Introduction

Quercus liaotungensis is an essential deciduous tree in warm-temperate deciduous broad-leaved forests. It is a suitable species for establishing mixed broadleaf-conifer forests in the middle and low elevations of North China1. The regeneration of the Q. liaotungensis forest depends on sprouting. However, the growth of sprout seedling degraded, and the soil seed bank of Q. liaotungensis was lost due to predation, insect moth, and physiological death2,3. Therefore, the regeneration and establishment of Q. liaotungensis forest mainly rely on artificial breeding, so provenance selection is essential. Provenance selection will help to discover excellent germplasm resources of Q. liaotungensis for regeneration and establishment. More recent attention has focused on providing resource distribution, phenotypic traits, and the economic value of Q. liaotungensis seeds from different provenances4,5. A key aspect of sustainable forestry is to cultivate plants with efficient photosynthetic system6,7. Photosystem II (PSII) is driven by light energy and provides assimilation power to synthesize plant carbohydrates8. It is generally believed that PSII, an important component of the photosynthetic system, is most sensitive to environmental changes9,10. Currently, there is a lack of work for evaluating differences in the PSII activity between Q. liaotungensis provenances.

Chlorophyll a fluorescence is a potential and simple tool to analyze the performance of Photosystem II (PSII). This technology is able to discriminate differences due to the influence of environmental factors on the photochemical activity of PSII, such as temperature11,12, water13,14, salinity15,16, light17,18, and insect feeding19,20. Different parameters have been used to describe the photochemical activity of PSII. Butler and Kitajima first constructed the maximum quantum yield of primary PSII photochemistry (Fv/Fm) based on the characteristics of the OJIP curve in 197521. However, this parameter was shown to be nonspecific22 and often insensitive14. In 1999, Strasser et al. developed performance index of overall photochemistry (PIABS) by using three independent OJIP curve parameters (φPo—maximum quantum yield of primary PSII photochemistry, ψEo—efficiency with which a PSII trapped electron is transferred from \({\text{Q}}_{\text{A}}^{-}\) to PQ: and RC/ABS—the density of PSII reaction centers)23. PIABS could reflect the state of plant photosynthetic apparatus more accurately than Fv/Fm24, whereas PIABS was related only to the electron transport to the PQ pool25. Tsimilli-Michael and Strasser proposed the performance index of overall photochemistry (PItotal)26. PItotal calculates by PIABS and δRo (the efficiency of the electron from PQH2 is transferred to final PSI acceptors), which can fully describe the photochemical activity of the linear photosynthetic electron transfer chain25. Data from numerous studies indicated that PItotal decreased significantly in response to high PAR dose, high ambient temperature, low soil water content, K+-deficiency stress, Mg-deficiency stress, shade stress, and heat stress27,28,29,30,31. Simultaneously, PItotal increased significantly during the light-induced plasticity of plant growth32 and was considered the most salinity-sensitive parameter33,34. Widespread plant species often show extensive variation in morphological and growth characteristics as well as the substantial difference in stress resistance due to different individual selection pressures35. Existing studies have shown significant differences in the PSII photochemical activity from different provenances36,37. However, the mechanism and the trend of change for the interspecies differences of PSII photochemical activity are still unclear. Investigating PSII interspecies differences can help implement the OJIP-test to provenance trials, and obtain excellent germplasm resources with high and stable photosynthetic inorganic carbon assimilation ability.

In this paper, 2-year-old Q. liaotungensis seedlings coming from 10 different provenances were used as test materials. The difference in PItotal between provenances from internal factors (φPo, ψEo, δRo, and RC/ABS) and external factors (temperature, precipitation, solar radiation, wind speed, transpiration pressure, and soil properties) in provenances was analyzed. Several studies have investigated that dry climate threatened Q. liaotungensis forest growth38,39. In order to verify the experimental results, two separate experiments were conducted in semi-arid and sub-humid distribution zones of Q. liaotungensis. We tested three different hypotheses: (1) There are differences among the Q. liaotungensis provenances of PItotal. (2) These changes can be explained by the parameters used to calculate PItotal and the environmental factors of provenance. (3) The difference trend may be predicted. Our expected results were to 1) evaluate the degree of PSII interspecific difference of Q. liaotungensis. (2) Understand the patterns of variation observed.

