Excreting and non-excreting grasses exhibit different salt resistance strategies

Salt marsh grasses are adapted to thrive under saline conditions by various combinations of traits. Some researchers suggested that salt excreting grasses would differ from non-excreting ones in these traits. However, little is known about the differential responses between these plant types. Here, we compared the growth and physiology of salt excreting and non-excreting grasses. Differences were found between the two grass types in leaf water content, accumulation of organic compounds and Na+ distribution which appeared to be linked with salt excretion. Additional studies on a number of halophytic grasses could help to identify key traits for salt resistance.


Introduction
Halophytic grasses in arid and semi-arid areas dominate salt marsh vegetation where soil salinity varies between 1 and 30 dS m 21 and occasionally at or above seawater salinity (Gulzar and Khan 1994). Considerable variations in salinity, moisture and nutrients allow different species to occupy distinct ecological niches. Phylogenetic studies indicate that salt tolerance evolved more than 70 times in grasses and is supported by studies on intraspecific variations in physiological, morphological and biochemical responses under saline conditions (Bennett et al. 2013).
Salt resistance is a complex trait (Flowers and Colmer 2008;Flowers et al. 2010) and represents a continuum between glycophytes and halophytes (Bonales-Alatorre et al. 2013). Decreased water uptake, ion toxicity, nutrient imbalance, reduced photosynthesis and subsequent production of reactive oxygen species usually lead to growth reduction under saline conditions. Reduced biomass allocation to shoot versus root is reported to ensure longterm survival by improved water and ion balance and reduced salt accumulation in the rhizosphere (Passioura et al. 1992). Grasses generally tend to maintain low Na + in aboveground tissues in comparison with dicotyledonous halophytes by restricting Na + uptake (Flowers and Colmer 2008). Salt tolerance in grasses was found to be negatively correlated with shoot [Na + ] (Marcum et al. 1998). The higher K + /Na + selectivity and potassium-use efficiency of grasses compared with most dicotyledonous species (Flowers and Colmer 2008) is essential for survival under saline conditions. More than 30 grass species from about 16 genera mostly in the tribes Chlorideae, Sporoboleae and Aeluropodeae are known to remove salt from the leaf surface by means of bi-cellular epidermal salt glands (Liphschitz and Waisel 1974;Marcum 2001;Kobayashi and Masaoka 2008;Lefè vre et al. 2009). Salt resistance has been associated often with efficient salt excretion from leaves (Marcum et al. 1998) and also with increased water-use efficiency (WUE), a typical feature of C 4 grasses (Naidoo et al. 2012).
Plants need to maintain water uptake through osmotic adjustment under saline conditions and are known to accumulate carbon (C)-rich (sugars, sugar alcohols) and nitrogen (N)-rich (proline, glycinebetaine, trigonelline) organic osmolytes to counterbalance salt entry into the plant (Munns 2002(Munns , 2011Koyro et al. 2011). Accumulation of N-rich compounds also makes aboveground tissues more palatable for herbivores (Elser et al. 2000). However, increased salinity may hinder N uptake because of competition with Cl 2 and Na + accumulation (Taiz and Zeiger 2006) and reduce C uptake because of reduction of CO 2 /H 2 O gas exchange rates. Therefore, salinity resistance of halophytic species was also associated with efficient N allocation towards synthesis of organic compounds (Geissler et al. 2009) which in turn is linked with processes for C assimilation into biomass (Hussin et al. 2013).
Salt tolerance is a complex phenomenon which is a manifestation of multigenic traits at the whole-plant level. However, little information is available on small-scale differences among salt-excreting (Aeluropus lagopoides and Sporobolus tremulus) and non-excreting (Paspalum paspalodes and Paspalidium geminatum) C 4 salt marsh grasses. The salt-excreting species are characteristically found in more saline sandy soils (ECe . 1.5 dS m 21 ) while the non-excreting ones appear to prefer clayey, less salty substrates (ECe , 1.0 dS m 21 ). To the authors knowledge no salt tolerance studies have been carried out on latter three grass species and previous work on salt tolerance of A. lagopoides (Gulzar et al. 2003) was related only to growth and water status but not to its nutrient status or synthesis of compatible osmolytes under saline stress. The present research is an attempt to scale up from the physiological/biochemical level to understand the whole-plant salt resistance mechanisms of four C 4 grasses growing naturally in saline marshes. We hypothesize that salt-excreting grasses will differ from nonexcreting ones in terms of (i) growth, (ii) water relations, (iii) ion regulation and (iv) nitrogen-use efficiency (NUE). The results should also provide some explanations about their distribution in natural populations. Therefore, we compared the relative salt resistance in terms of growth, water and ion relations, C/N ratios and NUE of C 4 salt excreting (A. lagopoides and S. tremulus) (Chloridoideae) and non-excreting (P. paspalodes and P. geminatum) (Panicoideae) grasses.

