Light Energy Partitioning and Photoprotection in an Exotic Species ( Salix Psammophila ) Grown in a Semi-Arid Area of Northwestern China

Thermal dissipation of excess excitation energy is an important photoprotective mechanism that plants have evolved to cope with surplus illumination. However, light-energy-partitioning dynamics in an exotic sand-dune willow (Salix psammophila) commonly used in restoring and/or stabilizing sand lands in northwestern China is largely unknown. In this study, chlorophyll fluorescence (ChlF) of photosystem II (PSII) was continuously monitored in situ in Salix psammophila to investigate plant acclimation processes driven by excessive solar radiation and extreme air temperatures (Ta). As part of a heat-regulation mechanism, energy partitioning is shown to vary with prevailing environmental conditions. In this investigation, energy absorbed during periods of moderate photosynthetically active radiation (PAR < 1200 μmoL·m−2·s−1) was largely allocated towards photochemistry (ΦPSII) with nominal amounts to thermal dissipation through reversible thermal dissipation (ΦNPQr). In extremely high solar radiation (PAR > 1500 μmoL·m−2·s−1) or in a cold temperature (Ta < 0 ◦C), more energy was dissipated by way of non-regulated thermal energy (Φf,D) and sustained thermal dissipation (ΦNPQs), leading to non-reversible photoinhibition or photodamage. This was mainly as a result of the low utilization and high absorption of light energy by PSII under cold conditions and physiologically-induced vulnerability. It was concluded that Salix psammophila had a clear tolerance to high temperatures and moderate solar radiation, but tended to be more vulnerable to high solar radiation and cold temperature. Based on species sensitivity to extreme environmental conditions, practical application and extension of Salix psammophila for land-restoration purposes should be approached cautiously, especially in high-latitude or high-altitude desert ecosystems commonly affected by events of high solar radiation and cold temperature.


Introduction
Chlorophyll fluorescence (ChlF) has been one of the most dependable and preferred methods of determining physiological responses in plants growing in stressed environments [1,2].Pulse-amplitude modulated (PAM) fluorescence provides a rapid and non-intrusive means of estimating ChlF-associated parameters in investigating vegetation response to stressful conditions [3][4][5][6].The maximal quantum yield of photosystem II (PSII) photochemistry (F v /F m ) derived from PAM-fluorometry is regularly used as an indicator of plant photosynthetic performance and health status [7,8].Low F v /F m values arise when plants are subject to biotic and/or abiotic stresses, pointing to possible damage to the PSII reaction centers (RC) [5,[9][10][11].
Light energy that plants absorb often exceeds the plants' needs for photosynthesis in the field, particularly in arid-and semi-arid areas of the world.Of numerous photoprotective mechanisms, the non-photochemical quenching coefficient (NPQ) of excess light energy within the light-harvesting antennae of PSII (LHCII) is believed to be the most effective.As a ratio of regulated and non-regulated heat dissipation, the increase in NPQ indicates elevated photoprotection in response to excessive sunlight [12,13].Indeed, even though the majority of excess light energy is safely dissipated through the NPQ pathway, photoinhibition is still inevitable, leading to photodamage as a result of the overproduction of reactive oxygen species (ROS) [14,15].
Monitoring energy partitioning with the ChlF-technique is a well-tested method of examining acclimation processes in stressed vegetation [16].Light energy absorbed by the light-harvesting complex in PSII (LHCII) is dispersed along three principal consumption pathways [17], i.e., (i) energy transfer to the photochemical portion of PSII (Φ PSII ), and (ii) through non-regulated and regulated thermal energy loss (Φ f,D and Φ NPQ , respectively).Allocation of light energy to the three energy consumption pathways yields Φ PSII + Φ NPQ + Φ f,D = 1.The three pathways are competitive, but function at a level of trade-off.Plants maintain equilibrium through the relative balance of energy consumption.Under certain environmental stress, the optimal regulation of maximal Φ PSII is maintained under high Φ NPQ :Φ f,D ratios (NPQ) [18,19].ChlF-associated parameter Φ NPQ is an important indicator of the actinic (photosynthetic) protective process.Energy dissipation associated with the xanthophyll cycle (Φ NPQ ) increases under moderate environmental stress, which plays an important role in protecting the photosynthetic apparatus from damage under strong daylight conditions [20].As an indicator of photodamage, the retention of Φ f,D is thought to be associated with a decrease in the photoprotective protein (PsbS), which reveals the inability of plants to protect themselves against damage from surplus illumination.
For improved understanding of seasonal acclimation in plants to environmental stress, two photoprotective thermal dissipation strategies (Φ NPQ ) were examined [21].They concern (i) the rapid reversible dissipation (∆pH-dependent and PsbS-controlled; Φ NPQr ) associated with photoprotection during warm, summer conditions on timescales of seconds to minutes [22]; and (ii) sustained thermal dissipation (∆pH-independent; Φ NPQs ) in several evergreens and over-wintering plants under unfavorable growing conditions.Sustained thermal dissipation persists over time and is maintained in non-daylight conditions under adverse environmental conditions, which leads to a reduction in F v /F m .This process is believed to be regulated by the sustained retention of zeaxanthin and antheraxanthin, as well as by the structural reorganization of protein complexes in the thylakoid membrane over protracted timescales [1,21,[23][24][25].
Extreme weather events are anticipated to occur more often under global climatic change [26][27][28].Plants develop many morphological features and physiological strategies to acclimate to environmental conditions [29][30][31][32].However, many plants may not be able to acclimate to the fluctuating environment at the same rate as the climate evolves [8,15,33,34].Dryland plant species are especially sensitive to climate change and understanding how these plants may cope with widely fluctuating environments is of special concern [35].Most studies conducted on plants in semi-arid areas focus on the indirect impacts of soil water deficits on light energy partitioning and/or photosynthetic capacity [36][37][38].Intense solar radiation accompanied by high-or low-temperature stress serves as a direct threat to PSII RC by affecting the light-harvesting complexes and enzyme activity.High temperatures adversely affect physiological processes that can reduce plant growth and survival in a number of ways [32,39].PSII is the site that damage is sustained from high temperatures >45 • C. Partial and reversible photoinhibition to PSII can result from moderate heat stress, i.e., air temperatures between 35-42 • C [40][41][42].While being exposed to the same high temperature treatment, relative inhibition varies significantly among plant species [43].Rapidly reversible downregulation of PSII quantum efficiency by increased Φ NPQ plays an important photoprotective role in leaves.This process involves the conversion of xanthophyll pigments and the development of a thylakoid proton electrochemical potential difference, and is clearly a crucial protective measure against the more harmful impacts of photodamage [44].
Salix psammophila C. Wang et C.Y. Yang, a non-native species, has been widely used in the construction of sand-breaks in the Mu Us desert sand lands since the 1980's.The species was introduced to the Mu Us desert from eastern Inner Mongolia (China) for afforestation due to its rapid growth and resistance to partial burial.However, as a mesophytic plant, Salix psammophila prefers to grow in moderate environmental conditions, including in moderate sunlight, temperature, and soil water availability.Dead branches and physical decay in the species (with an average lifespan of about 20 years) were regularly observed throughout the study area [45,46].We hypothesized that Salix psammophila has a limited capacity to acclimate to extreme (warm and cold) temperatures and excessive incident solar radiation common to the Mu Us desert.
The objective of this study was to: (i) examine the diurnal and seasonal dynamics of energy partitioning in environmentally-stressed plants through continuous measurement of Φ PSII , Φ f,D , and Φ NPQ , as key functional indicators of plant response; and (ii) explore species-specific acclimation strategies to anomalous temperatures and solar radiation intensities.The study has the potential to provide an improved characterization of Salix psammophila's response to changing climatic conditions.