Materials and methods

Test sites

The test was carried out simultaneously in Yangqu and Sanjiao test sites. The Yangqu test site is located in Yangqu County, Shanxi, China (38.0981° N, 112.7346° E, and 961 m above sea level), is a warm temperate semi-arid continental monsoon climate. The annual mean temperature is 8.5 °C, the annual rainfall is 430.4 mm, the annual average frost-free period is 164 days. The Sanjiao test site is located in Fushan County, Shanxi, China (35.9649° N, 112.0766° E, and 1197 m above sea level), is a warm temperate semi-humid continental monsoon climate. The annual mean temperature is 9.1 °C, the annual rainfall is 569.6 mm, the annual average frost-free period is 191 days. The soil type of the two test sites is loam.

Test materials

In the autumn of 2017, the seeds of Q. liaotungensis were collected from 10 provenances in the species' natural distribution range. The names, geographical locations, climate conditions, and soil types of all provenances were summarized in Fig. 1 and Table 1. For each provenance, seeds were collected from plus trees with plant spacing was greater than 50 m in the middle-aged forests. The fully mixed seeds were used as the provenance seed. Malathion was used to kill insects in seeds. The seeds were sown in the field in the autumn of the same year. The experiment had a randomized block design with three replications. At least 8000 seeds per replication were sown, and each seed was sowed at a distance of about 20 cm. Field management followed the normal agricultural practices in the two test sites. Two years later, vigorous seedlings from germinated seeds were evaluated for relative chlorophyll content (SPAD) and chlorophyll a fluorescence. All parameters were measured in August 2019.

Figure 1
figure 1

Geographical locations of Quercus liaotungensis provenances and test sites. The map is created using the Google Map Import plug-in of Origin 2018 (https://www.originlab.com/fileExchange/details.aspx?fid=344). Coordinate reference system is WGS-84. PS Pingsong provenance, LY Luyashan provenance, KC Kangcheng provenance, HH Henghe provenance, LK Lingkongshan provenance, DS Dongshan provenance, SJ Sanjiao provenance (and Sanjiao test site), SM Shanmen provenance, ZW Zhenwushan provenance, DT Diantou provenance, YQ Yangqu test site.

Full size image
Table 1 The names, geographical locations, climate conditions, and soil types of Quercus liaotungensis provenances.
Full size table

Parameter measurement methods

The mature fully expanded and unshaded leaves of 30 vigorous seedlings from each provenance were used to monitor the chlorophyll a fluorescence transient and SPAD. SPAD and chlorophyll a fluorescence transient were sequentially measured on the same leaf. One measurement per seedling was taken, resulting in 30 measurements per provenance.

The SPAD value of leaf was measured with a chlorophyll meter (SPAD-502Plus, KONICA MINOLTA, Japan) and chlorophyll a fluorescence transient was measured with a PAM-fluorometer (FluorPen FP110, Photon Systems Instruments, Czech Republic). The chlorophyll a fluorescence transient measurement was made from leaves that were dark adapted for 20 min using leaf clips. The mesophyll was illuminated with saturated blue light (2 100 μmol m−2 s−1) for 1 s, and the fluorescence signals at intervals of 10 μs (before 600 μs), 100 μs (between 600 μs and 15 ms), 1 ms (between 15 and 100 ms), and 10 ms (after 100 ms) were recorded. The OJIP-derived parameters (Table 2) were calculated with reference to Stirbet et al. 25 and Holland et al. 40.

Table 2 The OJIP-derived parameters.
Full size table

Environmental factors (11 temperature factors, 8 precipitation factors, 12 monthly solar radiation intensity factors, 12 monthly average wind speed factors, and 12 monthly transpiration pressure factors) were retrieved from the WorldClim database (https://www.worldclim.org/data/worldclim21.html) with a resolution of 10 arc-minutes based on the geographical location of the provenance. 32 soil factors were retrieved from the Harmonized World Soil Database (HWSD v1.2) provided by the National Tibetan Plateau Data Center. Eliminating collinearity was executed when the Pearson correlation coefficient between environmental factors was greater than 0.9. 22 factors were finally retained (Table 3).

