Response of Physiological Indicators to Environmental Factors under Water Level Regulation of Paddy Fields in Southern China

Agricultural water level control is the key to the combined control of water-saving irrigation and controlled drainage. It is easy to observe and master in farmland practice, and has important theoretical and practical significance. In this paper, a systematic study on stomatal regulation and photosynthetic-evapotranspiration coupling mechanism, as well as the effect of meteorological environmental factors on photosynthetic-evapotranspiration coupling mechanism of paddy rice in drought and flooding conditions was conducted by taking paddy field’s water level as a regulation indicator of paddy rice’s irrigation and drainage. Results showed that net photosynthetic rate (Pn) and evapotranspiration rate (Tr) had quadratic and linear relationship with stomatal conductance (Gs), respectively. The responding range of Gs in drought condition was lower than that in flooding condition. Drought stress aggravated the photosynthetic stomatal limitation of rice leaves at noon, and partial closure of stomata was the main reason resulting in the reduction of Pn suffering from drought at noon. Change curve of paddy rice’s leaf temperature difference (∆T) showed a W or V type daily change curve and its change rule was rightly contrary to that of Tr. Pn had quadratic curve relationships with photosynthetically active radiation (PAR) and surrounding CO2 concentration (Cs) both in the morning and in the afternoon, Pn had quadratic curve relationships in the morning and linear relationships in the afternoon with air temperature (Ta), respectively. Tr showed linear relationships with PAR, Ta, relative air humidity (RH) and barometric pressure saturation deficit (VPD).


Introduction
Rice irrigation area is important for regional economic development, social harmony and stability, improvement of ecological environment and so on [1].However, there are still some problems in rice irrigation areas, such as low efficiency of water and fertilizer utilization, and serious agricultural non-point source pollution, which restrict the sustainable development [2,3].With the development of the theory of crop irrigation and drainage, water-saving irrigation and controlled drainage technology attracted great attention [4,5].Water level regulation is crucial for water-saving irrigation-drainage, which is meaningful for water saving, pollution reduction and high rice yield in south China [6][7][8].

Experimental Site
This study was carried out at the Key Laboratory of Efficient Irrigation-Drainage and Agricultural Soil-Water Environment in Southern China, Ministry of Education in 2011 and 2012.The region has a subtropical humid monsoon climate, with an average annual evaporation of 900 mm, annual average temperature of 15.4 • C, maximum and minimum air temperatures of 43.0 • C and −14.0 • C, mean annual rainfall is 1041 mm, and there are 220 frost-free days per year.The soil in the area is loamy clay.The area adopts a five-year rice-wheat rotation system.There are 32 fixed lysimeter plots (28 with closed bottom and 4 without) with the specifications of 2.5 m length, 2 m width, and 2 m depth.The lysimeter layout is divided into two groups, with each group containing 16 plots.The underground corridors and underground equipment rooms are built between the two groups.The irrigation system is an automatic irrigation system controlled by the host-electromagnetic valve.The topsoil (0-30 cm) with pH value of 6.97 in the lysimeter contained 2.40% soil organic matter, 0.9048 g•kg −1 of total nitrogen, 27.65 mg•kg −1 of available nitrogen, 0.32 g•kg −1 of total phosphorus, and 12.5 mg•kg −1 of available phosphorus.

Experimental Design
The paper conducted a series of experiments in paddy fields between 2011 and 2012 with the combination of water level regulation and water control duration, as is shown in Table 1.It was only considered as the "flood suffered" when there was deep flooding of rice and a certain leakage intensity was maintained.Two kinds of leakage intensity (2 mm/day and 4 mm/day) were designed in 2-years experiments.At the tillering stage, rice plant was short, and the flooding water level was designed for 120 mm, while flooding water level was 250 mm at the jointing to milky stage.The flooding duration was kept for 10 days at all growth stages without water supply even when water level decreased.This process was a simulation of dynamics of paddy field water levels after a short-term heavy rainfall.The groundwater depth was taken as the controlled water level during drought test, and two water levels were designed at each growth stage.For the tillering stage, the water level was respectively −300 mm and −500 mm, while for jointing to milky stage, the water level was −400 mm and −700 mm, respectively.
Notes: (1) During the water control period, water level and duration was controlled according to

