Is foliar Cl- concentration the cause of photosynthetic decline in grapevine during mild salinity?

Moderate levels of Clhave been associated with grapevine salt tolerance. The hypothesis to be tested in this work is: photosynthesis in grapevine is negatively correlated with foliar Clconcentration. To further test this hypothesis, multiple mild salinity experiments on four different Vitis genotypes (Cabernet-Sauvignon, Riparia Gloire, Ramsey and SC2) were conducted and photosynthesis, ion concentrations and gene expression responses were quantified. The salt-tolerant rootstock Ramsey had greater Clexclusion capabilities than V. vinifera cultivars both during rooted cutting greenhouse experiments and three years of field-grafted experiments; SC2 also excluded Cl-. Differential gene expression indicated that salinity affected transcript abundance more in salt-sensitive genotypes (97.7 % of DEGs in the dataset), especially chloroplast-related transcripts. The transcript abundances of known anion transporters were determined and a family of putative B transporters was associated with the Clexclusion phenotype. Photosynthesis and growth were maintained in Ramsey and SC2 under mild salinity. However, photosynthesis declined in Cabernet-Sauvignon with isosmotic 20 mM salt concentrations of NaCl, KCl or NaNO3, independent of the salt type. While foliar Clconcentrations did correlate with salt tolerance during control and NaCl conditions, it was not found to be the cause of photosynthetic decline in Vitis during mild salinity.


INTRODUCTION
Salinity is a prevalent abiotic stress affecting agricultural land throughout the world (Rengasamy, 2006). It is estimated that approximately 20 % of cultivated land and 33 % of irrigated land is affected by salinity throughout the globe (Shrivastava and Kumar, 2015). Salinity stress is characterised by high ion concentrations in soils. The USDA defines salinity as soils with an electrical conductivity (EC) above 4 dS m -1 , an exchangeable Na + percentage below 15 and a soil pH below 8.5 (Allison et al., 1954). While ions such as Mg 2+ , Ca 2+ , K + , SO 4 2and HCO 3 are known to contribute to salinity, the majority of research into the subject has involved Na + and Cl -. This is not only because they are two of the most soluble and widespread salts, but also because many crops are particularly sensitive to these ions (Munns and Tester, 2008).
High levels of salt in soils trigger both osmotic and ionic disruption in plant function (Munns, 2002;Munns and Tester, 2008). The initial osmotic effect is similar to the symptoms of water deficit (Munns, 2002), characterised by reduced water potential in the soils due to the increased solute concentration. This reduction in soil water potential inhibits water uptake into the plant, decreasing growth and nutrient uptake (Acosta-Motos et al., 2017). To overcome this initial osmotic effect, many plants respond through the closure of stomates, increasing accumulation of organic osmolytes and hormonal regulation (Munns and Tester, 2008). All of these mechanisms are energetically costly and can lead to slowed or stunted growth (Munns and Gilliham, 2015). Another response is to increase the uptake of salt ions, for use as osmoticums, from the soil, lowering the water potential in the plant and driving water uptake (Munns and Gilliham, 2015). Uncontrolled ion uptake over time will exasperate the secondary effect of salinity-ionic toxicity (Munns and Tester, 2008). Ionic toxicity occurs when cellular ionic concentrations exceed the optimal range and lead to detrimental effects on plant function such as reduced ribosomal function, photosynthetic inhibition and decreased uptake of other essential macro and micronutrients (Geilfus, 2018). The symptoms of ion toxicity are not observed until the ion concentrations rise to toxic levels for the specific plant. This is dependent on the individual plant's tolerance and developmental stage, the ion concentrations in the soil, as well as the length of exposure (Greenway and Munns, 1980). Crops have different sensitivities to salinity in their growth environment, while some are tolerant to high levels, others, such as Vitis, are considered salt-sensitive (Greenway and Munns, 1980;Munns and Tester, 2008).
Vitis vinifera is a relatively drought-tolerant plant species (Cardone et al., 2019), but one of the few crop species considered to be sensitive to Cl -, along with citrus (Citrus ssp. L.), avocado (Persea americana L.), soybean (Glycine max L.) and faba bean (Vicia faba L.) (Li et al., 2017). Particularly, photosynthetic decrease in V. vinifera during salinity exposure has been associated with increased Clconcentrations in leaves (Downton, 1977a;Downton et al., 1990;Prior et al., 1992;Walker et al., 1981;Walker et al., 1997). Researchers have shown in Vitis that Clconcentrations in various tissues (leaf, petiole, stems and roots) reach higher concentrations than Na + when exposed to equal concentrations in the growth medium (Downton, 1977a;Martin et al., 2020;Walker et al., 1981;Walker et al., 2010). Notably, in V. vinifera cultivars (the primary grapevine species used for wine production (Zhou-Tsang et al., 2021)), there exists a prevalence of the dominant HKT1;1 allele controlling high rates of Na + exclusion from the shoot (Henderson et al., 2018), potentially explaining the underlying reason why Vitis shows a higher accumulation of Clthan Na + when the two ions are present in equal concentrations in the growth medium. Within the genus of Vitis, there are different tolerances and ion accumulation patterns during saline conditions among genotypes (Downton, 1977b;Fort et al., 2015;Heinitz et al., 2020;Sykes, 1987).

