Effects of drought stress and rehydration on physiological and biochemical of four oak species in China

Background: Quercus fabri Hance, Quercus serrata Thunb, Quercus acutissima Carruth, and Quercus variabilis BL are four Chinese oak species commonly used for forestation. In recent years, with the global warming caused by the greenhouse effect, seedlings in mountainous areas after afforestation often suffer seasonal drought stress, which seriously affects their survival and growth. In order to ensure the survival of seedlings, we need to select oak species with strong drought resistance. Therefore, we rst need to understand the differences in drought resistance of the four oak tree species at the seedling stage, and comprehensively evaluate their drought resistance capabilities by studying the changes in the physiological and biochemical characteristics of the seedlings under continuous drought and rehydration conditions. Methods: The four oak seedlings were divided into drought-rewatering treatment group and normal watering control group. For the seedlings of the drought-rewatering treatment group, drought stress lasting 31 days was used, and then re-watering and recovering for 5 days. The water parameters, osmotic adjustment substance content, antioxidant enzyme activity and photosynthesis parameters of the seedlings in the two groups were measured every 5 days. Principal component analysis, correlation analysis and membership function were used to analyze the physiological and biochemical characteristics of the seedlings of the four oak in two groups. Results and conclusions: Compared with the control group, the relative water content, water potential, net photosynthetic rate, transpiration rate, and stomatal conductance levels of the four oaks all showed a downward trend under continuous drought stress, and showed an upward trend after rehydration. The soluble protein, soluble sugar, proline, peroxidase, superoxide dismutase and catalase content of the four oaks increased rst and then decreased under drought stress, and then increased after rehydration. The content of glycine betaine and malondialdehyde continued to increase, and gradually decreased after rehydration. The weight of each index was calculated by PCA, and then the comprehensive evaluation of each index was carried out through the membership function method. The drought resistance levels of the four oak species were as follows: Q. serrata > Q. fabri > Q. >


Introduction
With rising global temperatures and frequent extreme phenomena, 64% of the world's land is already under drought stress, which has seriously affected the normal growth and development of many plants and crops (Li and Geng 2013). Half of China's land is arid or semiarid, and drought is a serious problem in northwest and southwest China (Piao et al. 2010). Moreover, even southern regions with abundant rainfall are often affected by seasonal drought, which is a major factor that limits the normal growth and development of plants (Zhai et al. 2005).
When plants are under drought stress, the plant water balance is destroyed, and normal physiological processes are affected (Zhu 2002). The relative water content (RWC) and water potential of plant leaves decrease (Farooq et al. 2009). In addition, plants close their stomata to reduce water loss, which can limit the supply of CO 2 and lead to reduced photosynthesis Reactive oxygen species (ROS) metabolism is the primary response of plants to stress (Ben-Gal et al. 2009). Drought stress can break the dynamic balance of production and elimination of ROS in plants, and excessive accumulation of ROS causes oxidative stress (Gechev et al. 2006). To protect themselves from oxidative stress, plants produce antioxidant enzymes and nonenzymatic substances that scavenge ROS (Huang et al. 2019). The antioxidant enzymes mainly include peroxidase (POD), superoxide dismutase (SOD), ascorbate oxidase (APX) and catalase (CAT), and the nonenzymatic antioxidant substances mainly include reduced glutathione (GSH), reduced ascorbic acid (AsA) and carotenoids (Ahmad et al. 2010). Many studies have shown that the responses of different protective enzymes to drought stress vary (Ashraf 2009, Anjum et al. 2011). The protective enzymes that play a dominant role in plants may also differ between periods of stress (Seki et al. 2007). In addition, plants also reduce their osmotic potential by accumulating osmoregulatory substances, such as soluble sugars (SS), proline (Pro) and glycine betaine (GB), to obtain the water needed to maintain normal growth under drought stress, thus alleviating the damage caused by drought stress (Chaves et al. 2009).
The leaf is an important organ that is in close contact with the external environment and is the main organ for photosynthesis and transpiration (Koyro 2006). The leaf tissue structure and physiological characteristics are susceptible to environmental factors and are the best re ection of a plant's adaptability to its environment (Xu et al. 2015). The morphological structure (leaf area, leaf thickness and stomatal density) and physiological properties (RWC, antioxidant system and osmoregulatory substances) of leaves are often used as important indicators in drought studies (Medina et al. 1990, Kruskopf and Flynn 2006). At present, many studies have been carried out on the physiological and biochemical responses and drought resistance mechanisms of plants under drought stress (Toscano et al. 2016, Bandeppa et al. 2019. A pot experiment with arti cially controlled water was performed to simulate the natural drought process in the soil and to observe the plant resistance response to different levels of drought stress (Rivest et al. 2013). Pot experiments more objectively re ect the response characteristics of plants to natural drought stress. There are many methods to evaluate the drought resistance of plants, such as the a liation function method, hierarchical evaluation, grey correlation analysis and principal component analysis , Du et al. 2020). Among them, the a liation function method can make a more comprehensive and integrated evaluation based on the determination of multiple indicators, avoiding the limitations and inaccuracies of a single evaluation indicator, and is the most widely used in drought resistance evaluation (Sio-Se Mardeh et al. 2006). In recent years, there has been a preference for combining multiple evaluation methods to improve the accuracy of the results (Lu et al. 2011). Seedlings are sensitive to drought stress at the beginning of their development, which not only threatens their survival but also affects their later growth, biomass formation and ability to overwinter (Ditmarová et al. 2010). Therefore, by studying the physiological and biochemical characteristics of oak seedling leaves under drought stress and combining various evaluation methods, we can evaluate drought resistance with greater reliability.
The SOD activity was determined by the nitrogen blue tetrazole (NBT) method (G0101W Assay Kit, Suzhou Geruisi, China). The CAT activity was determined by sodium thiosulfate titration (G0105W Assay Kit, Suzhou Geruisi, China). The POD activity was determined by the peroxidase chromatographic method (G0107W Assay Kit, Suzhou Geruisi, China). The MDA content was determined by thiobarbituric acid colorimetry (G0109W Assay Kit, Suzhou Geruisi, China).