Method
At each growth stage, 4 and 3 representative plants were chosen and labeled for each treatment, and the Pn, Tr, Gs were measured.The measuring interval was one time before water control, and twice between water control period, then measured every 2 or 3 days after water level returning to normal, with measuring time at 9:30 am to 10:30 am.Photosynthetic day change was measured every 2 h within 1 to 2 days after the end of water control at each growth stage from 8:00 a.m. to 18:00 p.m. Leaf photosynthetic parameters were measured by Li-6400 portable photosynthetic instrument and light-emitting diode (LED) red/blue light source (LI-COR, Lincoln, NE, USA).Through gas exchange method, the Pn, Tr, Gs were analyzed and calculated according to the parameters (CO 2 , H 2 O concentration, gas flow rate, leaf area, temperature and pressure, etc.) of the reference chamber and sample chamber.At the same time, some important environmental parameters such as light intensity, temperature and humidity were recorded.Simple data calculation and diagramming was completed by Excel 2010.Correlation analysis and regression analysis was carried out by IBM SPSS Statistics 19.The synchronous day change law between CO 2 concentration (Ci) and stomatal limitation value (Ls), net photosynthetic rate (Pn) and stomatal conductance (Gs) of paddy rice in drought condition was shown in Figure 1.It showed that Pn and Gs in H1 and H2 decreased from 11:00 a.m. to 12:00 p.m. at tillering stage, and Ci decreased too, but Ls increased, which illustrated that Gs reduction was the reason why photosynthetic rate decreased.As for contrast treatment (CK) treatment, Pn decreased from 12:00 p.m. to 13:00 p.m., and the corresponding Ci was tended to increase.Ci increase was the result of Pn decrease, i.e., non-stomatal limitation was dominant.The above results showed that drought stress aggravated the photosynthetic stomatal limitation in rice leaves at noon, and partial closure of the stomata was the main reason resulting in the decrease of photosynthetic rate of drought paddy rice at noon.At jointing-booting stage, Pn decreased from 10:30 to 12:00 in H3 and CK treatment, while Gs, Ci and Ls all increased, which illustrated Pn decrease was caused by non-stomatal factors.From 12:00 to 13:00, Pn in H3 continued to decrease, and Gs and Ci changed from the increasing trend to the decreasing trend, while the increase of Ls showed that stomatal factor was the main reason for Pn decrease.Thus the factors causing decrease to Pn at noon were not constant, but were caused by the mutual action of stomatal limitation factors and non-stomatal limitation factors, and both of them interacted with each other, where one factor dominated at a certain time.

Results and Discussion
At heading-flowering stage, Pn and Ci decreased from 9:00 to 11:00 for H5 and H6, whereas Gs and Ls increased; certainly it can be judged that the main reason causing Pn decrease was the stomatal factors.It illustrated that stomatal limitations could happen when Gs increased.Under this condition, the diffusion rate of CO 2 from leaf pores into leaf cells was the consumption rate of cell's photosynthetic reaction, therefore Ci kept decreasing and stomata was still the limiting factor of Pn decrease.As for CK, Pn decreased from 9:00 to 11:00.Ci was basically constant, although Ls value decreased.Thus, it was difficult to judge whether it was a stomatal limitation or non-stomatal limitation.At milky stage, Pn and Gs in H7 and H8 decreased simultaneously from 11:00 to 12:00, Ci and Ls decreased, which showed non-stomatal factors were the main reason of Pn decreasing.Referring to the judgment basis for stomatal and non-stomatal limitations of photosynthesis, Huang [24] found that the change trend of Ci and Ls were two essential reliable criterions.Ci decreasing and Ls increasing showed the stomatal conductance was the main reason, while Ci increasing and Ls decreasing showed the non-stomatal factor was the main reason, which was consistent with the conclusions in this paper.Referring to the judgment basis for stomatal and non-stomatal limitations of photosynthesis, Huang [24] found that the change trend of Ci and Ls were two essential reliable criterions.Ci decreasing and Ls increasing showed the stomatal conductance was the main reason, while Ci increasing and Ls decreasing showed the non-stomatal factor was the main reason, which was consistent with the conclusions in this paper.