Plant material
Four grapevine genotypes were used for the greenhouse experiments: Cabernet-Sauvignon clone 8 (V. vinifera) abbreviated: CS, Riparia Gloire (V. riparia) abbreviated: RI, Ramsey (V. champinii) abbreviated: RA, SC2 (V. vinifera × V. girdiana) abbreviated: SC. Original vines used for clonal propagation and grafted vines for the field experiment were obtained from Dr Andrew Walker and the Foundation Plant Services at the University of California, Davis, CA, USA. Single nodal vegetative cuttings were rooted in water aerated by stone diffusion aeration pumps (referred to as rooted cuttings throughout). When cuttings were rooted, plants were transferred to pots as described below.  (Cochetel et al., 2020) daily until experiments began, then all control groups were irrigated every day with Cramer's Solution for the entirety of the experiments. All salt-treated plants were irrigated daily with their respective salinity solutions described below.

Greenhouse salinity experiments
All greenhouse experiments (one-week, two-week and big-pot photosynthesis), except the multiple salt photosynthesis experiment, were conducted with four genotypes (CS, RA, RI and SC) and two treatments (control and salt) in a completely randomised design. The one-week experiment (n (number of replicates) = 5-8 individual vines) was conducted first, and the two-week experiment (n = 10 individual vines) was designed following the one-week experiment. In these experiments, the control group plants were irrigated daily with 100 mL Cramer's Solution for the entire treatment period. Salt group plants were irrigated with 100 mL of Cramer's Solution + 20 mM NaCl. Leaf area and root length were measured at each harvest. At the harvest date (day 7 or 14), shoot tissue (leaf, petiole and stem) was harvested and then dried for 48 hours in a 60 °C oven.
For the large pot photosynthesis experiment (n = 5 individual vines), the control group plants were irrigated with 2 L of Cramer's Solution daily for the treatment period. Salt-treated plants were irrigated with 2 L of Cramer's Solution + 20 mM NaCl. The experiment was conducted over 16 days with 4 days of pre-treatment observation. Lateral shoot length was measured every odd day in the morning with a meter stick. Photosynthesis and stomatal conductance were measured every even day starting at ~10:00 AM. At harvest, leaf tissue (without petiole) was collected and then dried for 48 hours in a 60 °C oven.
The multiple salt photosynthesis experiment was conducted with one genotype (CS) and four treatments (control, NaCl, KCl or NaNO 3 ) in a completely randomised design. For the multiple salt photosynthesis experiment (n = 3 individual vines), salt treatments included either NaCl, KCl, or NaNO 3 at 20 mM solution added to the Cramer's Solution; the control group was irrigated only with Cramer's Solution. All plants received 400 mL of Cramer's Solution daily, an excess amount to flush the pots to prevent salt accumulation. The experiment was conducted over 26 days, with 4 days of pre-treatment analysis each morning, measuring photosynthesis and lateral shoot length on alternative days. Auxiliary shoot length was measured every odd day in the morning with a meter stick. Photosynthesis measurements were taken every even day starting at ~10:00 AM. At harvest, leaf tissue (without petiole) was sampled and then dried for 48 hours in a 60 °C oven.

Field site
Grafted vines were planted in the Las Vegas Cooperative Extension Research Center and Orchard in the Spring of 2016 in Las Vegas, NV, USA. These vines were grown in the same location of planting, in mildly saline soil with no additional application of salinity during the experimentation period (2017, 2018 and 2019). Two scion materials were used: Colombard (V. vinifera) abbreviated: CO and Syrah (V. vinifera L.) abbreviated: SY. Graft combinations include the following scion/rootstock (6 graft types) combinations: CO/CO, CO/RA, CO/RI, SY/SY, SY/RA and SY/RI planted in a completely randomised design (n = 6-12 individual vines). The control vines grafted onto their own genotype's rootstocks (CO/CO and SY/SY) are referred to as selfgrafted throughout. During the Fall seasons of 2017, 2018 and 2019, one mature healthy front-facing leaf (without petiole), free of insect pressure, was collected from each of the four cordons of each vine. The four leaves from each year from individual plants were grouped as one composite biological replicate. Leaves were dried for 48 hours in a 60 °C oven. Leaves (without petioles) were ground and then analysed for Cl -. The soil at the vineyard planting site in Las Vegas, NV, USA were analysed for soil salinity and ionic concentrations using 10 sub-samplings equally spaced along each outer row compiled for a single composite sample for each side.

Greenhouse growth measurements
Fresh and dry weights of plant tissue were measured at the time of harvest and after drying, respectively. The tissue was dried in an oven at 60 °C for 48 hours before weighing dry plant tissue. Root lengths were obtained by scanning with an Epson Perfection V800 Photo and quantified in WinRHIZO™ from Regent Instruments Inc. (Quebec, Canada). Excised leaves (without petioles) were photographed and leaf area was quantified using the ImageJ software (Schneider et al., 2012). Mean relative growth rate, net assimilation rate and leaf area ratio equations used are from Hunt, 1982. Leaf area ratio is calculated using the variables: L A (leaf area) and W (whole plant dry weight). Mean relative growth rate and net assimilation rate are calculated from the one-and two-week data using the variables:

Lateral shoot length
All vines in the big-pot photosynthesis and multiple-salt photosynthesis experiments were trained as a single cane on a bamboo stake. One lateral shoot at the top of the vine was allowed to grow, all other lateral buds were removed daily. Every other day, the lateral shoot length was measured in the morning (starting at ~10:00 AM) with a meter stick from the base of the cane to the tip of the lateral meristem. The growth rate was calculated as growth change from the previous measurement over the day. In the big-pot photosynthesis experiment, one SC control and one SC salt lateral shoots broke during treatment and were not included in lateral shoot length or lateral shoot growth rate analysis (n = 4), all other genotypes (CS, RI and RA) and their treatment groups (control and salt) had all replicates included in the analysis (n = 5).

