Canopy Structure, Light Intensity, Temperature and Photosynthetic Performance of Winter Wheat under Different Irrigation Conditions

A high-quality canopy architecture is central to obtaining high crop yields. A field experiment was carried out at the Wuqiao Experimental Station from 2015 to 2019 under four irrigation schemes (W0, no irrigation after sowing; W1, 75 mm irrigation at jointing stage; W2, 75 mm irrigation at jointing and anthesis stages, respectively; W3, 75 mm irrigation at tillering, jointing and anthesis stages, respectively) to investigate the canopy structure, canopy apparent photosynthesis (CAP), canopy temperature (CT), yield and water use efficiency (WUE). The results showed that increasing irrigation times improved the leaf area index (LAI), non-leaf area index (NLAI) and light interception (LI) of the spike and total canopy but decreased the canopy temperature (CT) after anthesis. The CAP in the W3 treatment was consistently lower than that in the W1 treatment, suggesting lower effective utilization of light energy under the W3 treatment. Increasing irrigation times improved wheat yield, but the W2 treatment had no significant difference in yield compared to the W3 treatment. In addition, the W1 and W2 treatments had higher WUEs. The CT, organ temperature and LI were closely positively associated with each other, but they were all strongly negatively related to the yield. Overall, the W2 treatment was the best irrigation scheme for constructing a reasonable canopy architecture for winter wheat, obtaining more efficient water use and yield in the North China Plain (NCP). CT and organ temperature can be used as proxy parameters to estimate the canopy structure.


Introduction
Winter wheat (Triticum aestivum L.) is the most widely planted and important global grain crop [1]. Improving wheat production is imperative for ensuring an adequate supply of food for the expanding global population [2]. A favorable crop population architecture is the foundation for obtaining high crop yields and water productivity. Crop population architecture will affect canopy light distribution and light interception, which is determined by many factors, including genotype, water, density, nitrogen and their interactions [3][4][5]. However, increases in frequent and irregular water stress have become the main limiting factor for wheat productivity in the North China Plain (NCP). In this region, the average water consumption in winter wheat growing season is approximately 450 mm, which is two to three times the average rainfall [6]. Irrigation is required to guarantee the water demand for winter wheat production [7,8]. But, overexploitation of groundwater has caused the rapid decline of groundwater levels, inducing serious environmental problems such as land subsidence and threats to sustainable agricultural development [9,10]. Thus, developing a high-quality population structure with efficient light energy utilization and WUE under water-saving conditions is important for thoroughly resolving water scarcity.
Droughts alter plant morphological (e.g., declines in leaf area, length, and width) and physiological traits (e.g., decrease in leaf chlorophyll concentrations and maximum

Soil Water Content
Irrigation is one of the key factors affecting soil water. Compared with the W 0 treatment, the W 1 treatment significantly improved the soil water content (SWC) of the 0-180 cm soil layer at the anthesis stage in 2015/2016, 2016/2017 and 2018/2019. It also significantly improved the SWC of the 0-140 cm soil layer in 2017/2018, and it had no effect at the deeper layers ( Figure 1). Compared with the W 1 treatment, the W 3  improves population photosynthesis and ultimately improves crop yield and water use efficiency. Therefore, the objectives of this study were: (1) to compare the area and ratio of the photosynthetic organs in single plants; (2) to study the light interception for different organs; (3) to determine how the canopy structure parameters were correlated with yield; and (4) to test the use of CT and NDVI as a proxy to estimate canopy architecture.

