Leaf stiffness of two Moraceae species based on leaf tensity determined by compressing different external gripping forces under dehydration stress

ABSTRACT Leaf water status determination based on mechanical and electrophysiological properties helps determine the inherent as well as instantaneous leaf dehydration tolerance synchronously. The leaf water potential (ΨL), physiological capacitance (CP) and gripping force (F) were determined with leaves of Broussonetia papyrifera (L.) Vent. and Morus alba L. Real-time leaf stiffness (LCSrt) and maximum leaf stiffness (LCSmax) were investigated by compressing a leaf with external gripping force. Results indicated that LT displayed good correlation with F. Compared to M. alba, a better instantaneous dehydration tolerance or pressure resistance in B. papyrifera was correlated to its persistent stronger LCSrt or LCSmax, respectively. B. papyrifera showed better flexibility and tolerance to wider range of pressure than M. alba. The higher leaf mechanical strength helped to maintain a higher outward pulling force of cell walls; thus, the subsequent negative pressure effectively inhibited cellular water loss. B. papyrifera exhibited better drought resistance than M. alba.


Introduction
Water maintains cell tension and helps branches and leaves stand upright. Furthermore, it generates turgor that contributes to the movement of stoma or other plant structures and cellular growth (Kroeger et al. 2011). Plant growth is often affected by biotic or abiotic stresses, one of which is drought (Grossi et al. 2016). However, drought resistance differs amongst plant species, and the rapid determination of plant drought resistance helps in the implementation of appropriate irrigation strategies on the plant (Egea et al. 2017). The assessment of leaf traits is one of the methods for studying plant drought resistance (Petrov et al. 2018), a slower water loss rate in the leaf is associated with better drought resistance of plants (Zhou et al. 2018).
Leaf water status could be rapidly determined by analyzing the variation of leaf tensity (LT) during the water loss process (Zhang et al. 2015). Calculated according to the coupling relationship between leaf water potential (Ψ L ) and physiological capacitance (CP), LT reflected water status better than Ψ L , and Ψ L was easily affected by the environment (Javed et al. 2017). The online monitoring and prediction of leaf water status can be realized through the determination of LT. However, pressure variation caused by the gripping force (which was used for clamping leaf during the CP determination) changes the concentration of the cytosol solute and the elasticity and plasticity of the cells in the leaf, which causes dielectric constant variation in the cytosol solute. Thus, the values of LT change (Zhang et al. 2015). Therefore, the subsequently assessed plant drought resistance varies as gripping force changes. As such, the plant drought resistance determined by using a specific gripping force was defined as instantaneous drought resistance. Measurements of CP should be conducted under the same gripping force, in order to make the comparison of drought resistance among different plant species credible. However, it is difficult to keep the gripping force consistent at each measurement (Xing et al. 2018). Therefore, it is unreasonable to compare the drought resistance among different plant species only based on the electrophysiological properties.
Mechanical properties, which are closely related to the internal architectures, proved to be another useful tool to investigate drought resistance of plants (Balsamo et al. 2015). Leaf mechanical properties were affected by water status, and a positive correlation between leaf internal architecture, tensile strength and tolerance to dehydration for grasses were observed (Balsamo et al. 2006;Rascio et al. 2015). Plant cells are composed of the cell wall and intracavitary substances (mainly protoplasm), and cell turgor plays an important role in the mechanical stability (Malgat et al. 2016). The wall, located outside the membrane, is a thick and tough layer with a slight elasticity. It not only forms a strong network that functions like a boundary, preventing unconstrained flow of water or nutrition and maintaining internal pressures of cells but also plays key roles in bearing external load and maintaining the mechanical strength of the plant body (Cosgrove 2016). Lignin, which fills in the cellulose cytoskeleton, is a necessary element for the cell walls of all vascular plants. Lignin strengthens the hardness of the cell walls and enhances the mechanical support and compressive strength of the cells (Boerjan et al. 2003). Generally, plants with higher drought resistance could enhance the mechanical strength of cell walls by rapidly improving lignin biosynthesis under drought stress conditions (Fan et al. 2006;Yin et al. 2017). High stiffness is correlated to higher mechanical strength of the wall, which subsequently yields negative pressure and effectively inhibits water loss (Deng and Zhang 1998;Charrier et al. 2016). However, mechanical properties are always measured at the point of failure load. Therefore, tissue water status determined based on mechanical properties is an inherent value and implies the inherent drought resistance of plants. Nevertheless, the plant tissues will be damaged irreversibly during the measurement. Therefore, it is hard to continuously monitor plant drought resistance through mechanical properties.
Considering the above drawbacks, the aims of this study were to investigate the coupling relationship between the gripping force and electrophysiological parameters, synchronously determine and assess the inherent and instantaneous plant drought resistances, and combine with the variability of their difference, then comprehensively compare the drought resistance among different plant species. As such, the continuous monitoring of drought resistance of plants can be realized and the influence of inconsistent gripping force on leaf water determining can be reduced or avoided.
Broussonetia papyrifera (L.) Vent. and Morus alba L., which belong to the Moraceae family, are characterized by a rapid growth rate and greater adaptability to adversities than other members of the family (Wu et al. 2009). These species are always cultivated as medicinal or ornamental plants. Researches revealed that B. papyrifera exhibited better drought resistance than M. alba due to its higher bicarbonate use capacity and better water status (Wu and Xing 2012). The latter maintained the LT of B. papyrifera (Zhang et al. 2015). In this study, B. papyrifera and M. alba were selected as experimental materials, researches on the behaviors of leaf mechanical combined with electrophysiological properties under dehydration stress help understand the biomechanics mechanisms of plants in adapting to drought adversity. And a method for rapidly and comprehensively determining plant drought resistance based on the mechanical and electrophysiological properties could be developed.