Response Relationship between Ci and Influencing Factors of Paddy Rice under Water Level Control
The relationships between Ci and Gs, Pn and Cs (CO2 concentration around the leaves) were shown in Figure 2. It was clear that Ci distribution was widest in the morning, ranging from 200 to 400 μmol•mol −1 .Gs increased first and then decreased as Ci decreased under drought and flooding treatments in the morning.Gs reached peak in the morning; the one of drought treatment was slightly higher than that of flooding treatment when Ci was 270~300 μmol•mol −1 .Pn also increased first and then decreased with Ci decreasing for drought treatment, while it increased with Ci decreasing for flooding treatment and did not decrease after reaching peak.Pn reached the peak when Ci was 200 and 250 μmol•mol −1 for flooding and drought treatments, respectively.At 11:00~14:00, Ci ranged from 200 to 300 μmol•mol −1 , and the corresponding Gs and Pn values were higher.In the afternoon, Ci mainly ranged from 250 to 350 μmol•mol −1 and it decreased as Gs and Pn decreased.Ci increased as Cs increased both in the morning and afternoon.Where Cs ranged from 330 to 370 μmol•mol −1 , and Ci ranged from 200 to 300 μmol•mol −1 from at noon, certainly there was

Response Relationship between Ci and Influencing Factors of Paddy Rice under Water Level Control
The relationships between Ci and Gs, Pn and Cs (CO 2 concentration around the leaves) were shown in Figure 2. It was clear that Ci distribution was widest in the morning, ranging from 200 to 400 µmol•mol −1 .Gs increased first and then decreased as Ci decreased under drought and flooding treatments in the morning.Gs reached peak in the morning; the one of drought treatment was slightly higher than that of flooding treatment when Ci was 270~300 µmol•mol −1 .Pn also increased first and then decreased with Ci decreasing for drought treatment, while it increased with Ci decreasing for flooding treatment and did not decrease after reaching peak.Pn reached the peak when Ci was 200 and 250 µmol•mol −1 for flooding and drought treatments, respectively.At 11:00~14:00, Ci ranged from 200 to 300 µmol•mol −1 , and the corresponding Gs and Pn values were higher.In the afternoon, Ci mainly ranged from 250 to 350 µmol•mol −1 and it decreased as Gs and Pn decreased.Ci increased as Cs increased both in the morning and afternoon.Where Cs ranged from 330 to 370 µmol•mol −1 , and Ci ranged from 200 to 300 µmol•mol −1 from at noon, certainly there was no obvious difference between drought and flooding treatments.When Cs is same, Ci in the afternoon was slightly higher than the one in the morning.
no obvious difference between drought and flooding treatments.When Cs is same, Ci in the afternoon was slightly higher than the one in the morning.

Analysis on Daily Change Law of Temperature Difference (ΔT) and Tr under Water Level Control
The daily change of ΔT and Tr was shown in Figure 3.It could be seen that ΔT was basically negative in the daytime, and its change curve showed W type or V type, which was opposite to the change curve of Tr.When Tr was higher, |ΔT| was also higher.Tr reached the lowest level around 12:00 at noon for rice under water level control at tillering stage, while ΔT reached the peak value.This was mainly because stomata were partially closed to avoid excessive loss of leaf moisture due to the high temperature and high light intensity at noon, which resulted in a temporary phenomenon of "noon break" to Pn of paddy rice [25].At this time the leaf temperature would be increasing quickly as ΔT decreased significantly, namely the |ΔT| decreased.At jointing-booting stage, Tr was not at the lowest value at noon, but there was no significant difference in Tr at the end of drought

Analysis on Daily Change Law of Temperature Difference (∆T) and Tr under Water Level Control
The daily change of ∆T and Tr was shown in Figure 3.It could be seen that ∆T was basically negative in the daytime, and its change curve showed W type or V type, which was opposite to the change curve of Tr.When Tr was higher, |∆T| was also higher.Tr reached the lowest level around 12:00 at noon for rice under water level control at tillering stage, while ∆T reached the peak value.This was mainly because stomata were partially closed to avoid excessive loss of leaf moisture due to the high temperature and high light intensity at noon, which resulted in a temporary phenomenon of "noon break" to Pn of paddy rice [25].At this time the leaf temperature would be increasing quickly as ∆T decreased significantly, namely the |∆T| decreased.At jointing-booting stage, Tr was not at the lowest value at noon, but there was no significant difference in Tr at the end of drought and flooding treatment; it was mainly caused by the cloudy weather by that time.At tillering and jointing-booting stages, the time with the lowest value of ∆T at the end of flooding treatment was delayed by 1~1.5 h more than the time with Tr peak value, signifying that ∆T had a certain hysteresis in response to transpiration.
Water 2018, 10, x FOR PEER REVIEW 7 of 15 and flooding treatment; it was mainly caused by the cloudy weather by that time.At tillering and jointing-booting stages, the time with the lowest value of ΔT at the end of flooding treatment was delayed by 1~1.5 h more than the time with Tr peak value, signifying that ΔT had a certain hysteresis in response to transpiration.

