The application of mixed nitrogen increases photosynthetic and antioxidant capacity in rice (Oryza sativa) under heat stress at flowering stage

In recent years, heat stress has become a serious threat to rice production in China. It is well documented that nitrogen forms can regulate the heat tolerance of plants. However, its effect on the rice plants has not been described before. In this study, a pot experiment was performed with two restorer lines of indica rice, heat-tolerant 996 (HT 996) and heat-sensitive 343 (HS 343), to investigate whether nitrogen forms could mitigate heat stress at flowering stage and their physiological mechanism. The experiment received five treatments: (1) NN0: natural temperature level (32 ℃) with urea-N; (2) HN0: high temperature level (38 ℃) with urea-N; (3) HN1: high temperature level with NH4+-N; (4) HN2: high temperature level with mixed N [NH4+: NO3− at 50: 50 (w:w)]; and (5) HN3: high temperature level with NO3−-N, respectively. The results showed that heat stress reduced seed setting, photosynthetic characteristics, leaf nitrogen content and antioxidant enzyme activities of rice. However, the application of mixed N significantly increased the 1000-grain weight and seed setting rate under heat stress, thus increasing yield. The increased yield was mainly caused by higher leaf N content, photosynthetic rate, transpiration rate and stomatal conductance, which led to a reduction in panicle and leaf temperatures. Meanwhile, the application of mixed N also improved the activities of antioxidant enzymes, and reduced the amount of reactive oxygen species, thereby mitigating the damage caused by exposure to high temperature. In conclusion, the application of mixed N can effectively alleviate the adverse consequences of high temperature, and could be further applied to provide strategies for the prevention and control of high temperature-prone areas during the rice flowering period.


Introduction
Rice (Oryza sativa L.) is one of the most important food crops in the world, and more than half of the world's population uses rice as a staple (Khush 2005). According to the 5th report of the IPCC (IPCC 2013), by the end of the twenty-first century, the global average surface temperature will increase by 1.1 to 6.4 ℃. Under the background of global warming, extreme high temperature weather occurs frequently, leading to the intensification of rice heat damage (Wang et al. 2017). Peng et al. (2004) reported that rice yield decreased by 10% when the average nighttime minimum temperature increased by 1 ℃ during rice growth period. Rice is the most sensitive crop to high temperature stress during the heading and flowering periods (Jagadish et al. 2010). However, the rice flowering stage in the main ricegrowing areas of China is in the high temperature period in summer, and in the past two decades, rice yield reduction or even failure due to high temperature stress occurred frequently (Tian et al. 2009;Luo et al. 2015), which has seriously threatened rice production and food security of China. Therefore, exploring effective measures to mitigate heat damage at the flowering stage of rice is an important theoretical problem to be solved urgently.
Nitrogen (N) is the most important nutrient element that affects plant growth and development (Ercolano et al. 2015). The two main forms of inorganic nitrogen absorbed by plants are nitrate nitrogen (NO 3 − -N) and ammonium nitrogen (NH 4 + -N), among them, NO 3 − -N consumes more energy than NH 4 + -N (Guo et al. 2007a). Exposing plants to high levels of NO 3 − -N will easily lead to N loss and NO 3 − accumulation in plants, while excessive application of NH 4 + -N will cause various metabolic disorders in plants and lead to NH 4 + poisoning (Zebarth et al. 2012). Previous studies have suggested that rice prefers NH 4 + (Arth et al. 1998;Kronzucker et al. 1998). However, Raman et al. (1995) found that rice roots can absorb NO 3 − and leaves can reduce NO 3 − . During the reproductive period, NO 3 − -N application improves metabolism of hybrid rice (Yang and sun 1991), and some rice varieties grow better under stressful conditions with NO 3 − -N as their main N source instead of NH 4 + -N (Ta and Ohira 1981). Tan et al. (2002) reported that the application of mixed N can increase the biomass and yield of rice compared with the application of NH 4 + -N or NO 3 − -N alone. These results indicate that rice not only prefers NH 4 + , but also absorbs a considerable amount of NO 3 − . Nitrogen plays an important role in improving the tolerance of plants to high temperature stress, and the appropriate nitrogen management can partially mitigate the damage caused by high temperature to crops (Waraich et al. 2012). For example, applying nitrogen fertilizers at panicle differentiation or flowering stage can improve rice yield under high temperature conditions (Dai et al. 2009;Yang et al. 2014). Moreover, increasing the total amount of nitrogen fertilizer can reduce the yield loss caused by high temperature at flowering stage of super hybrid rice (Liu et al. 2019). Furthermore, nitrogen forms play a vital role in regulating plant resistance to high temperature. Li et al. (2007) showed that the application of mixed N increased the activities of antioxidant enzymes, reduced the effect of oxidative stress, and thus improved heat tolerance in Festuca arundinacea experiencing heat stress. However, whether nitrogen forms could mitigate heat stress in rice remains unclear.
This study investigated the mitigate effect of nitrogen forms on heat stress at flowering stage of rice. The pot experiment was conducted using two different heat-tolerant rice varieties HT 996 and HS 343. Rice plants, which applying with different nitrogen forms, were treated by high temperature during flowering period, and the subsequent changes, like seed setting, panicle and leaf temperatures, leaf nitrogen content, photosynthetic characteristics and antioxidant enzyme activities were examined, and then the relationship between these physiological changes and seed setting characteristics was also analyzed to reveal the underlying causes of rice heat tolerance.