Table 3 Environmental factors for the geographical location of Quercus liaotungensis provenances after eliminating collinearity.
Full size table

Data processing

ArcGIS 10.2 (ESRI, USA) was used to retrieve environmental factors. Calculations for the differences of SPAD and chlorophyll a fluorescence parameter between provenances were done in Excel 2016 (Microsoft Office, USA) and R version 4.0.2 (R Core Team). Data were tested for normality and homogeneity of variance before the analysis and were log-transformed prior to analysis to provide normality when required. Pearson correlation coefficient was calculated using the R. The curves of chlorophyll a fluorescence transient were normalized to the interval [0, 1]. Differences between means of provenances were evaluated with the LSD test at 0.05 and 0.01. The coefficient of variation (CV = (Standard deviation/mean) × 100%) was calculated in order to investigate the degree of dispersion of φPo, ψEo, δRo, and RC/ABS. The relationship between PItotal (the mean of the provenances) and provenance’s environmental factor was examined using the Maximal Information Coefficient correlation analysis (MIC). MIC was calculated using the “Minerva” of R package. The relationship between PItotal and the warmest quarter precipitation of provenance was determined using regression analyses through the Origin 2018 (Origin Lab Corporation, USA). The interactions among the provenance, test site, and test site:provenance were analyzed by the Generalized Linear Mixed Model (GLMM), calculated using the “nlme” of R package.

Ethics approval

The experimental research and field studies on plants, including the collection of plant material, complied with relevant institutional, national, and international guidelines and legislation. The appropriate permissions and licenses for the collection of plant or seed specimens were obtained for the study. Prof. Xinping Li identify the plant material. The plant materials were deposited in the herbarium of Shanxi Academy of Forestry and Grassland Sciences, China. The voucher ID of the specimen was from LKY-LZY01 to LKY-LZY10.

Results

Relative chlorophyll content analysis

Differences in SPAD were not significant between the 10 provenances of Q. liaotungensis in the two test sites, respectively. SPAD in Yangqu test site ranged from 43.70 to 37.02 with a mean of 40.66 ± 2.22, and SPAD in Sanjiao test site ranged from 47.23 to 42.77 with a mean of 44.86 ± 1.16. It indicated that the material basis for light energy absorption of the tested leaves were similar in the same test site (Fig. 2).

Figure 2
figure 2

Relative chlorophyll content analysis of Quercus liaotungensis seedlings between provenances. YQ represents the Yangqu test site, and SJ represents the Sanjiao test site. Different lowercase letters indicate a significant difference at the 0.05 level, and uppercase letters indicate a significant difference at the 0.01 level.

Full size image

Chlorophyll a fluorescence transient analysis

Both normalized chlorophyll a fluorescence transient showed the typical OJIP shape in two test sites. Differences in each step's relative fluorescence intensity between provenances were not evident, indicating that the electron transfer status between PSII and PSI of all provenances were similar (Fig. 3).

Figure 3
figure 3

Normalized chlorophyll a fluorescence transient of Q. liaotungensis provenances. The O, J, I, and P steps are marked in the figure, where: O is for origin, J and I are for the intermediary fluorescence levels at 2 ms and 30 ms, and P is for the peak20. YQ represents the Yangqu test site, and SJ represents the Sanjiao test site.

Full size image

Chlorophyll a fluorescence parameter analysis

PItotal reflected the photochemical activity of the electron transfer chain from the PSII oxygen-evolving complex to the final electron acceptors of PSI. Figure 4 showed that there were significant differences in PItotal between provenances in each test site. In Yangqu test site, SJ provenance had the highest PItotal, with an average value of 4.07, reaching 2.21-fold higher (p < 0.01) than that of the lowest PItotal (KC provenance) in the same test site. In Sanjiao test site, the average PItotal value of LK provenance was 5.21, reaching 1.83 -fold higher (p < 0.05) than that of the lowest PItotal (ZW provenance) in the same test site.

