An investigation on possible effect of leaching fractions physiological responses of hot pepper plants to irrigation water salinity

Background The modification effect of leaching fraction (LF) on the physiological responses of plants to irrigation water salinity (ECiw) remains unknown. Here, leaf gas exchange, photosynthetic light–response and CO2–response curves, and total carbon (C) and nitrogen (N) accumulation in hot pepper leaves were investigated under three ECiw levels (0.9, 4.7 and 7.0 dS m− 1) and two LFs treatments (0.17 and 0.29). Results Leaf stomatal conductance was more sensitive to ECiw than the net photosynthesis rate, leading to higher intrinsic water use efficiency (WUE) in higher ECiw, whereas the LF did not affect the intrinsic WUE. Carbon isotope discrimination was inhibited by ECiw, but was not affected by LF. ECiw reduced the carboxylation efficiency, photosynthetic capacity, photorespiration rate, apparent quantum yield of CO2 and irradiance–saturated rate of gross photosynthesis; however, LF did not influence any of these responses. Total C and N accumulation in plants leaves was markedly increased with either decreasing ECiw or increasing LF. Conclusions The present study shows that higher ECiw depressed leaf gas exchange, photosynthesis capacity and total C and N accumulation in leaves, but enhanced intrinsic WUE. Somewhat surprisingly, higher LF did not affect the intrinsic WUE but enhanced the total C and N accumulation in leaves.


Background
In many countries, the shortage of fresh water is a principal factor restricting the development of irrigated agriculture. The use of saline water is a possible alternative to meet the increased water demands for irrigation [1]. A prototypical case is the cultivation of pepper (Capsicum annuum L.), which is now one of the most widely grown crops in the world. In 2016, global pepper production (fresh and dry) from some 4 million ha was estimated at some 39 million tonnes, increasing by some 30% in the last decade [2]. Increasing demand for pepper is perhaps not surprising for high nutritional value of pepper. However, the total water requirement for pepper cultivation is by no means small ranging from 500 to 900 mm and up to 1250 mm in some areas [3]. In arid and semi-arid regions where much of the pepper cultivation occurs, fresh water resources are scarce necessitating the use of recycled (and often saline) water. In some areas, up to 1200-1400 mm of saline water with salinity levels ranging from 2.2 to 3.7 dS m − 1 have been successfully used to meet pepper water requirements [4]. Unsurprisingly, as with many other crops, irrigation with saline water can result in the accumulation of salt in the root zones, leading to the reduction in pepper growth and yield [5,6]. Such reduction is the consequence of several physiological responses including lower CO 2 uptake, intercellular CO 2 concentration, and availability of intercellular CO 2 for carboxylation by decreasing stomatal conductance (g s ), as well as the reduction in photosynthesis capacity, photosynthesis rate (P n ), and depression in both the photochemical and Calvin cycle reactions [7,8]. To maintain the minimum salinity in the root zones and enhance crop growth, a considerable amount of water is needed to drain salinity when the field is irrigated with saline water [9]. Leaching fraction (LF) is the volume of drainage water passing through the root-zones divided by the volume of irrigation water. Crop yield with saline water irrigation depends on plant evapotranspiration as well as soil salinity leaching [10]. Previous studies have focused on the effects of LF on root growth [11], root-zone salinity, evapotranspiration and yield [10,[12][13][14]. However, little information is available on the physiological response of hot pepper leaves to LF.