Study Site
The study was carried out at the Yanchi Research Station (37 • 42 N, 107 • 13 E; 1530 m a.s.l.) of Beijing Forestry University, Ningxia province, along the southern edge of the Mu Us sand lands, northwestern China.The site is located in the mid-temperate zone and is characterized by a semi-arid continental monsoon climate.The mean annual temperature is approximately 8.1 • C (based on 1954-2004 data) and the frost-free season lasts on average for about 165 days [47,48].The vegetation in the area consists of a mixture of deciduous shrub species, including Artemisia ordosica Krasch, Salix psammophila, Hedysarum mongolicum Turez, and Hedysarum scoparium Fisch.et Mey.

Long-Term Monitoring of ChlF and Hydrometeorological Variables
Chlorophyll fluorescence of Salix psammophila was monitored in situ with a MONITORING-PAM Multi-Channel Chlorophyll Fluorometer (Walz, Effeltrich, Germany) from 15 April to 13 October 2012.The MONI-PAM system is equipped with up to seven cylindrical-shaped, robust, and weather-resistant measuring heads (MONI-HEAD/485) recording PAM fluorescence, ambient PAR, and temperature.A portion of healthy sun leaf of the current year was affixed to the MONI-HEAD's leaf clip consisting of two aluminum frames (35 × 25 mm).Three shrubs, representing three replicates, were monitored concurrently.The MONI-HEADs emitter-detector unit was secured to an iron stand to keep the unit stable in moderate wind speeds and was positioned to prevent self-shading.The leaf in the MONI-HEAD's leaf clip was replaced every two weeks to minimize the impact of leaf injury on ChlF measurements.Modulated blue LED lights of different intensities were used to measure the fluorescence emitted from a leaf sample.A saturating pulse of 3500 µmoL•m −2 •s −1 was supplied every ten minutes with a pulse duration of <1 s for obtaining a reading of maximum fluorescence yield in the light-adapted state (or F m ).Actinic light (1500 µmoL•m −2 •s −1 ) was supplied to obtain the actual fluorescence yield (or F t ).The measuring light was switched off between every two consecutive measurements.ChlF-associated parameters F t and F m were automatically recorded for each pulse by the PAM system.Prior-dawn (3:00-4:00) maximum ChlF was estimated as the maximum ChlF in the dark-adapted state (F m ), and the minimum F t at night was estimated as F o [49].