Plant material
The two restorer lines of indica rice, heat-tolerant line 996 (HT 996, which is recognized as a high temperature resistant line) and heat-sensitive line 343 (HS 343, which has been shown to be sensitive to high temperature for many years) were used as the plant materials. The growth period of these two rice lines is similar, and the whole growth period is about 130 days.

Experimental design
This experiment was conducted in 2016 at the Lujiang Experimental Base of the Anhui Agricultural University, Anhui Province, China (31°48′55″N, 117°23′55″E) during the rice growing season. Rice plants were planted in pots, and the pot was 35 cm in height and 20 cm in diameter, filled with 13 kg sieved paddy soil. The soil type was sandy loam soil, containing 28.76 mg/kg organic matter, 62.45 mg/kg available N, 14.66 mg/kg available phosphorous (P), 222 mg/kg available potassium (K) and with soil PH of 6.78. Seeds were sown on May 26, and seedlings were transplanted to the pots on June 15. Each pot contained three hills, and with only one seedling per hill. All pots were kept in the field under routine field management, the weak tillers were cut off during the vegetative growth period, and 5-6 single stalks with identical growth were retained in each pot.
In this study, nitrogen form and temperature treatment were designed as follows.

Treatment of nitrogen form
In this experiment, all treatments were equivalent to the same amount of nitrogen, which was 2.0 g pure nitrogen per pot. Under the premise of the same total pure N content (2.0 g N/pot), ammonium sulfate ((NH 4 ) 2 SO 4 ), ammonium nitrate (NH 4 NO 3 ) and sodium nitrate (NaNO 3 ) were applied as fertilizer sources to regulate the nitrogen form in the soil. Four treatments with different nitrogen forms were conducted as follows: (1) N0: control, urea was used as the N source (4.29 g/pot CH 4 N 2 O); (2) N1: sole NH 4 + was used as the N source (9.43 g/pot (NH 4 ) 2 SO 4 ); (3) N2: the mixed N source, NH 4 + : NO 3 − at 50:50 (m:m) (5.72 g/ pot NH 4 NO 3 ); (4) N3: sole NO 3 − was used as the N source (12.15 g/pot NaNO 3 ). In order to inhibit nitrification, the nitrification inhibitor dicyandiamide (C 2 H 4 N 4 ) was added at 5% of the nitrogen fertilizer to all treatments. The application ratio of base fertilizer:panicle fertilizer was 6:4, of which phosphorus fertilizer (1.0 g/pot P 2 O 5 ) and potassium fertilizer (2.4 g/pot K 2 O) were used as the base fertilizer. Base fertilizer was applied before transplanting, while panicle fertilizer was used when the leaf-age remainder was 3.5.