Plant materials
The experiment was performed on the campus of Jiangsu University, Jiangsu Province, China (32.20°N, 119.45°E). Eight years old B. papyrifera and M. alba plants which grew in the yellow-brown soil on a sunny slope were selected as the experimental materials. The area receives a mean annual precipitation of approximate 1100 mm and has a mean annual air temperature of about 15.6°C. Fresh branches from the two plants were picked in July. Leaves growing uniformly were taken from the fourth and fifth leaf positions of each branch. The fresh leaves removed from the branches were placed in double distilled water immediately and soaked for 30 min in order to keep all the leaves in the same initial state (water-saturated), which could make the comparison of drought resistance among different plant species credible. After soaking, water on the surface of the leaves was removed. Finally, the detached leaves were placed on a dry ventilated desktop in the laboratory for 5 h, the temperature was 26°C , Photosynthetic Photon Flux Density (PPFD) was 160 μmol m −2 s −1 , and relative air humidity was 40%. Measurements were done in triplicate at 0 (baseline), 1, 2, 3, 4 and 5 h after water loss.

Physical model of LT and F
The gravity equation is: where F is the gravity (gripping force; unit: N), M i is the mass of iron (unit: kg), m is the mass of the foam board and electrode (unit: kg), and g is the acceleration of gravity with a value of 9.80 N kg −1 . Cytosol solute in the leaf was taken as the dielectric. The leaf was clipped between the two electrodes of the parallelplate capacitor, which formed a parallel-plate capacitor sensor. LT of the plant under different F could be determined by changing the mass of iron in the parallel-plate capacitor (Figure 1(a)). Pressure variation changed the concentration of the cytosol solute and the elasticity and plasticity of the cells in the leaf, which caused dielectric constant variation in the cytosol solute and leaf effective thickness (d L ) between the two electrodes of the parallel-plate capacitor. Thus, the LT changed.
The elasticity of the cell in the leaf was correlated to the water content in the cells. LT differed with plant species under the special F.
The equation for Gibbs's free energy is: The equation for the energy of the capacitor is: where W C is the energy of the capacitor, which is equal to the work converted from Gibbs's free energy (ΔG, W C = ΔG); ΔH is the internal energy of the system composed of cells in the plant leaves; P is the pressure imposed on the plant cells, V is the volume of plant cells; and U is the test voltage. P can be calculated using the following equation: The CP of the leaf was expressed using Equation (5): where ɛ 0 is the vacuum dielectric constant with a value of 8.854 × 10 −12 F m −1 and ɛ r is the relative dielectric constant of the cytosol solute. LT was calculated according to Equation (6). The unit was cm 2 cm −1 (Zhang et al. 2015): where A CP is the effective area of the leaf in contact with the capacitor plates (unit: cm 2 ); d L is the leaf effective thickness (unit: cm); i is the dissociation coefficient (with value of 1); R is the gas constant (with value of 8.30 × 10 −3 L MPa mol −1 K −1 ); T is the thermodynamic temperature (T = 273 + t°C, unit: K); ɛ 0 is the vacuum dielectric constant (with value of 8.854 × 10 −12 F m −1 ); a is the relative dielectric constant of the cytosol solute; M is the relative molecular mass of the cytosol solute (unit: g mol −1 ); and 81 is the relative dielectric constant of water at normal temperature. In this study, the sugar C 12 H 22 O 11 was identified as the solute in the cytosol; therefore, a was 3.30, M was 342 g mol −1 , and t was 20°C. Equation (6) could be rewritten as: According to Equations (2), (3), (4), (5) and (7), the relationship between LT and F could be expressed as follows: Incorporating 2DH 1 0 1 r U 2 = y 0 and 2V 1 0 1 r A CP U 2 = k into Equation (8) changes this equation to: where y 0 and k are the model parameters. Determination of LT under different F was conducted on fresh leaves taken from fresh branches. The CP of these leaves was measured using an LCR tester (model 3532-50, Hioki, Nagano, Japan). The frequency and voltage used were 3 kHz and 1 V, respectively. Each leaf was clipped onto the custom-made parallel-plate capacitor (a) with a diameter of 10 mm (Figure 1(a)). With a dew point microvoltmeter in a universal sample room (C-52-SF, Psypro, Wescor, Logan, Utah), Ψ L was measured at the same position of the leaves with the above CP testing.
The relationship curve between LT and F for B. papyrifera or M. alba was established using Sigmaplot (ver. 12.5, Systat Software, Inc., San Jose, Cal.). The relationship between LT and F was fitted, respectively. The model parameters y 0 and k of B. papyrifera or M. alba were estimated, respectively.