Soil Water Content
Irrigation is one of the key factors affecting soil water. Compared with the W0 treatment, the W1 treatment significantly improved the soil water content (SWC) of the 0-180 cm soil layer at the anthesis stage in 2015/2016, 2016/2017 and 2018/2019. It also significantly improved the SWC of the 0-140 cm soil layer in 2017/2018, and it had no effect at the deeper layers ( Figure 1). Compared with the W1 treatment, the W3 treatment significantly increased the SWC in the 80-180 cm soil layer in 2016/2017 and improved it within a 0-100 cm soil depth in 2018/2019 at the anthesis stage, while the W3 and W1 treatments exhibited no significant differences in 2017/2018. At maturity, compared with the W0 treatment, the W1 treatment improved the SWC of the 80-140 cm soil layer. The W2 treatment significantly increased the SWC in the 0-120 cm soil layer in 2015/2016 and 2016/2017 and had no effect on it in 2017/2018 or 2018/2019 when compared with the W1 treatment. The W3 treatment had a higher SWC in the 0-200 cm than the W2 treatment in 2016/2017. Soil water content at a depth of 0-200 cm with different irrigation treatments at anthesis (a-d) and maturity (e-h). The red asterisks indicate significant differences between W0 and W1, the green asterisks indicate significant differences between W3 and W2, and the blue asterisks indicate significant differences between W2 and W1.

The Area and Ratio of the Photosynthetic Organs
The leaf blade had the maximum area and ratio at anthesis as the main photosynthetic organ, followed by the stem and spike ( Table 1). The leaf area and spike area increased with increasing irrigation times, and the area of the leaf increased much more quickly than the stem and spike areas. Compared with W0, the W1 and W3 treatments Soil water content at a depth of 0-200 cm with different irrigation treatments at anthesis (a-d) and maturity (e-h). The red asterisks indicate significant differences between W 0 and W 1 , the green asterisks indicate significant differences between W 3 and W 2 , and the blue asterisks indicate significant differences between W 2 and W 1 .

The Area and Ratio of the Photosynthetic Organs
The leaf blade had the maximum area and ratio at anthesis as the main photosynthetic organ, followed by the stem and spike ( Table 1). The leaf area and spike area increased with increasing irrigation times, and the area of the leaf increased much more quickly than the stem and spike areas. Compared with W 0 , the W 1 and W 3 treatments increased the leaf area by averages of 19.3% and 59.6%, respectively, and increased the spike area by averages of 7.4% and 19.8%, respectively. The impact of irrigation on the total green organ area showed a very similar trend to the effects of irrigation on leaf area. The leaf ratios were greater under the W 1 and W 3 treatments, while the ratios of non-leaf organs (stem and spike) were higher under the W 0 treatment. The trends in irrigation effects on the total green organ area were consistent with the leaf area.

Canopy Green Organs' Area Index
At anthesis, the NLAI approached or even exceeded the LAI, especially in the W 0 treatment ( Table 2). The GAI, LAI and NLAI increased with water input, while the NLAIs under the W 1 and W 3 treatments were similar, and both were significantly higher than those under the W 0 treatment.

Light Interception
Irrigation increases the total light interception ratio of the wheat canopy, which ranged from 86.5% to 97.9% ( Figure 2). The spikes intercepted the most light due to their topside advantage, and the light interception ratio increased with an increasing water supply. Irrigation increased the light interception ratio of the flag leaf, especially under the W 1 treatment in 2016/2017 and the W 3 treatment in 2017/2018. In 2016/2017, increasing the irrigation times had no effect on the light interception ratio of the penultimate leaf and the rest of the leaves, but there was a significant decrease under the W 3 treatment during the 2017/2018 season. topside advantage, and the light interception ratio increased with an increasing water supply. Irrigation increased the light interception ratio of the flag leaf, especially under the W1 treatment in 2016/2017 and the W3 treatment in 2017/2018. In 2016/2017, increasing the irrigation times had no effect on the light interception ratio of the penultimate leaf and the rest of the leaves, but there was a significant decrease under the W3 treatment during the 2017/2018 season.

Canopy Temperature and Canopy Temperature Depression
The CT and CTD values under different irrigation treatments after anthesis are presented in Table 3. The CT after anthesis decreased significantly with increasing irrigation times. In the two growth seasons, the CT was generally lower than the air temperature, resulting in positive CTD values. The change in trend of the CTD was opposite to that of the CT; the CTD after anthesis increased under the irrigation treatment. In 2016/2017, there were significant differences in the CTD between the W1 and W2 treatments, but in 2017/2018, there were no significant differences.   The CT and CTD values under different irrigation treatments after anthesis are presented in Table 3. The CT after anthesis decreased significantly with increasing irrigation times. In the two growth seasons, the CT was generally lower than the air temperature, resulting in positive CTD values. The change in trend of the CTD was opposite to that of the CT; the CTD after anthesis increased under the irrigation treatment. In 2016/2017, there were significant differences in the CTD between the W 1 and W 2 treatments, but in 2017/2018, there were no significant differences.