Treatment of temperature
Pots were moved to the artificial climate chambers when rice reached the end of booting stage (At around 12:00 on August 15, 2016, it was found that nearly 50% of the panicles of HT 996 and HS 343 had been pulled out from the flag leaf sheaths, and a few spikelets at the top of the panicles began to open). 40 pots were set up for each treatment, and a total of 200 pots were set up for the whole experiment. Rice plants were exposed to high temperature (H, from 8:00 to 18:30 (Beijing time) at 38 ℃, and from 18:30 to the next morning 8:00 at 30 ℃), and natural temperature (N, from 8:00 to 18:30 at 32 ℃, and from 18:30 to the next morning 8:00 at 25 ℃), respectively. Other controlled environmental conditions (light intensity of 14,000 LX in the day and 0 LX in the night, and relative humidity of 75% in the day and 80% in the night) in each climate chamber were consistent. After 7 consecutive days of treatment, rice plants were moved to the field for natural growth, and conventional field management was carried out until maturity.
The five treatments of combining temperature and nitrogen form were conducted as follows: (1) NN0: natural temperature level (32 ℃, as the control) with urea-N (as the control); (2) HN0: high temperature level (38 ℃) with urea-N; (3) HN1: high temperature level with NH 4 + -N; (4) HN2: high temperature level with the mixed N (NH 4 + : NO 3 − at 50:50 (m:m)); and (5) HN3: high temperature level with NO 3 − -N, respectively. Detailed data of these five treatments were shown in Table 1. Each treatment had three replicates, and 20 pots were randomly placed for each replicate.

Determination of panicle and leaf temperatures
On the 7th day of high temperature treatment, the temperature of rice flag leaf and panicle was measured by a Raytek infrared thermometer at 10:00-11:00. Leaf temperature was measured on both sides of the leaf, avoiding the veins, and panicle temperature was also measured on both sides of the middle part of the panicle. Using these temperatures, the leaf temperature depression (LTD) and panicle temperature depression (PTD) were determined as follows:

Determination of photosynthetic characteristics in flag leaves
The net photosynthetic rate (Pn), intercellular CO 2 concentration (Ci), stomatal conductance (Gs) and transpiration rate (Tr) of each flag leaf was measured by LI-6400 portable photosynthetic apparatus (LI-COR Company, USA) from LTD = Air temperature − Leaf temperature PTD = Air temperature − Panicle temperature 9:00 to 11:30 on the 7th day of temperature treatment. The photosynthetic apparatus was equipped with red and blue light source, which intensity was fixed at 1200 umol·m −2 ·s −1 .

Determination of chlorophyll fluorescence parameters in flag leaves
The modulated chlorophyll fluorescence imaging system MINI-IMAGING-PAM (Walz Company, Germany) was used to determine chlorophyll fluorescence parameters at 19:00 on the 7th day of temperature treatment. The rice plants were first treated with dark adaptation for 20 min, and the initial fluorescence (Fo) was measured with a detection light intensity of 0.5 μmol·m −2 ·s −1 and a pulse frequency of 1 Hz. The maximum photosynthetic efficiency (Fv/Fm) of PS II was detected by exciting the sample using saturated pulsed light with a light intensity of 2500 μmol·m −2 ·s −1 and a pulsed light time of 0.8 s. The fluorescence parameters were measured every 20 s with actinic light (intensity of 156 μmol·m −2 ·s −1 ). After the actual initial light capture efficiency ΦPSII was stabilized, the light was turned off and the far infrared light was used immediately to measure the non-photochemical quenching (NPQ), which was calculated automatically by the instrument.

Determination of antioxidant enzyme activity and MDA content in flag leaves
At 18:30 on the 7th day of high temperature treatment, rice flag leaves were frozen in liquid nitrogen for 1 min before storing at −80℃. These leaves were used to determine nitrogen concentration, the activities of superoxide dismutase (SOD), peroxidase (POD) and catalase (CAT), as well as the content of malondialdehyde (MDA). The assay methods of above parameters were modified from Zhang et al. (2016).