Cltitrations
Depending on tissue type Clconcentration relative to an instrument detection threshold, ~0.5-2 grams of tissue were weighed into 50 mL conical tubes. A solvent of 20 mL of 0.5 M HNO 3 was added to the tube. Tubes were vortexed for 3 min. Then tubes were incubated for 30 min in a 65 °C water bath. Tubes were re-vortexed for 3 min. Tubes were then centrifuged for 5 min at 3500 × g. The supernatant was filtered through Whatman™ #1 filter paper. Finally, 10 mL of filtered supernatant was transferred to the titration cup and diluted with distilled-deionised H 2 O to 30 mL. Titrations were performed using a Mettler Toledo (Columbus, OH, USA) EasyPlus™ Titrator equipped with a Silver Ring Sensor (EM45-BNC) using 0.1 M AgNO 3 as titrant.

Soil analysis was performed by A & L Western
Laboratories (Modesto, CA, USA) using the North American Proficiency Testing (NAPT) procedures. The saturation percentage and pH were quantified using the saturated paste procedure (NAPT S-1.00 and 1.10). The EC, Cl -, CO 3 2-, HCO 3 and B were analysed using the saturated paste extract procedure (NAPT S-1.20, 1.30, 1.40 and 1.50). The Ca 2+ , Mg 2+ and Na + were measured using the saturated paste extract-inductively coupled plasma spectrometry procedure (NAPT S-1.60). The Na + absorption ratio and exchangeable Na + percentage were calculated from the collected data.
Foliar Na + and K + ion analysis was performed by A & L Western Laboratories (Modesto, CA, USA) through HNO 3 /HCl digestion using a microwave and analysis by inductively coupled plasma spectrometry (NAPT P-4.30). NO 3 was quantified by A & L Western Laboratories (Modesto, CA, USA) using automated Cd 2+ reduction (NAPT P-3.10).

Transcriptomics
Small potted plants from all four genotypes were exposed to either two weeks of salinity (20 mM NaCl + Cramer's solution) or control (Cramer's solution) conditions. Harvests were conducted after one and two weeks of treatment of root and leaf tissue (n = 2-5 individual vines). The RNA extraction, sequencing, processing, quantification, differential expression, gene ontology and principal component analysis were performed as described in Cochetel et al. (2020) except the V2/V3 assembly and annotation of the Vitis genome were used (Canaguier et al., 2017).

Graphing and statistics
Graphing of transcriptomic data was performed using the R software (R Core Team, 2018). RNA-Seq data is available at the Sequence Read Archive (SRA) database from NCBI (https://www.ncbi.nlm.nih.gov/) as BioProject: PRJNA516950. All non-transcriptomic figures were created in Prism GraphPad (Version 6.0h). Statistics were done in Prism GraphPad (Version 6.0h) or using the R software (when performing three-way repeated-measures ANOVAs) (R Core Team, 2018). Statistics were done using multiple t-tests, linear regressions and ANOVAs. The fixed effects were based on the experiment (i.e., genotype, treatment, time, rootstock and year). Succeeding the ANOVAs, multiple comparisons, followed by either Tukey's (when comparing every mean with every other mean) or Sidak's (when comparing only control versus salt within genotype) post hoc tests were used. All ANOVA tables are included in Supplementary Table 1. All other statistical tests and their results for the figures are contained in Supplementary Table 2. The symbols used for statistical significance are: ns, *, **, *** and **** refer to P > 0.05, P ≤ 0.05, P ≤ 0.01, P ≤ 0.001 and P ≤ 0.0001, respectively. The letters used for statistical significance within figures denote statistical significance of P ≤ 0.05 (a is significantly different than b). Overlapping letters within figures denote statistical significance of P > 0.05 (ab is not significantly different than b). Letters without an apostrophe (a) and letters with an apostrophe (a') within a single figure are the results of separate statistical comparisons (a is not comparable to a').