Figure 4
figure 4

PItotal analysis of Quercus liaotungensis seedlings between provenances. P-value indicates the differential significance of PItotal between provenances in the same test site. YQ represents the Yangqu test site, and SJ represents the Sanjiao test site. Different lowercase letters indicate a significant difference at the 0.05 level, and uppercase letters indicate a significant difference at the 0.01 level.

Full size image

PItotal is based on four independent parameters: φPo, ψEo, δRo, and RC/ABS. φPo, ψEo, and δRo represent the probabilities that electron is transferred to the QA, PQ, and final PSI acceptor side. RC/ABS represents the density of PSII reaction centers. Table 4 showed the analysis of φPo, ψEo, δRo, and RC/ABS of Q. liaotungensis seedlings between provenances. The variation analysis (using CV) showed the data-sparse in descending order were δRo, ψEo, and φPo between provenances. The differences of δRo and ψEo between provenances were highly significant (p < 0.01) in both test sites. The CV of the PSII reaction center was 7.53 (in Yangqu test site) and 7.61 (in Sanjiao test site), implying the data-sparse of RC/ABS between provenances were prominent in both test sites. These results suggest that the difference in probability of electron transfer from PSII to PSI and the number of active reaction centers in PSII were gradually increased between provenances, causing a significant difference in PItotal between provenances in the same test site.

Table 4 φPo, ψEo, δRo and RC/ABS analysis of Quercus liaotungensis seedlings between provenances.
Full size table

Environmental factor analysis

MIC correlation analysis is used to find linear and nonlinear correlations between variables. The MIC value ranges from 0 to 1. The closer the MIC value to 1.0, the higher the correlation of the variable is. As shown in Fig. 5, the warmest quarter precipitation (B18) had the highest MIC value in both test sites. The results showed that the warmest quarter precipitation (B18) in provenance was closely related to the PSII photochemical activity of Q. liaotungensis seedlings. Additionally, the solar radiation intensity in July (R7) and the annual precipitation (B12) also significantly affected the PSII photochemical activity in the Sanjiao test site.

Figure 5
figure 5

MIC correlation coefficient of Quercus liaotungensis seedlings between PItotal and environmental factor. YQ represents the Yangqu test site, and SJ represents the Sanjiao test site. See Table 3 for the meaning of environmental factor code. Black bars indicate the same environmental factor with the highest MIC between test sites.

Full size image

The regression analysis was used to predict the precipitation of the warmest quarter (B18) of provenance and PItotal in the two test sites. The results showed that the precipitation of the warmest quarter (B18) and PItotal could be fitted by the Sine function in the two test sites (R2 was 0.90 in Yangqu test site and was 0.77 in Sanjiao test site). However, the peak location of the fitted curves differed between the two test sites. The fitting curve of Yangqu test site had a peak of around 310 mm, while in Sanjiao test site, the peak was around 350 mm. Compared with the fitted curve of Yangqu, the peak location of the Sanjiao’s fit curve was shifted to the right. Interestingly, Fig. 6 showed that the precipitation of Sanjiao was shifted to the right compared with Yangqu.

Figure 6
figure 6

Regression analysis of Quercus liaotungensis seedlings between PItotal and precipitation of the warmest quarter in provenance. YQ represents the Yangqu test site, and SJ represents the Sanjiao test site. The dotted lines indicate the precipitation of the warmest quarter in two test sites. R2 of Sine fit of YQ is 0.90, P < 0.01, and R2 of Sine fit of SJ is 0.77, P < 0.01.

Full size image

Inter-location PItotal difference analysis

Comparison of SPAD, chlorophyll a fluorescence parameter, and environmental factor

Table 5 showed the influence of provenance and test site on SPAD and chlorophyll a fluorescence parameter. Test site had significant effect on parameters and test site:provenance had significant combined effect on φPo and RC/ABS, while provenance had no effect. For PItotal, only test site had a significant influence (P = 0.04).

