Impact of salinity on biomass allocation and transpiration
There was a significant reduction in the total biomass of all three cultivars under the 30 mM treatment (P < 0.05), and, for Paulsen and SO4, also under the 10 mM treatment (P < 0.01; Table 1). All three cultivars had a greater reduction in the bottom part of the roots at the 10 mM treatment (P < 0.01), while in the top part of the roots, a significant reduction was observed only under the 30 mM treatment (P < 0.05). A significant decline in stem mass was observed in Paulsen and SO4 only under the 30 mM treatment (P < 0.05) but was not observed in R-110. Leaf mass was significantly reduced in Paulsen under the 10 mM treatment (P < 0.01). However, R-110 and SO4 did not have significant changes. Consequently, root mass allocations decreased under the 30 mM treatment, mainly in the bottom part of the roots, with 53%, 67%, and 86% reductions in Paulsen, R-110, and SO4, respectively, compared to control. Furthermore, the top part of the root system decreased by 39%, 44%, and 49% under the 30 mM treatment in Paulsen, R-110, and SO4, respectively. The ratio between leaf mass and total mass increased in all three cultivars, while the stem mass ratio increased in R-110 and SO4 but did not change in Paulsen. Therefore, the ratio of root to shoot significantly decreased under the 10 mM treatment, compared to the control, for all three cultivars (P < 0.05) and under the 30 mM treatment only in R-110 and SO4 (P < 0.01; Table 1).
All three cultivars reduced their cumulative transpiration under the 30 mM treatment compared to control, 30% reduction for Paulsen and SO4 and 40% reduction for R-110. Paulsen and SO4 also had a significant reduction in their cumulative transpiration under the 10 mM treatment, while for R-110, the reduction was not significant (P<0.05; Fig. 1a). This pattern of reductions was similar between the 10 mM and 30 mM treatments. Paulsen and SO4 had a significant reduction, while R-110 did not. Paulsen transpired more than SO4 Under the control treatment and more than R-110 under the 30 mM treatment. Only R-110 increased its WUE under the 30 mM treatment compared to control (one-way ANOVA, P < 0.01; Fig. 1b). This increase in R-110's WUE is due to its relatively low reduction in shoot biomass combined with a relatively high reduction in its cumulative transpiration. SO4's WUE under the control treatment was higher than R-110's WUE and higher but not significantly than Paulsen's WUE. Under the 30 mM treatment, R-110's WUE increased while Paulsen's did not. Therefore, its WUE was lower than SO4's WUE.
Root system response to salinity
Root formation efficiency increased for all three cultivars, displaying a significant increase in SRA in response to salinity (P < 0.01). However, R-110 had a higher SRA only under the 30 mM treatment than the control, while Paulsen and SO4 had a significant increase under the 10 mM treatment compared to the control (P < 0.05; Fig. 2). The root formation efficiency improved from 8.7, 7.6, and 7.6 m2/kg, under the control treatment, to 13.8, 12, and 10.9 m2/kg, under the 30 mM treatment for SO4, Paulsen, and R-110, respectively. Paulsen's and SO4's root formation efficiency also increased under the 10 mM treatment compared to the control, while the increase in R-110 was not significant.
Reductions of 26%, 23%, and 16% in root tissue density were observed under the 10 mM treatment in Paulsen, R-110, and SO4, respectively (Fig. 3). Also, under the 30 mM treatment, 35%, 32%, and 23% reductions were observed in Paulsen, R-110, and SO4, respectively. Furthermore, under the 10 mM treatment, 4%, 4%, and 16% reductions of root diameter were observed in Paulsen, R-110, and SO4, respectively, though for Paulsen and R-110, the reductions were not significant. Under the 30 mM treatment, 10%, 12%, and 24% reductions were observed in Paulsen, R-110, and SO4, respectively, while for R-110, it was not significant.
