High Leaf Vein Density Promotes Leaf Gas Exchange by Enhancing Leaf Hydraulic Conductance in Oryza sativa L. Plants

Six cultivated rice genotypes showing different stomatal conductance (gs) values were used to investigate the influence of leaf vein traits on leaf gas exchange and leaf hydraulics. The results showed that gs was the main determinant of the varietal difference in the net photosynthetic rate (PN), whereas the area-based leaf nitrogen content (Narea) and mesophyll conductance (gm) were not main factors. gs and PN were both positively correlated with leaf hydraulic conductance (Kleaf). A high density of leaf veins (vein length per leaf area, VLA), especially minor leaf veins (VLAminor), was of benefit for improving the Kleaf. The proportion of the minor leaf vein length to the total leaf vein length did not impact the leaf hydraulics or leaf gas exchange. Overall, these findings suggested that a high density of leaf veins, especially minor leaf veins, enhances Kleaf and promotes gs and PN in cultivated rice genotypes and a high VLA can be regarded as a high photosynthetic capacity trait in rice plants.


INTRODUCTION
Under the current ambient atmospheric conditions, CO 2 diffusional conductance from air to carboxylation sites is regarded as one of the main limiting factors of net photosynthetic rate (P N ) in C 3 plants (Evans et al., 2009;Li et al., 2009;Yamori et al., 2011;Flexas et al., 2012;Adachi et al., 2013;Gago et al., 2019). To reach carboxylation sites, CO 2 in the air must first overcome the air-leaf boundary resistance to reach the surroundings of the stomata, and it then enters the stomatal pores to reach the substomatal cavity, diffuses to the surroundings of the cell wall, and successively passes through the cell wall, cell membrane, cytoplasm, chloroplast envelope, and the stroma (Terashima et al., 2011;Tholen et al., 2012). The CO 2 diffusional resistance from air to the surface of the leaf is called boundary layer resistance, the CO 2 diffusional resistance from air to the substomatal cavity is called stomatal resistance, and the CO 2 diffusional resistance from the substomatal cavity to the carboxylation sites is called mesophyll resistance. Mesophyll resistance can be as important as stomatal resistance under many conditions (Terashima et al., 2011), although it was ignored in earlier studies (Farquhar et al., 1980;Kodama et al., 2011). The reciprocals of stomatal resistance and mesophyll resistance are called stomatal conductance (g s ) and mesophyll conductance (g m ), respectively. Many previous studies have shown that P N is positively correlated with both g s and g m (Giuliani et al., 2013;Carriquí et al., 2015;Liu and Li, 2016).
Stomatal pores are the common pathway for CO 2 entering the leaf and H 2 O evaporating from the leaf. g s is mainly determined by the stomatal size, stomatal density and distribution, and especially by the stomatal aperture (Xu and Zhou, 2008;Ocheltree et al., 2012;Ouyang et al., 2017). g s has been found to be closely linked to plant hydraulics (Hirasawa et al., 1992;Brodribb et al., 2007;Xiong and Nadal, 2020). Tabassum et al. (2016) reported that the P N , g s , and the transpiration rate (E) values in Oryza sativa L. plants were all positively correlated with whole-plant hydraulic conductance.
As leaf hydraulic conductance (K leaf ) is one of the key components of whole-plant hydraulic conductance (Sack et al., 2003;, it is vitally important for determining leaf gas exchange parameters. Brodribb et al. (2007) found that P N was positively correlated with K leaf across diverse terrestrial plants. Hirasawa et al. (1992) found that g s was positively correlated with K leaf in rice plants across different light and chemical treatments. Taylaran et al. (2011) found that the P N of the high-yield rice cultivar Takanari was higher than that of the other common rice cultivars, which was partly due to its higher K leaf . In higher plants, transpiration drives water from the stem into the petiole, and then, the water enters the midrib, flows in an orderly fashion to different vein branches, passes through the vascular bundle sheath, enters the mesophyll tissue, evaporates into the intercellular airspace, and finally diffuses out of the leaf from stomatal pores (Rockwell et al., 2014;Xiong et al., 2017). Thus, leaf hydraulic transport can be divided into xylem hydraulic transport and outside-xylem hydraulic transport.
According to the water transport route inside leaves, K leaf is mainly determined by the leaf vein traits such as the leaf vein density, the size of the xylem conduits within the bundle sheath (Flexas et al., 2013), and the leaf anatomical traits such as the fraction of intercellular airspace (Xiong et al., 2017). Many recent studies have focused on the relationships between K leaf and leaf vein density (vein length per leaf area, VLA) (Xiong et al., 2015;Tabassum et al., 2016). Although a few studies have demonstrated that K leaf showed no or even negative relationships with VLA, most studies illustrated that a high VLA is beneficial for improving K leaf and promoting g s and P N (Sack and Frole, 2006;Brodribb et al., 2007;Boyce et al., 2009;Brodribb and Field, 2010;McKown et al., 2010;Field et al., 2011;Walls, 2011;Nardini et al., 2012;Flexas et al., 2013). High VLA can improve the parallel transport pathways per leaf area, thus improving the xylem hydraulic conductance (K x ), and it can also shorten the distance from veins to evaporating sites, thus improving the outside-xylem hydraulic conductance (K ox ) (Sack and Frole, 2006;Brodribb et al., 2007;Buckley et al., 2015).
Nevertheless, VLA was found to affect K x but not K ox or K leaf across Oryza species, including cultivated and wild genotypes (Xiong et al., 2015(Xiong et al., , 2017. The irrelevance of K ox and K leaf with VLA may be due to the low ratio of xylem hydraulic resistance (R x ) to leaf hydraulic resistance (R leaf ), which is only 40% (Xiong et al., 2017). Thus, although a high VLA improved K x , the improvement in K x contributed to K leaf very slightly.
Although the average R x /R leaf in Oryza species was only 40%, the ratios in cultivated rice genotypes were much higher than 40% (Xiong et al., 2017), which may be due to the increase in VLA during domestication. Thus, although no correlations were found between K leaf and VLA across Oryza species, these correlations may exist when only considering cultivated rice genotypes because cultivated genotypes have relatively high R x /R leaf (Xiong et al., 2017). Therefore, we hypothesized that in cultivated rice genotypes in which R x /R leaf is relatively high, high VLA can enhance K leaf and promote leaf gas exchange.
In fact, the correlations among leaf gas exchange, leaf hydraulics, and leaf vein traits have already attracted a lot of attention. For example, Brodribb et al. (2007) reported that leaf maximum photosynthetic rate and venation are linked by hydraulics across 43 species. Brocious and Hacke (2016) found that stomatal conductance was positively correlated with the transport capacity of the petiole, estimated from the diameter and number of xylem vessels, whereas the variation in stomatal conductance and leaf hydraulic conductance was not linked to vein density or other anatomical lamina properties across different populus genotypes. However, most previous studies were done in wood species, and rare studies were done in cereal species, especially in rice.
In the present study, six rice genotypes that showed different g s values in a previous study (Ye et al., 2019) were selected to verify the hypothesis that high VLA can enhance K leaf and promote leaf gas exchange in cultivated rice genotypes. The results may provide insights for high photosynthetic capacity rice breeding.