ChlF-Parameter Calculation
Seven ChlF-associated parameters were calculated for this analysis.Calculations were based on Equations ( 1)-( 7) listed in Table 1.ChlF-associated parameters F o and F m are the minimum and maximum fluorescence measured at night and just prior to sunrise, respectively.Parameters F o and F m were estimated nightly over the duration of the experiment.Parameter F mR was estimated from F m when the photochemistry yield was greatest [49,51].The degree of stress in Salix psammophila was quantified using five classes, i.e., no stress (with F v /F m > 0.75), minor stress (0.70-0.74), fair stress (0.65-0.69), moderate stress (0.60-0.65), and severe stress (<0.59).
Actual quantum yield of PSII photochemistry Non-regulated thermal energy loss Regulated thermal energy loss

Statistical Analyses
Raw data of three replicate measurements were processed using the batch-file feature of the WinControl-3 software (pre-installed on the data logger).Chlorophyll fluorescence values were screened using limit checking.Half-hourly values of F t < F m (non-dimensional) were considered abnormal and were removed from the dataset.ChlF-data during rainy days with totals > 5 mm were removed.All analyses were performed with the SPSS v. 20.0 software package (SPSS, Chicago, IL, USA).Student t-tests were used to test the statistical significance of ChlF-associated parameters between summer and growing-season data.One-way analysis of variance (ANOVA) was conducted on seasonal variation in hydrometeorological variables (PAR and T a ) and ChlF-parameters (i.e., Φ NPQr and Φ NPQs ).Non-linear regressions were performed between F v /F m and the environmental variables, i.e., maximum PAR, averaged T a , SWC, and VPD.Normality and homogeneity of variance were verified using Levene's test.Statistical significance was identified at a significance level of 0.05.

Environmental Dynamics
There were periods of intense solar radiation and high temperature differences between daytime and nighttime periods (Figure 1), which are typical in semi-arid areas of northwestern China.Cold temperatures accompanied high PAR in spring and autumn, and high temperatures and high PAR in summer.Most months had a high midday PAR > 1200 µmoL•m −2 •s −1 , with the exception of October (PAR = 1121 µmoL•m −2 •s −1 ), particularly on May 2 with PAR = 1610 µmoL•m −2 •s −1 , with a daily mean of 18.7-682.8µmoL•m −2 •s −1 (Figure 2a).Daily mean T a ranged between 2.8-31.5 • C (Figure 2b).The seasonal pattern of VPD followed that of T a , with a daily mean VPD ranging from 0.15-1.92kPa (Figure 1c).Daily SWC ranged from 0.06-0.15m 3 •m −3 with two rapid increases in 30 June and 4 September, reaching a level of about 0.15 m 3 •m −3 each time.

Environmental Dynamics
There were periods of intense solar radiation and high temperature differences between daytime and nighttime periods (Figure 1), which are typical in semi-arid areas of northwestern China.Cold temperatures accompanied high PAR in spring and autumn, and high temperatures and high PAR in summer.Most months had a high midday PAR > 1200 μmoL•m −2 •s −1 , with the exception of October (PAR = 1121 μmoL•m −2 •s −1 ), particularly on May 2 with PAR = 1610 μmoL•m −2 •s −1 , with a daily mean of 18.7-682.8μmoL•m −2 •s −1 (Figure 2a).Daily mean Ta ranged between 2.8-31.5 °C (Figure 2b).The seasonal pattern of VPD followed that of Ta, with a daily mean VPD ranging from 0.15-1.92kPa (Figure 1c).Daily SWC ranged from 0.06-0.15m 3 •m −3 with two rapid increases in 30 June and 4 September, reaching a level of about 0.15 m 3 •m −3 each time.

Seasonal and Diurnal Dynamics in Energy Partitioning Components
Diurnally, Φ PSII exhibited a bell-shaped curve (Figure 3a), being highest at night (Φ PSII > 0.7) irrespective of season, and lowest (<0.7) during daytime conditions, with the lowest value occurring between 10:00-13:00 LST (Local Standard Time), when PAR was highest (Figures 1 and 3).By contrast, Φ NPQ increased from morning, peaking at noon, when radiation was highest (Figure 3c).ChlF-associated parameter Φ f,D changed marginally compared to Φ PSII , with diurnal fluctuations ranging in between 0.13-0.28(Figure 3b).The non-photochemical quenching coefficient (NPQ) had a similar diurnal pattern to that of Φ NPQ (Figure 3d).There were clear seasonal patterns in energy partitioning, i.e., ΦPSII, ΦNPQ, and Φf,D (Figure 3), with ΦPSII being highest during the cold-temperature seasons (i.e., spring and autumn) and the daily mean ranging between 0.4-0.8.ChlF-parameter ΦNPQ demonstrated a contrasting temporal pattern.The highest daily mean Φf,D was observed in September, followed closely by Φf,D-values in July and October.Lowest night values of ΦPSII were found to occur during the same three months.The photosynthetic capacity, denoted by ΦPSII, decreased as PAR increased.Salix psammophila demonstrated a high quenching rate when PAR < 300 μmoL•m −2 •s −1 (Figure 4a).Accordingly, photoprotection, denoted by ΦNPQ, rapidly increased as solar radiation increased (Figure 4b).Seasonally, ΦNPQ was apparently higher in summer days (June-August).While passive thermal dissipation (i.e., Φf,D) remained fairly stable when PAR < 1200 μmoL•m −2 •s −1 , averaging about 0.24 and 0.18 during the growing season and summer periods, respectively.However, Φf,D rapidly increased when PAR > 1200 μmoL•m −2 •s −1 , peaking at 0.3 (Figure 4c).Parameter NPQ continued to increase with increasing PAR, until PAR reached 1200 μmoL•m −2 •s −1 .During the summer, Salix psammophila exhibited a decreased photochemical efficiency and relatively constant passive thermal dissipation under moderate illumination, with strengthened thermal dissipation by way of ΦNPQ (solid symbols, Figure 4).