Temperature of Organs
The temperature of the spike was significantly higher than the temperature of the other organs at anthesis and DAA10 (Table 4). Irrigation tended to significantly decrease the temperature of all the wheat organs. Among all the wheat organs, the spike experienced the largest decrease due to irrigation, ranging from 0.92 to 2.59 • C. The variation trends in the CAP were similar across the different irrigation treatments ( Figure 3). The CAP values increased during the booting stage, reached a peak at the anthesis stage, and then decreased during the grain filling stage. The CAP under the W 3 treatment at the booting stage was significantly lower than that in the W 0 and W 1 treatments. At anthesis and the grain filling stage, the CAP under the W 1 treatment was significantly higher than in the other treatments. Compared to the CAP at anthesis, the CAPs at 14 DAA in the W 1 , W 3 and W 0 treatments were decreased by 4.85%, 15.41% and 17.49%, respectively.  The red asterisks indicate significant differences between W0 and W1, and the green asterisks indicate significant differences between W3 and W0.

Diurnal Variation of Canopy Apparent Photosynthesis and Respiration Rate
The diurnal variation of the CAP rate and respiration rates were measured with an average temperature of 16.7 ℃ and a maximum temperature 23.5 °C at 1 DAA ( Figure 4). The CAP increased to the maximum close to noon and decreased in the afternoon. The maximum value appeared no later than 14:00. The diurnal variations in CAP under the W0 and W1 treatments reached their maximum from 12:00 to 14:00, and the diurnal variations in CAP under the W3 treatment peaked between 10:00 and 12:00. The diurnal variations in canopy respiration rates followed a similar trend to the CAP. The maximal respiration rates were 0.89, 1.05 and 1.01 mg CO2 s −1 m −2 under the W0, W1 and W3 treatments, respectively. Compared to the W0 treatment, the irrigation significantly increased the CAP from 8:00 to 18:00 and significantly increased the respiration rates from 8:00 to 20:00. The red asterisks indicate significant differences between W 0 and W 1 , and the green asterisks indicate significant differences between W 3 and W 0 .

Diurnal Variation of Canopy Apparent Photosynthesis and Respiration Rate
The diurnal variation of the CAP rate and respiration rates were measured with an average temperature of 16.7°C and a maximum temperature 23.5 • C at 1 DAA ( Figure 4). The CAP increased to the maximum close to noon and decreased in the afternoon. The maximum value appeared no later than 14:00. The diurnal variations in CAP under the W 0 and W 1 treatments reached their maximum from 12:00 to 14:00, and the diurnal variations in CAP under the W 3 treatment peaked between 10:00 and 12:00. The diurnal variations in canopy respiration rates followed a similar trend to the CAP. The maximal respiration rates were 0.89, 1.05 and 1.01 mg CO 2 s −1 m −2 under the W 0 , W 1 and W 3 treatments, respectively. Compared to the W 0 treatment, the irrigation significantly increased the CAP from 8:00 to 18:00 and significantly increased the respiration rates from 8:00 to 20:00. growth stages in 2017/2018. Vertical bars represent the mean SE of three replications per treatment. The red asterisks indicate significant differences between W0 and W1, and the green asterisks indicate significant differences between W3 and W0.