Intrinsic water use efficiency (WUE), defined as the ratio of P n to g s at leaf level, can explain instantaneous responses to environmental factors [15]. Intrinsic WUE can be enhanced either by lowering g s , or by maintaining or enhancing the P n [16,17]. As salinity stress simultaneously decreases g s and P n , the intrinsic WUE varies under different salinity levels. Assessing the Brazilian pepper tree (Schinus terebinthifolius Raddi), Ewe and Sternberg (2005) [18] reported that the intrinsic WUE did not statistically differ among their salinity treatments, ranging from 0 to 21.4 dS m − 1 . Likewise, Yarami and Sepaskhah (2015) [19] noted that the intrinsic WUE of saffron (Crocus sativus) was not affected when irrigation water salinity (EC iw ) was lower than 3.0 dS m − 1 . However, for some crop species, including water melon (Citrullus lanatus) [20], henna (Lawsonia inermis) [21] and plantain (Plantago coronopus) [22], high salinity improved the intrinsic WUE as the sensitivity of g s to salinity increased relative to P n . Further investigation is therefore necessary to assess whether EC iw and LF can affect intrinsic WUE for hot pepper.
Stable carbon isotope composition (δ 13 C), which is frequently expressed as carbon isotope discrimination (Δ 13 C), has been correlated with gas exchange responses in the plant growth cycle. δ 13 C in plants therefore provides a time-integrated measurement of intrinsic WUE to environmental stress, such as water and salinity stresses [16,23]. Consequently, the variation of Δ 13 C has been suggested as an indicator of intrinsic WUE since there is a negative relationship between leaf Δ 13 C and intrinsic WUE [15,24].
Crop nitrogen (N) is important for plant growth. The natural variation of the N isotope composition (δ 15 N) in plants under salinity stress is useful as it is related to N metabolism [23]. Isotope fractionation may occur during the N enzymatic assimilation of nitrate, recycling, translocation, exudation, or volatilization [25,26]. Salinity-induced impacts on metabolism may cause a substantial change in the isotopic content of metabolites. For instance, increased salinity results in a significant reduction of δ 15 N in wheat shoots, which may result from reduction in the loss of ammonia and nitrous oxide [27]. Many studies have also shown that δ 15 N in plants can be used as an indicator to assess the mineralization rate of soil organic N [28]. Higher δ 15 N in plants indicates more N is absorbed from soil organic N pools than from inorganic mineral N. In addition, the uptake and assimilation of ammonium, plant growth and root length density or surface area may also affect plant N accumulation. Previous studies showed that increasing salinity leads to a reduction in the N content and total N accumulation [23,27,29,30]. However, the modification effect of LF on the uptake of hot pepper N uptake to EC iw remains unclear. In addition, the salinity-induced reduction in hot pepper N may affect C retention in the plant.
Therefore, the objectives of this study are (1) to analysis the response of photosynthetic capacity, intrinsic WUE and total C and N accumulation of hot pepper leaves exposed to different EC iw treatments, and (2) to assess the modification effect of LF on leaf gas exchange, intrinsic WUE, and total C and N accumulation to EC iw .

Results
Gas exchange, intrinsic WUE, photosynthetic lightresponse and CO 2 -response curves Higher EC iw induced the lower P n and g s . Compared to the EC iw of 0.9 dS m − 1 , the treatment with EC iw of 7.0 dS m − 1 decreased P n and g s by 37.7 and 60.5%, respectively, showing that P n declined slower than g s , which led to a higher intrinsic WUE (i.e. P n / g s ) with higher EC iw (Table 1). Interestingly, high LF did not affect P n and g s significantly. As a consequence, the intrinsic Table 1 Photosynsthis (P n , μmol m − 2 s − 1 ), leaf stomatal conductance (g s , mol m − 2 s − 1 ), intercellular to ambient CO 2 concentration ratio (C i / C a ) and intrinsic water use efficiency (WUE) (μmol CO 2 mol − 1 H 2 O) in hot pepper leaves subjected to varying levels of irrigation water salinity (EC iw , dS m − 1 ) and two leaching fractions (LF). The gas exchange parameters were measured with a fixed PPFD level of 1200 μmol m − 2 s − 1 (under light saturate condition). The values for each treatment were the averages of three measurements (23, 39 and  WUE had no statistical difference between the two LFs treatments (Table 1). There were