Plant Materials
A pot experiment was conducted at the Huazhong Agricultural University (114.37 • E, 30.48 • N), Wuhan, Hubei Province, China. Six cultivated rice genotypes (O. sativa L.) found worldwide were used, including Kirmizi Celtik, Huayou 675, Teqing, Huanghuazhan, Champa, and N22. The g s of these six genotypes varied greatly in the study by Ye et al. (2019), thus indicating feasibility for studying the underlying mechanisms for different g s and P N values in rice. Rice plants were grown from September to December 2014. After germination on moist filters, seeds were transferred to nursery plates. When the seedlings had developed an average of three leaves, they were transplanted to 11 L pots with a density of three hills per pot and two seedlings per hill. There were five pots per genotype, and each pot was filled with 10 kg of soil. The soil used for the experiment had a clay loam texture, with pH 6.63, organic matter 6.42 g/kg, total nitrogen (N) 0.07 mg/kg, available phosphorus (P) 8.21 mg/kg, and available potassium (K) 126.95 mg/kg. P and K were applied as basal fertilizers at an amount of 1.5 g/pot. N was applied at an amount of 2.0 g(N)/pot, with 40% applied as a basal fertilizer, and 60% applied at the mid-tillering stage. Plants were watered daily, and a minimum 2 cm water layer was maintained to avoid drought stress. Pests were intensively controlled using chemical pesticides.
Rice plants were grown outdoors. All genotypes were arranged in a random design with five replicates. The radiation intensity, average temperature, and relative humidity during rice growth were 12.6 ± 5.7 MJ/m 2 /d, 23.3 ± 3.4 • C, and 74.4 ± 10.4%, respectively. Gas exchange measurements were conducted in a growth chamber (Conviron GR48, Controlled Environments Limited, Winnipeg, MB, Canada) [photosynthetic photon flux density (PPFD), 1,000 µmol(photo)/m 2 /s at the leaf level; temperature, 28 • C; relative humidity, 60%; and CO 2 concentration, 400 µmol/mol] to avoid the influence of a changing environment on gas exchange parameters. The vapor pressure deficit between leaf and air (VPD leaf−air ) of the six rice genotypes during the measurements is shown in Supplementary Table 1. The measurements were conducted 2 weeks after mid-tillering fertilization, and all measurements were conducted on the newly expanded leaves from three different pots of each genotype.