Experimental conditions
Ramets of P. paspalodes and P. geminatum were collected from Korangi, Karachi (24851 ′ 03.2 ′′ N; 67805 ′ 60.4 ′′ E), while A. lagopoides and S. tremulus were collected from Manora Creek near Sandspit, Karachi (24849 ′ 06.70 ′′ N; 66856 ′ 06.80 ′′ E). Tillers were potted in plastic pots (26 cm high×20 cm diameter) in sand culture and watered daily to a constant volume by sub-irrigation with half strength Hoagland solution (Epstein 1972) to establish for 1 month. Salt treatments (0, 200 and 400 mM NaCl) were introduced gradually with 100 mM NaCl increments every 24 h to avoid osmotic shock. Final concentrations in trays used for sub-irrigation were maintained daily by adding distilled water to compensate for evaporation. In plastic pots this was achieved by flushing soil from above with the respective nutrient solution at 3-4-day intervals allowing them to drip from below. The treatment solutions were replaced every fifth day. Plants were cut at 15 cm above the soil surface when final salt concentrations were reached which were maintained for another 45 days before the final harvest.

Growth parameters
Plants were carefully removed from the soil and roots were washed with the respective solution and then dipped twice in distilled water for a few seconds and wiped with tissue paper. Roots were separated from shoots and then measured for fresh weight. Plant samples were placed in a microwave oven for about 6 min to determine dry weight (Popp et al. 1996). A half-filled 100-mL beaker with distilled water was also placed alongside the samples to avoid burning. Plant samples were allowed to cool down in a desiccator before measuring the dry weights. Relative growth rate (RGR) was calculated using the formula: where W 1 and W 2 are the initial and final dry weights whereas t 1 and t 2 are the initial and final time in days from the start of salinity treatments.

Water relations
Leaf succulence on a dry weight basis was measured using the equation: Relative water content (RWC) was found out with the help of the formula: where FW is the fresh weight; DW the dry weight and TW the turgid weight of leaves after rehydration in distilled water for 24 h at room temperature ( 25 8C). Leaf osmotic potential was calculated using van't Hoff equation (Kramer and Boyer 1995) on osmolality of expressed leaf sap measured by a vapour pressure osmometer (VAPRO-5520; Wescor Inc., Logan, UT, USA) (Gucci et al. 1991). Xylem pressure potential (XPP) was measured on excised stems with a plant water status console Model 1000 (PMS Instrument Co., Albany, NY, USA). Instantaneous WUE was calculated from the rate of CO 2 fixation (A) which was measured using a Li-6400XT portable photosynthesis system (LICOR Biosciences) per amount of water transpired (E) from the leaf surface as Cation contents (Na 1 , K 1 Ca 21 and Mg 21 ) Hot-water extracts were prepared with homogenized finely ground dry plant material in deionized water at 100 8C (Khan et al. 2000) in capped Pyrex test tubes. Soluble Na + , K + , Ca 2+ and Mg 2+ in shoot and root were determined on dilutions of the hot water extracts by atomic absorption spectrometry (AA-700; Perkin Elmer, Santa Clara, CA, USA).
Selectivity of K 1 , Ca 21 and Mg 21 over Na 1 Selective absorption (SA) and selective transport (ST) of K + , Ca 2+ and Mg 2+ over Na + were calculated according to Wang et al. (2005) as follows: where 'X' stands for K + , Ca 2+ or Mg 2+

