Mutant Population
The fast-neutron library was developed from the photoperiod-insensitive, purple rice variety Jao Hom Nin (JHN), by Rice Science Center, Kasetsart University, Thailand, as previously described (Ruengphayak et al., 2015).
Stomatal variation screening
The rice plants used in this study were 216 mutant lines of the JHN Mutant Core Collection. All rice lines were screened by observing the stomatal characteristic under a compound light microscope. Candidate lines identified in the study were used to evaluate drought tolerance.
Phenotypic variation in stomatal traits
To identify plants with stomatal phenotypes from the mutant population, we screened the flag leaves of 216 Mutant Core Collection members by measuring the stomatal size (guard cell length) and density on 3 biological replicates per line (Fig. 1, Supplementary Fig. 1). Stomatal traits of the 216 mutants derived from an indica rice cv. Jao Hom Nin were screened, yielding four candidate lines that have specific stomata traits. Fully expanded leaves were collected from the 216 Mutant Core Collection members grown in the experimental field at the Kasetsart University, Kamphaeng Saen campus. Stomata were localized in rows parallel to leaf veins (Figure 3).
Microscopic observation of stomatal traits
Leaf sections from the adaxial and abaxial leaf surfaces were imprinted on glass slides by applying dental resin (Coltene Whaledent, Switzerland) or nail varnish peels over a fully expanded leaf surface to achieve leaf imprinting. Cell counts were captured from the middle area of leaves from at least eight plants per genotype and using six fields of view per leaf. Leaf epidermal imaging was conducted using a camera fitted on a light microscope (Leica, DM750-ICC50 HD) and ImageJ software (Fiji v. 1.51u). Stomatal density within each field of view was counted manually and total stomatal density per mm2 of leaf was calculated by applying a set conversion factor. Stomatal size was determined by measuring the length of the guard cell from tip to base; a minimum of 10 randomly selected stomata were measured. Total stomatal complex area was calculated by manually tracing around the outside of both the guard cells and subsidiary cells using the polygon tool in ImageJ. Further stomatal measurements, such as pore length and guard cell width, were also measured and used to calculate the maximum pore aperture and the anatomical maximum stomatal conductance using the Franks and Farquhar (2001) equation.
Plant material and growth conditions
For the plant growth of the four selected FNB lines and the JHN control line, rice seeds from the M5 generation were placed into sealed Petri dishes containing sterile water and kept at room temperature (25–30 °C) with ample light. One-week-old seedlings were transplanted into pots containing soil originating from the same field as the initial phenotyping. We carefully added the same amount of soil to each pot. Plants were cultivated in the greenhouse with temperatures averaging 30–40 °C. Eight plants of each line were grown, and phenotypic measurements were performed on flag leaves between November and December 2017. Organic fertilizer was applied twice during plant development, at the late seedling stage and the maximum tillering stage. This plant set was prepared for the short-term drought experiment and leaf gas exchange measurement.
Leaf gas exchange measurement at steady state and gsmax calculation
Stomata conductance and light response assays were performed using a LI-6400 portable photosynthesis system (Licor, Lincoln), conducting infrared gas analysis (IRGA) on the first fully expanded flag leaf from the primary tiller using an LI-6400XT portable photosynthesis system. For heat experiments, an LI-6800 (Licor, Lincoln) was used with a chamber flow rate of 400 µmol s−1 and leaf temperature of 32 °C. Reference CO2 was maintained at 480 ppm and light intensity at 2000 µmol m−2 s−1. The relative humidity inside the IRGA chamber was kept at 65–75% using a self-indicating desiccant. Steady-state measurements under these conditions were recorded for carbon assimilation (A) and stomatal conductance (gs) for the restricted-water experiment; A and gs were measured between 9 a.m. and 12 p.m. on 7th and 14thdays after water stress treatments. Intrinsic water-use efficiency (iWUE) was calculated from carbon assimilation divided by stomata conductance (A/gs). Abaxial anatomical gs max was calculated using the double end corrected version of the Franks and Farquhar (2001) equation, from Dow et al. (2014):
where (m2 s−1) is the diffusivity of water in air and (m3 mol−1) is the molar volume of air. Assuming equal stomatal densities on both sides of the leaf, this value was doubled to provide the total anatomical gsmax.
Rhythmic response to dark/light transition
The experiments were conducted in the greenhouse, where plants were acclimatized at saturating light (2000 µmol m−2 s−1), followed by 10 minutes of complete darkness (0 µmol m−2 s−1), and then saturating light was re-applied for the final 10 minutes. At each light intensity, 20 measurements for A and gs were recorded. Stomatal closure was instigated by 10 minutes of complete darkness (0 m-2 s-1 PAR) following steady-state conditions of saturating light (2000 m−2 s−1 PAR). To re-open stomata, saturating light was re-introduced. The rate of gs per minute of the stomatal model lines were analyzed in 5-minute segments during initial stomatal closure, late stomatal closure, initial stomatal opening, and late stomatal opening.