Determination of nitrogen concentration in flag leaves
The nitrogen concentration in rice flag leaves was determined by the Kjeldahl nitrogen method.

Determination of panicle characters
Approximately 10 pots of rice panicles from each treatment were harvested at maturity period. Samples were naturally dried, and then measured for the yield per pot, number of grains per panicle, grain weight, seed setting rate and panicle enclosure rate. Seed setting rate and panicle enclosure rate were calculated as follows: Seed setting rate = plump grain number/total grain number

Statistical analysis
For all statistical analyses, at least three biological replicates were used for each treatment and control. Statistical analyses of the data were accomplished by the standard analysis of variance (ANOVA), and the mean values were tested by the least significant difference (LSD) at the 5% level using SPSS16.0 (SPSS, Inc., Chicago, IL, USA).

Rice yield and its yield components
Rice yield was dramatically affected by nitrogen forms. Heat stress led to a significant reduction in the yield and yield components of HT 996 (Table 2). Compared with the application of urea-N, the application of mixed N and NH 4 + -N remarkably increased the yield, seed setting rate and 1000grain weight of HT 996 plants when exposing to heat stress, and the yield increased by 50.00 and 26.32%, seed setting rate increased by 30.85% and 20.57%, as well as 1000grain weight increased by 7.55 and 4.72%, respectively. In addition, under heat stress, the application of mixed N and NO 3 − -N significantly increased spikelets per panicle of HT 996, which were 6.92% and 5.38% higher than that of urea-N, respectively. The panicle characteristics of HS 343 were similar to those of HT 996 under the different temperature and nitrogen treatments, however, the yield and yield components of HS 343 had larger range of changes than those of HT 996.

Panicle and leaf temperature
Panicle and leaf temperatures of rice were significantly affected by nitrogen forms. Under heat stress, the panicle temperature was higher than the leaf temperature for both varieties (Fig. 1). Compared with the application of urea-N, the application of mixed N and NH 4 + -N decreased the panicle and leaf temperatures of HT 996 and HS 343 when exposing to heat stress, and the leaf temperature decreased by 3.56% and 2.71%, panicle temperature reduced by 2.92% and 2.55%, respectively. Despite showing similar trends, the difference was only significant in mixed N treatment. Interestingly, the panicle and leaf temperatures were higher in the NO 3 − -N treatment when compared to that of urea-N. These findings suggested that under high temperature stress, panicle and leaf temperatures of rice were the lowest when mixed N was applied and were the highest when NO 3 − -N was applied. The comparison among varieties showed that the panicle and leaf temperatures of HS 343 were lower than those of HT 996 under normal temperature, but they were higher than those of HT 996 under heat stress.
Leaf temperature depression (LTD) was higher than panicle temperature depression (PTD) for both varieties under different treatments (Table 3), indicating that leaf temperature was lower than panicle temperature, consistent with results from Fig. 1. LTD and PTD of HT 996 and HS 343 under heat stress was higher in plants that were treated with mixed N compared to urea-N, and LTD increased by 48.46% and 82.21%, PTD increased by 82.31% and 320.0%, respectively. However, when NO 3 − -N was applied, the LTD and PTD were lower than the urea-N treatment. These results suggest that mixed N is more conducive to reducing the panicle and leaf temperatures of rice than plants with a single N source.

Nitrogen content of rice leaves
Leaf nitrogen content of both varieties under heat stress was lower than that under natural temperature (Fig. 2). When exposing to high temperature stress, the nitrogen content in leaves of HT 996 and HS 343 was the highest under mixed N treatment, while it was the lowest under NO 3 − -N treatment, where the values were 7.13 and 7.18% higher and 1.47% and 9.53% lower than the urea-N treatment for HT 996 and HS 343, respectively.