Non-Photochemical Quenching in Response to Solar Radation Stress
ChlF-parameters ΦNPQr and ΦNPQs demonstrated varying sensitivity and patterns to solar radiation in different months (Figure 5).ΦNPQr accounted for a larger proportion of ΦNPQ and it was generally more responsive to solar radiation than ΦNPQs, increasing with increasing PAR.The reduction in ΦNPQr in response to environmental stress varied seasonally, with ΦNPQr decreasing and gradually approaching ΦNPQs at the beginning and end of the growing season (Figure 5).There were clear seasonal patterns in energy partitioning, i.e., Φ PSII , Φ NPQ , and Φ f,D (Figure 3), with Φ PSII being highest during the cold-temperature seasons (i.e., spring and autumn) and the daily mean ranging between 0.4-0.8.ChlF-parameter Φ NPQ demonstrated a contrasting temporal pattern.The highest daily mean Φ f,D was observed in September, followed closely by Φ f,D -values in July and October.Lowest night values of Φ PSII were found to occur during the same three months.The photosynthetic capacity, denoted by Φ PSII , decreased as PAR increased.Salix psammophila demonstrated a high quenching rate when PAR < 300 µmoL•m −2 •s −1 (Figure 4a).Accordingly, photoprotection, denoted by Φ NPQ , rapidly increased as solar radiation increased (Figure 4b).Seasonally, Φ NPQ was apparently higher in summer days (June-August).While passive thermal dissipation (i.e., Φ f,D ) remained fairly stable when PAR < 1200 µmoL•m −2 •s −1 , averaging about 0.24 and 0.18 during the growing season and summer periods, respectively.However, Φ f,D rapidly increased when PAR > 1200 µmoL•m −2 •s −1 , peaking at 0.3 (Figure 4c).Parameter NPQ continued to increase with increasing PAR, until PAR reached 1200 µmoL•m −2 •s −1 .During the summer, Salix psammophila exhibited a decreased photochemical efficiency and relatively constant passive thermal dissipation under moderate illumination, with strengthened thermal dissipation by way of Φ NPQ (solid symbols, Figure 4).

Non-Photochemical Quenching in Response to Solar Radation Stress
ChlF-parameters Φ NPQr and Φ NPQs demonstrated varying sensitivity and patterns to solar radiation in different months (Figure 5).Φ NPQr accounted for a larger proportion of Φ NPQ and it was generally more responsive to solar radiation than Φ NPQs , increasing with increasing PAR.
The reduction in Φ NPQr in response to environmental stress varied seasonally, with Φ NPQr decreasing and gradually approaching Φ NPQs at the beginning and end of the growing season (Figure 5).

Non-Photochemical Quenching in Response to Temperature Stress
In order to eliminate the effect of the diurnal dynamic, the effects of temperature on energy partitioning (ΦPSII, ΦNPQ, Φf,D) and on two regulated thermal energy loss components (ΦNPQr, ΦNPQs) were analyzed based on values at night (PAR < 5 μmoL•m −2 •s −1 , Figure 6a-c) and at noon (PAR ≥ 1000 μmoL•m −2 •s −1 , Figure 6b-d), respectively.ChlF-parameter ΦPSII was positively correlated to Ta at night, with an obvious depression when Ta ≤ 0 °C.Meanwhile, ΦNPQs stayed above ΦNPQr at night, and showed a negative relationship with temperature when Ta > 0 °C.Under the extreme cold condition

Non-Photochemical Quenching in Response to Temperature Stress
In order to eliminate the effect of the diurnal dynamic, the effects of temperature on energy partitioning (ΦPSII, ΦNPQ, Φf,D) and on two regulated thermal energy loss components (ΦNPQr, ΦNPQs) were analyzed based on values at night (PAR < 5 μmoL•m −2 •s −1 , Figure 6a-c) and at noon (PAR ≥ 1000 μmoL•m −2 •s −1 , Figure 6b-d), respectively.ChlF-parameter ΦPSII was positively correlated to Ta at night, with an obvious depression when Ta ≤ 0 °C.Meanwhile, ΦNPQs stayed above ΦNPQr at night, and showed a negative relationship with temperature when Ta > 0 °C.Under the extreme cold condition