Salt sensitivity was associated with elevated shoot Cl -
The small pot salinity experiments were conducted in the greenhouse to test the hypothesis that: RA and SC are salt-tolerant and will maintain growth while reducing shoot Claccumulation relative to the salt-sensitive CS during mild salinity. To test this hypothesis, one and two-week mild salinity (20 mM NaCl) experiments were performed with plant growth measurements (tissue weight, leaf area and root length), transcriptomic expression analysis and shoot Clquantified after the experiments. *Presentation of physiological data (root length, leaf area, leaf area ratio) and growth data (mean relative growth rate and net assimilation) from the two-week experiment. Salt treatment is normalised for week one (W1) and week two (W2) time points by percent of control. Physiological and growth data are from plants whose Cldata is presented in Figure 1B. Asterisks indicate significance between salt and control treatment per respective genotype and week by multiple t-tests. Data are mean and SD (standard deviation), n = 10 individual small potted vines.  Figure 1A was a one-week experiment and Figure 1B was a separate two-week experiment. Week 1 (W1) and Week 2 (W2) groups were harvested after 7 and 14 days of treatment respectively. Significance calculated by A) two-way ANOVA or B) three-way ANOVA multiple comparisons with a Tukey's post hoc test. Letter differences denote statistical significance of P ≤ 0.05 within respective experiment. Letters with and without apostrophes are results from separate statistical comparison. Data are means ± SE (standard error of the mean), A) n = 5-8, B) n = 10, individual small potted vines.
From the plant tissue growth data, leaf area ratio, mean relative growth rate and net assimilation rate were calculated (Table 1).
CS's root length decreased under salt treatment relative to control-treated plants during week one (19.2 %) and two (38.5 %), as well as the decreased mean relative growth rate (29 %) over the treatment period (Table 1). Leaf area, leaf area ratio and net assimilation rate for salt-treated CS were no different than the control group. The salt-tolerant RA was the only genotype to have significant increases with net assimilation rate increasing 72 % and mean relative growth rate increasing 80 % in salt versus control for this genotype while having no significant changes in root length, leaf area, or leaf area ratio (Table 1). During one week of salinity treatment, no differences were observed in any growth measurements for RI, RA or SC. Following two weeks of salinity treatment, no significant decreases were observed in RI, relative to the control (Table 1). SC decreased leaf area ratio (13.8 %) at the week two harvest, with all other plant growth measurements unaffected by the treatment.
In Figure 1A there was a significant increase in shoot Clconcentrations from control to salt that was equal for all genotypes, shown by the interaction coefficient being non-significant ( Figure 1A, Supplementary Table 1 and  Supplementary Table 2). In the control group of the one-week experiment, the two saltsensitive genotypes: CS and RI had mean shoot Clconcentrations of 75.2 ± 7.2 (mean ± SE) and 70.8 ± 4.5 µMol Cl -• gDW -1 , respectively, and the two salt-tolerant genotypes: RA and SC were 40.7 ± 3.5 and 51.2 ± 9.4 µMol Cl -• gDW -1 , respectively ( Figure 1A). The salt-treated plants, in the one-week experiment, had significantly increased shoot Clfor CS, RI and RA but not SC, relative to the control plants of each genotype (Supplementary Table 2). In this experiment, salt-treated CS, RI, RA and SC had a shoot Clconcentration of 134.4 ± 17.4, 118.8 ± 13.1, 86.3 ± 6.5 and 86.9 ± 13.8 µMol Cl -• gDW -1 , respectively ( Figure 1A). In the salt-treated group, both RA and SC had significantly lower Clthan CS, while RI did not (Supplementary Table 2).
In Figure 1B, genotype, time, salt and all of the interactions between them significantly impacted shoot Clconcentration (Supplementary Table 1). In the first week, RI had the only significant difference (an increase) in control shoot Clfrom CS, with 110.6 ± 6.7 (RI) versus 62.8 ± 2.7 (CS) µMol Cl -• gDW -1 .
During the second week, RA (38.5 ± 4.3 µMol Cl -• gDW -1 ) and SC (39.7 ± 3.0 µMol Cl -• gDW -1 ) maintained a lower Clconcentration than CS (86.4 ± 4.9 µMol Cl -• gDW -1 ) which had a relative increase from week one ( Figure 1B). In the first week of salinity, the RI shoot Clconcentration was at its highest with a Clconcentration of 161.1 ± 10.4 µMol Cl -• gDW -1 , consistent with the increased shoot Clconcentration in the control at this time point, compared to 113.9 ± 5.0, 109.1 ± 4.2 and 75.5 ± 8.5 µMol Cl -• gDW -1 for CS, RA and SC, respectively ( Figure 1B). In CS there was a significant increase in Clfrom week one (113.9 ± 5.0 µMol Cl -• gDW -1 ) to week two (192.4 ± 10.9 µMol Cl -• gDW -1 ) coupled with a decrease in mean relative growth rate (29 % decrease) in the salt treatment group relative to the control (Table 1, Figure 1B and Supplementary Table 2). SC was the only genotype with lower Clrelative to CS in the week one salt group, although this was not significant ( Figure 1B and Supplementary Table 2). During the second week, RI's Clconcentration lowered to 126.2 ± 13.5 µMol Cl -• gDW -1 , leading RI to have significantly lower shoot Clrelative to CS along with SC ( Figure 1B and Supplementary Table 2). While RI's shoot Cllowered from week one to week two in the salt group, the opposite trend can be seen in RA. During the two weeks of salt treatment, RA's shoot Clconcentration significantly increased from 109.1 ± 4.2 µMol Cl -• gDW -1 in week one to 160.1 ± 12.5 µMol Cl -• gDW -1 in week two ( Figure 1B and Supplementary Table 2).

Salt sensitivity was associated with increased differential gene expression
The transcriptomic experiment was conducted to test the hypothesis: mild salinity (20 mM) will induce a transcriptomic response in saltsensitive Vitis genotypes. To test this hypothesis, CS, RA, RI and SC were treated with control or salt (20 mM NaCl) conditions and root and leaf tissue were harvested after one and two weeks with evaluation by RNA-sequencing. The transcriptomic analysis detected significant expression differences among the four genotypes during salinity treatment (Supplementary Figure 1). A principal component analysis (PCAs) of the roots and leaves in both week one and week two, indicated that genotype explained the majority of the variation in the data set (Supplementary Figure 1A, 1B, 1E, 1F). Only RI in the leaves had separation based on treatment in the PCAs (Supplementary Figure 1B and F).
Overall, there were low numbers of differentially expressed genes (DEGs) in all four genotypes in the root tissue at week one (47) and week two (357) (Supplementary Figure 1C, 1G). There were more DEGs in the leaf tissue, compared to the root tissue, during week one (2824) and week two (3383), with the majority in RI (1644) and CS (1187) in week one and RI (3369) in week two (Supplementary Figure 1D, 1H). Very low numbers of DEGs in root and leaf tissue during both harvests were found in RA (20) and SC (134) (Supplementary Figure 1C, 1D, 1G, 1H).
Salt-sensitive genotypes had more DEGs than salt-tolerant genotypes (97.7 % of DEGs). In the DEGs shared by the salt-sensitive genotypes (CS and RI) in the week one leaf harvest, there was an enrichment of DEGs with a gene ontology of chloroplast-related transcripts. Gene ontology enrichment identified two significant enrichments in the week one leaf harvest for increased DEGs shared by the salt-sensitive (CS and RI) genotypes as chloroplast accumulation movement and chloroplast avoidance movement (P adjusted values = 0.0069). Additional analyses of known anion transporters amongst species or induced by salinity did not identify any clear candidates for Clexclusion. However, differences in constitutive expression of putative B transporters (VIT_04s0008g04980, VIT_11s0016g04740, VIT_09s0002g05810, VIT_07s0005g02060, VIT_17s0000g08530, VIT_05s0020g00440) were associated with the Clexclusion phenotype (data not shown), with higher Clexclusion associated with higher B transporter transcript abundance. These transporters are all members of the ion transporter BOR family (Takano et al., 2002). One putative B transporter (VIT_11s0016g04740) had very high levels of gene expression in SC roots (~600 TPM [transcripts per million]) during both weeks in salt and control vines with minimal to no expression measured for the other three genotypes; there was no expression of this gene in the leaves for any genotype. Another putative B transporter (VIT_04s0008g04980) was significantly expressed (~10 TPM) in the roots in SC and RA, and only expressed in the leaves of RA, but with little or no expression in CS or RI in roots or leaves.