Diurnal Variation of Canopy Apparent Photosynthesis and Respiration Rate
The diurnal variation of the CAP rate and respiration rates were measured with an average temperature of 16.7 ℃ and a maximum temperature 23.5 °C at 1 DAA (Figure 4). The CAP increased to the maximum close to noon and decreased in the afternoon. The maximum value appeared no later than 14:00. The diurnal variations in CAP under the W0 and W1 treatments reached their maximum from 12:00 to 14:00, and the diurnal variations in CAP under the W3 treatment peaked between 10:00 and 12:00. The diurnal variations in canopy respiration rates followed a similar trend to the CAP. The maximal respiration rates were 0.89, 1.05 and 1.01 mg CO2 s −1 m −2 under the W0, W1 and W3 treatments, respectively. Compared to the W0 treatment, the irrigation significantly increased the CAP from 8:00 to 18:00 and significantly increased the respiration rates from 8:00 to 20:00.  The red asterisks indicate significant differences between W 0 and W 1 , and the green asterisks indicate significant differences between W 3 and W 1 . Figure 5 displays the time courses of NDVI for different irrigation treatments from the booting stage to maturity. The NDVI values demonstrated that W 2 > W 1 > W 3 > W 0 , and in general, the NDVI values decreased gradually from booting to maturity. The NDVI values were higher for the W 2 treatment than the other treatments during the middle to later grain filling stage. The NDVI decreased at a slower rate from anthesis to mid-grain filling under the W 0 treatment, but with a more rapid decline and lower NDVI values in the late season.  Figure 5 displays the time courses of NDVI for different irrigation treatments from the booting stage to maturity. The NDVI values demonstrated that W2 > W1 > W3 > W0, and in general, the NDVI values decreased gradually from booting to maturity. The NDVI values were higher for the W2 treatment than the other treatments during the middle to later grain filling stage. The NDVI decreased at a slower rate from anthesis to mid-grain filling under the W0 treatment, but with a more rapid decline and lower NDVI values in the late season.

Yield, ET and WUE
Generally, the yield and ET increased with increasing irrigation amounts ( Table 5). The wheat under W3 achieved the highest yield (averaged 8077 kg ha −1 ) and ET (averaged 474.0 mm), but there was no significant difference in grain yield between the W3 and W2 treatments except for the yield in 2018/2019. The WUE of winter wheat showed an increas-

Yield, ET and WUE
Generally, the yield and ET increased with increasing irrigation amounts ( Table 5). The wheat under W 3 achieved the highest yield (averaged 8077 kg ha −1 ) and ET (averaged 474.0 mm), but there was no significant difference in grain yield between the W 3 and W 2 treatments except for the yield in 2018/2019. The WUE of winter wheat showed an increasing and then decreasing trend with increasing irrigation times, and the W 1 and W 2 treatments had the highest WUE.

Correlation Analysis
A principal component analysis (PCA) was conducted to explain the relationship between proxy indicators (i.e., CT and NDVI), canopy photosynthetic parameters and canopy architecture and yield ( Figure 6). The yield was very significantly positively correlated with the GAI, LAI and LI of spike at anthesis. The organ temperatures and light intercept of the lower layers were strongly negatively related to the yield. The organ green area and ET were closely positively related, but negatively related with the WUE. Nevertheless, there was no clear correlation between NDVI and the other indicators.

Correlation Analysis
A principal component analysis (PCA) was conducted to explain tween proxy indicators (i.e., CT and NDVI), canopy photosynthetic p opy architecture and yield ( Figure 6). The yield was very significan lated with the GAI, LAI and LI of spike at anthesis. The organ tem intercept of the lower layers were strongly negatively related to the yie area and ET were closely positively related, but negatively related wi theless, there was no clear correlation between NDVI and the other in

Discussion
Irrigation is the most widely used way to combat soil water defi winter wheat yields in semi-arid and arid regions [6]. However, poor