Non-Photochemical Quenching in Response to Temperature Stress
In order to eliminate the effect of the diurnal dynamic, the effects of temperature on energy partitioning (Φ PSII , Φ NPQ , Φ f,D ) and on two regulated thermal energy loss components (Φ NPQr , Φ NPQs ) were analyzed based on values at night (PAR < 5 µmoL•m −2 •s −1 , Figure 6a-c) and at noon (PAR ≥ 1000 µmoL•m −2 •s −1 , Figure 6b-d), respectively.ChlF-parameter Φ PSII was positively correlated to T a at night, with an obvious depression when T a ≤ 0 • C.Meanwhile, Φ NPQs stayed above Φ NPQr at night, and showed a negative relationship with temperature when T a > 0 • C.Under the extreme cold condition (T a ≤ 0 • C), the excess energy partitioning dissipated by increasing Φ f,D .At noon, Φ PSII remained constant and equal when T a < 30 • C. Obvious depression of Φ PSII corresponding with a high degree of regulated thermal energy loss occurred under a high temperature (T a > 30 • C), whereas Φ f,D remained constant with increasing temperature.There was an inverse pattern of the two components of regulated energy loss where Φ NPQr played a dominant role in regulated thermal dissipation when T a > 20 • C, with a positive regulation with increasing temperature.

Seasonl Dynamics of Daily Fv/Fm
Depressions in Fv/Fm were identified at different levels throughout the growing season (Figure 7).Minor stress in Salix psammophila was recorded on April 25 under low temperatures and intense solar radiation (Fv/Fm = 0.72) and severe stress on 30 June to 4 July under intense solar radiation (0.14-0.52) and 5 October under low temperatures and intense solar radiation (~0.6;Figures 2 and 7).Parameter Fv/Fm recovered much more slowly in autumn, taking several more days, than in spring or summer, when temperatures were low at night.High resolution monitoring described the details of the photosynthetic depression dynamic with the fluctuation of environmental factors.
Diurnal dynamic of energy partitioning components (ΦPSII, ΦNPQ, Φf,D), along with temperature and PAR in three typical photoinhibition periods, are presented in Figure 8.The sudden acceleration of PAR (Figure 8a), a combination of high temperature and high PAR (Figure 8b), and low temperature (Figure 8c) would cause long lasting photoinhibition.Specifically, at the end of June (Figure 8a), continuous extreme high PAR (exceeded 1500 μmoL•m −2 •s −1 ) along with high temperature (around 30 °C) after a few cloudy days resulted in severe and sustaining depression in ΦPSII (below 0.5).In late July (Figure 8b), relatively high PAR (exceeded 1200 μmoL•m −2 •s −1 ) and high temperature caused minor photoinhibition, with ΦPSII being around 0.7 for several days (Figure 8b).At the end of

Seasonl Dynamics of Daily F v /F m
Depressions in F v /F m were identified at different levels throughout the growing season (Figure 7).Minor stress in Salix psammophila was recorded on April 25 under low temperatures and intense solar radiation (F v /F m = 0.72) and severe stress on 30 June to 4 July under intense solar radiation (0.14-0.52) and 5 October under low temperatures and intense solar radiation (~0.6;Figures 2 and 7).Parameter F v /F m recovered much more slowly in autumn, taking several more days, than in spring or summer, when temperatures were low at night.High resolution monitoring described the details of the photosynthetic depression dynamic with the fluctuation of environmental factors.
Diurnal dynamic of energy partitioning components (Φ PSII , Φ NPQ , Φ f,D ), along with temperature and PAR in three typical photoinhibition periods, are presented in Figure 8.The sudden acceleration of PAR (Figure 8a), a combination of high temperature and high PAR (Figure 8b), and low temperature (Figure 8c) would cause long lasting photoinhibition.Specifically, at the end of June (Figure 8a), continuous extreme high PAR (exceeded 1500 µmoL•m −2 •s −1 ) along with high temperature (around 30 • C) after a few cloudy days resulted in severe and sustaining depression in Φ PSII (below 0.5).In late July (Figure 8b), relatively high PAR (exceeded 1200 µmoL•m −2 •s −1 ) and high temperature caused minor photoinhibition, with Φ PSII being around 0.7 for several days (Figure 8b).At the end of the growing season (Figure 8c), the low temperature (~0 • C) led to moderate photoinhibition, with PSII being around 0.6.

Effect of Abiotic Factors on F v /F m
As an indicator of the photochemical response to stress and overall plant health status [7,54], F v /F m has been shown to have an optimal range between 0.75 and 0.85, decreasing whenever plants are under stress.Some studies had reported a decrease in F v /F m under water-deficit conditions [55,56].However, reductions in F v /F m at low SWC and high VPD were not readily discernible from our field measurements (Figure 9c,d), consistent with the previous studies [57,58].The discrepancies in these studies could be attributed to the different growing conditions of the plants.Plants in favorable soil water conditions are generally more sensitive to drought, while Salix psammophila has a long-term acclimation to dry conditions in arid land.In contrast, severe stress was found to occur in non-water-limited conditions, i.e., SWC > 14%.This particular condition may have developed because rain was distributed uniformly throughout the growing season (Figure 2d).Furthermore, the apparent negative correlation between F v /F m and T a and maximum PAR, when F v /F m < 0.75, provides some preliminary evidence of the photo-impairment in Salix psammophila due to extreme temperatures and PAR.Our results are consistent with those reported in previous studies for Salix psammophila growing in Inner Mongolia, China [45].

