Physiological and ecological characteristics of Periploca sepium Bunge under drought stress on shell sand in the Yellow River Delta of China

This study investigated the physiological and ecological changes in P. sepium Bunge and elucidated the physiological regulatory mechanisms underlying the adaptation of P. sepium to drought stress in shell sand. Drought stress led to a significant decrease in the net photosynthesis rate (Pn) and respiration rate of leaves and a decrease in low-intensity light-use efficiency (LUE) and light ecological amplitude. An increase in drought stress led to a considerable decrease in the photosynthetic electron transport rate in the P. sepium leaves and a significant increase in the amount of light energy dissipated as heat. In addition, the photosynthesis process suffered from severe photoinhibition. P. sepium plants counteracted the effects of drought stress primarily by increasing their peroxidase (POD) activity and by regulating membrane lipid peroxidation by secreting greater numbers of osmotic adjustment substances (proline (Pro) and soluble sugars (Ss)) and malondialdehyde (MDA). As drought stress increased, both the stem sap flow rate and the cumulative sap flow of P. sepium decreased considerably. P. sepium Bunge adapts to drought stress through interregulatory activity between photosynthesis, water-related physiological activities, and physiological and biochemical processes, and this species exhibits relatively high adaptive plasticity to drought.

Shell ridges are unique beach ridge landforms near the high tide line that forms from the accumulation of residual shells from dead mollusks and their fragments, which are transported by waves. The shell ridges in the Yellow River Delta compose one of the three largest groups of ancient shell ridges in the world, and this group is the only one composed of both new and old shell ridges. Owing to its close proximity to a muddy coastal zone, the groundwater below these shell ridges is relatively shallow and is easily affected by natural factors, e.g., tides and high evaporation-precipitation ratios 1 . The soil in the beach ridge area of the shell ridges, which sits at a relatively high altitude, undergoes frequent dry/wet cycles but has been experiencing an increasing frequency of severe drought stress events. The soil in the beach ridge area of the shell ridges has a high porosity, a low capacity to store and retain water, and low capillary action. These properties have led to a lack of freshwater resources in the shell sand habitat. Soil water content is a key factor that limits plant growth and community succession in shell sand habitats 2 . Periploca sepium Bunge is a dominant shrub species on the shell ridges of the Yellow River Delta. Owing to its exceptional capability to disrupt wind, hold sand, resist drought, and retain water, P. sepium Bunge has become a primary tree species studied for its ability to persist under drought conditions 3,4 . During drought stress, the soil water content substantial affects plants in terms of their physiological, biochemical, and photosynthesis characteristics as well as their water-related physiological processes. Dominant plants can withstand stress by regulating their metabolic rhythm 2,5 . Therefore, studying the physiological and ecological characteristics of the dominant shrubs in shell sand on the Yellow River Delta under drought stress is of great importance for elucidating the mechanism through which xerophytic vegetation in the beach ridge area of shell ridges adapts to drought stress.

Results
Light responses according to the physiological photosynthesis indices. As demonstrated in Fig. 1a, the rectangular hyperbolic correction model had a high goodness of fit with the P n -PAR response process in the leaves of P. sepium Bunge from the four shell sand habitats whose water conditions differed. The fitted correlation coefficients for the control (CK), T1, T2, and T3 groups were 0.998, 0.999, 0.995, and 0.932, respectively. As the drought stress increased, the P n of P. sepium Bunge decreased significantly (P < 0.05). When the PAR was 1200 µmol•m −2 • s −1 , the P n values of P. sepium Bunge in the T1, T2, and T3 groups were 23.50%, 42.44%, and 95.12% lower than that in the CK, respectively. Figure 1b shows that as the drought stress increased, the G s of the leaves decreased significantly (P < 0.05). When the PAR was 1200 µmol•m −2 • s −1 , the G s of the leaves in the CK was the greatest, at 243.56 mmol•m −2 •s −1 , and the G s values in the T1, T2, and T3 groups were 41.42%, 30.88%, and 6.89% of that in the CK, respectively.   groups were 40.10%, 61.44%, and 119.88% greater than that in the CK, respectively. The effects of drought stress on the ETR and Φ PSII of the leaves exhibited the same pattern. As the drought stress increased, there was a significant decrease in the leaf ETR (P < 0.05). The ETRs in the leaves in the T1, T2, and T3 groups were 27.84%, 32.14%, and 57.65% lower than that in the CK, respectively. physiological and biochemical characteristics. As the drought stress increased, the SOD activity first increased but then decreased, and the POD activity increased (Fig. 3a). The POD activity in the T1, T2, and T3 groups was 2.80, 3.01, and 4.19 times that in the CK, respectively. As the drought stress increased, there was a significant increase in both the proline (Pro) and soluble sugar (Ss) (two types of osmotic adjustment substances) contents in the leaves (P < 0.05) (Fig. 3b). The MDA contents in the T1, T2, and T3 groups were 1.50, 2.02, and 1.80 times that in the CK, respectively (Fig. 3c).