Discussion
Irrigation is the most widely used way to combat soil water deficiency and improve winter wheat yields in semi-arid and arid regions [6]. However, poor irrigation management results in unreasonable canopy structures, leading to disruptions in WUE [35]. Optimal irrigation scheduling can promote root growth into deeper layers of the soil and water absorption from a wide soil rhizosphere, relatively optimizing wheat canopies and regulating water consumption [36]. In this study, the canopy architecture and microenvironment were affected by the irrigation treatment. As the irrigation amount increased, the leaf area and spike area increased, while the proportion of non-leaf area to total green area per stem decreased (Table 2). In this study, we did not measure the area of the awn directly; instead the spike surface area was calculated as a cylinder using a compromise method. The spike surface area measured using this method was close to the observed spike surface area (with awns). This method may be applicable only to wheat varieties with the same spike type as this study. The non-leaf area comprised approximately half of the total green area per stem. However, the results of our study were inconsistent with the findings of Zhang [18], who reported that the ratio of the non-leaf area to the total green area of wheat was 63%. This is due to the different methods of determining the area of non-leaf organs. The previous method measured the total surface area of non-leaf organs rather than the effective surface area for light interception, and this may overestimate the non-leaf organs' areas. In spite of this variability, the results obtained using different methods support the importance of the non-leaf organs as major photosynthesis organs.
The green area index is an important index of canopy architecture. In this study, increasing irrigation times improved the LAI and NLAI at anthesis in the successive 4 years ( Table 2). Thus, the amount of light intercepted by the canopy was increased, especially by the spike in the upper strata ( Figure 2). The improvement in the photosynthetic rate that was observed with increasing irrigation was mainly due to the change in the canopy architecture, the increase in the GAI and canopy light interception, and the decrease in light leakage loss [37,38]. Previous studies have indicated that drought and weak incident light accelerates leaf senescence [12,39]. This phenomenon was found in the current study, which was reflected in the NDVI, as shown by the lower values under the W 0 and W 3 treatments during post-anthesis ( Figure 5). The wheat plants' NDVI declined rapidly due to increased drought conditions under the W 0 treatment, while a deterioration in the light environment in the lower layers led to faster NDVI declines under the W 3 treatment. Although NDVI is a quick and easy way to measure the stay-green and plant status, it cannot reflect vertical changes in the canopy structure [32].
Photosynthesis is the primary determinant of crop yield, and CAP is closely related to the canopy architecture and positively correlated with light interception [40]. The wheat canopy is complex and heterogeneous, and it was difficult to accurately measure the amount of light intercepted by each organ. In this study, we divided the wheat canopy into different layers from top to bottom, and depending on the dominant organ in the different layers, we estimated the light interception of the different organs. The CAP accurately describes the photosynthetic capacity per unit area of a crop and integrates the characteristics of the canopy structure [41]. In this study, the CAP under the W 3 treatment was consistently lower than under the W 1 treatments during the post-anthesis period (Figure 3), while there was no significant difference in the SWC at 0-200 cm between the W 1 and W 3 treatments at the anthesis stage (Figure 1c,g). This is because the light microclimate deteriorated quickly from the top to the bottom of the canopy and was aggravated with an increasing water supply. The light redundancy in the upper layers resulted in a poorer light environment in the lower layers, in which respiration exceeded photosynthesis (below the light compensation point), causing light starvation in the lower layers. The differences in CAP between the W 3 and W 1 treatments were due to irrigation-induced changes in the canopy structure and not the soil moisture. The diurnal photosynthesis amount was highest in the W 1 treatment, followed by W 3 treatment, and were lowest in the W 0 irrigation treatment. The diurnal respiration amounts were 2.9, 3.6 and 4.2 mg CO 2 s −1 m −2 under the W 0 , W 1 and W 3 treatments, respectively. The W 1 treatment reduced the leaf area relative to the W 3 treatment, improved the canopy architecture, reduced canopy respiration and maintained a higher daily CAP, and its yield and WUE were also higher.
Canopy structure can affect canopy temperature, and canopy temperature differs between irrigation treatments [42]. In this study, the CT after anthesis decreased significantly with increasing irrigation times (Table 3). Previous studies have reported that temperature differences of up to 5 • C were observed among the spikes between irrigated and rainfed crops on a hot day [43]. In our study, irrigation tended to significantly decrease the temperature of all wheat organs, and the spike experienced the largest decrease due to irrigation, ranging from 0.92 to 2.59 • C ( Table 4). The position of the spikes ensures a minimum amount of shading, and the transpiration of the spike is lower than that of the flag leaves. Therefore, the spike temperature was higher than the leaf temperature and slightly lower than the CT. In our study, the spikes had the highest temperature among the plant organs at anthesis and the early grain filling stage. This study found significant correlations between CT and yield, and the organ temperature of central and base of the stem were closely positively related with the light intercept of lower layers. The CT and organ temperature can be important parameters reflecting the ecological and physiological state of the plant [44].
The correlations among the canopy architecture, microclimate and yield disclosed that the GAI, LAI and LI of the spike were negatively correlated with the yield. The spike played an important role in grain yield, and its canopy characteristics can be proxy measures of grain yield [17]. Increasing the number of irrigations increased the yield, but there was no significant difference in yield between the W 2 and W 3 treatments ( Table 5). The effect of irrigation on winter wheat yield and WUE have been well studied. Appropriate increases in irrigation can improve the yield, but excessive irrigation will reduce the wheat yield and WUE [45]. In our study, the WUE first increased and then decreased as the amount of irrigation time increased from 2016 to 2018, and the WUE was highest under the W 2 treatment (1.80 and 1.95 kg m −3 , respectively). However, the WUE was highest under the W 1 treatment in 2018/2019 due to a deficiency in precipitation (56.2 mm) during the growing period of the winter wheat. However, this study only explored the effects of irrigation on canopy structure and light intensity and did not accurately measure the photosynthetic areas of non-leaf organs and light interception of different organs. Further research is needed to more accurately characterize canopy structure and examine the interactions between soil nutrients (e.g., nitrogen) and varieties.