Gas Exchange and Chlorophyll Fluorescence Measurements
A portable photosynthesis system (LI-6400XT, LI-CORInc., Lincoln, NE, United States) with an integrated fluorescence leaf chamber  was used to measure gas exchange and chlorophyll fluorescence on leaves between 08:00 and 16:00. Measurements began after the plants had acclimatized to the chamber for ∼2 h. In the LI-6400XT cuvette, the ambient CO 2 concentration was controlled and set to 400 µmol/mol, the leaf temperature was maintained at 28 • C, the PPFD was 1,500 µmol(photo)/m 2 /s, and the flow rate was 500 µmol/s. After reaching a steady state, which usually takes 25 min, the gas exchange parameters, steady-state fluorescence (F s ), and maximum fluorescence (F m ') were recorded with a light saturating pulse of 8,000 µmol(photo)/m 2 /s. The actual photochemical efficiency of photosystem II ( PSII ) was calculated as follows: The electron transport rate (J) was calculated as follows: where α is the leaf absorptance and β is the partitioning of absorbed quanta between photosystem II and photosystem I. The product αβ was determined from the slope of the relationship between PSII and the quantum efficiency of CO 2 uptake ( CO2 ), which was obtained by varying the light intensity under nonphotorespiratory conditions at <2% O 2 (Valentini et al., 1995). The variable J method described by Harley et al. (1992) was used to calculate the chloroplastic CO 2 concentration (C c ) and g m . C c and g m were calculated as follows: where Ŵ * represents the CO 2 compensation point in chloroplasts without day respiration. The day respiration (R d ) and the apparent CO 2 photocompensation point (C * i ) were determined using the Laisk method (Brooks and Farquhar, 1985). Briefly, A/C i (A, net photosynthetic rate; C i , intercellular CO 2 concentration) curves were measured over the linear portion of the response curve (at 100, 80, 50, and 25 µmol CO 2 /mol air) over three PPFDs (150, 300, and 600 µmol/m 2 /s) with an LI 6400-02B chamber (Li-Cor), and then linear regressions to the responses for each PPFD were fitted for individual leaves. The intersection point of three A/C i curves was considered as C * i (x-axis) and R d (y-axis) (von Caemmerer et al., 1994). Ŵ * was calculated as follows: The Ŵ * and R d of the six rice genotypes are shown in Supplementary Table 2.

Leaf Hydraulic Conductance
The evaporative flux method (EFM) was used to determine the K leaf (Sack et al., 2002;Brodribb et al., 2007;Guyot et al., 2012;Sack and Scoffoni, 2012;Tabassum et al., 2016). A leaf of each genotype was excised in water, and the base of the leaf was placed in a test tube filled with distilled water under favorable conditions for transpiration (in a growth chamber under a PPFD of 1,000 µmol(photo)/m 2 /s and air temperature of 28 • C). Immediately after placing the leaf in a test tube filled with distilled water, the leaf was attached to a Li-COR 6400XT portable infrared gas analyzer (IRGA) (LI-COR, NE, United States) to record the leaf transpiration rate (E). A Dewpoint Potential Meter WP4C (Decagon, Pullman, WA, United States) was used to measure the leaf water potential (Ψ leaf ) of the leaf that was used to measure E. K leaf was calculated as follows (Taylaran et al., 2011):