Changes in stem sap flow.
There was a significant difference in the daily stem sap flow rate dynamics in P. sepium Bunge under the different water treatments (P < 0.05) (Fig. 4a). Under mild drought stress, the stem sap flow activity in P. sepium Bunge occurred primarily between 4:30 and 19:00. Under the other water conditions, stem sap flow activity occurred throughout the day. As the drought stress increased, the stem sap flow rate decreased overall. The daily cumulative stem sap flow was the greatest in the CK, followed by the T1, T2, and T3 groups (Fig. 4b). The daily cumulative stem sap flow rates in the T1, T2, and T3 groups were 23.98%, 33.3%, and 63.94% lower than that (205.21 g·d −1 ) in the CK, respectively. The daily cumulative stem sap flow in the T3 group increased slowly and linearly. The slope of the daily cumulative stem sap flow curve for the T2 group decreased over time, and the curve exhibited an inverted "Z" shape; the daily cumulative stem sap flow exhibited a typical "S"-shaped pattern in the CK and T1 groups.
changes in the T r and WUE. There was a significant difference in the T r values of the P. sepium Bunge leaves under different drought stress levels (P < 0.05). The T r was the greatest in the CK, followed by the T1, T2, and T3 groups (Fig. 5). When the PAR was 1200 µmol•m −2 •s −1 , the T r results for the T1, T2, and T3 groups were 50.82%, 37.16%, and 10.11% of that (3.66 mmol•m −2 •s −1 ) in the CK. As the drought stress increased, the water-use efficiency (WUE) of the leaves first increased but then decreased significantly (P < 0.05). The WUE remained at a relatively high level in the T1 and T2 groups but was the lowest in the T3 group. When the PAR was 1200 µmol•m −2 •s −1 , the WUE values of the leaves in the T1 and T2 groups were 37.87% and 38.09% greater than that in the CK, respectively, whereas the WUE of the leaves in the T3 group was 59.84% lower than that in the CK.

Discussion
There was a significant difference in the extent of the decreases in the P n and G s between the P. sepium leaves in the shell sand habitats under different drought stress conditions. The decrease in the photosynthesis efficiency of the leaves is related to their stomatal state 12 . Research has shown a significant decrease in both the P n and G s of Populus × euramericana 'Neva' and Phaseolus vulgaris L., a difference in the stomatal limitation mechanism between these two species, and a significant positive correlation between the P n and G s under various water stress conditions 22,23 . These findings are similar to those of the present study. The leaves of P. sepium Bunge adapt to drought stress in shell sand via stomatal regulation. This mechanism varies significantly with the drought stress level. As the drought stress level increased, the P n of the leaves of the four-year-old P. sepium Bunge seedlings decreased significantly, whereas the P n of the leaves of one-year-old P. sepium Bunge seedlings first increased but then decreased 24 . This may be because young P. sepium Bunge seedlings are more sensitive to drought stress. The response of the low-intensity light-use efficiency (LUE) of the leaves to drought stress is nonsignificant. However, the low-intensity LUE of the leaves can decrease significantly under severe drought stress. Under natural conditions, the AQY of plant leaves ranges from 0.03 to 0.05 mol·mol −125 . In the present study, the AQY of the leaves from the shell sand habitats was significantly lower than this range. This result suggests that the photosynthesis capability of P. sepium Bunge is relatively weak under low-intensity light. The difference between the LSP and LCP of the leaves under severe drought stress was only 40.10% of that in the CK. This finding indicates that drought stress can significantly reduce the LUE and light ecological amplitude of leaves during photosynthesis. Consistent changes in the R d and AOY of the P. sepium Bunge leaves under drought stress were detected. This finding suggests that leaves can adapt to drought stress by decreasing respiration. Waterlogging and drought stress resulted in a  www.nature.com/scientificreports www.nature.com/scientificreports/ significant decrease in the AQY, P max , and LSP of leaves of T. chinensis growing in a shell sand habitat, and there was a significant increase in their LCP 12 . These findings are similar to those of the present study. The AQY and R d of Aralia elata first increased but then decreased as drought stress increased, and both of its AQY-drought stress and R d -drought stress curves exhibited an "n" shape 26 . While the AQY and R d of A. elata 26 start to decrease at a lower water stress level than did those of P. sepium Bunge on shell sand ridges, the AQY and R d of A. elata decrease more sharply as the water stress level increases. Thus, P. sepium Bunge exhibits a greater LUE under drought stress than under no drought stress.