Photosynthetic characteristics of rice leaves
The net photosynthetic rate (Pn), stomatal conductance (Gs), transpiration rate (Tr) and stomatal limitation (Ls) of rice leaves were significantly changed in plants cultivated  with different nitrogen forms and treated with heat stress (Table 4). Heat stress led to a reduction in the Pn, Gs and Tr of both varieties. Compared with the application of urea-N, the application of mixed N and NH 4 + -N increased the Pn, Gs and Tr of HT 996 when exposing to heat stress, of which Pn increased by 9.74% and 6.08%, Gs increased by 11.76% and 5.88%, and Tr increased by 3.94% and 10.99%, respectively. However, when NO 3 − -N was applied, the Pn, Gs and Tr of HT 996 treated with NO 3 − -N were similar to these treated with urea-N. It's worth noting that Ls showed an increasing trend under heat stress. Pn, Gs and Tr of HS 343 also showed similar patterns as those of HT 996, suggesting that exposure to high temperature at the flowering stage led to decreased photosynthetic characteristics, which was mitigated by the application of mixed N and NH 4 + -N. Under heat stress, HS 343 showed more dramatic decreases in photosynthetic capacity than HT 996.   14.14 ± 0.56c 0.34 ± 0.04bc 3.57 ± 0.30c 0.14 ± 0.03b

Chlorophyll fluorescence characteristics of rice leaves
The effects of temperature and nitrogen form on chlorophyll fluorescence parameters of rice were examined (Fig. 3). The primary light energy conversion efficiency (Fv/Fm) of PS II is the most effective index to determine whether PS II is injured or not. Heat stress led to a reduction in Fv/Fm of flag leaves in HT 996 and HS 343, but the difference of Fv/Fm was significant between natural and heat conditions for HS343. Moreover, applying different forms of nitrogen did not significantly impact Fv/Fm on both varieties under heat stress. Non-photochemical quenching (NPQ) reflects the excessive light energy captured by the antenna pigments of the PS II reaction center and dissipated by thermal energy, and its value characterizes the extent to which the photosynthetic apparatus is destroyed (Gharbi et al. 2018). Heat stress significantly increased the NPQ values of flag leaves in HT 996 and HS343 (Fig. 3), while the application of different nitrogen forms had insignificant effect on the NPQ values of both varieties under heat stress, suggesting that there was no significant relationship between chlorophyll fluorescence kinetic parameters of rice leaves when plants were cultivated with different nitrogen forms and heat stress.

Antioxidant enzyme activities and MDA content
Superoxide dismutase (SOD), peroxidase (POD) and catalase (CAT) are the main components of active oxygen scavenging system in plants. They play an important role in preventing membrane lipid peroxidation, mitigating membrane damage caused by stress and delaying plant senescence. Under heat stress, the SOD, POD and CAT activities decreased significantly in flag leaves of both varieties (Fig. 4). Compared with the application of urea-N, the application of mixed N and NH 4 + -N increased SOD, POD and CAT activities of HT 996 when exposing to heat stress, of which SOD increased by 9.08% and 11.08%, POD increased by 13.94% and 15.66%, and CAT increased by11.65% and 13.04%, respectively. However, when NO 3 − -N was applied, the SOD, POD and CAT activities of HT 996 treated with NO 3 − -N were similar to these treated with urea-N. Malondialdehyde (MDA) is a product of membrane lipid peroxidation in plant organs undergoing stress or aging. It is usually used as an index of lipid peroxidation to indicate the degree of cell membrane lipid peroxidation and the strength of plant response to stress. Changes in MDA content showed the opposite pattern as the activity of antioxidant enzymes (Fig. 4), where heat stress led to significantly increased MDA content. Under heat stress, the MDA content of HT 996 leaves treated with mixed N and NH 4 + -N was significantly lower than urea-N. The activities of antioxidant enzymes and MDA content of HS 343 were similar to HT 996. In general, the results from this study show that leaves treated with mixed N and NH 4 + -N maintained high antioxidant enzyme activity and low MDA content when exposing to high temperature at the flowering stage. Moreover, the increased scavenging of active oxygen free radicals may mitigate the damage caused by high temperature to rice in plants treated with mixed N and NH 4 + -N.

