaria). During the first 3 days of NaCl treatment (100 and 200 mM), proline concentration in fresh tissue increased with increase in NaCl concentration in all grasses except Axonopus, where levels of proline declined (Table 3; Figure 3b). The largest increases (on a percentage basis) were recorded in Cynodon, Arundinella and Imperata. Similarly after 6 and 9 days of treatment, proline concentrations increased as NaCl concentration increased in all grasses except Axonopus, Chrysopogon, Thysanolaena and Oplismenus, where concentrations declined with increasing NaCl concentration. The largest percentage increases in proline concentration were observed in Cynodon and Arundinella (1.8-3-fold increase).
1LSD (P550.05) Species = 40.82; Treatment = 20.41. 2LSD (P^0.05) Species = 41.82; Treatment = 20.91. 3LSD (P^0.05) Species = 40.15; Treatment = 20.07. Values represent Mean ± SD, where n = 3.
Total sugar concentration. Concentration of sugars in untreated fresh leaves varied from 16.1 mg/g (Capillipedium) to 56.9 mg/g (Eragrostis). Changes in concentration followed no consistent pattern across the various grasses subjected to NaCl treatments (Table 4; Figure 3c), with some showing decreases while a few showed increases. Those showing greatest decreases were Capillipedium (69% decrease) and Oplismenus (45% decrease), with most of the grass species showing little change in sugar concentration over the 9 days, even at 200 mM NaCl.
1LSD (P550.05) Species = 4.96; Treatment = 2.48. 2LSD (P^0.05) Species = 7.24; Treatment = 3.62. 3LSD (P^0.05) Species = 6.92; Treatment = 3.46. Values represent Mean ± SD, where n = 3.
Membrane lipid peroxidation. MDA concentration in untreated fresh leaves varied from 2.2 mM/g (Chrysopogon) to 11.9 mM/g (Arundo). Concentrations showed a consistent pattern, increasing across all concentrations and durations of NaCl treatment in all grasses with greater responses to increasing concentration than to increasing duration of exposure (Table 5; Figure 3d). After 9 days, greatest increases in MDA concentration occurred in Chrysopogon (5-fold), Capillipedium (3-fold) and Axonopus (2.4-fold).
1LSD (P550.05) Species = 3.34; Treatment = 1.67. 2LSD (P^0.05) Species = 5.07; Treatment = 2.53. 3LSD (P^0.05) Species = 6.25; Treatment = 3.12. Values represent Mean ± SD, where n = 3.
Electrolyte leakage. Electrolyte leakage levels in untreated fresh leaves varied from 5.1% (Arundinella) to 15.5% (Setaria) and increased across all concentrations and durations of NaCl treatment in all grasses (Table 6; Figure 3e). Arundo and Capillipedium showed the greatest increases in electrolyte leakage with exposure to NaCl treatment with a much greater response to increasing concentration (80-90%) than to duration of exposure (10-24%). The lowest responses occurred with Cynodon and Imperata.
1LSD (P550.05) Species = 2.6; Treatment = 1.3. 2LSD (P^0.05) Species = 2.53; Treatment = 1.27. 3LSD (P^0.05) Species = 2.82; Treatment = 1.41. Values represent Mean ± SD, where n = 3.
H2O2 concentration. Concentrations of H2O2 in untreated fresh leaves ranged from 2.4 umol/g (Chrysopogon) to 11.8 umol/g (Digitaria and Thysanolaena) and increased across all concentrations of and durations of exposure to NaCl solutions for all grasses (Table 7; Figure 3f). The most responsive grasses were Chrysopogon, Capillipedium and Arundo, while the least responsive were Cynodon and Imperata.
1LSD (P550.05) Species = 2.1; Treatment = 1.05. 2LSD (P^0.05) Species 2.42; Treatment = 1.21. Values represent Mean ± SD, where n = 3.