Organic osmolytes
Total soluble sugars were determined using the method of Yemm and Willis (1954). A 1-ml dilution of hot water extract in distilled water was added to 5 mL Anthrone's reagent in a Pyrex test tube and incubated in a water bath at 100 8C for 30 min. The reaction was terminated in an ice bath and the absorbance was recorded at 630 nm with a spectrophotometer (Beckman DU-530 spectrophotometer; Beckman Coulter Inc., USA). Proline was determined according to Bates et al. (1973) on 200 mL of hot water extract made up to 2 mL with distilled water which was mixed with ninhydrin : glacial acetic acid (1 : 1 v/v) and incubated at 100 8C for 1 h. The reaction was terminated on an ice bath and the chromophore was extracted with 4 mL toluene. Proline content was measured at 520 nm on a spectrophotometer (Beckman DU-530 spectrophotometer, Beckman Coulter Inc.). Glycinebetaine was determined by using dilutions of hot water extract in 50 mM potassium dihydrogen phosphate buffer (4.5 pH), the mobile phase for the high-performace liquid chromatography. The samples were filtered through a 0.45-mm membrane filter before injecting into a Nucleosil HPLC Column (4.6 × 250 mm) on a JASCO Intelligent UV/ VIS HPLC system (JASCO, Japan) at a flow rate of 1.2 mL min 21 at 25 8C. Estimations were based on standard curves of 1, 10 and 100 mmol dilutions of glycinebetaine (Khan et al. 1999).

Carbon, nitrogen analyses and NUE
Carbon and nitrogen were determined on 0.4 mg fine ground dry shoot material mixed with tungsten trioxide (catalyst) wrapped in a tin foil boat. The sample boats were loaded on the auto sampler of a CNS elemental analyzer Vario EL III equipped with a thermal conductivity

Statistical analyses
Statistical analysis was carried out using SPSS Ver. 11.0 for Windows (SPSS Inc., Chicago, IL, USA) (SPSS 2006). Twoway analysis of variance (ANOVA) was used to test for significant differences among species, salinity and their interactions. One-way ANOVA was used to reveal significant differences across salinity treatments within individual species while a post hoc Bonferroni test was used to test for significant differences between individual treatment means.

Growth parameters
Two-way ANOVA indicated a significant decrease in shoot and root biomass, shoot and root length and RGRs in all grass species (P , 0.001) with increase in salinity (P , 0.05) (Figs. 1-3). Shoot length was affected more than root growth in all grasses (Fig. 3). Dry mass of P. geminatum was reduced by 60 % in 400 mM NaCl compared with those grown in 0 mM NaCl. Plant biomass reduction was 35 and 50 % for the salt-excreting grasses (A. lagopoides and S. tremulus), respectively, but only 23 % for P. paspalodes (Fig. 2). The root-to-shoot biomass allocation ratio was quite variable among species (F ¼ 9.2; P , 0.001; Table 1) which registered a salt-stimulated increase in P. paspalodes (P , 0.05). Salt-excreting grasses displayed lower and generally invariable root-to-shoot biomass allocation ratios than non-excreting ones, whereas a marked reduction (50 %) in the root-to-shoot biomass ratio was noted under saline conditions in P. geminatum (P , 0.05;