Response Relationship between ΔT and Influencing Factors under Water Level Control
The relationships between ΔT and Gs, Tr and Ta (air temperature) of paddy rice's leaf under water level control were shown in Figure 4.It showed that the response of Gs and Tr to ΔT were slightly different in the two conditions of drought and flooding treatments.The distribution of Gs and Tr was more distinct for flooding treatment, while it was concentrated for drought treatment in three time sections.For flooding treatment with the same ΔT, Gs and Tr reached the highest at noon, followed by the morning, while it was lowest in the afternoon.As for drought treatment, the smaller |ΔT|, the closer the Gs and Tr values in three time sections.The leaf temperature was a bit higher than Ta when the Ta was lower in the morning (27~33 °C), and ΔT was positive (0~1 °C).As the Ta raised, the leaf evapotranspiration accelerated, and |ΔT| gradually increased.When the Ta was at 35~37 °C at noon, the ΔT was positive for flooding treatment.This was because of weaker evapotranspiration for partially closed stomata and which prevented excessive water loss of leaves.The leaf temperature was higher than Ta at this time.ΔT was not positive at noon for drought treatment, which was mainly because of the weather conditions.When the Ta exceeded 37 °C, in order to prevent leaves from being burnt by high temperature, the stomata would open again to increase evapotranspiration and lower the leaf temperature, and |ΔT| increased at this moment.In the afternoon, |ΔT| gradually became zero with the temperature dropping.The reduction rate of

Response Relationship between ∆T and Influencing Factors under Water Level Control
The relationships between ∆T and Gs, Tr and Ta (air temperature) of paddy rice's leaf under water level control were shown in Figure 4.It showed that the response of Gs and Tr to ∆T were slightly different in the two conditions of drought and flooding treatments.The distribution of Gs and Tr was more distinct for flooding treatment, while it was concentrated for drought treatment in three time sections.For flooding treatment with the same ∆T, Gs and Tr reached the highest at noon, followed by the morning, while it was lowest in the afternoon.As for drought treatment, the smaller |∆T|, the closer the Gs and Tr values in three time sections.The leaf temperature was a bit higher than Ta when the Ta was lower in the morning (27~33 • C), and ∆T was positive (0~1 • C).As the Ta raised, the leaf evapotranspiration accelerated, and |∆T| gradually increased.When the Ta was at 35~37 • C at noon, the ∆T was positive for flooding treatment.This was because of weaker evapotranspiration for partially closed stomata and which prevented excessive water loss of leaves.The leaf temperature was higher than Ta at this time.∆T was not positive at noon for drought treatment, which was mainly because of the weather conditions.When the Ta exceeded 37 • C, in order to prevent leaves from being burnt by high temperature, the stomata would open again to increase evapotranspiration and lower the leaf temperature, and |∆T| increased at this moment.In the afternoon, |∆T| gradually became zero with the temperature dropping.The reduction rate of |∆T| for flooding treatment was faster than that for drought treatment, and the Ta range for drought treatment in the afternoon was 33~40 • C, which was wider than that for flooding treatment (33~37 • C).From the results above, the soil surface temperature was very high for drought condition, resulting in the higher temperature inside the rice canopy.The leaves could maintain the high evapotranspiration for a longer time and |∆T| decreased slowly finally.
Water 2018, 10, x FOR PEER REVIEW 8 of 15 |ΔT| for flooding treatment was faster than that for drought treatment, and the Ta range for drought treatment in the afternoon was 33~40 °C, which was wider than that for flooding treatment (33~37 °C).From the results above, the soil surface temperature was very high for drought condition, resulting in the higher temperature inside the rice canopy.The leaves could maintain the high evapotranspiration for a longer time and |ΔT| decreased slowly finally.