RA rootstock decreased foliar Claccumulation regardless of scion in field conditions
The field site experiment was conducted to test the hypothesis: commercially used RA decreases foliar Clconcentrations in grafted scions (CO and SY) under field conditions relative to self-grafted (CO or SY) and salt-sensitive (RI) rootstocks. To test this hypothesis, SY and CO scions grafted onto either their own genotype's roots, RA, or RI were evaluated for foliar Clconcentrations over three years at the Las Vegas Field site.
In Figure 2, the graft type and year both had significant impacts on the foliar Clconcentrations, but the interaction between the two was not significant (Supplementary Table 1). The scions grafted to RA rootstocks had decreased foliar Clconcentrations versus V. vinifera self-grafted rootstocks (CO and SY) over three years, consistent with the decreased accumulation of shoot Clobserved in rooted cuttings of RA versus CS in the greenhouse (Figures 1 and 2). This lower Clconcentration was significant for all years for SY/RA and in 2017 for CO/RA. In 2018 and 2019 CO/RA and CO/CO had no significant differences, with foliar Clconcentrations in CO/CO of 19.2 ± 1.7 and 18.2 ± 1.4 µMol Cl -• gDW -1 versus 39.7 ± 3.5 and 29.9 ± 4.1 µMol Cl -• gDW -1 , respectively ( Figure 2). Scions on RI also had decreased foliar Clconcentrations during 2018 and 2019 for SY/RI versus SY/SY, but no significant decreases for any years for CO/RI versus CO/CO (Figure 2).

Genotypes varied in photosynthetic response to mild salinity
This experiment was conducted to test the hypothesis: mild (20 mM) salinity decreases photosynthesis, stomatal conductance and lateral shoot growth in salt-sensitive Vitis genotypes (CS and RI), but not in salt-tolerant (RA and SC). To test this hypothesis, all four genotypes (CS, RI, SC and RA) had their photosynthesis and lateral shoot growth measured over 16 days of daily irrigation treatment with either mild salinity (Cramer's solutions + 20 mM NaCl) or control conditions (Cramer's solution) in five replicated big pots. In Figure 3A, the genotype, salt, time and genotype × time interaction had significant impacts on photosynthesis, while the other interactions did not (Supplementary Table 1).
During 16 days of salinity treatment in big pots, CS had the only significant decrease in photosynthesis, starting the day of salinity treatment ( Figure 3A). RA and SC had no significant differences in photosynthesis over the 16 days of 20 mM NaCl treatment relative to their controls. At the end of treatment (day 16), photosynthesis was lowered for CS 47 % in the salt treatment relative to the control treatment; RI had a lowered photosynthesis of 40 % that was not statistically significant ( Figure 3A). Stomatal conductance was affected by genotype, time and genotype × time interaction, with the salt treatment only having a significant effect in CS with no significant differences between control and salt-treated RA, SC or RI (Supplementary Table 1 and Supplementary  Figure 2A). During the experiment, no genotype had a significant difference in lateral shoot length or growth rate over the 15 days of treatment, and only time and the genotype × time interaction had a significant effect on growth rate (Supplementary Table 1, Supplementary Figure 2B, 2C). Foliar Na + concentration in Supplementary Figure  2D was increased by the salt treatment with a genotype × salt interaction, but no genotypespecific effect (Supplementary Table 1). Na + was increased two-fold in SC salt-treated relative to its control, with no significant increases observed in CS, RI or RA (Supplementary Figure 2D). No significant increase was observed in any genotype FIGURE 3. Photosynthesis and Cldata from big pot photosynthesis experiment.
A) Photosynthesis was measured during every other day of treatment and B) foliar Clwas quantified after harvest. A) The solid lines correspond to the control treatment (Cramer's solution) and the dash to the salinity treatment (Cramer's solution + 20 mM NaCl). A) Day 0 is when treatment was started, prior to day 0 all groups were irrigated with Cramer's solution. Significance calculated by A) repeated measures three-way ANOVA multiple comparisons followed by a simple main effect comparison against control treatment per genotype for days 2-16 and displayed as treatment effect between salt and control or B) by two-way ANOVA with multiple comparisons and a Tukey's post hoc test displayed using letters identifying significant differences of P ≤ 0.05. Data are means ± SE, n = 5 individual big potted vines.
for Clin salt-treated groups relative to their control but RI, RA and SC all had significantly lower Clthan CS in the control and salt treatment groups ( Figure 3B and Supplementary Table 2).