Leaf Morphological Traits
Leaf width was measured in vivo, and leaves were excised to count the major veins (midrib vein and large veins) and minor veins separately under 40× magnification using a light microscope (SA3300, Beijing Tech Instrument Co., Ltd, Beijing, China). Four leaves were counted for each genotype. As rice leaf veins are parallel to each other, the major leaf vein density (VLA major ), minor leaf vein density (VLA minor ), and total leaf vein density (VLA) were calculated as follows: VLA major = Numbers of leaf major veins × leaf Length Leaf width × leaf length VLA minor = Numbers of leaf minor veins × leaf Length Leaf width × leaf length (8) VLA = VLA major + VLA minor (9)

Leaf Nitrogen Content
The leaf used for gas exchange measurements was detached, and another two leaves were also detached, then the leaf  area of these three leaves was measured using a LI-Cor 3000C (LI-COR Inc., Lincoln, NE, United States) leaf area analyzer. Leaves were then oven-dried at 80 • C until they achieved a constant weight. Afterward, the leaf dry mass was weighed and the leaf mass per area (LMA) was calculated as the ratio of the leaf dry mass to leaf area. The leaf N content based on leaf mass (N mass , %) was measured with an Elementar Vario MAX CN analyzer (Elementar Analysesysteme GmbH, Hanau, Germany), and the area-based leaf N content (N area ) was calculated by multiplying N mass with LMA.

Statistical Analysis
One-way analysis of variance (ANOVA) was used to assess the effects of the genotypes on each parameter using Statistix 9.0 software (Analytical Software, Tallahassee, FL, United States). Parameters were compared between genotypes based on the least significant difference (LSD) test level at the 0.05 probability level. Graphs were created, and a linear regression analysis was performed to test the correlations between the parameters using SigmaPlot 10.0 (Systat Software Inc., CA, United States).

Varietal Differences in Leaf Gas Exchange Parameters and N area
As shown in Table 1, P N , N area , and g s showed significant varietal differences among the six rice genotypes. P N ranged from 21.2 to 36.2 µmol/m 2 /s, and the highest and the lowest values were found in Champa and N22, respectively. Kirmizi Celtik showed the highest N area of 2.03 g/m 2 and N22 showed the lowest N area of 1.23 g/m 2 . g s ranged from 0.18 to 0.46 mol/m 2 /s and the highest and lowest values were found in Champa and N22, respectively. g m showed no significant varietal difference among the six rice genotypes. The highest g m was 0.33 mol/m 2 /s, which was found in Teqing, whereas the lowest g m was 0.23 mol/m 2 /s, which was found in Kirmizi Celtik. As shown in Figure 1, P N was not correlated with N area or g m but was positively correlated with g s across the six rice genotypes.

Varietal Differences in the Leaf Morphological Traits and Leaf Hydraulics
As shown in Table 2, leaf width showed significant varietal differences. Champa possessed the widest leaf of 14.2 mm, whereas Kirmizi Celtik possessed the narrowest leaf of 9.4 mm. VLA, VLA minor , and the proportion of minor leaf vein length to the total leaf vein length all showed no significant varietal differences. VLA major showed significant varietal differences. Huayou 675 had the largest VLA major of 0.82 mm/mm 2 , and N22 had the smallest VLA major of 0.66 mm/mm 2 . E, Ψ leaf , and K leaf all showed no significant varietal differences.

Relationships Among the Leaf Gas Exchange Parameters, Leaf Hydraulics, and Leaf Morphological Traits
As shown in Figure 2, g s was positively correlated with both E and K leaf across the six rice genotypes and P N was positively correlated with K leaf but had no relationship with E across the six rice genotypes. As shown in Figure 3, E, K leaf , g s , and P N were all positively correlated with VLA and VLA minor across the six rice genotypes; P N was positively correlated with VLA major , but E, K leaf , and g s had no relationship with VLA major across the six rice genotypes.