significant relationships (i.e., a typical logarithmic correlation) between P n and g s under different EC iw levels and LF treatments (Fig. 1a, b), showing that partial stomatal closure would result in an increase in intrinsic WUE [31]. A clear logarithmic decrease of intrinsic WUE with increasing of g s was also found based on the pooled data from all treatments (Fig. 1c). Collectively, based on these results, it is suggested that EC iw reduced g s more than P n , resulting in an increase in intrinsic WUE; in contrast LF had no marked effect on g s and P n , leading to an identical intrinsic WUE. ANCOVA analyses also show that the EC iw × g s or LF × g s interactions were not significant, indicating that the slopes of the regression lines between P n and g s under different levels of EC iw and LFs were not significantly different. These results also further suggest that at a certain g s , the differences in P n among the EC iw or LF were consistent ( Fig. 1) The effects of EC iw and LF on gas exchange were further investigated by measuring the photosynthetic light-response (P n -PPFD) and CO 2 -response (P n -C i ) curves. Figure 2 shows the P n -PPFD and P n -C i curves of hot pepper leaves under varying EC iw and LF treatments. The photosynthetic characteristics inculding α, P n max , κ and R d derived from P n -PPFD curve and ε, P n sat , and R p derived from P n -C i curve are shown in the Table 2. There were no significant interactions between EC iw and LF in terms of the parameters derived from the P n -PPFD and P n -C i curves. κ was also not influenced by EC iw and LF, indicating P n increased identically to P n max as increasing PPFD. The identical R d under various levels of EC iw and LFs indicate steady early symptom of carbon metabolism [32]. However, salinity-induced reductions in P n max , α and P n sat were observed in this study (Table 2).
In agreement with the prior analysis for P n , g s and intrinsic WUE in this study, the improvement of carboxylation capacity, electron transport, P n max and P n sat in the higher LF were not observed on the P n -PPFD and P n -C i curves (Fig. 2, Table 2), indicating that the higher LF treatment did not enhance g s , which ultimately affected photosynthesis capacity and intrinsic WUE. Δ 13 C, δ 15 N and total C and N accumulation in leaves Although no significant interaction between EC iw and LF was found for the Δ 13 C of leaves, Δ 13 C decreased by 2.4 and 6.1% in the EC iw treatments of 4.7 and 7.0 dS m − 1 , respectively, when compared to the EC iw of 0.9 dS m − 1 ( Table 3). This suggests that higher EC iw had greater stomatal closure. A significantly negative linear relationship between the Δ 13 C and electrical conductivity of soil saturated paste extract measured at the end of the experiment was observed regardless of the LF treatments (Fig. 3), indicating that soil salinity restricted CO 2 diffusion in P n [33]. A previous study has shown that salinity-induced reductions in Δ 13 C accompany decreases in C i / C a [34]. In this study, the decline in Δ 13 C as EC iw increased from 0.9 to 7.0 dS m − 1 corresponded to a reduction of C i /C a from 0.8 to 0.7 (Table 1). In addition, a significant positive relationship between the Δ 13 C and C i / C a between the LF treatments was also found (R 2 = 0.92, n = 6, P < 0.01). Partial stomatal closure or higher photosynthetic capacity or a combination of both could lead to a decrease in C i / C a Fig. 1 Photosynthesis (P n ) and intrinsic water use efficiency (i.e. P n / g s ) (c) expressed as a function of stomatal conductance (g s ) in the leaves of hot pepper plants under different levels of irrigation water salinity (EC iw , a) and two leaching fractions (LF, b). The data points used were obtained from the pooled data of three measurements of leaf gas exchange (23, 39 and 76 days after transplanting) [35]. In this study, a significantly positive relationship between C i / C a and g s represents partial stomatal closure caused by salinity as a result of lower C i / C a levels ( Fig. 4, Table 1) Previous studies have shown that salinity markedly reduced the δ 15 N in leaves of broccoli and barley plants [36,37]. However, the δ 15 N in leaves of hot pepper plants was not affect by EC iw (Table 3), indicating that the similar soil organic N mineralization and therefore the identical soil N bioavailability under different levels of EC iw [16]. However, total C and N accumulation in leaves decreased with increasing EC iw (Table 3).