Experimental Site and Experimental Design
The field experiments were carried out in 2015/2016, 2016/2017, 2017/2018 and 2018/2019 at the experimental station at the China Agricultural University, China, located in Wuqiao, Hebei Province. The climate is summer monsoon climate with an average annual precipitation of 562 mm, but with only 30% precipitation concentrated on the winter wheat growth season (October to July). The solar radiation, air temperature and rainfall from 2015 to 2019 are shown in Figure 7. The precipitation received during the winter wheat seasons from 2015 to 2019 was 153.2, 87.8, 191.5 and 56.2 mm, respectively. The soil was clay-loam with 1.25% organic matter, 1.02 g kg −1 total nitrogen (N), 31.6 mg kg −1 available phosphorus (P) and 48.1 mg kg −1 available potassium (K). est under the W1 treatment in 2018/2019 due to a deficiency in precipitation (56.2 mm) during the growing period of the winter wheat. However, this study only explored the effects of irrigation on canopy structure and light intensity and did not accurately measure the photosynthetic areas of non-leaf organs and light interception of different organs. Further research is needed to more accurately characterize canopy structure and examine the interactions between soil nutrients (e.g., nitrogen) and varieties.

Experimental Site and Experimental Design
The field experiments were carried out in 2015/2016, 2016/2017, 2017/2018 and 2018/2019 at the experimental station at the China Agricultural University, China, located in Wuqiao, Hebei Province. The climate is summer monsoon climate with an average annual precipitation of 562 mm, but with only 30% precipitation concentrated on the winter wheat growth season (October to July). The solar radiation, air temperature and rainfall from 2015 to 2019 are shown in Figure 7. The precipitation received during the winter wheat seasons from 2015 to 2019 was 153. 2, 87.8, 191.5 and 56.2 mm, respectively. The soil was clay-loam with 1.25% organic matter, 1.02 g kg −1 total nitrogen (N), 31.6 mg kg −1 available phosphorus (P) and 48.1 mg kg −1 available potassium (K).   (Table 6). In 2015/2016, three irrigation treatments were designed, i.e., W 0 ., no irrigation during the growing stage; W 1 , 75 mm of total irrigation with 75 mm irrigated at the jointing stage; and W 2 , 150 mm of total irrigation with 75 mm irrigated at the jointing and anthesis stages, respectively. In 2016/2017, 2017/2018 and 2018/2019 an additional irrigation treatment was established: W 3 , 255 mm of total irrigation with 75 mm irrigated at the tillering, jointing and anthesis stages, respectively. W 3 is the conventional irrigation for winter wheat and is used by local farmers, W 2 is currently promoted by the local government and allows for a high yield and water use efficiency (WUE) and W 1 and W 0 are irrigations with a more reduced frequency to obtain basic data to further conserve irrigation water. All experiments received 225 kg N ha −1 (as urea), 300 kg P ha −1 (as superphosphate) and 225 kg K ha −1 (as potassium sulfate) before sowing. Pests and diseases were properly controlled. At anthesis, two adjacent 50 cm row samples were collected to count the number of culms and spikes for each plot. A total of 20 stems with spikes were randomly selected from the sample to measure and record the green organ areas. For the leaf lamina, an automatic planimeter (LI-3000 area meter, LI-COR, Lincoln, NE) was used to measure the leaf area. Non-flat surfaces organs, spike length, peduncle