DISCUSSION
In the present study, P N was not correlated with N area and g m but was correlated with g s and K leaf . VLA and VLA minor showed no varietal differences in the present study, which was perhaps due to the few genotypes used, and their vein traits ranged very narrowly. Rice leaf veins are distributed in parallel, and the leaf vein density is strongly affected by the leaf width (Baird et al., 2021). For example, Tabassum et al. (2016) reported that polyethylene glycol-induced water deficit caused rice leaf narrowing and led to an increase in leaf density. Xiong et al. (2015) investigated the leaf morphological traits of 11 rice genotypes, including seven wild genotypes and four cultivated genotypes, and found that leaf morphological traits showed great varietal differences because the leaf area showed ∼seven times the variation, ranging from 18.4 to 127.3 cm 2 , and leaf width showed ∼six times the variation, ranging from 0.38 to 2.20 cm. The leaf width in the present study only showed 1.5 times the variation, ranging from 9.4 to 14.2 mm. Thus, leaf vein densities in the present study did not vary greatly. More genotypes and varying environmental conditions should be included in future studies to investigate the varietal differences in rice leaf vein traits and their impacts on leaf gas exchange.
In the present study, K leaf , E, g s , and P N all substantially increased with VLA and VLA minor across the six cultivated rice genotypes, thus verifying the hypothesis that high VLA can enhance K leaf and promote leaf gas exchange in cultivated rice genotypes in which R x /R leaf is relatively high. By sorting the correlations between leaf gas exchange parameters, leaf hydraulic parameters, and leaf vein densities, we drew the conclusion that leaf vein density affected leaf gas exchange parameters by regulating leaf water transport. However, we still lack the analyses of many other vein traits such as the diameter of xylem conduits and the outside xylem anatomical traits on leaf hydraulics and gas exchange. More studies should be done to explore the coordination between the physiological and structural traits of the leaf. Furthermore, VLA and K leaf data from other studies (Xiong et al., 2015(Xiong et al., , 2017 were extracted to verify our hypothesis. As shown in Figure 4, K leaf was positively correlated with VLA and VLA minor across the cultivated rice genotypes, whereas these correlations disappeared across the wild rice genotypes. Therefore, we conclude that in cultivated rice genotypes, high VLA and high VLA minor can enhance K leaf and promote leaf gas exchange, and a high VLA can be regarded as a feature of rice materials having high photosynthetic capacity. Variations in VLA drove the correlation between K leaf and VLA, which was probably due to the determining role of VLA on K leaf , as other features of hydraulic architecture may vary slightly in cultivated rice genotypes. However, this correlation may be FIGURE 2 | Relationships of stomatal conductance (g s ) and net photosynthetic rate (P N ) with leaf transpiration rate (E) (A,C) and leaf hydraulic conductance (K leaf ) (B,D) across the six rice genotypes. Data are shown as the means ± SD of three replicates. obscured when other features of hydraulic architecture vary. For example, the positive effects of high VLA on K leaf may be attenuated by the negative effects of small xylem conduits in a plant. Besides, as one of the few studies on cereal crops, the present study showed different results with previous studies on wood species, for example, populus plants in Brocious and Hacke's (Brocious and Hacke, 2016) study. A positive correlation was found between K leaf and VLA in the present study but not in Brocious and Hacke (2016) may also be attributed to the different importance of VLA to K leaf in the species. The comprehensive effects of hydraulic architecture on leaf hydraulics should be summarized.
Correlations between K leaf and VLA existing in cultivated but not wild rice genotypes illustrate the importance of leaf vein traits during rice domestication or genetic improvement. For example, Wu et al. (2020) revealed that an increase in leaf vein dimensions during rice genetic improvement enhanced leaf gas exchange. In addition to leaf vein traits, leaf morphological traits associated with leaf gas exchange, such as stomatal density and stomatal size in rice, have also been improved during crop genetic improvement (Panda et al., 2018;Wu et al., 2020). There is a need for establishing the underlying morphological and physiological changes for crop improvement.