It should be noteworthy that LF did not affect Δ 13 C with values ranging from 22.87 ‰ to 23.09 ‰. Additionally, in accordance with similar Δ 13 C values in two LF treatments, the C i / C a was also identical for two LFs, which may attribute to slimiar stomatal opening and photosynthetic capacity as discussed earlier (Tables 1 and 2). Furthermore, LF also did not influence the δ 15 N in leaves of hot pepper plants. However, higher LF enhanced total C and N accumulation in leaves (Table 3).

Discussion
Pepper is considerate moderately sensitive to salinity (generally no yield loss when EC iw was lower than 1.5-2.0 dS m − 1 [14,38]). Hence higher EC iw in this study markedly inhibited the P n and g s , leading to a higher intrinsic WUE. In addition, a significant linear positive correlation between intrinsic WUE and EC iw was observed within the range of EC iw levels considered here regardless of LF treatments (R 2 = 0.993, n = 6, P < 0.001). However, additional data on more severe EC iw levels are necessary to assess the aforementioned correlation. For instance, when  Table 2 Effects of irrigation water salinity (EC iw , dS m − 1 ) and leaching fraction (LF) on maximum apparent quantum yield of CO 2 (α, mol CO 2 mol − 1 photons), irradiance-saturated rate of gross photosynthesis (P n max , μmol m − 2 s − 1 ), dark respiration rate (R d , μmol CO 2 m − 2 s − 1 ), and dimensionless convexity term (κ) derived from the photosynthetic light-response curve and on carboxylation efficiency (ε, mol m − 2 s − 1 ), photosynthetic capacity (P n sat , μmol CO 2 m − 2 s − 1 ), photorespiration rate (R p , μmol CO 2 m − 2 s − 1 ) derived from the photosynthetic CO 2response curve. The light-response curves were measured at a fixed CO 2 concentration of 400 μmol mol − 1 . Measurements of CO 2 -response curves were conducted at a fixed PPFD of 1200 μmol m − 2 s − 1  Table 4 from Chartzoulakis and Klapaki (2000) [6], only a small increase in intrinsic WUE was found when EC iw higher than 12.6 dS m − 1 , showing that intrinsic WUE did not appreciably increase for the aforementioned correlation. Salinity-induced reductions in P n max and α from P n -PPFD curves were observed in this study, revealing a comparatively lower capacity of the biochemical reactions responsible for CO 2 fixation and lower photochemical efficiency of photosystem in hot pepper leaves in higher EC iw [39]. Similarly, P n sat derived from P n -C i curves also restricted in the EC iw of 7.0 dS m − 1 treatment as shown by the decline in the initial slope and the level of the upper plateau in the P n -C i curve (Fig. 2b) [40]. Brugnoli and Lauteri (1991) [41] observed similar results in bean and cotton plants, with the effect more marked in bean plants. A decline in carboxylation efficiency (ε) was a major component among those inhibiting P n by mesophyll limitations in higher salinity (e.g. EC iw of 7.0 dS m − 1 in this study); this was likely produced by a reduction in enzyme activities in the carbon reduction cycle [42]. In addition, owing to the decreases in the CO 2 /O 2 ratio in the mesophyll, an increase in salinity may increase the rate of photorespiration (R p ) in C 3 plants [8,43]. However, analysis of the P n -C i curves of hot pepper leaves in this study suggested that R p decreased significantly when EC iw was higher than 4.7 dS m − 1 Table 3 Carbon isotope discriminaion (Δ 13 C, ‰), C content (% DW), total C accumulation (g plant − 1 ), nitrogen isotope composition (δ 15 N, ‰) and total N accumulation (g plant − 1 ) in hot pepper leaves as affected by varying levels of irrigation water salinity (EC iw , dS m − 1 ) and two leaching fractions (LF). The values for each treatment measured at the end of the experiment were the averages of four replications   Table 2). Similar findings have also been reported in mallow [44] and mangrove [45] leaves based on the measurements of gas exchange. The enhanced PEPCase may account for the reduction in R p [45], however further research is needed to explore the physiological mechanisms of reduced R p within hot pepper leaves under high salinity levels.