CO2 Assimilation rate (A) response to different ambient CO2 concentrations
A was measured for the following carbon dioxide values: 100, 200, 340, 480, 600, 800, 1000, 1200, and 1500 ppm. At each step, a pause of 10 min was allowed for A to stabilize before the measurements were automatically recorded.
Water stress experiments
Experiment I: Panicle-initiation stage (R1) 5D-W-5D
The first experiment was conducted during the reproductive stage, when flag leaves were fully expanded and the panicle was initiated. Initially, plants were kept sufficiently watered for 82 days after transplantation. Then, a subset of plants were restricted-water treated by withholding water for 5 days, followed by re-watering to saturate the soil, and then withholding water for another 5–10 days. The normal control groups were kept well-watered throughout the experiment. Once the plant had reached maturity, all the plants were harvested to determine the aboveground wet biomass, grain yield, and total weight of filled seeds for each plant.
Experiment II: R2 reproductive water stress (R1-2 5D-14W)
The water was drained before the R2 stage for 3 days or at 71 days after transplantation, withholding water for 5 days, and then adding 200 mL of water every 2 days until 14 days. All physiological parameters and plant phenotypes were measured 7 and 14 days after treatment. Leaf dryness percentage (%DL) scorings were divided into four levels as: highly tolerant (0–25%), tolerant (26–50%), moderate (51–75%), and sensitive (76–100%) to drought.
High-resolution phenotyping
HRPP was performed at nighttime for chlorophyll fluorescence, thermal imaging, and hyperspectral imaging.
- Estimation of quantum yield of photosystem II (ΦPSII): The quantum yield was measured using a PSI FluorCam FC-800MF pulse amplitude modulator installed on the system. A FluorCam was used to capture chlorophyll fluorescence images and to estimate the maximum quantum yield of PSII (Fv/Fm) of the control and treated leaves. Fluorescence chlorophyll imaging was set up without dark adaptation daily in the early morning (6 a.m.) during the five days of drought measurement. Raw data were analyzed and images were acquired using the Plant Data Analyzer program. Plant vegetative biomass was also measured using the wet tissue of the plants, and the tiller and panicle numbers of the plant were recorded as well as the yield harvested. PSI Open FluorCam FC 800-O (PSI, Brno, Czech Republic) was used to capture the chlorophyll fluorescence images and to estimate the maximum quantum yield of PSII (Fv/Fm) in the leaves of control and treated plants. For chlorophyll fluorescence, the minimum fluorescence (Crawford et al.) and maximum fluorescence (Fm) in a dark-adapted state were used to calculate the maximum PSII quantum yield (QY) using the equations:
QY = Fm − Fo/Fm
QY = Fv/Fm
- For all image acquisitions, observed leaves were maintained horizontally and the images were acquired and analyzed using the Fluorcam 7 program (PSI).
- Canopy temperature: One day before the drought experiments, canopy temperature was evaluated using specially designed FLIR thermal cameras on the Plants Screen System (PSI, Brno, Czech Republic) with a resolution of 640 × 480 pixels.
Water-Use Efficiency: Rice seeds were germinated on 200-well plastic plates. At 21 days, seedlings were transferred into 1.2 m (diameter) × 0.8 m (height) round concrete blocks (CBs), which held 750 kg of paddy soil mix per block. We grew 24 seedlings with 20 cm spacing on each CB until harvest (Figure 2.2.6.). Soil temperature was measured 12 cm below the soil surface. A split-plot design with the main-plot, sufficient-water, and restricted-water treatments were set with four replicates using a randomized complete block (RCB) design. Phenotypic data were collected weekly. Soil moisture and temperature were monitored daily at 3:00 p.m. at a 12 cm soil depth in each CB using a soil moisture meter (Field Scout TDR150). From germination to maximum tillering, irrigation for all treatments provided sufficient water until the initiation of the reproductive stage. Five liters of irrigation were applied daily to each CB until the initiation of the reproductive stage. The average total amount of water use during the vegetative stage under both conditions was 340 L/CB. Then, the irrigation was increased to 10 L/CB daily during the reproductive period when rice uses more water for grain filing in the sufficient-water treatment. No irrigation was applied to the restricted-water plants. However, 8 L of rainfall was added to each CB during the experiment. The average volumetric soil moisture content (VMC) of sufficient-water and restricted-water treatments were 35–45% and 10%, respectively. The average soil temperatures of sufficient-water and restricted-water treatments were 35–37 and 40–43 °C, respectively, throughout the growing season. When rice is approaching maturity, the number of panicles per cm2 and 1000 seed weight were collected for each line. Seed set percentage is defined as:
% seed set = (seed-filled florets/total florets) × 100
Biomass is the dried weight of the whole aboveground plant.
Water use efficiency (WUE) is the dried weight of biomass (kg) per liter of water used. The total amount of water used in the sufficient-water and restricted-water treatments was 908 and 348 L/CB, respectively.
Statistical analysis
One-way ANOVA was used to determine if there were interactions between treatments. Treatment means were compared using least significant difference (LSD) to determine whether they were significantly different at the 0.05 probability level.