Photosynthetic parameters
After the experiment, the photosynthetic physiological parameters of the leaves were measured at 9:30 − 11:30 AM every ve days. The upper, healthy, sunny-side leaves of the seedlings were observed using a LI-COR 6400 system (Li-COR, Lincoln, NE, USA). The air ow rate was set at 500 µmol/s, and the data were read after 2 min of stabilisation to record the net photosynthetic rate (Pn), transpiration rate (Tr) and stomatal conductance (Gs).

Statistical analyses
All the data were tested for normality and Leneve's test. The data were recorded and processed using Excel, and statistical analysis was carried out using SPSS 19.0. One-way analysis of variance (ANOVA) was used, and differences between treatments were analysed using Duncan's multiple comparisons (p < 0.05). GraphPad Prism 9 and JASP 0.14.1 were used for plotting data. For principal component analysis (PCA), the number of principal components is usually selected as the variable containing more than 80% information; that is, the cumulative contribution rate of characteristic roots should be greater than 80%.
The drought tolerance of four oak species was evaluated by the membership function method. When the indicator was positively correlated with drought resistance, the formula was U(X ij )=(X ij -X jmin )/(X jmax -X jmin ). When the indicator was negatively correlated with drought resistance, the formula was U(X ij ) = 1-(X ij -X jmin )/(X jmax -X jmin ). The overall evaluation value was calculated as X i = ΣU(p m ×X ij )/n. In the formula, X ij is the measured value of an index of a certain oak tree. X jmax and X jmin are the minimum and maximum values of the index. p m is the weight coe cient of the m-th principal component, and n is the number of indicators (Xiu et al. 2019). These values were expressed as the mean ± standard error (SE) of three replicate samples.

Soil moisture content
The soil moisture content of the four oak control groups ranged from 28.0-31.9%. With continuous drought stress, the soil moisture content of the treatment group decreased gradually, and the range of the decrease was similar. The soil moisture content reached the lowest point of 3.1% ~ 3.5% at day 31 of drought stress and returned to the control level after rewatering (Fig. 1).