Calculation of leaf stiffness
The effective thickness per leaf area (d LA , cm cm −2 ) could be calculated using Equation (10): During the process where the leaf was compressed by an external force, the moving distance of the force was defined as d m (unit: cm cm −2 ). According to Equations (9) and (10), d m could be calculated as follows: where d LA0 is d LA under 0 N external force and d LAF is d LA under F N external force. Meanwhile, as the leaf was compressed by an external force, pressure per leaf area was presented as stress (σ, N mm −2 ). σ was calculated as follows: where A F is the area of the leaf in contact with the probe implementing the gripping force. According to Equations (11) and (12), leaf stiffness (LCS) was calculated as follows: The unit of LCS was N mm −1 .
Determination of real-time leaf stiffness LT and F at each moment of water loss were defined as realtime LT (LT rt ) and real-time F (F rt ), respectively. The CP of leaves at each moment of water loss was measured using the LCR tester with a frequency and voltage of 3 kHz and 1 V, respectively. The leaf was clipped onto the custommade parallel-plate capacitor (b) with a diameter of 10 mm (Figure 1(b)). With the same dew point microvoltmeter in a universal sample room, Ψ L was also measured at the same position of the leaves with the above CP testing. LT rt was calculated according to Equation (7) based on the above values of CP and Ψ L , and F rt was calculated according to Equation (9). The diameter of the probe implementing F rt was 10 mm. A F was calculated as follows: A F = 25p, where the value of π was 3.14. Then LCS rt was calculated by using the following equations: LCS rt = 10 × y 0 × (y 0 + kF rt ) k × 25p .

Determination of maximum leaf stiffness
The maximum gripping force (F max ) of leaf at each moment of water loss was measured with the texture analyzer TA.XT-Plus (Stable Micro System, United Kingdom) using the P/2n probe with a diameter of 2 mm. The instrument working parameters were determined by the test mode compression; with pretest speed at 2 mm s −1 , test speed at 1 mm s −1 , post-test speed at 2 mm s −1 and trigger force at 100 N (ensure that the leaf is crushed and the cell is broken up). The diameter of the probe implementing F max was 2 mm, A F was calculated as follows: A F = p, where the value of π was 3.14. Then LCS max was calculated by using the following equations: LCS max = 10 × y 0 × (y 0 + kF max ) k × p .

Statistical analysis
All collected data were analyzed using SPSS software (version 13.0, SPSS Inc., New York). The differences between the stress levels were assessed using the least significant difference posthoc test at 5% significance level (P ≤ .05). The data were shown as the means ± standard errors determined using the one-sample T-test. The confidence interval was 95%.

Relationship between LT and F
The model parameters y 0 and k of B. papyrifera and M. alba were estimated using Equation (9). The relationship curves ( Figure 2) between LT and F for B. papyrifera and M. alba was obtained using Sigmaplot (ver. 12.5, Systat Software, Inc., San Jose, Cal.). The relationship between LT and F was fitted. The fitting equations between LT and F for B. papyrifera and M. alba were LT = 0.43 + 0.46F(R 2 = 0.97, P < .0001, n = 16) and LT = 0.06 + 0.19F(R 2 = 0.96, P < .0001, n = 16), respectively. Higher F values were correlated with higher LT values of B. papyrifera and M. alba.