Response of Pn to Environmental Factors under Paddy Field Water Level Control
As shown in Figure 5, the Pn value in the morning was higher than that in the afternoon under the same photosynthetically active radiation (PAR).Compared to flooding treatment, the Pn value decreased obviously in the morning for drought treatment, reaching the peak value at the PAR of around 900 μmol•m −2 •mol −1 .At the noon, the PAR value concentrated in the range of 900~1800 μmol•m −2 •mol −1 for both flooding and drought treatments.Also, the change of Pn value for flooding treatment (10~25 μmol•m −2 •mol −1 ) was greater than that for drought treatment (15~25 μmol•m −2 •mol −1 ).Ta is a major factor for photosynthesis process and the physical exchange of CO2 and water vapor between leaves and air.It could also affect the Pn level according to the effect on enzyme activity.The Pn value reached the peak at the temperature of 37 °C and 35 °C respectively for flooding and drought treatments in the morning.The Ta at the noon ranged from 35 °C to 42 °C,

Response of Pn to Environmental Factors under Paddy Field Water Level Control
As shown in Figure 5, the Pn value in the morning was higher than that in the afternoon under the same photosynthetically active radiation (PAR).Compared to flooding treatment, the Pn value decreased obviously in the morning for drought treatment, reaching the peak value at the PAR of around 900 µmol•m −2 •mol −1 .At the noon, the PAR value concentrated in the range of 900~1800 µmol•m −2 •mol −1 for both flooding and drought treatments.Also, the change of Pn value for flooding treatment (10~25 µmol•m −2 •mol −1 ) was greater than that for drought treatment (15~25 µmol•m −2 •mol −1 ).Ta is a major factor for photosynthesis process and the physical exchange of CO 2 and water vapor between leaves and air.It could also affect the Pn level according to the effect on enzyme activity.The Pn value reached the peak at the temperature of 37 • C and 35 • C respectively for flooding and drought treatments in the morning.The Ta at the noon ranged from 35 • C to 42 • C, and the Pn ranged from 15 to 35 µmol•m −2 •mol −1 .The Pn in the afternoon was lower than that in the morning under the same Ta.CO 2 is the raw material for photosynthesis.The Pn values for flooding and drought treatments reached the peak in the morning when the Ci levels were 380 and 350 µmol•mol −1 , respectively.When the Ci ranged from 330 to 360 µmol•mol −1 at the noon, the Pn value ranged from 15 to 35 µmol•m −2 •mol −1 .At the same Ci value, the Pn value in the afternoon was lower than that in the morning.Relative humidity (RH) and vapor pressure difference (VPD) could affect Pn according to the effect on Gs and Tr.Moreover, the Pn trend was not clear at the three periods; however, the distribution area of scatter points was obviously divided in each time interval.Xia found that PAR was the most important ecological factor affecting Pn, followed by air CO 2 concentration; with the intensification of water stress, the Ta evidently restrained Pn by using path analysis [26].The relative water content regimes were different to our study.Pn according to the effect on Gs and Tr.Moreover, the Pn change trend was not clear at the three periods; however, the distribution area of scatter points was obviously divided in each time interval.Xia found that PAR was the most important ecological factor affecting Pn, followed by air CO2 concentration; with the intensification of water stress, the Ta evidently restrained Pn by using path analysis [26].The relative water content regimes were different to our study.

Response of Tr to Environmental Factors under Paddy Field Water Level Control
The relationship between Tr and environmental factors under water level control was shown in Figure 6.Under the same PAR condition, Tr was higher in the afternoon compared to the morning, which was mainly because of the higher accumulated temperature after the noon.The leaf stomata need to enhance evapotranspiration to avoid being burnt.The highest Tr value in the afternoon for drought treatment was obviously higher than that for flooding treatment, while the highest Tr value at noon for flooding treatment was higher.This explained the hysteresis effect of Tr on light intensity for drought treatment, which was mainly related to the micro-environment of the farmland formed under two conditions.The effect of temperature on evapotranspiration was achieved by affecting the water vapor pressure difference between inside and outside the leaves.Leaf temperature increased with Ta increasing, resulting in much more water vapor concentration increasing of stomata rather than the air [27].The water vapor pressure difference inside and outside the leaves also increased, and Tr accelerated.The RH could affect crop Tr through affecting the evapotranspiration capacity of atmosphere.The more humid the air, the smaller the difference in vapor pressure between the inside and outside of the leaf, which is more detrimental to the evapotranspiration of the leaves.The RH, VPD and Tr showed a good linear relationship in the morning.The Tr increased with the RH decreasing or VPD increasing in the morning, while it decreased with RH increasing.