Isosmotic salts with and without Cldecreased photosynthesis in CS equally
This experiment was conducted to test the hypothesis: Cl --based salinity is the cause of the photosynthetic decrease in Vitis. To test this hypothesis, CS were treated in triplicate with either salt-containing Cl -(NaCl or KCl), a salt solution without Cl -(NaNO 3 ) but of equal isosmotic molarity (20 mM salt + Cramer's solution), or Cramer's solution control. In Figure  4A and Supplementary Figure 3A, photosynthesis and stomatal conductance were affected by time and the time x salt interaction, but salt alone did not have a significant effect (Supplementary Table  1). While there was a significant effect for all three salt treatments (NaCl, KCl and NaNO 3 ) relative to the control for photosynthesis on the final day of treatment (day 26), there were no observable differences between the salt treatment groups (Figures 4A and Supplementary Table 2). The only significant difference in stomatal conductance over the treatment period for any salt treatment (NaCl, KCl and NaNO 3 ) relative to the control, existed on day 20 for the NaNO 3 treatment (*), with no other significant differences (Supplementary Figure 3A).
The NaCl, KCl and NaNO 3 treatment groups led to decreases in photosynthesis of 32 %, 31 % and 35 % ( Figure 4A) relative to the control group on day 26, respectively. The growth rate in Supplementary Figure 3B was significantly affected by salt but not the time or the salt × time interaction (Supplementary Table 1). The growth rate was significantly different from the control for the NaNO 3 treatment after two weeks of treatment (day 15-27) but not for the NaCl or KCl treatments (Supplementary Figure 3B). On day 27, for the NaNO 3 treatment, the growth rate decreased 73 %, relative to the control (Supplementary Figure 3B). Lateral shoot length was significantly different from the control for NaCl and NaNO 3 but not for the KCl treatment ( Figure 4B). The decrease in lateral shoot length at day 27 was 18 % and 49 % of the control for NaCl and NaNO 3 treatments, respectively ( Figure 4B). The KCl treatment was the same as the control for lateral length over the treatment period, with significant differences from both the NaCl and NaNO 3 treatments ( Figure 4B and Supplementary Table 2). The multiple salt treatments had significant effects on Na + , K + and Clbut not NO 3 -(Supplementary Table 1). Na + concentrations increased two-fold in the NaNO 3 treatment relative to the control, while NaCl and KCl were not significantly different in Na + concentration from the control (Supplementary Figure 4A). K + concentrations increased in the KCl and NaNO 3 treatments but not in the NaCl A) Photosynthesis and B) lateral shoot length were measured alternatively every other day of treatment. Day 0 is when treatment was started, prior to day 0 all groups were irrigated with Cramer's solution. Significance calculated by A) two-way ANOVA using repeated measures multiple comparisons with a Tukey's post hoc test (day 2-26) and displayed using letters as final day (26) significance B) or linear regression (day -3-27) displayed as slope's significance using different letters relating to a P ≤ 0.05 within respective figures. Data are means ± SE, n = 3 individual small potted vines. relative to the control (Supplementary Figure 4B). Clconcentrations increased three-fold over the control in the KCl treatment, with no significant increases in the other treatments (NaCl or NaNO 3 ) (Supplementary Figure 4C). NO 3 concentrations were not significantly changed in any of the treatments (Supplementary Figure 4D).

DISCUSSION
In this study, we found genotypic differences in foliar Clconcentrations as well as photosynthetic, transcriptomic and growth responses to salinity; however, not all of the responses were as expected. The relationship between Cland photosynthetic decline in Vitis genotypes under mild salinity treatment was not present as in previous studies of V. vinifera (Downton, 1977a;Downton et al., 1990;Prior et al., 1992;Walker et al., 1981Walker et al., , 1997. It was expected that increased foliar Clconcentrations would co-occur with decreased photosynthesis during mild salinity and only Clcontaining salts would negatively impact the shoot. Osmotic effects of the different salts (NaCl, NaNO 3 and KCl) were not as previously shown under water deficit (Cochetel et al., 2020), and this was validated by the 20 mM KCl treatment having no effect on lateral shoot length or growth rate. SC and RA were found to be good Clexcluders, which was as expected due to previous reporting (Heinitz et al., 2020;Southey & Jooste, 1991;Stevens et al., 1996;Walker et al., 1997).