Ψ l and LT rt at each moment of water loss
Leaf water potential of B. papyrifera was the highest at 1 h and the lowest at 5 h (Table 1). Ψ L values of B. papyrifera at 0, 2 and 3 h exhibited no significant difference. Ψ L of M. alba decreased significantly under drought conditions compared with that at 0 h. The value of Ψ L at 5 h was the lowest. Values of Ψ L in M. alba at 1 and 2 h or 3 and 4 h exhibited no significant difference. LT rt of B. papyrifera at 0, 1 and 2 h showed no significant difference. The LT rt values at 0 and 1 h were the highest, whereas the value at 5 h was the lowest. LT rt of M. alba at 0 h was the highest. Low LT rt values were correlated with increasing water loss moments. LT rt values at 1, 2 and 3 h exhibited no significant difference, whereas the value at 5 h was the lowest. Table 2 shows the values of F rt and LCS rt at each moment of water loss. F rt exhibited the same variation with LT rt within the same plant species. LCS rt of B. papyrifera at 1 and 2 h showed no significant difference compared with that at 0 h. These values were higher than those at other water loss moments, and the value at 5 h was the lowest. Low values of LCS rt in M. alba were correlated with increasing water loss moment, and the values at 1, 2 and 3 h exhibited no significant difference. Value of LCS rt in B. papyrifera at 3 h was still 78% of that at 0 h, while the value of LCS rt in M. alba at 1 h decreased to 72% of that at 0 h. Values of LCS rt in B. papyrifera were significantly higher than that in M. alba at each water loss moment.  −2.78 ± 0.06d −3.18 ± 0.10d 0.51 ± 0.02c 0.13 ± 0.01d Note: The mean ± SE (n = 5) followed by different letters in the same column differ significantly at P ≤ .05, according to one-way ANOVA and t test.  Table 3 shows the values of F max and LCS max at each moment of water loss. High values of F max in B. papyrifera and M. alba were correlated with increasing water loss moment. The increase in B. papyrifera was more significant than that in M. alba. F max of M. alba at 2, 3 and 4 h exhibited no significant difference. LCS max exhibited the same variation with F max within the same plant species. Value of LCS max in B. papyrifera at 2 h was already 131% of that at 0 h, while the value of LCS max in M. alba at 4 h increased to 132% of that at 0 h. Values of LCS max in B. papyrifera were significantly higher than that in M. alba at each water loss moment.

Variability of the difference between LCS rt and LCS max
High values of LCS vc in B. papyrifera and M. alba were correlated with increasing water loss moment (Table 4). The increase in B. papyrifera was more significant than that in M. alba. B. papyrifera exhibited higher average variability of LCS vc (24.67) than M. alba (12.67) during the water loss treatment period.