Effect of Abiotic Factors on Fv/Fm
As an indicator of the photochemical response to stress and overall plant health status [7,54], Fv/Fm has been shown to have an optimal range between 0.75 and 0.85, decreasing whenever plants are under stress.Some studies had reported a decrease in Fv/Fm under water-deficit conditions [55,56].However, reductions in Fv/Fm at low SWC and high VPD were not readily discernible from our field measurements (Figure 9c,d), consistent with the previous studies [57,58].The discrepancies in these studies could be attributed to the different growing conditions of the plants.Plants in favorable soil water conditions are generally more sensitive to drought, while Salix psammophila has a long-term acclimation to dry conditions in arid land.In contrast, severe stress was found to occur in non-waterlimited conditions, i.e., SWC > 14%.This particular condition may have developed because rain was distributed uniformly throughout the growing season (Figure 2d).Furthermore, the apparent negative correlation between Fv/Fm and Ta and maximum PAR, when Fv/Fm < 0.75, provides some preliminary evidence of the photo-impairment in Salix psammophila due to extreme temperatures and PAR.Our results are consistent with those reported in previous studies for Salix psammophila growing in Inner Mongolia, China [45].

Energy Partitioning in Response to Low Temperatures
Under extreme sub-zero temperatures, high photoinhibition characterized by a low photosynthetic efficiency and quenching was shown to occur in over-wintering coniferophytes [59,60].These species have cold-hardening mechanisms conferring freeze tolerance with specific physiological and morphological features [59,61].Similar to many other temperate regions, occasional short chilling episodes are typical for our study area.Our ChlF-data consistently indicated a decrease in photochemical efficiency at low temperatures.Salix psammophila generally exhibited elevated Φ f,D and low Φ NPQ (Figure 6a), which is explained by a reduction in the repair of RC's [62], leading to a suppression in Φ PSII at night (Figure 6a).A possible further explanation for this would be that the inhibition of antioxidant enzyme activity by chilling may have led to the production of ROS, which may have imparted serious oxidative damage to the photosynthetic apparatus [63].A decrease in temperature was proved to reduce the activity of Calvin-Benson cycle enzymes, which would reduce PSII light utilization through reducing the NADPH demands and causing the subsequent accumulation of reductants [15].A similar result was found in a high latitude area, where freezing impaired dark reactions and inhibited the capacity of light reactions to dissipate excess energy as heat [64].Elevated Φ f,D and low Φ NPQ in cold air conditions suggest an inappropriate level of photoprotection, resulting in photodamage.Moreover, Φ NPQs was mainly functional in cold conditions <0 • C, where Salix psammophila had a limited protection against upregulated Φ NPQs .Low photochemical efficiency was also observed in shrubs in the Mediterranean region, when temperatures fell below 10 • C [65].
Given the same temperatures and similar solar radiation intensities during the cold months, May had higher Φ NPQ and lower Φ f,D than in September, and a higher photochemical efficiency (Figures 1a and 2b).This can be explained by a heightened sensitivity in senescent leaves during the late growing season.In addition, reversible thermal dissipation was degenerative in September and sustained thermal dissipation increased when PAR was high (Figure 5), indicating a reduction in NPQ by variable thermal dissipation in autumn [3,21].Increased Φ f,D combined with decreased photoprotection instituted the point that physiologically-induced vulnerability in autumn reduced Salix psammophila's tolerance to stress.

Energy Partitioning in Response to High Temperatures
PSII is very sensitive to light fluctuations.High temperatures have the potential to accelerate photoinhibition, even photodamage in plants, by inducing an imbalance between light energy absorption and partitioning, while light-induced photodamage to PSII and the repair of photodamaged PSII occur simultaneously [15].While in this study, although high radiation combined with high temperature resulted in a minor photoinhibition in Salix psammophila (Figure 8b), the photodamage was not caused by enhanced temperature (Φ f,D 's were persistently low, Figure 6b).Correspondingly, Φ NPQ , especially the reversible form of thermal dissipation, was the major energy loss under high temperatures when air temperatures were greater than 30 • C.Under these conditions, the proportion of NPQ exceeded photochemical quenching (Figure 6b).This harmless ∆pH-dependent quenching associated with the conversion of violaxanthin via antheraxanthin to zeaxanthin decreases the rate of PSII photodamage under intense radiation [66][67][68].NPQ regulation of energy partitioning within PSII plays an important role in the acclimation mechanism to withstand high temperatures in existing field conditions.Similar findings were obtained for an endangered Cistaceae spp., i.e., Tuberaria major, which had the ability to protect itself from irreversible photodamage under high temperatures, even under soil water shortage [37].In general, net photosynthesis in Salix psammophila was not measurable when leaf temperatures exceeded 40 • C [45].The current results are consistent with those presented in a previous study [40], which demonstrated that stress under moderately-high temperatures causes a partial inhibition in PSII.In the future, with global warming, more intense, more frequent, and longer heat waves are anticipated [69], which poses a considerable threat to the survival of the species in the Mu Us desert.