Water potential and RWC
The RWC and water potential of the leaves of all four oak species tended to decrease under sustained drought stress compared to those of the control group.The values of the treated groups were not signi cantly different from those of the control group on the rst day,but all decreased and reached the lowest values, which were signi cantly lower than those of the control group, on day 31 (P<0.05).After rewatering, the leaf water potential increased rapidly, and the recovery rate of the leaf water potential was faster than that of RWC on day 36.The RWC and water potential of Q. fabri leaves decreased from 81.13% and −1.52 MPa (day 1) to 41.10% and −2.84 MPa (day 31), with decrease rates of 49.47% and 87.25%, respectively. After rewatering, the RWC and water potential recovered to 73.02% and −1.67 MPa (day 36), respectively, but did not recover to the normal level. The RWC and water potential of Q. serrata leaves decreased from 81.67% and −1. .16%, respectively. After rewatering, the RWC and water potential recovered to 73.67% and −1.57 MPa (day 36), respectively.The leaf water potential of Q. serrata, Q. acutissima and Q. variabilis returned to normal levels, but the RWC remained lower than the normal level (Fig.2). Fig.2 Effects of drought stress and rewatering on relative water content and water potential of four oak species (a, relative water content; b, water potential). DR-F: drought-rewatering group of Q. fabri; CK-F: normal watering group of Q. fabri; DR-S: drought-rewatering group of Q. serrata; CK-S: normal watering groupof Q. serrata; DR-A: droughtrewatering group of Q. acutissima; CK-A: normal watering groupof Q. acutissima; DR-V: droughtrewatering group of Q. variabilis; CK-V: normal watering group of Q. variabilis. Different lowercase letters indicate signi cant (P<0.05) differences among the four oak species subjected to the same treatment time.

Osmotic regulating substances
Under continuous drought stress, the levels of soluble protein, SS and Pro in the four oak species rst increased, then decreased, and then increased after rehydration. However, GB level increased continuously and decreased after rehydration. The SS levels of Q. fabri, Q. acutissima and Q. variabilis (Day 6) peaked earlier than that of Q. serrata (day 11). Compared with the control group, the SS level of the four oak species increased by 1.50, 1.28, 1.15 and 1.11 times, respectively. Then, the SS levels of the four oak species were the lowest on day 31 and were signi cantly lower (P < 0.05) than that of the control group, decreasing by 43.74%, 48.75%, 42.52% and 51.60%, respectively. After rewatering, only Q. fabri returned to a normal level, while the levels in the other three oak species were signi cantly lower than normal (P < 0.05) (Fig. 3a). There were no signi cant differences in the SS levels when comparing between the four oak species and the control group at the early stage (day 1 and day 6) (P > 0.05). On day 26, the SS level in each of the four oak species reached a peak, which was 1.46, 1.48, 1.35 and 1.37 times higher than that of the control group, respectively. Then, the levels decreased and were the lowest on day 31, decreasing by 15.44%, 12.79%, 10.62% and 14.22%, respectively. After rewatering, the SS level of Q. serrata was signi cantly lower than the normal level (P < 0.05), but the other three oak species returned to normal levels (Fig. 3b). The Pro level in the four oak species reached a peak on day 26 and was signi cantly higher than that in the control group (P < 0.05), increasing by 1.89, 1.66, 1.98 and 1.87 times, respectively. Then, the Pro level decreased but was still signi cantly increased compared with the level in the control group (day 31) (P < 0.05) and returned to the normal level after rehydration (Fig. 3c). The GB level in the four oak species showed a continuous increasing trend, which was signi cantly higher than that in the control group (P < 0.05), increasing by 1.56, 2.05, 1.96 and 2.29 times, respectively. After rehydration, the GB level decreased sharply but was still signi cantly higher than that in the control group (P < 0.05), increasing by 1.28, 1.50, 1.65 and 1.98 times, respectively (Fig. 3d).