RLD was estimated for the entire root system and was separated based on different diameters. RLD showed that narrow roots (between 0 and 1 mm) did not substantially change in response to salinity in any of the three cultivars (Fig. 4). Paulsen also did not have significant changes between treatments in any of the root diameters (Fig. 4a), while R-110's and SO4's thick roots (between 1 and 4 mm) had significant reductions in RLD under the 30 mM treatment, compared to the control (P < 0.05; Fig. 4b, c). R-110's 1–2 mm roots' RLD significantly decreased under the 10 mM treatment but not between the 10 mM and 30 mM treatments. R-110's thicker roots' RLD (2–4 mm) decreased significantly only under the 30 mM treatment, compared to the control, while the RLD of the 3–4 mm roots did not have significant changes between treatments (P < 0.05; Fig. 4b). SO4's 1–2 mm roots' RLD reduced only under the 30 mM treatment, compared to the control, with no changes under the 10 mM treatment, but SO4's 2–3 mm and 3–4 mm roots' RLD was significantly reduced under the 10 mM and the 30 mM treatments, compared to the control (P < 0.05; Fig. 4c).
All three cultivars had changes in their root diameters, sharpening the curve in response to increased salinity levels (Fig. 5). This result indicates that thinner roots contributed more to the surface area as salinity increased. Only SO4 had significant changes between the control and salinity treatments (P < 0.05). SO4's D-50 and D-90 were reduced by 0.5 and 1 mm, respectively, between the control and the 30 mM treatment (Fig. 5c), but no significant differences were observed between the 10 mM and the 30 mM treatments. However, Paulsen and R-110 had significantly lower root diameters in the bottom part of their root system under the salinity treatments (P < 0.05; SI table 1), the differences in their D-50 and D-90 were not significant. The reduction of the average root diameter contributing to 50% and 90% of the cumulative surface area reached 0.14 and 0.2 mm, respectively, in Paulsen under the 30 mM treatment (Fig. 5a) and 0.15 and 0.37 mm in R-110 under the 30 mM treatment (Fig. 5b).
Root slices taken 15 mm from the root tip revealed that the stele area significantly decreased in Paulsen, but not in R-110 and SO4, under the 30 mM treatment, compared to the control (P < 0.05; Fig. 6a). The cortex area did not change (Fig. 6b), and the cortex to stele ratio significantly increased in response to salinity in Paulsen (P < 0.05) and R-110 (P < 0.01), while in SO4, the increase was close to statistically significant (P = 0.051; Fig. 6c). Also, the stele area, 40–50 mm from the root tip, significantly decreased in R-110 under the 30 mM treatment, compared to the control (P < 0.01; Fig. 6d). The cortex area did not change (Fig. 6e), and only R-110's cortex to stele ratio significantly increased under the 30 mM treatment, compared to the control (P < 0.05; Fig. 6f). No significant differences were observed in root area, xylem area, and xylem density measured 15 mm and 40–50 mm from the root tip (SI fig. 1).
Cl- and Na+ accumulations in the roots
Cl- accumulation in the fine roots showed a significant increase for all three cultivars as salinity increased (P < 0.01). Cl- accumulation in Paulsen's, R-110's, and SO4's fine roots significantly increased in response to both salinity treatments (P < 0.05), but no significant differences were observed between the 10 mM and 30 mM treatments (Fig. 7a). Paulsen had 0.43, 0.55, and 0.67 mmol/g under the control, 10 mM, and 30 mM treatments, respectively. R-110 had similar results with 0.22, 0.6, and 0.68 mmol/g under the control, 10 mM, and 30 mM treatments, respectively. However, SO4 had lower levels of Cl- than R-110 under the 10 mM treatment and lower than both Paulsen and R-110 under the 30 mM treatment, with 0.19, 0.4, and 0.49 mmol/g under the control, 10 mM, and 30 mM treatments, respectively. The fine roots' Na+ concentration increased between the control and the 10 mM treatment for all three cultivars and increased between the 10 mM and 30 mM treatments in Paulsen's and SO4's roots but not in R-110's roots (Fig. 7b). SO4 had a lower root Na+ concentration than R-110 under the 10 mM treatment and compared to Paulsen under the 30 mM treatment. All three cultivars had lower K+/Na+ ratios under the 10 mM and 30 mM treatments compared to control. Under the control treatment the K+/Na+ ratios were 2.9, 2.8 and 3.6 for Paulsen, R-110 and SO4, respectively, but with no significant differences between them (data not shown). All three cultivars had lower K+/Na+ ratios under the 30 mM treatment compared to the 10 mM treatment (Fig. 7c). SO4 had a higher K+/Na+ ratio compared to Paulsen and R-110 under the 10 mM treatment. Under the 30 mM treatment, SO4's K+/Na+ ratio was slightly higher but not significantly compared to Paulsen and R-110.