Discussion
The water loss rate can be used to characterize the dehydration resistance of a detached leaf. Water regulation caused by enzymes, such as carbonic anhydrases (CAs, EC 4.2.1.1), changes the leaf water status under drought conditions (Wu and Xing 2012). And it becomes difficult to diagnose the plant water deficit only by using Ψ L . The decrease of Ψ L in B. papyrifera at each water loss moment was slighter than that in M. alba, which indicated that B. papyrifera exhibited better water status than M. alba under dehydration stress conditions. However, the moderate decrease of Ψ L in B. papyrifera could help improve water absorption capacity and reduce natural water loss in a short time. This result was consistent with the research by Ren et al. (2015) in rice. LCS rt , which was measured mainly based on the determination of electrophysiological parameters, represented the instantaneous water status determined by using a specific gripping force. When the gripping force used for CP determination was far lower than the force of failure load, the variations of LT rt were always correlated to the vacuolar concentration and the elasticity and plasticity of the cells. Water loss in leaves decreased leaf thickness and increased the vacuolar concentration ( Figure 3). LT rt and F rt decreased when the leaves suffered continuous water loss. As a result, the calculated values of LCS rt decreased correspondingly. In fact, a slower water loss speed of leaf caused slighter decrease of LT rt in B. papyrifera than that in M. alba. The LCS rt in B. papyrifera decreased by only 22% in the third hours, while the value in M. alba decreased by 28% in the first hour. The persistent stronger LCS rt at water loss moment meant better instantaneous dehydration tolerance in B. papyrifera. However, some researches have demonstrated that the pressure variation caused by the gripping force also changed the vacuolar concentration, which influenced the LT rt . As a result, the subsequently assessed drought resistance varied as gripping force changed (Zhang et al. 2015;Xing et al. 2018). Therefore, it was unreasonable to compare the drought resistance only based on the electrophysiological properties.
LCS max , which was measured mainly based on the determination of mechanical parameters, represented the inherent water status. It was closely related to the leaf internal architectures. Plant cells need to maintain the dynamic balance of a variety of forces under adversity. For example, it is necessary to have a certain elasticity so as to stretch the cells, or an appropriate stiffness to maintain turgor (Malgat et al. 2016). Water adversity provokes lignin synthesis and cell-wall thickening (Eynck et al. 2012), which provide mechanical support for the wall and prevents the cell wall from collapsing caused by negative pressure during plant transpiration (Lewis and Yamamoto 1990). At the moment the cell lost water and shrank, the wall with stronger stiffness pulled out the plasma membrane and produced negative pressure. As a result, the continuous water loss in the cell was effectively inhibited, thus decreasing the water loss rate (Deng and Zhang 1998). When the gripping force implemented on leaves was high enough (equal to failure load), leaf cells could no longer shrink, and F max was mainly influenced by the substance composition of leaf cells (Figure 3). Since lignin synthesis and cell wall thickening of plants could be provoked by water deficit, F max increased when the leaves suffered continuous water loss. As a result, the calculated values of LT max and LCS max increased correspondingly. A stronger Notes: The mean ± SE (n = 5) followed by different letters in the same column differ significantly at P ≤ .05, according to one-way ANOVA and t test.
a This column stands for the percent value after water loss treatment with reference to that of the 0 h. lignin synthesis and cell-wall thickening capacity in droughttolerant plants caused more significant increase of LCS max in B. papyrifera than that in M. alba. Higher LCS max values at each water loss moment and more significant increase of LCS max with increasing water loss time indicated that B. papyrifera exhibited better pressure resistance under dehydration conditions. Besides, the above response trait could also be used for evaluating the quality of B. papyrifera tissues when they were used as raw materials for paper-making. However, since the LCS max needed to be measured at the point of failure load, the plant tissues would be damaged irreversibly during the determination, therefore, it became hard to realize the online monitoring of plant drought resistance. The difference between LCS max and LCS rt was defined as leaf stiffness variable capacity (LCS vc ) in this study. The average variability of LCS vc during the water loss treatment period could represent the flexibility and range of external forces tolerated by leaf. The average variability of LCS vc in B. papyrifera was higher than that in M. alba, which indicated that the leaves of B. papyrifera showed better flexibility and tolerance to a wider range of external forces under dehydration conditions. The higher leaf mechanical strength in B. papyrifera helped to maintain a higher outward pulling force of cell walls in the leaf; thus, the subsequent negative pressure could effectively inhibit cellular water loss. Therefore, B. papyrifera exhibited better drought resistance than M. alba. This result is consistent with the results of the studies by Wu et al. (2009) andZhang et al. (2015).

Conclusions
Leaf water potential could not completely reflect the plant water status due to the influences of the surroundings and water regulation caused by enzymes, such as CA in plants.
Although the determination of LT was less influenced by the surroundings, the results were easily affected by the area of the leaf in contact with the plates of custom-made parallel-plate capacitor or the gripping force. In this study, the variations of leaf mechanical and electrophysiological properties were synchronously analyzed by determining LT together with F; thus, LCS rt , LCS max and the variability of LCS vc were investigated to analyze the leaf dehydration tolerance. The influence caused by the variation of the leaf area in contact with the plates of the custom-made parallel-plate capacitor and inconsistent gripping force could be reduced or avoided.
LT displayed good correlation with F. The persistent stronger LCS rt , more significant increase in LCS max , together with the higher average variability of LCS vc indicated that B. papyrifera exhibited better pressure resistance and better flexibility and tolerance to wider range of external forces under dehydration conditions. The higher leaf mechanical strength in B. papyrifera helped to maintain a higher outward pulling force of cell walls in the leaf; thus, the subsequent negative pressure could effectively inhibit cellular water loss. Therefore, B. papyrifera exhibited better drought resistance than M. alba. The synchronous investigation of LCS rt , LCS max and the variability of LCS vc could not only realize the online monitoring of plant drought resistance but also reduce the influence of inconsistent gripping force on leaf water determination. The above-mentioned method could also be used for detecting the plant responses to other environmental stresses.

Disclosure statement
No potential conflict of interest was reported by the authors.