Hierarchical cluster analysis for the evaluation of NaCl tolerance
Based on the variable effects of NaCl treatment on biochemical parameters, the grasses were grouped according to their NaCl tolerance through hierarchical cluster analysis, where the fold change values of all parameters were taken into consideration (Figures 3a-3f).
The ranges of fold change values in the clusters are represented by the colored bars. Results suggested the probable interrelations among biochemical parameters subjected to NaCl stress and variable salt tolerance between all grass genera.
Based on their salt sensitivities, the grasses formed 2 distinct groups (Figure 4). One group was comprised of Axonopus, Chrysopogon, Oplismenus and Thysanolaena. The remaining grasses with varying response patterns to NaCl solutions formed the second group and were classified into 3 subgroups: Arundo and Capillipedium; Arundinella and Setaria; and Digitaria, Cynodon, Eragrostis and Imperata.
Discussion
This rapid screening for salinity tolerance in the forage grasses has been attempted as a simple method of identifying the most salt-tolerant grasses for introduction into areas with increasing soil salinity and decreasing productivity. Previously, Zulkaliph et al. (2013) in their studies with turfgrasses ranked the different species of grasses for salinity tolerance on the basis of shoot and root growth, leaf firing, i.e. yellowing of leaves resulting from cell death due to osmotic imbalances, turf color and turf quality. We estimated salinity tolerance of the grasses primarily by a salt sensitivity index (SSI), determined by evaluating the effects of NaCl solutions on leaf discs over 96 hours. This type of bioassay has been used previously in several transgenesis experiments to evaluate the tolerances of transgenic plants relative to the wild type plants from which they were bioengineered (Bhaskaran and Savithramma 2011; Yadav et al. 2012).
The amount of chlorophyll leached out from the leaf discs into the NaCl solution was used as an indicator of the effect of NaCl on leaf tissues. The decrease in chlorophyll concentration in plants subjected to NaCl treatment has been inversely correlated with salinity tolerance. For instance, the decrease in Chlorophyll a: Chlorophyll b ratio in salt-tolerant Najas graminea was lower than in Hydrilla verticillata and Najas indica (Rout et al. 1997). In the present study, we quantified the amount of chlorophyll in the leaf discs in both control and treatment sets and the values were used to reciprocate the sensitivity of grasses towards NaCl treatment. Greater salt sentivity index values denoted greater susceptibility of the grasses towards NaCl. Overall, the results of the bioassay indicated that among the grasses tested, Imperata, Cynodon and Digitaria could be considered as less sensitive or resistant on the basis of SSI values at 100 and 200 mM NaCl. SSI therefore presents an easy and rapid technique to screen out the potential salt-tolerant forage grasses.
The 6 biomarkers we selected to analyze the salt-tolerance potential of the forage grasses, namely relative water content (RWC), proline and soluble sugar concentrations, membrane lipid peroxidation, electrolyte leakage and H2O2 concentration, proved useful in indicating differences between species in ability to tolerate saline conditions both simply and rapidly.
While RWC of any plant always decreases with the increase in NaCl concentration, a lower decrease in RWC is a valuable marker in the selection of salt-tolerant species (Ziaf et al. 2009). In our study, lowest decreases in RWC were observed in Cynodon, Eragrostis and Imperata across all concentrations and durations of NaCl treatments, identifying them as salt-tolerant species. In contrast, accumulation of proline and soluble sugars is considered to be positively correlated with salinity tolerance (Karsensky and Jonak 2012; Hayat et al. 2012). Accumulation of higher levels of proline has been reported in the halophytes, Mesembryanthemum crystallinum and Sporobolus virginicus when compared with the glycophytes carrot and rice (Thomas et al. 1992; Tada et al. 2014). In the present study, apart from Axonopus, Chrysopogon and Oplismenus, proline accumulation increased in all grasses subjected to NaCl treatment. We also observed that soluble sugar accumulation decreased in Arundo, Axonopus, Capillipedium, Oplismenus, Setaria and Thysanolaena across all concentrations of NaCl and durations of exposure. In contrast, accumulation of soluble sugars increased in Digitaria, Imperata and Arundinella subjected to NaCl treatments for 3, 6 and 9 days. Nedjimi (2011) also correlated the accumulation of greater amounts of soluble sugars in the forage grass Lygeum spartum with osmotic adjustment and protection of membrane stability that conferred salinity tolerance.