It is well established that Δ 13 C analysis in leaf samples is one of the most versatile methodologies in assessing the environmental effects on the efficiency of photosynthesis in plants [32]. For instance, variation of Δ 13 C was found when plants were subjected to water and salinity stresses [33,46], which was confirmed by salinity stress in this study. Variation in Δ 13 C relies not only on changes within C i / C a , but also the variation in intrinsic WUE [26]. This is confirmed by the negative correlation between the intrinsic WUE and Δ 13 C regardless of LF treatments in this study (R 2 = 0.92, n = 6, P < 0.01).
LF did not affect the gas exchange, photosynthesis capacity and hence intrinsic WUE, which further confirmed by the identical value of Δ 13 C. The possible reason is that no creditable soil salinity may leach from root zone in high LF in this study, as indicated by that the electrical conductivites of soil saturated paste extract measured at the end of the experiment were no more than 2.5 dS m − 1 between two LFs, especially for lower salinity levels [47].
Higher EC iw induced lower total C accumulation in leaves (Table 3). A lower leaf biomass or a decreased C content in the biomass could retain less C in plant [48]. In this study, lower leaf dry biomass and C content might account for lower total C accumulation in leaves in the higher EC iw treatments (Tables 3 and 4). It is noteworthy that the reduction in leaf dry biomass in higher EC iw levels could result from lower P n sat and limited root water uptake ability ( Table 2). Root water uptake is mainly depended on soil's matric and osmotic potentials [49,50]. The salinity reduces the osmotic potential [51], causing the plant to spend more energy in taking up water from the soil solution, leading to a reduction in root water uptake [52,53]. Salinity-induced reduction of root growth and excessive Na + absorption also limited the root water uptake rate ( Table 4).
As expected, high LF enhanced total C accumulation in leaves because of high leaf dry biomass and C content (Table 4), where the enhanced leaf dry biomass in high LF may result from the reduction in Na + uptake and increased osmotic potential (Table 4). However, the reasons for the reduction in C content in higher EC iw and lower LF treatments remain unclear. Wang et al. (2010) [48] suggested that the C content in the plant is affected by the ability of C utilization in the plant. Plant N nutrition is one of the essential factors regulating C metabolism in plants because N is an important element for enzymes concerning metabolism, carbohydrate transport, and utilization in plants [54].
Based on literature surveys, at least four factors may determine plant N uptake from the soil. Firstly, the decreased leaf N accumulation in higher EC iw or lower LF could be attributed to a decrease in plant available N in the soil [28]. If this was the case, the δ 15 N in the high EC iw or low LF treatment should be low because the source of N taken up by plants could be reflected by variations in δ 15 N [55]. However, neither the EC iw nor LF affects δ 15 N in this study (Table 3). Alternatively, the reduced leaf N accumulation may result from the inhibited uptake and assimilation of ammonium as a result of competitive inhibition of Na + [30]. We observed that the Na + content in roots was greater in the higher EC iw and lower LF treatments (Table 4), which might imply that the uptake and assimilation of ammonium was restricted by higher Na + in the higher EC iw and lower LF, and reduced leaf total N accumulation. Thirdly, the reduction in N accumulation in the higher EC iw treatment may result from the decrease in the root surface area for N uptake [28]. Even though the root length density or surface area was not investigated in this study, the root dry biomass declined with increasing EC iw or was not affected by LF (Table 4). This might indicate the lower root density in higher EC iw and similar root density between the two LF treatments. This implies that the lower root length density and root surface area in the higher EC iw might account for the reduction in leaf N accumulation. Lastly, plant N uptake is also affected by plant growth, as shown by significant positive linear Table 4 Dry biomasses of leaves and roots (g plant −1 ) and Na + content (mg g −1 DW) in hot pepper leaves measured at the end of the experiment subjected to varying levels of irrigation water salinity (EC iw , dS m − 1 ) and two leaching fractions (LF *, ** and *** represent significant differences between means at 0.05, 0.01 and 0.001 level of probability, respectively; NS, no significant. Different letters within a column indicate significant difference at P < 0.05 by Duncan's multiple range tests correlation between total N content and dry biomass of leaves, regardless of the LFs in this study (R 2 = 0.98, n = 6, P < 0.001), indicating leaf total N accumulation was in accordance with the dry biomass accumulation of leaves.