Effects of nitrogen forms on photosynthetic characteristics of rice
Photosynthesis is one of the most temperature-sensitive processes in plants, and higher temperature often reduces photosynthesis efficiency (Biswal et al. 2011). In this study, nitrogen content and net photosynthetic rate (Pn) of rice  Table 4). Nitrogen is the main component of chloroplast in plant leaves, and thus nitrogen significantly affects photosynthesis (Ciompi et al. 1996). Moreover, the Pn is linearly related to its leaf nitrogen content (Evans. 1983). In this study, an increase in the Pn under mixed N treatment was observed, which corresponded to the increased leaf nitrogen content. Nitrogen forms can also significantly affect photosynthesis (Guo et al. 2007b;Song et al. 2007;Ali et al. 2013), where NH 4 + -N leads to higher Pn and NO 3 − -N can inhibit photosynthesis (Song et al. 2007;Raab and Terry 1994). However, previous studies (Guo et al. 2007b;Xu et al. 2012) suggested that mixed N promotes photosynthesis greatly. In this study, the stomatal conductance (Gs), transpiration rate (Tr) and Pn were higher or lower in plants that were applied with NH 4 + -N or NO 3 − -N compared to urea-N, respectively. While, the above parameters were the highest in the application of mixed N sources (Table 4). These results indicate that rice not only prefers NH 4 + , but also absorbs a considerable amount of NO 3 − , and compared with single NH 4 + -N or NO 3 − -N, rice had greater photosynthetic capacity under mixed N treatment.
Chlorophyll fluorescence parameters are a group of variables used to describe the photosynthetic status of plants. These values are regarded as intrinsic parameters that can be used to study the relationship between photosynthesis and the environment (Schreiber et al. 1994). The primary light energy conversion efficiency (Fv/Fm) of PS is the most effective index to determine whether PS is injured or not. When plants are in the normal conditions, the Fv/Fm value generally ranges from 0.75 to 0.85 (Yamazakiet et al. 2011). In this study, the Fv/Fm of HT 996 leaves remained above 0.75 under heat stress (Fig. 3), which indicated that the stress did not destroy the PSII system of HT996 leaves. However, the Fv/Fm of HS 343 leaves was about 0.70 under heat stress (Fig. 3), while the stomatal limitation values (Ls) remained the same (Table 4). This indicated that the decreased Pn of HS 343 leaves under high temperature was not caused by Gs, but by the destruction of the PSII system. Nitrogen forms significantly affect chlorophyll fluorescence kinetics parameters of crop leaves (Qiao et al. 2013;Yang et al. 2015). There was no significant difference in the Fv/Fm and NPQ among different nitrogen forms treatments, and this was consistent with the findings in Zhou et al. (2011). In general, the results in this study show that there is no significant relationship between chlorophyll fluorescence kinetics parameters and applied nitrogen forms in rice.

Effects of nitrogen forms on leaf and panicle temperature of rice
There is a linear relationship between plant temperature and air temperature (Nelson and Bugbee 2015). In this study, panicle temperature was generally higher than the leaf temperature, but both were lower than the air temperature, which was consistent with the findings of Ayeneh et al. (2003;Yan et al. 2008). The external reason for higher panicle temperature than leaf temperature may be that the panicle is at the top of the canopy experience less shade. When solar radiation reaches the surface of rice, it first contacts the panicle, resulting in a rapid rise in panicle temperature. Moreover, the differences between panicle and leaf temperature can also be driven by the transpiration of leaves. The transpiration was also influenced by the stomatal openness of leaves. Stomatal transpiration is higher when the stomatal opening is larger, leading to faster heat dissipation and the lower leaf surface temperature (Olivo et al. 2009). Amani et al. (1996) found a significant correlation between canopy temperature and Gs of leaves. In this study, the panicle and leaf temperatures of rice under mixed N treatment was lower compared with the application of urea-N or single NH 4 + -N or NO 3 − -N. This difference mainly due to the maximum Gs and Tr of leaves treated with mixed N source (Table 4). Therefore, when transpiration is strengthened, the leaf temperature decreases significantly.