length and stem length were measured with a ruler; spike thickness, spike width, the diameter of the peduncle and each node were measured at the middle of the organ using Vernier calipers. We measured the planar area rather than the total green area. The total surface area of the spike was calculating as a cylinder: where L is length from the rachis to the top of the spike, excluding the length of the awns, and D is the average spike thickness and width. We did not use this method to measure the area of the awn directly, and since the area of the awns is proportional to the spike length and width, the spike surface area (excluding the awns) is overestimated, compensating for the underestimation due to neglecting the awns. The spike area (effective surface area) was calculated as the spike surface area multiplied by 0.5, Spike area = Spike surface area × 0.5, which is better correlated with light interception.
The total surface area of the stem with the leaf sheath was calculated as: where L1 is the length of the peduncle, D1 is the diameter of thevpeduncle; L2 is the length of the stem and sheath excluding the length of the peduncle, and D2 is the average diameter of the culm and sheath. The stem area was calculated the same as the spike: Stem area = Stem surface area × 0.5. The LAI was calculated using only one side of the leaf tissue area per ground unit. The non-leaf area index (NLAI) was calculated as: The green area index (GAI) was calculated as:

Light Interception
The photosynthetically active radiation (PAR) intercepted by the structures in different layers was calculated under different irrigations at anthesis in 2016/2017 and 2017/2018. For this, measurements were taken between 11:00 to 13:00 using a hand-held ceptometer (Model CI-150, CID, Inc., Vancouver, WA, USA). The PAR was measured at defined levels within the crop canopy (i.e., just above the plant canopy, below the spike, below the flag leaf, below the penultimate leaf, at the soil level). The height of the selected spike, flag leaf, and penultimate leaf was based on the height of most plants in the plot. Light interception (LI) was calculated using the following equation: where I a is the PAR above the layer and I b is the PAR below the layer.

The Canopy Temperature and Temperature of Wheat Organs
The CT and temperature of wheat organs (e.g., spike, flag leaf) were measured at anthesis, 10 days after anthesis (DAA) and 20 DAA using a handheld infrared thermometer (Sixth Sense LT 300, Instrument, South Burlington, VI, USA). For each plot,10 plants were selected to measure the temperature of the spike, flag leaf, center of the stem and base of the stem. The CT measurements were made at ground level along each of the plots from 0.5 m above the canopy, angled to avoid bare soil and directed specifically at the part of the plot most exposed to the sun. A sensor was installed at a height of 1.5 m in a nearby uncropped plot to collect air temperature data.
Canopy temperature depression (CTD) is expressed as the difference between the air temperature (T air ) and canopy temperature:

The Normalized Difference Vegetation Index
For each plot, the normalized difference vegetation index (NDVI) was measured using a handheld crop sensor (GreenSeeker RT100, NTech Industries, Ukiah, CA, USA) at 70-90 cm above the plant canopy every day after the booting stage. The NDVI was calculated using the following equation: where NIR and R are the measured reflectance near the infrared and red spectral bands, respectively.