Antioxidant enzyme activities and MDA
Under continuous drought stress, the activities of POD, SOD and CAT of the four oak species rst increased, then decreased, and then increased after rehydration. The MDA level increased continuously and decreased after rehydration. The POD activity of Q. fabri, Q. acutissima and Q. variabilis (day 16) reached a peak later than that of Q. serrata (day 11), which was 1.36, 1.11, 1.34 and 1.18 times higher than that of the control group, respectively. The POD activity levels of the four oak species were the lowest on day 31 and were signi cantly lower than that of the control group (P < 0.05). The POD activity levels were decreased by 18.34%, 32.21%, 34.74% and 36.65%, respectively, and did not return to normal levels after rehydration (Fig. 4a). The SOD activity level of the four oak species was signi cantly different from that of the control group at day 5 (P < 0.05) and peaked at day 21. The SOD activity levels were 2.14, 2.33, 1.93 and 2.28 times higher than that of the control group, respectively. Then, the SOD activity levels were the lowest at day 31 and were signi cantly lower than that in the control group (P < 0.05). The SOD activity levels were reduced by 32.55%, 29.26%, 37.10% and 20.73%, respectively, and did not recover to normal levels after rehydration (Fig. 4b). The CAT activity levels of Q. fabri, Q. acutissima and Q. variabilis (day 21) reached a peak later than that of Q. serrata (day 16). The CAT activity levels were 1.54, 1.48, 1.52 and 1.39 times higher than that of the control group, respectively. The CAT activity levels of the four oak species were the lowest on day 31 and were signi cantly lower than that of the control group (P < 0.05). The CAT activity levels decreased by 25.13%, 20.61%, 8.12% and 11.43%, respectively. After rehydration, only Q. fabri returned to the normal CAT activity level, but the levels of the other three oaks were signi cantly lower than normal (P < 0.05) (Fig. 4c). The MDA levels of the four oak species showed a continuous growth trend, and there were no signi cant differences in the levels at the early stage (day 1 and day 6) when comparing between the four oak species and the control group (P > 0.05). The MDA levels peaked on day 31 and were signi cantly higher than that in the control group (P < 0.05), increasing by 2.71, 2.15, 2.23 and 2.06 times, respectively. The MDA levels decreased sharply after rehydration but were still signi cantly higher than that in the control group (P < 0.05), increasing by 1.38, 1.41, 1.36 and 1.45 times, respectively (Fig. 4d).

Photosynthetic parameters
Compared with the control group, the Pn, Gs and Tr values for the four oak species showed a decreasing trend under continuous drought stress. The values were the lowest on day 31 and rose rapidly after rehydration. As drought stress was prolonged, the Pn values of the four oak species decreased continuously. The Pn of Q. serrata at day 6 was signi cantly different from that of the control group (P < 0.05). The Pn values of the four oak species decreased to the lowest level on day 31 and were signi cantly lower than that of the control group (P < 0.05). The Pn values decreased by 72.10%, 63.29%, 63.67% and 60.16%, respectively. After rehydrating, the Pn of Q. serrata was signi cantly lower than normal (P < 0.05), but the Pn values of the other three oak species returned to normal (Fig. 4a). There were no signi cant differences in the Gs values when comparing between the four oak species and the control group at the early stage (day 1 and day 6) (P > 0.05). The Gs values of the four oak species were the lowest on day 31 and were signi cantly lower than that of the control group (P < 0.05). The Gs values decreased by 71.14%, 57.45%, 50.03% and 45.24%, respectively, but returned to normal after rehydration (Fig. 4b). There were signi cant differences in the Tr values when comparing between the four oak species and the control group at day 5 (P < 0.05). The Tr values were the lowest on day 31 and were signi cantly lower than that of the control group (P < 0.05). The Tr values decreased by 62.37%, 60.08%, 61.28% and 60.72%, respectively, and did not return to normal after rehydration (Fig. 4C).