Effects of nitrogen forms on activities of antioxidant enzymes and MDA content of rice leaves
When plants are exposed to abiotic stresses such as high temperature and drought, reactive oxygen species (ROS) in the plant will increase. The accumulation of ROS leads to membrane lipid peroxidation, resulting in senescence and death of plant tissues (Gill and Tuteja 2010;Faisal et al. 2013). Decreased antioxidant enzyme activity under heat stress leads to accumulation of ROS free radicals and damage the cell membrane, therefore, regulating the levels of free radicals is important physiological mechanism against heat stress in plants (Matsui et al. 2001;Liao et al. 2013). In this study, it was found that the physiological function of rice flag leaf was damaged under high temperature at the flowering stage, with lower SOD, POD and CAT activity, higher MDA content, increased peroxidation of membrane lipids, as well as permeability of the plasma membrane, and decreased photosynthetic capacity. In addition, HT 996 was less damaged by high temperature than HS 343. Under heat stress, plants treated with NH 4 + -N or mixed N had higher activity of antioxidant enzymes and lower MDA content, which likely mitigated the effect of ROS on the cell membrane. In addition, lower MDA contents likely led to delays in leaf senescence, and thus mitigating the damage of high temperature. (Hu and Wang 2011;Cao et al. 2009;Jagadish et al. 2009) reported that canopy temperature was negatively correlated with soluble protein content, SOD and CAT activity, and positively correlated with MDA content. In conclusion, lower leaf temperature increases the activity of antioxidant enzymes in plants treated with NH 4 + -N and mixed N, protecting rice from heat stress.

Effects of nitrogen forms on rice yield
Global warming will lead to the intensification of heat stress caused by high temperature, which could lead to serious yield reduction of rice and other crops (Espe et al. 2017). It has been found that heat stress at the flowering stage significantly reduced yield, spikelets per panicle, seed setting rate and 1000-grain weight of HT996 and HS343 (Table 2). Nitrogen forms can affect crop biomass by regulating the growth and development as well as photosynthesis of leaves, and thus affect its yield (Guo et al. 2007b;Ding et al. 2015). In most crops, application of mixed NH 4 + -N and NO 3 − -N leads to higher yield. For example, when the ratio of NO 3 − -N:NH 4 + -N is 50:50 (m:m), the yield of wheat is nearly 78% higher than that of single NO 3 − -N (Heberer and Below 1989). The NO 3 − -N:NH 4 + -N ratio of 3:7(m:m) to 5:5 (m:m) results in higher spring cauliflower yield (Liu et al. 2013). This phenomenon was also found in this study, the seed setting rate and 1000-grain weight as well as the yield were the highest in rice plants experiencing heat stress when mixed N was applied. The application of mixed N may first increase nitrogen accumulation in rice leaves (Fig. 2), improving photosynthetic efficiency (Table 4) and promoting the growth of rice plants. Secondly, it can increase transpiration rate and stomatal conductance, reducing panicle and leaf temperatures (Fig. 1). In addition, it also can increase the activities of antioxidant enzymes (Fig. 4), and eliminate peroxides such as reactive oxygen species. Through a combination of these benefits, application of mixed N likely mitigates damages caused by exposure to high temperature and maintain higher rice yield.

Conclusions
The application of NH 4 + -N and mixed N could mitigate the effects of heat-induced damage at flowering stage of rice, and the improvements of mixed N on yield and physiological parameters was more than those of NH 4 + -N. Application of mixed N can significantly increase N accumulation in rice leaves and improve photosynthetic efficiency, which promotes plant growth. In addition, appling mixed N can also improve Gs and Tr of leaf, reduce panicle and leaf temperatures, increase antioxidant enzymes activities, and thus effectively scavenging peroxides. The combined action of these physiological changes can mitigate damage by heat stress and increase rice yield. The results of this study provide a new choice for ameliorating heat damage at the flowering period of rice in high temperature-prone areas.