Water relations
Leaf succulence increased in salt-excreting grasses and it decreased in non-excreting grasses with increase in NaCl concentrations (Table 2). Relative water content increased in salt-excreting grasses, remained unchanged in P. paspalodes but decreased sharply in P. geminatum.
Water-use efficiency of salt-excreting grasses was unaffected while it increased (P , 0.0001) in P. paspalodes and declined (P , 0.0001) in P. geminatum (Table 2) with increase in NaCl concentrations. Leaf osmotic potential (c s ) and XPP showed significant differences among species (P , 0.0001) and were particularly lower in the salt-excreting grasses (Fig. 4). Effects of salinity and species × salinity interactions were highly significant for c s (P , 0.001, P , 0.05 respectively) and XPP (P , 0.001).
Cation contents (Na 1 , K 1 Ca 21 and Mg 21 ) Na + content of shoot in all test grasses increased substantially (F ¼ 16.89; P , 0.001) compared with root with increasing salinity. K + was generally higher in shoots than in roots but declined sharply in S. tremulus (F ¼ 7.10; P , 0.01) shoots at 400 mM NaCl (Fig. 5). Shoot/root Na + ratios also increased with concomitant decreases in shoot/ root K + ratios in salt-excreting grasses but remained unchanged in non-excreting grasses. Shoot Na + /K + ratios increased significantly (P , 0.05) in salt-excreting grasses at 400 mM NaCl but did not vary in the non-excreting grasses (Fig. 6). K + , Ca 2+ and Mg 2+ content in shoot and root of our test grasses under saline treatments was not lower than their respective non-saline controls.
Selectivity of K 1 , Ca 21 and Mg 21 over Na 1 Selective absorption of K + , Ca 2+ and Mg 2+ against Na + increased in all test species (P , 0.001) and with increasing NaCl concentrations (P , 0.001; Table 3). Selective transport of K + and Mg 2+ against Na + towards shoot varied among test species (P , 0.001) and salinity treatments (P , 0.05) except for ST of Ca 2+ which was unaffected by salinity (Table 3).

Carbon, nitrogen analyses and NUE
Median nutrient (N and C) concentrations were measured at 38.5 % for C and 1.4 for N with considerable variations in S. tremulus (lower C, higher N) and P. paspalodes (lower C and N) [see Supporting Information]. In general, nonexcreting grasses had a high (.30) C/N ratio compared with salt-excreting grasses (Fig. 7) which decreased due to increasing shoot N from 0.8 % in P. paspalodes to about the median value of 1.4 % in P. geminatum (Fig. 8A). Sporobolus tremulus had the highest N at 3 % of dry weight (Fig. 8A). Nitrogen-use efficiency (on dry weight basis) was substantially (P , 0.001) higher in nonexcreting grasses than in the salt-excreting ones but did not vary with salinity treatments except for a decline in P. geminatum. Shoot C/N ratios were linearly correlated (R 2 ¼ 0.81) with NUE across all species and salinity treatments. Salt-excreting grasses displayed lower values for C/N and NUE relative to non-excreting grasses ( Fig. 8B; Table 4).

Growth and biomass allocation
We found species-specific variations in eco-physiological traits for salt resistance in four perennial C 4 grasses. However, traits related to ion transport and nutrient allocation on dry weight basis appeared to be related to their excreting versus non-excreting nature. Salinity inhibited the growth of test species in the following order: P. geminatum . S. tremulus ¼ A. lagopoides . P. paspalodes. The upper limit of salt resistance in salt-excreting grasses  varied from 300 mM NaCl (Halopyrum mucronatum; Khan et al. 1999) to 1000 mM NaCl (A. lagopoides and Urochondra setulosa; Gulzar and Khan 2006;Ahmed et al. 2013) and in non-excreting grasses it was up to 500 mM NaCl (Panicum turgidum, Koyro et al. 2013; Phragmitis karka, Abideen et al. 2014). The growth of P. geminatum was Figure 6. Shoot/root ratios of Na + and K + and Na + /K + ratios of the salt-excreting (A. lagopoides, S. tremulus) and non-excreting (P. paspalodes, P. geminatum) grasses grown under increasing salinity treatments (0, 200 and 400 mM NaCl). Symbols indicate means (+SE) at each salinity level (n ¼ 3). Asterisks (** and ***) represent significant differences among cation contents at P , 0.01 and 0.001 respectively; NS indicates nonsignificant differences among salinity treatment by one-way ANOVA.  inhibited and leaf senesced earlier by salinity compared with other grasses. Reduced biomass production without apparent signs of injury or senescence of salt-excreting grasses in our experiment was recorded in up to 400 mM NaCl. This indicated a plastic response for long-term survival by diverting available energy to ensure survival rather than higher growth and reproduction. Paspalum paspalodes proved to be the most salt-resistant grass among the four test species. Shift in biomass allocation towards root in halophytic grasses could improve their water and nutrient uptake under saline conditions (Donovan and Gallagher 1985;Marcum 1999) and therefore improve their salt resistance (Lee et al. 2004). Variation in biomass allocation among species could also be related to their intrinsic growth rate (Hermans et al. 2006). The sharp reduction in root-to-shoot biomass ratios in P. geminatum with rise in salinity could result in poor resource acquisition under saline conditions to support aboveground tissues and decreased salt exclusion from root surface. The salt-excreting grasses could tolerate saline conditions whereas salinity stimulated root growth of P. paspalodes.