Multivariate statistical analysis
The eigenvalues and contribution rates of principal components were the basis for selecting principal components. The 13 physiological and biochemical indexes of the leaves from the 4 oak species were analysed by PCA. Two principal components with eigenvalues greater than 1 were obtained, and their contribution rates were 64.24% and 19.70%, respectively. The cumulative contribution rate was 83.94%, and most of the information on the original characteristics was retained (Table 1). Therefore, the rst two principal components could be selected as the important principal components of the drought resistance of the four oak species. The factors with higher loading capacity in the rst principal component were Ψw, RWC, Pro, SP, GB, MDA, Pn, GS and Tr, which were mainly related to leaf water status, osmoregulatory substances and photosynthesis. The second principal component was mainly related to antioxidant enzymes. The treatments of the 4 oaks under drought stress and during rehydration were completely separated. In addition, Q. acutissima, Q. serrata and Q. variabilis highly overlapped under drought stress and during rehydration. In contrast, Q. fabri varied greatly with the other three species (Fig. 5). To further understand the relationship between leaf water status, osmotic regulatory substances, photosynthesis and antioxidant enzymes, Pearson correlation analysis was used to analyse the data. The results showed that Pn, GS and Tr positively correlated with each other, negatively correlated with MDA, GB and Pro levels and positively correlated with SP, Ψw and RWC. Ψw and RWC levels were signi cantly negatively correlated with Pro, GB and MDA levels. MDA levels were signi cantly positively correlated with GB and Pro levels and negatively correlated with SP levels and POD activities. Pro levels were signi cantly negatively correlated with SP levels and positively correlated with SS and GB levels. In addition, SOD, POD and CAT activities were signi cantly positively correlated (Fig. 6). To comprehensively evaluate the drought resistance of the four oak species, 13 physiological and biochemical indexes of the four oak species under continuous drought stress and rehydration conditions were analysed by membership functions ( Table 2). The membership function value of each index was calculated according to a formula. The degree of correlation between different indexes and drought resistance was different. The arithmetic mean of the membership value of each index did not fully re ect the level of drought resistance in the four oak species. Therefore, the proportion of eigenvalues corresponding to each principal component to the sum of the total eigenvalues of the extracted principal components was taken as the weight. The comprehensive evaluation value of the four oak species was calculated. Higher comprehensive evaluation values were positively correlated with stronger drought-resistance ability. The results showed that the order of drought resistance of the four oak species were as follows: Q. serrata > Q. fabri > Q. variabilis > Q. acutissima.  It is generally believed that smaller decreases in the RWC and water potential of plant leaves correlate with higher water-retaining ability and stronger adaptability of the leaves to drought stress (Schonfeld et al. 1988). In our research, under continuous drought stress, the relative water content and water potential of the four oak species decreased to different degrees, which indicated that the four oak species could absorb more water by reducing Ψw to resist drought stress. Among them, the RWC of Q. serrata declined the least and recovered the fastest after rehydration. Q. variabilis showed the fastest decrease in RWC but the smallest decrease in Ψw. The Ψw of Q. fabri decreased the most and recovered the slowest after rehydration. These results indicate that Q. serrata can maintain a higher water balance than the other three oak species. Q. serrata leaves may be more leathery than the leaves from the other species, which would be bene cial for reducing water transpiration during drought stress. The results herein were similar to those from studies of soybean (Du et al. 2020) and Cyamopsis tetragonoloba (L.) Taub. (Upreti et al. 2021). This study also found that Ψw and RWC were highly signi cantly positively correlated with Pn, Gs and Tr. This indicates that under continuous drought stress, the water potential of plant leaves decreases with a constant decrease in water content, leading to a decrease in photosynthesis.

Osmotic regulating substances
Plants under drought stress are often damaged by osmotic stress. When cells lose water, it can decrease turgor pressure and lead to death (Xiong and Zhu 2002). Under a certain degree of drought stress, plants We also found that the SOD activity increased signi cantly at day 6. The results showed that SOD was the rst to respond to the early stages of drought stress, and then POD and CAT enzymes responded.
Compared with the control group, there were no signi cant differences in the MDA levels of the four oak species at day 6. The results indicated that the three enzymes could effectively eliminate the production of reactive oxygen species in the early stages of drought stress. However, in the late stages of drought stress, the MDA levels in the four oak species were signi cantly higher than that of the control group (P < 0.05). It is possible that the four oak species suffered from long-term drought stress, which damaged the antioxidant enzyme system to different degrees. ROS overaccumulation may have destroyed the cell membranes, which may have caused MDA to accumulate in large quantities.