A significant decrease in the F v /F m ratios of leaves of P. sepium Bunge growing in the shell sand habitat under severe drought stress was detected. This finding suggests that under these conditions, there was a decrease in the photochemical quantum yield of PSII in the electron transport chain in the leaves, and the photosynthesis process suffered from severe photoinhibition 21,27 . Under severe drought stress, the leaves suffered from severe photoinhibition, and there was a significant decrease in the number of electrons for the light reactions or CO 2 assimilation and a decrease in photosynthesis efficiency 21,28 . A significant positive correlation was detected between the changes in the ETR and Φ PSII 28 . Under mild drought stress (relative water content (RWC) > 40%), there was a relatively insignificant decrease in the NPQ of the leaves, which had a relatively low heat dissipation capacity. Drought stress at this level did not cause damage to the photosynthesis organs of the leaves. Under mild drought stress, the leaf photosynthesis activity was relatively insignificantly inhibited and could be restored by measures such as rehydration 12,29 . Under severe drought stress (RWC ≤ 40%), there was a significant decrease in the NPQ of the leaves and a significant increase in their heat dissipation. Additionally, under severe drought stress, the photosynthesis ability of the leaves could be easily inhibited and could not be restored 12,29 . The P n of the leaves was extremely and significantly negatively correlated with the NPQ values but significantly positively correlated with Ф PSII , ETR, and R d ( Table 2). This finding suggests that orderly electron transport along the electron transport chain is the determinant for photosynthesis output and that light energy dissipation is a key protective mechanism of photosynthesis under stress conditions 27,28 . In addition, respiration also regulates normal photosynthesis output to a certain extent. A decrease in the soil water content led to a significant decrease in the F v /F m ratio, Φ PSII , and ETR of four wolfberry species and a significant increase in their NPQ 30 . However, compared with that of P. sepium Bunge, the decrease in the F v /F m ratio of these four wolfberry species was more pronounced. Silva et al. reported that, under various water stress conditions, there was no significant difference in the F v /F m ratio of Jatropha curcas, a significant decrease in its ETR, and a significant increase in its NPQ 31 . Cecilia et al. 32 reported that, increasing in excessive electron transport and non-photochemical quenching was shown on Celtis australis due to photosynthesis and gas exchange processes inhibited, and C. australis showed conserved water utilization behavior and was sensitive to drought stress 32 . These changes are consistent with those observed in this study. The effects of drought stress on the physiological photosynthesis processes of various plant species across various habitats can evidently differ relatively significantly.
Under moderate drought stress, the SOD activity in the P. sepium Bunge leaves was relatively highly adaptable and adjustable. However, as the drought stress level increased, the ability to alter the SOD activity in the leaves was inhibited. As a result, there was a decrease in SOD activity. In comparison, there was a significant increase in POD activity. P. sepium Bunge evidently eliminates drought stress-induced peroxide stress primarily by increasing POD activity 31,33 . Studies have indicated that there is a significant increase in SOD, POD, and catalase activities   34 and peanut seedlings 35 under various drought conditions. This finding suggests that, under drought stress, changes in the activities of plant protective enzymes are closely related to the habitat and drought resistance of the plants. As drought stress increases, the activities of protective antioxidant enzymes in plants increase or first increase but then decrease 33,36 . However, under severe drought stress, the ability of plants to alter their enzyme activities is limited, and consequently, their enzyme activities mostly decrease 33 37 .