Water relations
Higher salt resistance would require mechanisms to reduce osmotic and toxic ion effects related to salt accumulation (Neumann 1997(Neumann , 2011. Decreasing c s and XPP with increasing salinity and a parallel decrease in leaf succulence and WUE (poor stomatal regulation) resulted in a sharp decline in growth of P. geminatum. Growth inhibition in salt-stressed plants was attributed to decreased turgor (short term) and wall extensibility (long term) components of water relations (Neumann et al. 1988). Differences in WUE of C 4 grasses under saline conditions also appear to reflect their differential salt resistance (Carmo-Silva et al. 2009;Bennett et al. 2013). Saltexcreting grasses increased leaf succulence by decreasing c s and XPP more readily than non-excreting grasses to ensure water uptake under increasing substrate salinity (Touchette et al. 2009) and by maintaining WUEs similar to non-saline controls. Paspalum paspalodes managed leaf succulence and RWC by limiting salt uptake in shoots at moderate salinity (200 mM NaCl) but higher (400 mM NaCl) salinity reduced tissue water and osmotic potentials. In addition, increase in WUE of P. paspalodes appears to buffer adverse effects of salinity increments on its water relations.
Na 1 toxicity, ion homoeostasis and selective uptake Sodium concentrations increased in both root and shoot tissues of test grasses upon exposure to saline conditions except for P. paspalodes. Higher shoot Na + partitioning under saline conditions in salt-excreting grasses compared with non-excreting ones in our study appears partly due to their capacity to excrete salts. Naidoo et al. (2012) reported that an increase in biomass of Spartina maritima in 20 % seawater was mediated through an efficient saltexcreting mechanism in addition to improved photosynthetic efficiency and resource allocation. Salt-excreting grasses grown under warm ambient conditions in this experiment appeared to constitutively downregulate shoot growth possibly to minimize Na + accumulation in the rhizosphere and reduce shoot Na + uptake. Salt resistance of P. paspalodes could be attributed to its ability to minimize shoot Na + uptake and by maintaining essential minerals in metabolically active plant tissues (Gorham et al. 1986;Marcum and Murdoch 1992;Peng et al. 2004;Ahmed et al. 2013;Teakle et al. 2013).
In most plants, high Na + influx tends to reduce K + absorption and transport which otherwise has numerous roles in plant tissues such as osmotic adjustment, protein synthesis and enzyme activation (Evans and Wildes 1971;Flowers and Läuchli 1983). However, salt-resistant grasses such as Sporobolus virginicus (Marcum and Murdoch 1992), A. lagopoides, Sporobolus ioclados, U. setulosa (Gulzar and Khan 2006) and Zoysia japonica (Marcum et al. 1998) are known to maintain adequate shoot K + levels under saline conditions. Our test grasses did not appear to be K + deficient in spite of a more than five-fold increase in shoot Na + and maintained more than 125 mmol kg 21 DW of shoot K + up to 400 mM NaCl.
Potassium homoeostasis appeared to be achieved by high SA of K + over Na + (Bell and O'Leary 2003;Gulzar et al. 2003;Wang et al. 2009) in our test grasses and has also been related to stelar K + outward rectifiers (SKORs) and KUP-HAK protein channels (Santa-María et al. 1997;Gaymard et al. 1998). In addition, reduced ST of K + over Na + in salt-excreting grasses under saline conditions could help in maintaining membrane potential for transmembrane movement of essential macronutrients (Bonales-Alatorre et al. 2013).
The salt-excreting grasses A. lagopoides and S. tremulus displayed higher SA and ST of Ca 2+ over Na + which could result in the higher tissue Ca 2+ required for efficient salt excretion. Calcium maintains membrane and cell wall integrity (Marschner 1995), is a secondary messenger in many signal transduction pathways and supports growth under saline conditions via improved K + over Na + selectivity (Lä uchli and Grattan 2007). Excessive Na + concentration interferes with Ca 2+ uptake (Grieve and Fujiyama 1987;Dobermann and Fairhurst 2000) which could be offset by maintaining higher Ca 2+ selectivity at the root level, possibly through H 2 O 2 -activated Ca-permeable channels (Sun et al. 2010). Magnesium homoeostasis may help to sustain a number of enzymatic reactions under saline conditions (Shaul 2002).