Photosynthetic parameters
Drought stress can also affect the photosynthetic physiology of plant leaves, which is mainly re ected by the chlorophyll content, Pn and Tr . In this study, the Pn, Gs and Tr values of the four oak species showed a continuous decreasing trend as drought stress was prolonged and were signi cantly lower than those of the control group at day 31 (P < 0.05). This result shows that photosynthesis was inhibited in the four oak species under continuous drought stress, which might be due to the production and accumulation of a large number of reactive oxygen species in the leaves.
Accumulation of ROS results in damage to the mesophyll cell membrane, destruction of photosynthetic enzyme activity and a decrease in photosynthesis. In addition, at day 31, the Pn, Gs and Tr values of Q. fabri were the lowest among the four oak species, the Pn and Tr values of Q. serrata were the highest, and the Gs value of Q. variabilis was the highest. The results showed that the resistance of Q. fabri photosystems was the worst, and Q. serrata was the strongest. It may be that the osmoregulation (especially the GB content) of Q. serrata is stronger than that of the other three species, and the decrease in RWC is the smallest, which greatly reduces the damage to the photosystem. After rehydration, there was no signi cant difference in Gs, but the Tr value was still signi cantly lower than that of the control group. With the exception of Q. fabr, there were no signi cant differences in the Gs values when comparing the four oak species with the control group. The results showed that the recovery ability of the Q. serrata photosystem was weaker than that of the other three oak species. Q. serrata may have weak repair ability after severe drought, which was indicated by the osmotic regulatory substance recovery results.

Multivariate statistical analysis
The four oak species are distributed in different regions in China. In the long-term adaptation and evolution of the geographical environment, each species has a different level of drought resistance. The main differences are in the water parameters, osmotic regulation, antioxidant enzyme activity and photosynthesis. The changes in the physiological and biochemical indexes of the four oak species were complex and varied under drought stress, which indicated that different oak species have different ways to adapt to drought stress. The PCA results showed that the boundary between the drought rehydration group and the control group was clear, and the PC1 components of the four oak species showed the same trend during the drought rehydration period, indicating similar adaptation to drought stress.
Correlation analysis showed that there were complex and close relationships between the leaf water status, osmotic regulatory substances, antioxidant enzyme activities and photosynthesis, and they cooperated with each other to resist drought stress. Our results showed that there were signi cant correlations between osmotic regulatory substances and photosynthesis and between water parameters and antioxidant enzyme activities. In addition, there were signi cant correlations between photosynthesis and water parameters and between antioxidant enzyme activities and MDA levels. The sensitivity of different oak species to drought stress is different, so the response time is also different, and the trees may be in different response stages simultaneously (Fang and Xiong 2015

Conclusion
This study simulated continuous drought stress that occurs under natural conditions. Changes in water parameters, osmotic regulation substance content, antioxidant enzyme activity and photosynthesis were observed dynamically. It was discovered that four oak species showed adaptive changes to drought stress and resisted early stages of drought stress by increasing the level of osmotic regulation substances and regulating the activity of antioxidant enzymes. However, with prolonged drought stress, all four oak species reached a tolerance limit. Water content, osmotic regulatory substance content, antioxidant enzyme activity and photosynthetic parameters all decreased to different degrees. The recovery ability of the four oak species also differed after rehydration. Multivariate statistical analysis showed that there were complex and close relationships between the leaf water status, osmotic regulatory substances, antioxidant enzyme activities and photosynthesis, and they cooperated with each other to resist drought stress. PCA and the membership function method were used to analyse 13 physiological and biochemical indexes under continuous drought stress and rehydration. The drought resistance levels of the four oak species were as follows: Q. serrata > Q. fabri > Q. variabilis > Q. acutissima.

Availability of data and materials
Data sharing is not applicable to this article as no datasets were generated or analyzed during the current study.
Ethics approval and consent to participate Not applicable.

Consent for publication
Not applicable.  Effects of drought stress and rewatering on relative water content and water potential of four oak species (a, relative water content; b, water potential). DR-F: drought-rewatering group of Q. fabri; CK-F: normal watering group of Q. fabri; DR-S: drought-rewatering group of Q. serrata; CK-S: normal watering group of Q. serrata; DR-A: drought-rewatering group of Q. acutissima; CK-A: normal watering group of Q. acutissima; DR-V: drought-rewatering group of Q. variabilis; CK-V: normal watering group of Q. variabilis.
Different lowercase letters indicate signi cant (P < 0.05) differences among the four oak species subjected to the same treatment time.   variabilis. Different lowercase letters indicate signi cant (P < 0.05) differences among the four oak species subjected to the same treatment time.  stomatal conductance. * P < 0.05, ** P < 0.01, and *** P < 0.001.