In the present study, as the drought stress increased, the Pro content in the leaves increased to a significantly greater extent than did the Pro content. This finding suggests that P. sepium Bunge is more capable of improving its osmotic adjustability by altering the Pro content than the Ss content. Under drought stress, plants can reduce their osmotic potential by accumulating solutes such as Pro and Ss to improve the osmotic adjustability of their cells to ensure proper osmotic balance inside and outside their cells [37][38][39] . In the present study, in the leaves, the Pro content was extremely significantly negatively correlated with the P n ; significantly negatively correlated with the ETR, Ф PSII , and R d ; and significantly positively correlated with the NPQ. Additionally, in the leaves, the Ss content was significantly negatively correlated with the P n and R d ; significantly positively correlated with the NPQ; and extremely significantly positively correlated with the Pro and Ss contents ( Table 2). These findings suggest that, by secreting osmotic adjustment substances, P. sepium Bunge can adjust the osmotic pressure on both sides of the plasma membrane, and ensuring orderly physiological activity 37,38 , thereby affecting its photosynthetic electron transport ability. In the present study, as the drought stress increased, there was an increase in the contents of osmotic adjustment substances (e.g., Pro and Ss) in the P. sepium Bunge leaves as well as an increase in the activities of their antioxidant enzymes (SOD and POD). This trend may be the cause of the decrease in the MDA content 40,41 , which led to a reduction in cell membrane lipid peroxidation 36 . Abida et al. 42 reported that two different varieties of maize resisted effectively drought stress by increasing antioxidant enzyme activity, reducing cell membrane lipid peroxidation efficiency and accumulating osmotic adjustive substances. As a result, the toxic effects of cell membrane lipid peroxidation diminished under severe drought stress.
As the drought stress level increased, the time at which the stem sap flow in P. sepium Bunge peaked was delayed or advanced, and the sap flow activity diminished during the day and increased during the night. To adapt to drought stress, P. sepium Bunge improved its WUE by reducing its transpiration intensity and duration. Furthermore, P. sepium Bunge exhibited relatively high adaptive plasticity to the dry soil environment. Bhusal et al. reported a significant decrease in instantaneous sap flow rate under drought stress 43 . Similarly, Liu et al. reported, that under drought stress, there was a decrease in the instantaneous sap flow rate of several dominant species in a Quercus liaotungensis community in a hilly loess region, a significant decrease in their maximum sap flow rates, and an advance or delay in the time at which the sap flow process started 44 . Clearly, drought stress obstructs the water transport process in plants and reduces their usable water content. In the present study, the P n of the P. sepium Bunge leaves was extremely significantly positively correlated with the cumulative sap flow; the T r was extremely significantly positively correlated with G s , significantly positively correlated with the cumulative sap flow, and significantly negatively correlated with the NPQ; and G s was significantly positively correlated with the Ф PSII , ETR, and cumulative sap flow ( Table 2). These results suggest that transpiration is closely related to the state of stomata. The photosynthetic electron transport and light energy dissipation processes also affected the T r and G s while protecting the photosynthesis organs, thereby intervening in the water-related physiological processes of P. sepium Bunge. Mengesha et al. 45 reported that decreasing net photosynthetic rate and being limited gas exchange were shown on two types of wheat with different genotypes, which result in declining water transport and evapotranspiration rate, but improving significantly water use efficiency. Water transport and use in plants also affects their normal photosynthesis output. Certain relationships evidently exist between photosynthesis and water-related physiological processes in plants.
Under mild and moderate drought stress (RWC > 40%), there was a significant decrease in the T r and P n of the leaves (P < 0.05) (the P n decreased to a smaller extent than did the T r ) and an increase in their WUE. As the drought stress increased (RWC ≤ 40%), the T r and P n decreased to extremely low levels and were not significantly different (P > 0.05). Additionally, as the drought stress increased (RWC ≤ 40%), there was also a significant decrease in WUE (P < 0.05). Therefore, increasing its peroxidase activity and secreting more osmotic adjustment, the leaves of P. sepium Bunge exhibited a relatively high WUE. A decrease in G s reduces the water consumption of plants due to transpiration, obstructs water transport and use processes, and inhibits water-related physiological activities. A decrease in WUE leads to a decrease in the amount of raw materials for photosynthesis, thereby reducing photosynthesis. Liang et al. reported a significant decrease in the T r , P n , and WUE of Populus × euramericana 'Neva' as the drought stress level increased 22 . Similarly, Xia et al. reported that, as the drought stress level increased, the T r and P n of T. chinensis in a shell sand habitat decreased, whereas its WUE first increased but then decreased 46 . These findings are consistent with those of this study. Drought stress results in a significant decrease in the WUE and transpiration-related water consumption of plants, thereby inhibiting their water-related physiological processes. This trend ensures that plants will be able to use limited water efficiently to survive drought stress.