Energy Partitioning in Response to Solar Radiation
Salix psammophila provides morphological photoprotection by possessing lanceolate and revolute leaves that reduce light interception, limiting the likelihood that events of high insolation will exceed the physiological tolerance of the species [70].Exposure to high solar radiation, energy partitioning by Salix psammophila was shown to vary over the different seasons.Results from this study showed a rapid decrease in Φ PSII , a rapid increase in Φ NPQ , and the predominance of Φ NPQ under moderate solar radiation intensity (below 1200 µmoL•m −2 •s −1 ; Figure 4a,b), suggesting that Salix psammophila was capable of acclimating to typical summer conditions by NPQ of reversible thermal dissipation (Figure 5a).Meanwhile, Φ NPQr played a dominant role in excess light energy dissipation, showing a strong xanthophyll-cycle regulation.High PAR > 1200 µmol•m −2 •s −1 , led to a more excessive thermal beyond self-regulation ability, disabling acclimation to high summer sunlight loading, yielding to a severe suppression in F v /F m (Figure 9b).Many studies have reported PSII has an efficient and dynamic repair machinery where damaged PSII proteins (D1) are replaced with newly synthesized proteins [71][72][73].While under extreme intense radiation, PSI would be damaged when the electron flow from PSII exceeds the capacity of PSI electron acceptors.PSI lacks efficient repair with a characteristic of low recovery [74].The damaged PSI impaired PSII function by limiting electron transport between PSII and the Calvin-Benson cycle.Similar results were reported for full-sunlight-grown Arabidopsis thaliana with a higher chlorophyll content, where pronounced photo-oxidative and PSII-activity losses were detected.Additionally, during the late growing season, weakened NPQ-associated protection, indicated by a decrease in Φ NPQr , meant that more excess light energy dissipated via the Φ f,D form.This result supports our hypothesis that extreme high solar irradiation is an important stress factor affecting the afforestation quality of Salix psammophila, which also increases the sensitivity of Salix psammophila to biotic and abiotic stresses (e.g., senescence during the latter part of the growing season and low temperatures) at the level of gene expression [75].

Conclusions
ChlF-associated parameters F v /F m , Φ NPQ , Φ f,D , and NPQ are appropriate indicators to reflect environmentally-invoked stress in Salix psammophila.There exist different energy partitioning mechanisms in responding to different environmental stresses.Mechanisms of non-photochemical regulation explained the physiological plasticity of seasonal acclimation observed in Salix psammophila.Plants had a higher photochemical efficiency under moderate PAR and air temperature and allocated more energy to non-photochemical dissipation pathways.In extremely high PAR or in a cold temperature, more energy was dissipated by way of Φ NPQ and Φ f,D , leading to non-reversible photoinhibition.In high temperatures (>30 • C), but moderate radiation (<1200 µmoL•m −2 •s −1 ), the plants dissipated the energy more readily through Φ NPQ , showing a strong tolerance.Given the more frequent extreme weather events projected for northwest China under global change, large plantations of Salix psammophila in the region are at an extreme risk of declining in the near future.

Figure 4
Figure4illustrates energy partitioning with respect to ΦPSII, Φf,D, ΦNPQ, and NPQ as a function of PAR during the growing season (open symbols) and during summer (close symbols), respectively.The photosynthetic capacity, denoted by ΦPSII, decreased as PAR increased.Salix psammophila demonstrated a high quenching rate when PAR < 300 μmoL•m −2 •s −1 (Figure4a).Accordingly, photoprotection, denoted by ΦNPQ, rapidly increased as solar radiation increased (Figure4b).Seasonally, ΦNPQ was apparently higher in summer days (June-August).While passive thermal dissipation (i.e., Φf,D) remained fairly stable when PAR < 1200 μmoL•m −2 •s −1 , averaging about 0.24 and 0.18 during the growing season and summer periods, respectively.However, Φf,D rapidly increased when PAR > 1200 μmoL•m −2 •s −1 , peaking at 0.3 (Figure4c).Parameter NPQ continued to increase with increasing PAR, until PAR reached 1200 μmoL•m −2 •s −1 .During the summer, Salix psammophila exhibited a decreased photochemical efficiency and relatively constant passive thermal dissipation under moderate illumination, with strengthened thermal dissipation by way of ΦNPQ (solid symbols, Figure4).ChlF-parameters ΦNPQr and ΦNPQs demonstrated varying sensitivity and patterns to solar radiation in different months (Figure5).ΦNPQr accounted for a larger proportion of ΦNPQ and it was generally more responsive to solar radiation than ΦNPQs, increasing with increasing PAR.The reduction in ΦNPQr in response to environmental stress varied seasonally, with ΦNPQr decreasing and gradually approaching ΦNPQs at the beginning and end of the growing season (Figure5).