Conclusions
In summary, our results indicated that higher salinity impacted g s more than P n , which resulted in higher intrinsic WUE. High salinity also inhibited photosynthesis capacity and retained less C and N in leaves. The novelty of this study is that we found higher LF did not improve leaf gas exchange, photosynthesis capacity and intrinsic WUE. However, higher LF did enhanced C and N accumulation in leaves of hot pepper plants.

Experimental design
The experiment was conducted under a rain shelter from April 28 to All the pots were saturated with tap water before the transplanting. Five days after the transplanting, each plant was irrigated using tap water with an irrigation amount of 0.9 L pot − 1 (all pots observed drainage). Five days after this irrigation event, three different saline water treatments were initiated for two LFs treatments.
The three EC iw levels assessed were 0.9, 4.7 and 7.0 dS m − 1 and the two LFs treatments were 0.17 and 0.29; each treatment was replicated four times. The 24 pots were arranged as a randomized block design. Salinity was increased by adding 1:1 m equivalent concentrations of NaCl and CaCl 2 to fertilizers (half strength Hoagland solution, see Heeg et al. (2008) [56] and Qiu et al. (2018) [57] for detailed composition). The fertilizers added an electrical conductivity (EC) of 0.9 dS m − 1 to the irrigation water for each treatment. The characteristics of the irrigation water for each treatment were shown in Table 5. The evapotranspiration (ET, g) of each pot was calculated as follows: where W n and W n + 1 are the pot weights before the n th and (n + 1) th irrigation (g); AW and D are the amounts of applied irrigation and drainage water (L), respectively; and ρ is the water bulk density (1000 g L − 1 ). At each irrigation event, the plants were irrigated with 120 and 140% of ET for each EC iw treatment, which lead to an LF of 0.17 and 0.29 according to the method proposed by Letey et al. (2011) [1]: Therefore a different amount of water based on actual ET for each pot was applied to maintain the target LF. At the end of the experiment, the average actual LF based on the amount of seasonal drainage water and applied water was 0.17 and 0.27, respectively [47], showing that the amount of applied irrigation water is reasonable.
The drainage water of individual pots was collected with a glass bottle positioned beneath each pot, and the amount was collected after each irrigation event. Just before each irrigation event, each pot was weighed with an electronic scale of 20 kg with an accuracy of 0.1 g, afterwards the evapotranspiration and irrigation amounts were calculated. During the experimental period, the plants were irrigated every two to five days and a total of 24 irrigations were applied.