Increase in malondialdehyde (MDA) concentration, an indication of lipid peroxidation, is considered unfavorable for plant health, and plants, which show little increase in MDA concentration when exposed to NaCl, are considered to be salt-tolerant (Miller et al. 2010). Marked increases in MDA concentration were observed in Axonopus, Capillipedium, Chrysopogon and Thysanolaena, following exposure to salt. However, minimal increase was observed in Cynodon and Eragrostis across all concentrations and durations of treatment.
Similarly, low electrolyte leakage (EL) and limited increase in H2O2 concentration in response to NaCl treatment are also considered as markers of the salt tolerance of plants (Mostafa and Tammam 2012). Accumulation of H2O2 in plants interferes with the normal biochemical processes inside plants. In the present study, EL in all grasses increased with the increase in NaCl concentration and duration of treatment. Least EL was observed in Cynodon, Imperata and Arundinella, which could be considered salt-tolerant species in comparison with the other grasses. The high increases in H2O2 concentration observed in Arundo, Axonopus, Capillipedium and Chrysopogon indicate that these species can be considered susceptible to salination on the basis of this trait. Comparatively, low increases in H2O2 concentration observed in Imperata, Setaria and Cynodon indicate that they can be considered salt-tolerant.
Finally, hierarchical cluster analysis using the software CLUSTER 3.0 was used to represent the inter-relations among the physiological parameters and to align the grasses on the basis of their salinity tolerance as a similar type of hierarchical cluster analysis has been performed to evaluate the natural variation in drought tolerance in bermuda grass (Shi et al. 2012) and the variation in salt tolerance in rice cultivars (Chunthaburee et al. 2016). In the present study we utilized the relative fold change values of all the parameters in forming clusters. Based on the variations of the physiological parameters, all grasses were grouped according to their NaCl tolerance that could be interpreted with the aid of the fold change values denoted by colored bars. The relationships between the physiological parameters themselves was also illustrated in the cluster analysis. The grasses were clearly divided into 2 groups - a susceptible group (Axonopus, Chrysopogon, Oplismenus and Thysanolaena) and a relatively salt-tolerant group containing the remaining grasses. Critical analysis of the second group revealed 3 subgroups of less tolerant (Arundo and Capillipedium), moderately tolerant (Arundinella and Setaria) and tolerant grasses (Digitaria, Cynodon, Eragrostis and Imperata). These results are in accordance with the findings of other workers who reported the use of some of these and other related, tolerant grasses for the reclamation and utilization of saline soils and increased forage production (Kaffka 2001; Weber and Hanks 2006).
Based on the results of hierarchical clustering, we conclude that Imperata cylindrica, Eragrostis amabilis, Cynodon dactylon and Digitaria ciliaris were relatively salt-tolerant. SSI values individually pointed towards the superior salt-tolerance of Imperata, Digitaria and Cynodon, whereas proline concentration indicated marked tolerance in Cynodon, Arundinella, Imperata, Eragrostis and Setaria. If we consider the MDA concentrations, Cynodon, Arundinella, Imperata and Eragrostis could be considered salt-tolerant. Thus, while individual biochemical markers provide good indications of the degree of salt tolerance of a species, cluster analysis, which incorporates the results with several biomarkers, provides a much more reliable indication. However, SSI values can provide an easy and rapid tool for the screening of salt tolerance. Based on our screening results, we consider that the selective propagation of the most salt-tolerant species could be utilized for the rejuvenation of native grasslands and also for the reclamation of salinity infested wastelands.