conclusions Drought stress on shell sand can significantly affect the photosynthesis, water consumption, and physiological and biochemical processes of P. sepium Bunge. As the degree of drought in shell sand area increased, a significant decrease in the P n , T r , G s , and light energy for photosynthesis of P. sepium Bunge leaves occurred, and an increased amount of light energy was dissipated as heat. Moderate drought stress significantly improved the WUE, and severe drought stress led to severe photoinhibition and obstructed the water transport of P. sepium Bunge. Drought stress leads to a decrease in the amount of water available for other physiological activities. The decrease in photosynthesis is an important photoprotection mechanism that enables P. sepium Bunge to adapt to drought stress on shell sand. Moreover, P. sepium Bunge counteracts lipid peroxidation by increasing the contents of osmotic adjustment substances (Pro and Ss) and activities of antioxidant enzymes (SOD and POD) under severe drought stress on shell sand and adapts to drought stress through interactions and interregulatory activity between photosynthesis, water-related physiological activities, and physiological and biochemical processes. By reducing its photosynthesis output and WUE, P. sepium Bunge can reduce its photosynthetic electron transport efficiency, increase heat dissipation, increase both the activities of its protective enzymes and the content of its lipid peroxide molecules, and improve the osmotic adjustability of its cell membranes to withstand and adapt to drought stress on shell sand.

Experimental materials.
Four-year-old seedlings of P. sepium Bunge, a dominant shrub species that grows on the shell ridges of the Yellow River Delta, were selected as experimental materials. The P. sepium Bunge seedlings were uniformly cut to prepare samples that had a height of 1.18 ± 0.13 m, a rhizome thickness of 1.25 ± 0.06 cm and a root depth of 11.2 ± 0.3 cm. The seedling samples were subsequently trimmed such that the canopy size was 0.4 m (east-west) × 0.4 m (south-north).
Experimental design. This experiment was conducted in the Scientific Research Greenhouse (SRG) of the Shandong Key Laboratory of Eco-Environmental Science for the Yellow River Delta. Shell sand was collected from the area of the P. sepium Bunge community in the Binzhou National Shell Ridge and Wetland Nature Reserve. The collected shell sand was sieved through a 2.0 mm mesh and was subsequently transferred to a pot (diameter of the top, 40 cm; height, 50 cm). The basic chemical and physical properties of the shell sand were as follows: field capacity, 18.31%; bulk density, 1.29 g·cm −3 ; particle size, 0.2-2.0 mm; pH, 7.40; and salt content, 0.1%-0.4%. The primary environmental conditions of the SRG included the following: illumination intensity, 80%-84% of the natural external light intensity relative air humidity, 45%-65%; temperature, 20-35 °C; and atmospheric CO 2 concentration, 350-370 μmol·mol −1 .
The seedlings were planted in a total of 24 pots (three seedlings per pot) in the SRG on March 13, 2018, after which they were subjected to normal plant management practices for approximately 90 d. On June 15, 2018, the seedlings received one of four water treatments. The seedlings were randomly divided into four groups, with each group comprising six pots of seedlings. One of the four groups was selected as a CK (RWC of the shell sand: 77.72%). The remaining three groups were subjected to mild (RWC of the shell sand: 58.16%), moderate (RWC of the shell sand: 42.98%), or severe (RWC of the shell sand: 32.39%) drought stress treatments, which are referred to as the T1, T2, and T3 groups, respectively. The pots in the CK, T1, and T2 groups were watered once every 3 days with 4, 2, and 1 L of water, respectively. The pots in the T3 group were not watered. The RWC of the shell sand was calculated as the ratio of the water content of the shell sand to its field capacity, by weight. The RWC of the shell sand was monitored and controlled by weighing the pot and drying. A tray was placed underneath each pot to prevent water loss, thereby controlling the designed water treatment and ensuring experimental accuracy. Each group of seedlings was allowed to adapt to the water treatment for approximately 30 d. Afterward, from July 16 to July 23, 2018, the P. sepium Bunge seedlings were measured to determine the physiological photosynthesis indices and the chlorophyll fluorescence parameters of their leaves as well as their physiological and biochemical indices and stem sap flow parameters.