Organic osmolytes
All test grasses accumulated glycinebetaine with increase in salinity, a constitutive trait of plant species included in the Poaceae and Amaranthaceae (Chenopodiaceae) families (Albert 1975;McCue and Hanson 1990;Khan et al. 1998;Flowers and Colmer 2008). Proline content increased only in A. lagopoides (Sobhanian et al. 2010) and was constitutively higher in S. tremulus among test grasses. Proline and glycine betaine could be involved in cellular osmotic adjustment, i.e. a reduction of osmotic potential (c s ) in plant tissue as a result of vacuolar solute accumulation (Hassine et al. 2008). However the relative contribution of these and other osmolytes, such as low molecular carbohydrates and polyols, in cellular osmotic adjustment is debatable. Additional roles include protection and stabilization of thylakoid membranes (Jagendorf and Takabe 2001), PS-II complexes (Murata et al. 1992), enzyme structure and activity (Bohnert and Jensen 1996;Mäkelä et al. 2000) particularly at high tissue salt concentrations (Wyn Jones 1981). Decreased photosynthetic efficiency (lower WUE) and increased TSS under salt stress could indicate disturbed translocation of sugars and energy starvation in sink tissues of P. geminatum. Relative changes in leaf succulence appear to minimize species-specific variations in proline and glycinebetaine concentrations measured in this study; however, the contribution of other osmolytes (such as sugar and polyols not analysed here) in conferring salt resistance cannot be ruled out.

C/N ratios and NUE
Wide variations in nitrogen content (N) also accounted for proportional variations in NUE of our test grasses (Duarte 1990). Nitrogen deprivation repressed most of the regulatory genes involved in photosynthesis, chlorophyll production and plastid proteins synthesis (Scheible et al. 2004). Salt-excreting grasses displayed comparatively lower NUE and C/N ratios (higher N), which correlated well with higher proline and glycine betaine contents compared with non-excreting grasses. High NUE in P. paspalodes could reflect its preferential distribution in more stable, naturally anaerobic, saline environments with a low but sustained N supply. However, salt-excreting species appear to prefer well-drained, sandy soils with higher nutrient availability and turnover (Ghannoum 2009;Ghannoum et al. 2011).

Conclusions
Our hypothesis that there will be distinct variations between salt-excreting versus non-excreting grasses could not be completely proven. Differential responses between salt-excreting and non-excreting grasses occurred only in traits related to succulence, glycinebetaine content and root-to-shoot ion partitioning. This study highlights the relative importance of various combinations of traits which could ultimately help to improve the salt resistance of test grasses and closely associated conventional crops. Similar studies on a number of grasses could provide conclusive evidence for distinct patterns of salt resistance among grasses. Further investigations into drought and flooding stresses would elaborate differences in their responses to abiotic stresses.