Leaf gas exchange, δ 13 C and δ 15 N of hot pepper leaves and Na + content in roots Leaf gas exchange parameters, including P n and g s , were measured at 9:00-11:00 am on three sunny days (i.e. 23, 39, and 76 days after transplanting) using a portable photosynthesis system with a red-blue light source (LI The intercellular to ambient CO 2 concentration ratio (C i / C a ) were also obtained from the gas exchange measurements. As noted earlier, intrinsic WUE is defined as the ratio of P n to g s . The plants were harvested on July 22, 2015. The biomasses of the leaves were dried in an oven at 70°C for 72 h to obtain constant weight. Dry leaf samples were ground and used for δ 13 C and δ 15 N measurements. The values of δ 13 C and δ 15 N as well as the total C and N content in the leaves were measured using a MAT253 Stable Isotope Ratio Mass Spectrometer (Thermo Fisher Scientific, USA). The δ 13 C in leaf dry biomass can be calculated as: where R sample and R standard are the 13 C/ 12 C ratio of the sample and PDB (Pee Dee Belemnite) standard, respectively. The δ 15 N in the leaf biomass is calculated as: where R s and R b (= 0.3663 at % 15 N) are the N 15 : (N 14 + N 15 ) ratios of the leaf sample to standard, respectively. Δ 13 C in leaf dry biomass can be calculated as: where δ a and δ p are the carbon isotope composition of source air and plant material, respectively. The δ a was taken as − 8‰ [34]. The roots of each plant were washed with fresh water, and dried in an oven at 70°C to obtain constant weight. The dried roots were then ground into a powder, broken down with concentrated HNO 3 that was warmed with a heating block, and finally dissolved in 5% (v/v) highpurity HNO 3 . The sodium ion (Na + ) content in the dry roots was determined using an Inductively Coupled Plasma-Optical Emission Spectrometry (ICP-OES, Perkin Elmer Optima 8000). The electrical conductivity of soil saturated paste extract was determined at the end of the experiment by a dual channel pH/mV/Ion/Conductivity benchtop meter (MP522, Shanghai San-Xin Instrumentation Inc., China).
The P n -PPFD and P n -C i curves The P n -PPFD and P n -C i curves for different levels of EC iw and LFs were determined using a LI-6400 photosynthesis system (LI-COR, Lincoln, NE, USA). The P n -PPFD curves were measured at a fixed CO 2 concentration of 400 μmol mol − 1 on 2-4 plants per treatment. Measurements were made at PPFD levels of 2000, 1500, 1000, 700, 400, 200, 100, 50, 20 and 0 μmol m − 2 s − 1 . The nonrectangular hyperbola model was used to simulate P n -PPFD curve [58]: where P n is the rate of net photosynthesis (μmol CO 2 m − 2 s − 1 ); Q is the PPFD (μmol m − 2 s − 1 ); P n max is the irradiance-saturated rate of gross photosynthesis (μmol CO 2 m − 2 s − 1 ); R d is the dark respiration rate (μmol CO 2 m − 2 s − 1 ) at Q = 0; α is the maximum apparent quantum yield of CO 2 (mol CO 2 mol − 1 photons); and κ is a dimensionless convexity term [0, 1]. Measurements of P n -C i curves were made at CO 2 levels of 400, 250, 150, 100, 50, 500, 700, 1000 and 1500 μmol mol − 1 at a fixed PPFD of 1200 μmol m − 2 s − 1 . The P n were plotted against the respective C i . A nonrectangular hyperbola curve was used to simulate P n -C i curve [59,60]: where ε is carboxylation efficiency (mol m − 2 s − 1 ); P n sat is the photosynthetic capacity (μmol CO 2 m − 2 s − 1 ); and R p is the rate of photorespiration (μmol CO 2 m − 2 s − 1 ).

Statistic analysis
Two-way analysis of variation using the general linear model-univariate procedure was performed to assess the effects of the EC iw and LF on gas exchange parameters, intrinsic WUE, Δ 13 C, δ 15 N, C content and total C and N accumulation, dry biomass of leaves and roots, Na + content, the parameters obtained from the P n -PPFD and P n -C i curves. All analyses were conducted in the SPSS software package (Version 21.0, IBM Corp., Armonk, NY). Correlations between the measured parameters were determined with regression analyses. The slopes of the relationships between P n and g s under different EC iw levels and LFs were tested by a standard analysis of covariance (ANCOVA). P n was analyzed through a General Linear Model (GLM) of the natural logarithm of g s . The EC iw (or LF) and the interaction with the linear predictor were included to test for differences in slope.
If there was no significant interaction between EC iw (or LF) and linear predictor, the slopes were assumed to be the same.