Measurement of physiological photosynthesis indices.
The experiment was conducted from 9:00 to 11:30 on a sunny day. Three pots of P. sepium Bunge seedlings were randomly selected from each group as experimental samples. Three healthy, mature leaves were then selected from the middle-upper section of the seedlings in each pot. An LI-6400XT portable photosynthesis system (LI-COR, Inc., Lincoln, NE, USA) was used to measure the light response of the plants. A standard 6400-02B red and blue light-source leaf chamber was used. Photosynthetically active radiation (PAR) was subsequently applied at 0, 40, 80, 100, 150, 200, 400, 600, 800, 1000, and 1200 µmol•m −2 • s −1 , for a total of 11 illumination intensities. During the measurements, the PAR was decreased from the greatest level to the lowest level, with each level maintained for 120 s; these measurements were repeated three times. The instrument automatically recorded and calculated various physiological photosynthesis indices, such as the P n , transpiration rate (T r ), G s , and WUE (WUE = P n /T r ).

Measurements of chlorophyll fluorescence parameters.
An FMS-2 portable pulse-modulated fluorometer (Hansatech Instruments, Ltd., UK) was used to measure the chlorophyll fluorescence parameters of the P. sepium Bunge leaf samples as well as their photosynthesis indices. After the leaves were dark adapted for 30 min, the F 0 and the maximum fluorescence (F m ) of each leaf were measured. Subsequently, after each leaf was dark adapted to natural light for 50 min, it was measured to determine its steady-state and maximum fluorescence under light-adapted conditions (F s and F m ′). Additionally, the potential photochemical efficiency (F v /F m = (F m − F 0 )/F m ), actual photochemical efficiency (Φ PSII = (F m ′ − F s )/F m ′), noncyclic photosynthetic electron transport rate (ETR = Φ PSII × PAR × 0.5 × 0.84, where 0.5 is the partition coefficient between the two photosystems and 0.84 is the proportion of the amount of light absorbed by the leaf to the amount of incident light), and NPQ ((F m − F m ′)/F m ′) were calculated 47 .

Measurements of physiological and biochemical indices.
To measure physiological and biochemical indices, five to seven leaves were collected from the same section of each seedling from which the leaves were collected to measure and analyze the photosynthesis parameters 48 . The SOD activity was measured via nitro blue tetrazolium photoreduction, and the POD activity was measured via guaiacol colorimetry. In addition, the Pro content was determined via ninhydrin colorimetry, and the Ss content was determined according to the anthrone coloration method. Last, the MDA content was measured via thiobarbituric acid colorimetry. (2020) 10:9567 | https://doi.org/10.1038/s41598-020-66717-z www.nature.com/scientificreports www.nature.com/scientificreports/ Measurements of stem sap flow parameters. A Flow 32 heat-balance stem sap flow measurement system (Dynamax, Houston, TX, USA) was used to measure the P. sepium Bunge seedling stem sap flow rate and daily stem sap flow continuously. On the basis of the stem thickness for each P. sepium Bunge seedling selected for testing, an SGA-5 sensor with a suitable diameter (5-7 mm) was selected. The stem sap flow measurement system was installed according to the standard instructions provided by Dynamax. A Delta-T Logger data acquisition system was used to collect and record the instantaneous stem sap flow rates automatically at 30 min intervals.

Statistical analysis.
A rectangular hyperbolic correction model 49 was used to simulate the PAR-P n photosynthesis light-response process. Equation (1) shows the expression of the model as follows: Here, P n is the net photosynthesis rate; I is the PAR; I c is the LCP; and α, β, and γ are three coefficients unrelated to the light intensity (α is the initial slope of the light-response curve when the PAR is 0, which is considered the AQY, and β and γ are biological important photoinhibition and photosaturation terms, respectively).
On the basis of Eq. (1), three photosynthesis parameters, namely, R d , LSP, and P max , can be derived, as shown in Eqs. (2)-(4). where P n (I m ) is equal to P max . The data were processed and plotted in Microsoft Excel 2016. Correlation analysis, analysis of variance, and multiple comparisons of the data were performed by SPSS 19.0.