A previous study (Venturas et al., 2017) illustrated that minor leaf vein length accounts for more than 80% of the total leaf vein length in higher plants. In the present study, minor leaf vein length accounted for 83.5% of the total leaf vein length on average ( Table 2). Xiong et al. (2015) reported that the proportion of minor leaf vein length to total leaf vein length but not leaf vein density is the critical factor determining K leaf . However, in the present study, neither E, K leaf , g s , nor P N showed significant correlations with the proportion of minor leaf vein length to total leaf vein length (data not shown).
Except for the leaf vein density, other vein traits also greatly impacted K leaf . Previous studies documented that short distance from minor veins to stomata and large diameters of xylem conduit accompanied by thin xylem conduit walls were beneficial for improving K leaf (Brodribb et al., 2007;Venturas et al., 2017;Xiong and Nadal, 2020). Nardini et al. (2014) found that large leaves tend to have higher VLA major than small leaves and can promote K leaf . Simonin et al. (2012) demonstrated that K leaf increased with FIGURE 3 | Relationships of the leaf hydraulics and leaf gas exchange parameters with (A-D) total leaf vein density (VLA), (E-H) major leaf vein density (VLA minor ), and (I-L) minor leaf vein density (VLA minor ) across the six rice genotypes. Data are shown as the means ± SD. There were four replicates for leaf vein densities and three replicates for P N , g s , K leaf , and E. leaf area. However, a recent study exploring the development and biophysical determinations of grass leaf size worldwide revealed that small leaves have hydraulic benefits due to their higher VLA major (Baird et al., 2021). Optimal leaf morphological and anatomical traits for improving K leaf and P N should be defined to provide theories for high photosynthetic capacity rice breeding. K x is mainly determined by leaf vein traits, whereas K ox is mainly determined by leaf anatomical traits. Xiong et al. (2017) found that K ox is positively correlated with the fraction of intercellular airspace, the mesophyll cell surface area exposed to intercellular airspace per leaf area, and the surface area of chloroplasts exposed to intercellular airspace per leaf area, whereas it is negatively correlated with cell wall thickness. g m has been reported to be positively correlated with g s in recent years (Flexas et al., 2013). One reason for the positive relationship between g s and g m is that they are both positively correlated with K leaf (Xiong and Nadal, 2020). g s is positively correlated with K leaf due to the close relationship between stomatal opening and leaf water transport; g m is positively correlated with K leaf and K ox because g m and K ox are mediated by leaf anatomical traits (Xiong et al., 2017). However, K x , K ox , and leaf anatomical traits were not measured in the present study, and more studies should be done to investigate and explain the coordination among leaf gas exchange, leaf hydraulics, and leaf anatomy.
In addition, the water transport pathways inside leaves, i.e., where the evaporating sites occur, are currently unclear. Rockwell et al. (2014) reported that water dominantly evaporated from the vascular bundle sheath and its surroundings to FIGURE 4 | Relationships of leaf hydraulic conductance (K leaf ) with leaf vein densities across (A-C) cultivated rice genotypes and (D-F) wild rice genotypes. The solid circles are data for the six cultivated rice genotypes in the present study, and the empty circles are data from Xiong et al. (2015Xiong et al. ( , 2017 studies. the intercellular airspace in Quercus rubra. Moreover, the conversion of water between the vapor and liquid phases is unknown. For example, Buckley et al. (2015) found that vapor phase transport contributed 39.25-44.00% to K ox when the temperature difference between the xylem and leaf epidermis reached 0.2 K, suggesting that temperature greatly affects leaf hydraulic status. Evaporating sites and leaf hydraulic status inside rice leaves need more investigation.

CONCLUSIONS
In the present study, VLA major showed great varietal differences among cultivated rice genotypes, whereas VLA, VLA minor , and the proportion of minor leaf vein length to total leaf vein length showed no varietal differences. E, K leaf , g s , and P N were all positively correlated with VLA and VLA minor , and P N was positively correlated with VLA major . We concluded that high VLA and high VLA minor can enhance K leaf and promote leaf gas exchange in the cultivated rice genotypes and high VLA can be regarded as a high photosynthetic capacity trait in rice plants.

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author.

AUTHOR CONTRIBUTIONS
MY designed the research and wrote the paper. MY and ZuoZ conducted the experiments. MY and MW analyzed the data. HZ and ZujZ commented and revised the paper. All authors contributed to the article and approved the submitted version.