Figure 4
Figure 4 illustrates energy partitioning with respect to Φ PSII , Φ f,D , Φ NPQ , and NPQ as a function of PAR during the growing season (open symbols) and during summer (close symbols), respectively.The photosynthetic capacity, denoted by Φ PSII , decreased as PAR increased.Salix psammophila demonstrated a high quenching rate when PAR < 300 µmoL•m −2 •s −1 (Figure4a).Accordingly, photoprotection, denoted by Φ NPQ , rapidly increased as solar radiation increased (Figure4b).Seasonally, Φ NPQ was apparently higher in summer days (June-August).While passive thermal dissipation (i.e., Φ f,D ) remained fairly stable when PAR < 1200 µmoL•m −2 •s −1 , averaging about 0.24 and 0.18 during the growing season and summer periods, respectively.However, Φ f,D rapidly increased when PAR > 1200 µmoL•m −2 •s −1 , peaking at 0.3 (Figure4c).Parameter NPQ continued to increase with increasing PAR, until PAR reached 1200 µmoL•m −2 •s −1 .During the summer, Salix psammophila exhibited a decreased photochemical efficiency and relatively constant passive thermal dissipation under moderate illumination, with strengthened thermal dissipation by way of Φ NPQ (solid symbols, Figure4).ChlF-parameters Φ NPQr and Φ NPQs demonstrated varying sensitivity and patterns to solar radiation in different months (Figure5).Φ NPQr accounted for a larger proportion of Φ NPQ and it was generally more responsive to solar radiation than Φ NPQs , increasing with increasing PAR.

Forests 2018, 9 ,Figure 4 .Figure 5 .
Figure 4. Actual quantum yield of PSII photochemistry (ΦPSII, (a)), regulated thermal energy loss (ΦNPQ, (b)), non-regulated energy loss (Φf,D, (c)), and non-photochemical quenching coefficient (NPQ, Stern-Volmer quenching, (d)) in Salix psammophila as a function of photosynthetically active radiation (PAR) during the growing season of 2012.Data points of energy partitioning and NPQ represent binaveraged values.Here, the bins are set at 100 μmoL•m −2 •s −1 increments of PAR with a minimal sample size of six.The open and closed circles represent averages of data collected over the entire growing season and summer period, respectively.The error bars denote standard error.

Figure 4 .
Figure 4. Actual quantum yield of PSII photochemistry (Φ PSII , (a)), regulated thermal energy loss (Φ NPQ , (b)), non-regulated energy loss (Φ f,D , (c)), and non-photochemical quenching coefficient (NPQ, Stern-Volmer quenching, (d)) in Salix psammophila as a function of photosynthetically active radiation (PAR) during the growing season of 2012.Data points of energy partitioning and NPQ represent bin-averaged values.Here, the bins are set at 100 µmoL•m −2 •s −1 increments of PAR with a minimal sample size of six.The open and closed circles represent averages of data collected over the entire growing season and summer period, respectively.The error bars denote standard error.

Figure 4 .Figure 5 .
Figure 4. Actual quantum yield of PSII photochemistry (ΦPSII, (a)), regulated thermal energy loss (ΦNPQ, (b)), non-regulated energy loss (Φf,D, (c)), and non-photochemical quenching coefficient (NPQ, Stern-Volmer quenching, (d)) in Salix psammophila as a function of photosynthetically active radiation (PAR) during the growing season of 2012.Data points of energy partitioning and NPQ represent binaveraged values.Here, the bins are set at 100 μmoL•m −2 •s −1 increments of PAR with a minimal sample size of six.The open and closed circles represent averages of data collected over the entire growing season and summer period, respectively.The error bars denote standard error.

Figure 5 .
Figure 5. Seasonal variation in sustained and rapidly reversible thermal dissipation (i.e., Φ NPQs and Φ NPQr ) in Salix psammophila as a function of photosynthetically active radiation (PAR).Data points of Φ NPQs and Φ NPQr represent bin-averaged values.Here, the bins are set at 100 µmoL•m −2 •s −1 increments of PAR with a minimal sample size of six.The error bars denote standard error.

Figure 6 .
Figure 6.Actual quantum yield of PSII photochemistry (Φ PSII ), regulated thermal energy loss (Φ NPQ ), non-regulated energy loss (Φ f,D ), sustained and rapidly reversible thermal dissipation (Φ NPQs and Φ NPQr ) as a function of air temperature (T a ) at night (PAR < 5 µmoL•m −2 •s −1 ; (a,c)) and at noon (PAR ≥ 1000 µmoL•m −2 •s −1 ; (b,d)), respectively, during the 2012 growing season.Data points of energy partitioning represent bin-averaged values.Here, the bins are set at 1 • C increments of T a with a minimal sample size of six values.The error bars denote standard error.

Figure 7 .
Figure 7. Seasonal variation of daily maximal quantum yield of PSII photochemistry (F v /F m ) in Salix psammophila during 2012 growing season.

Table 1 .
Process yields and ChlF-associated yield equations.