Response of Tr to Environmental Factors under Paddy Field Water Level Control
The relationship between Tr and environmental factors under water level control was shown in Figure 6.Under the same PAR condition, Tr was higher in the afternoon compared to the morning, which was mainly because of the higher accumulated temperature after the noon.The leaf stomata need to enhance evapotranspiration to avoid being burnt.The highest Tr value in the afternoon for drought treatment was obviously higher than that for flooding treatment, while the highest Tr value at noon for flooding treatment was higher.This explained the hysteresis effect of Tr on light intensity for drought treatment, which was mainly related to the micro-environment of the farmland formed under two conditions.The effect of temperature on evapotranspiration was achieved by affecting the water vapor pressure difference between inside and outside the leaves.Leaf temperature increased with Ta increasing, resulting in much more water vapor concentration increasing of stomata rather than the air [27].The water vapor pressure difference inside and outside the leaves also increased, and Tr accelerated.The RH could affect crop Tr through affecting the evapotranspiration capacity of atmosphere.The more humid the air, the smaller the difference in vapor pressure between the inside and outside of the leaf, which is more detrimental to the evapotranspiration of the leaves.The RH, VPD and Tr showed a good linear relationship in the morning.The Tr increased with the RH decreasing or VPD increasing in the morning, while it decreased with RH increasing.

3. 1 .
Collaborative Analysis on Change in Intercellular CO 2 Concentration (Ci) and Stomatal/Non-Stomatal Limitation of Paddy Rice under Water Level Control

Figure 1 .
Figure 1.Stomatal and non-stomatal limitations relations on photosynthetic rate (Pn) under the drought condition.

Figure 1 .
Figure 1.Stomatal and non-stomatal limitations relations on photosynthetic rate (Pn) under the drought condition.

Figure 3 .
Figure 3. Daily change of temperature difference (ΔT) and evapotranspiration rate (Tr) under water level control.

Figure 3 .
Figure 3. Daily change of temperature difference (∆T) and evapotranspiration rate (Tr) under water level control.

Water 2018 ,
10, x FOR PEER REVIEW 9 of 15 and the Pn ranged from 15 to 35 μmol•m −2 •mol −1 .The Pn in the afternoon was lower than that in the morning under the same Ta.CO2 is the raw material for photosynthesis.The Pn values for flooding and drought treatments reached the peak in the morning when the Ci levels were 380 and 350 μmol•mol −1 , respectively.When the Ci ranged from 330 to 360 μmol•mol −1 at the noon, the Pn value ranged from 15 to 35 μmol•m −2 •mol −1 .At the same Ci value, the Pn value in the afternoon was lower than that in the morning.Relative humidity (RH) and vapor pressure difference (VPD) could affect

Figure 5 .
Figure 5. Response of photosynthetic rate (Pn)to environmental factors under water level control in paddy field.

Figure 5 .
Figure 5. Response of photosynthetic rate (Pn) to environmental factors under water level control in paddy field.

Figure 6 .
Figure 6.Response of evapotranspiration rate (Tr) to environmental factors under paddy field water level regulation.

Figure 6 .
Figure 6.Response of evapotranspiration rate (Tr) to environmental factors under paddy field water level regulation.

Table 1 .
Water level regulation scheme and water control stages division in 2011 and 2012.

Table
At water control stage, there was no water supplement when water table was decreasing.(3) Starting time for water flooding and drought at each growing stage was same.As for water drought test, it covered the one from no water layer decreasing to the set value.According to the water controlling days, once to twice drought was carried out at each growing stage.
. After that, water level was recovered to reasonable upper water level limit.(2) Water flooding test lasted for 10 days, and it was controlled according to the preset leakage amount.At other stages, the leakage amount was set at 2 mm•day −1 when there was a water layer, while it was 0 mm/d when there was no water layer.