Canopy Apparent Photosynthesis
The rates of CAP were measured using a custom-made chamber with an infrared gas analyzer (LI-6400, LI-COR, Lincoln, NE, USA). The chamber consisted of a 1 × 1 × 1 m aluminum frame covered with mylar film and two battery-powered 50 W fans. The chamber was tall enough for the wheat plants to be contained without affecting the canopy structure. Measurements with three replicates were taken between 9:00 and 11:30 at the booting stages, anthesis stages, 14 DAA, 24 DAA and 29 DAA. The diurnal changes in CAP were measured every two hours from 8:00 to 20:00 at 1 DAA. The decrease in carbon dioxide concentration was monitored within 5 min after the plants were enclosed in the chamber. After finishing the CAP measurement, the respiration rates (Rd) were measured by placing a black opaque cloth over the chamber. Soil respiration (R soil ) was measured in a nearby plot under the same treatment with all the plants removed at ground level. The CAP, Rd and R soil (mg CO 2 s −1 m −2 ) were calculated using the following equation: CAP(or Rd) = CAP or Rd + |R soil | where c 0 and c 1 are the initial and final concentrations of carbon dioxide (mg L −1 ); t 0 and t 1 are the start and end times (s); V is the volume of chamber (L); S is the ground area (m 2 ); P is the air pressure (Pa) and T is the air temperature ( • C).
4.2.6. Grain Yield, Soil Water Content, Evapotranspiration (ET), and WUE At physiological maturity, spikes from an approximately 2 m 2 area of each plot were harvested by hand, and samples were dried in an oven at 60-75 • C until a constant weight was obtained to investigate the yield.
Soil samples were taken from 0 to 200 cm in layer segments of 20 cm using a hand-held soil corer at the sowing stage, anthesis and maturity. The soil samples were dried in an oven at 105 • C for 48 h, then weighed, and the weight was recorded. The soil gravimetric water content (%; GWC) was measured as the water content (g) divided by the oven-dried weight (g). The soil volumetric water content (%; VWC) was calculated as the GWC multiplied by the soil bulk density. The soil water content (mm) was calculated as the VWC multiplied by the soil thickness (mm). The capillary rise and water drainage were negligible, and no surface runoff occurred from the experimental plots. The ET under varying irrigation regimes was calculated using the water balance equation: where ET is evapotranspiration (mm), SWD is the variation in moisture from sowing and maturity (mm), P is precipitation (mm) and I is irrigation amount (mm).
The WUE (kg m −3 ) values were defined as the ratio of the grain yield to the crop evapotranspiration (ET).

Statistical Analysis
Analyses of variances (ANOVAs) were performed using SPSS version 20.0 (IBM, Armonk, NY, USA) [46]. The least significant difference (LSD) test was used to evaluate differences among treatments, and the significance level was set to 0.05 or 0.01 confidence levels. The data were plotted using Origin Pro 2019 (Origin Lab Corporation, Northampton, MA, USA) [47], R version 4.1.1 [48] and Adobe Illustrator CC version 21.0.0 [49].

Conclusions
Increasing irrigation times improved the wheat yield, but the W 2 treatment had a comparable yield with the W 3 treatment and a higher WUE. The CAPs under the W 3 treatment were consistently lower than in the W 1 /W 2 treatments during the post-anthesis period, suggesting a lower effective utilization of light energy under the W 3 treatment. In addition, there were no significant differences in SWC at 0-200 cm between the W 1 /W 2 and W 3 treatments at the anthesis stage. This indicates that the lower effective utilization of light energy under the W 3 treatment is caused by a change in canopy structure, not soil moisture. Increasing irrigation times improved the LAI and NLAI at anthesis and increased the LI of the spike and canopy, but decreased the LI of the lower strata leaves and accelerated the rate of leaf senescence, which results in a decreased CAP under the W 3 treatment. As there is no clear correlation between the NDVI, other canopy photosynthetic parameters and canopy architecture indicators, it cannot reflect vertical changes in canopy structure. Significant correlations were found in this study between the CT and yield, while the organ temperature of the center and base of the stem were closely positively related with the LI of the lower layers. Overall, a reasonable canopy architecture played an important role in improving the photosynthetic potential later during grain filling. The W 2 treatment was the best irrigation scheme for saving water and obtaining a comparable yield and a higher WUE of winter wheat in the NCP. The CT and organ temperature can be important parameters used to reflect the canopy architecture.