Salt sensitive genotypes had higher shoot Cland differential gene expression during mild salinity
The 20 mM salinity treatment significantly decreased root length, as well as the mean relative growth rate in the salt-sensitive CS while stimulating differential gene expression relative to other grapevine species (Table 1 and Supplementary Figures 1C, 1D, 1G, 1H). Of particular intrigue, was the enrichment of chloroplast-related transcripts shared by CS and RI in the week one leaf harvest. This finding first indicated an effect on photosynthesis in the saltsensitive genotypes during mild salinity. In the salt-tolerant RA, this level of salinity significantly stimulated growth, the net assimilation rate and mean relative growth rate in the salt relative to the control treatment, with minimal differential gene expression (Table 1 and Supplementary  Figures 1C, 1D, 1G, 1H). This clear separation between the known salt-sensitive genotype's (CS) growth inhibition and high transcriptomic response versus the salt-tolerant genotype's (RA) growth stimulation and low transcriptomic response, as well as its agronomic relevance highlighted by the comparability to the field site (Supplementary Table 3), indicated that the mild salinity (20 mM) level was sufficient to identify genotypic differences for these experiments herein. These salinity concentrations represent realistic salt levels found in Nevada and California vineyards.
Dozens of papers have been published about salinity in grapevines Downton, 1977b;Henderson et al., 2014;Walker et al., 2018; to name just a few). To date, only one other publication has examined SC under saline conditions (Heinitz et al., 2020). Our findings are consistent with this study and demonstrate that SC consistently showed patterns of Clexclusion throughout all experiments (Figures 1 and 3B); SC appeared to be a good Clexcluder. The association of B transporters with the Clexclusion phenotype was of particular interest, especially the high constitutive expression of VIT_11s0016g04740 and VIT_04s0008g04980 in SC and RA. Overall, the differential gene expression was lower than what was observed in Cochetel et al. (2020), with only 23 % of the differential gene expression in the mild salinity dataset versus the water deficit dataset presented in that work. Multiple studies have examined the Clexclusion capabilities of CS, RI and RA (Downton, 1977b;Ismail et al., 2014;Southey and Jooste, 1991;Stevens et al., 1999;Suarez et al., 2019;Sykes, 1985;Tregeagle et al., 2006;Vincent et al., 2007;Walker et al., 1997;Walker et al., 2010). While CS, like many other V. vinifera cultivars, are shown to have a higher accumulation of aerial tissue Cl -, there are varying results for RI and RA that have been reported.
In this study, RA and SC (the salt-tolerant genotypes) had a shoot Clconcentration range lower than CS and RI (the salt-sensitive genotypes) during control conditions in week 1 (Figure 1) and this was consistent in week 2. Shoot Clwas used in the small pot experiments due to minimum dry tissue requirements for analysis, foliar Clwas not possible for all genotypes. During one week of saline conditions, RA and SC had a Clconcentration range below that of CS and RI (Figure 1). Following two weeks of salinity, CS, RA and SC's Clconcentration increased while RI's decreased ( Figure 1B and Supplementary Table 2). In RA, the increased Clconcentration was likely used as an osmolyte driving water flow into the plant, which would be consistent with the increased net assimilation rate and mean relative growth rate seen in salt versus control for this genotype (Table 1). It has been shown previously that RI can, during certain conditions, have lower foliar Clconcentrations than RA (Downton, 1977b), while RA is consistently described as salt-tolerant (Southey and Jooste, 1991;Stevens et al., 1999;Suarez et al., 2019;Sykes, 1985;Walker et al., 1997;Walker et al., 2010;Zhou-Tsang et al., 2021) and RI is described as salt-sensitive (Ismail et al., 2014;Zhou-Tsang et al., 2021). Also of note is that it has been shown that Clexclusion capabilities of RA can break down over prolonged exposure to salinity (Tregeagle et al., 2006), though this is likely not the case in our experiments due to the relatively short-term exposure to salinity. It is important to mention that experimental design, irrigation solution concentration, time of salinization and age of plants can all affect Claccumulation. In the two-week experiment, the increased accumulation of Clin RA, relative to RI, during the secondweek harvest of the salt group could have resulted in the increased plant growth in RA relative to RI. While RI had no significant changes in any growth measurements (root length, leaf area, leaf area ratio, mean relative growth rate and net assimilation rate) in salt-treated groups relative to the controls, RA had statistically significant increases (Table 1). As carbon assimilation, and ultimately growth, depend on water uptake and transpiration, which can result in increased uptake of ions, the higher Clconcentration in RA over RI may be a trade-off of increased growth in RA during these short-term exposures to mild salinity.

RA was a good Clexcluder in field-grafted conditions
RA was observed to be the best Clexcluding rootstock in the field experiments versus RI, SY and CO in the mildly saline soil. The soil Cland Na + concentrations were similar to the levels used in the greenhouse experiments, with average Cland Na + concentrations of 14.1 ± 5.3 and 21.1 ± 10.2 meq • L -1 at the Las Vegas Field Site, respectively (Supplementary Table 3). During each year of sampling, RA grafted-types, both SY/RA and CO/RA, had two of the lowest Cllevels for that year compared to all other graft types (Figure 2). Even though CO/RA foliar Clconcentration was at least 39 % lower than CO/CO in 2018 and 2019, due to biological variation and the low levels of Clin the CO/CO as well as the CO/RA, significant differences were not observed in these years. This salt tolerance of CO grafted to RA presented here in regards to foliar Clexclusion corroborated the findings of Southey and Jooste (1991), showing the salt tolerance of this graft type in regards to vine performance in mildly saline field conditions in South Africa (Southey and Jooste, 1991). Similarly, when Sultana (V. vinifera) scions are grafted to RA rootstocks under saline conditions in Australia (Walker et al., 1997), Clexclusion, photosynthesis and fruit yield in the grafted plants are improved compared to the rooted cutting of Sultana (Walker et al., 1997). In field-grown grafted plants in Australia, Stevens et al. (1996) also found that RA was able to decrease foliar Clin grafted SY scions versus rooted cuttings of SY during both saline and non-saline conditions (Stevens et al., 1996).
The field experiments were consistent across scions and years, RA decreased the Cltransported to the leaf tissue when compared to V. vinifera rooted plants (Figure 2). While RI also reduced the foliar Clconcentrations of both scions relative to V. vinifera rooted plants, RI never reduced concentrations lower than RA for a given scion during the same year ( Figure 2). Based on the results of the greenhouse experiments and the findings of the field site, our results confirm that RA is a good rootstock for reducing foliar Clconcentrations in mildly saline soils.

Photosynthetic decline was not associated with foliar Clconcentration during mild salinity
It has been noted in multiple publications in the grapevine that there is a correlation between increasing foliar Clconcentrations and decreasing photosynthesis (Downton, 1977a;Downton et al., 1990;Prior et al., 1992;Walker et al., 1981;Walker et al., 1997). This observed correlation has led to the predominant hypothesis in the field of grapevine salinity that specific ion toxicity, through increased foliar Clconcentration, is the cause of photosynthetic decline during salinity in grapevines (Zhou-Tsang et al., 2021). Our results did not indicate that this correlation was untrue, but did indicate that the foliar Clconcentration was not the cause of the observed photosynthetic decline. The lowered photosynthesis in CS on day 0 (the initial day of salinity treatment) ( Figure 3A), first indicated that foliar Clconcentration may not have caused the photosynthetic decrease in this genotype. Ion toxicity was unlikely to occur in such a short period. Munns (2002) notes that ionic toxicity typically occurs in days to weeks depending on the salinity level and other environmental factors that affect transpiration (Munns, 2002).
The foliar ion analysis of Cland Na + at the end of this experiment, supported this hypothesis ( Figures 3B and Supplementary Figure 2D). CS did not have a significantly increased Clconcentration during the treatment period in the salinity group relative to the control ( Figure 3B and Supplementary Table 2). RI and RA Clconcentrations in both control groups and salt groups were equivalent, with RA having no decrease in photosynthesis and RI having an observable, although non-significant lowered photosynthesis after 16 days (Figure 3 and Supplementary Table 2). These findings again indicated that the foliar Clconcentration was not the cause of the photosynthetic decrease during mild (20 mM) salinity in these experiments. The Na + concentrations also could not be the cause of the photosynthetic decreases, as the ion only increased in the SC salt-treated group (a 2-fold increase), but not in CS, RI or RA. Thus, it is clear that the decline in photosynthesis is more complicated than originally hypothesised. Instead, we postulate that the ionic strength of the irrigation solution or in the plant tissue (including cell or organelle compartmentation), rather than a specific ion toxicity, along with an osmotic adjustment in response to the salinity, may be responsible for the photosynthetic decrease occurring or not in a given genotype. While mild osmotic potential decreases in the soil solution could have caused declines in photosynthesis (e.g., decrease in stomatal conductance), this would most likely have been a temporary effect on the order of hours and not days or weeks. Supporting this argument, maize leaf elongation declines rapidly but recovers within hours due to osmotic adjustment with mild to moderate salinity (Cramer & Bowman, 1991), but photosynthesis is unaffected by salinities at this level (Cramer et al., 1994).

Na + containing salts reduced growth rate of CS
The largest and most rapid decreases in lateral shoot length and growth rate in CS occurred with Na + containing salts, with the biggest decreases seen with NaNO 3 . NaCl salinity reduced shoot growth in another V. vinifera cultivar (Sultana) (Prior et al., 1992). The K + and Clions each may have a stimulatory effect on growth, particularly in the short term, due to their use in osmoregulation by the plant (Munns and Gilliham, 2015). However, it appears that Na + had an inhibitory effect on growth in CS during mild salinity conditions. This could be due to the compartmentalization of Na + , an ionic strength effect, the energetic cost of Na + exclusion, or the production of compatible counter anions during Na + uptake down its electrochemical gradient (Munns & Gilliham, 2015). In the case of the NaNO 3 treatment, the NO 3 taken up would likely be assimilated into NH 4 + and amino acids and thus osmoregulation may be limiting (Supplementary Figure 4D). Further investigation into the cause of Na + induced shoot growth decrease is needed to support or reject these hypotheses.

Photosynthetic decline occurred in CS during isosmotic salinity regardless of Cl -
The largest decline in photosynthesis for all three salt-treated groups (NaCl, KCl and NaNO 3 ) began on day 14 of treatment and reached a new steady state of decreased photosynthesis, relative to the control group, on day 20 ( Figure 4A). This new steady-state photosynthesis for the three salttreated groups remained steady for the rest of the experiment. While the decrease that occurred after two weeks of treatment could be an indication of ion toxicity, the occurrence in all three salt treatment groups diminishes the possibility that this decrease was Cl --specific. If the decreases in photosynthesis were Cl --specific such as has been previously suggested (Downton, 1977a;Downton et al., 1990;Prior et al., 1992;Walker et al., 1981;Walker et al., 1997), the photosynthetic decrease should have been observed in the NaCl and KCl treatments, but not the NaNO 3 treatment. This was not what occurred, instead, all treatment groups declined in photosynthesis together and had significant differences from the control at the final day of treatment (day 26) with no significant differences between the salt treatments. This finding that foliar Clconcentration was not the likely cause of the photosynthetic decrease in this experiment was again corroborated by the fact that the foliar Clconcentration only significantly increased in the KCl treatment. These findings in conjunction with the findings of the big-pot photosynthesis experiment clearly indicated that foliar Clconcentration was not the cause of the photosynthetic decrease in Vitis species during mild salinity and takes it further in showing that 20 mM salinity, regardless of Cl -, resulted in a photosynthetic decrease in CS.

CONCLUSION
Genotypic differences in responses to mild salinity (20 mM) were observed in growth rate, ion accumulation, gene expression, stomatal conductance and photosynthesis. Supporting previous studies, our findings herein confirmed that both SC and RA had inherent Clexclusion capabilities and RA may be considered an effective Clexcluding rootstock for the scion; RA was more salt-tolerant than the other genotypes tested. The Clexclusion capabilities of RA rootstock observed in the greenhouse over multiple experiments, relative to V. vinifera (cv. CS) were reflected in the Las Vegas field site over all three years and with both grafted scions, V. vinifera (cvs. CO and SY). While Claccumulation, may be a useful proxy for salt tolerance in Vitis species, one must be careful to not overinterpret Cl -'s role in the decline of salt sensitivity-related photosynthesis. Clwas not attributed as the cause of photosynthetic decrease during mild salinity (20 mM) in the Vitis genotypes studied in this project. Instead, the decreased photosynthesis during salinity appeared to be more complicated and may be the result of the effects of ionic strength or osmotic adjustment of the tissue, cell type or organelle, regardless of the particular salt involved (NaCl, KCl or NaNO 3 ).