Table 5 The interaction among provenances, test site, and test site:provenance analysis of Quercus liaotungensis seedlings.
Full size table

The average values of SPAD, ψEo, and δRo of all provenances were significantly higher, and the average values of φPo and RC/ABS were lower in Sanjiao test site compared with Yangqu test site (Fig. 7). These results indicated that SPAD, ψEo, and δRo were the key parameters which caused the difference in PItotal between the two test sites.

Figure 7
figure 7

Ratio of the parameter mean value of Sanjiao test site to Yangqu test site. *The parameter is observed significant differences between Yangqu test site and Sanjiao test site at the 0.05 level. **Significant differences at the 0.01 level. The dotted lines indicate the 1.0×.

Full size image

In Sanjiao test site, the annual precipitation (B12) and the coldest quarter precipitation (B19) were 5.6 and 3.7 times higher than those in Yangqu test site, respectively. Additionally, the topsoil salinity (S7) and the topsoil sand fraction (S2) in Yangqu test site were 3.2 times higher than those in Sanjiao test site (Fig. 8). The result demonstrated precipitation, topsoil salinity, and sand fraction were the main environmental factors that cause differences in PItotal between the two test sites.

Figure 8
figure 8

Ratio of Sanjiao test site to Yangqu test site for each environmental factor. The dotted lines indicate the 1.0×.

Full size image

Discussion

In the two test sites, there were significant differences in PItotal between the 10 provenances of Q. liaotungensis when they were grown in the same environment. This finding was similar to the results of the research on interspecies differences in PIABS of European beech (Fagus sylvatica) by Kurjak, et al. 36, reflecting the difference in PSII photochemistry activity between provenances. PIABS is related to only the process of electron transport to the PQ pool25. In this study, PItotal was used to evaluate the state of the linear photosynthetic electron transfer chain between PSII and PSI. Therefore, the results of this study could be interpreted as the total PSII photochemical activity of Q. liaotungensis of different geographic provenance.

From the perspective of electron transfer from PSII to PSI, the CV of δRo was the highest, and the P-value of δRo was 0.001 (Table 4). This phenomenon indicated that electron transfer between PQ and PSI electron acceptors of Q. liaotungensis appeared to be most sensitive to environmental changes. On the other hand, this phenomenon might indicate that the PSI acceptor side of Q. liaotungensis's structure was relatively unstable and had higher diversity between different provenances. The data-sparse of RC/ABS between provenances of Q. liaotungensis was also rather large. Several reports have confirmed that environmental stress can cause decreased PSII active reaction centers41,42,43. In this study, the normalized OJIP curves of provenances did not show obvious signs of stress. Hence, a possible explanation for the data-sparse of RC/ABS is that some provenances encountered mild stress in the test site due to ecological distance (between the original field site and the common garden).

The study results showed that the precipitation of the warmest quarter of provenance location was closely related to the PSII photochemical activity of Q. liaotungensis in both test sites. Yangqu and Sanjiao test sites were both characterized by a temperate continental monsoon climate, with high temperatures and plentiful rainfall in the summer. Related studies have found that the net photosynthetic rate of Q. liaotungensis reached maximum in mid-July44. These results suggest that the PSII photochemistry activity of different Q. liaotungensis provenances was closely related to the water supply during the growing season. This speculation is similar to the conclusion of Wu et al. that sufficient water supply during the growing season can significantly increase the carbon assimilation rate of plant45.

The precipitation—PItotal fitted curves first raised and then fell in the two test sites. Studies have confirmed that drought stress46,47 and flooding stress48,49 will significantly reduce the photochemical activity of PSII. Ecological distance might lead to excess or lack of water for some provenance. As a result, the curves first raised and then fell. Furthermore, the peak location of the Sanjiao’s fitted curve was shifted to the right compared with the Yangqu’s curve. GLMM analysis revealed a significant effect of the test site on PItotal (Table 4) and the Yangqu test site with lower precipitation in the warmest quarter than the Sanjiao test site (Fig. 8). Based on this hypothesis, precipitation can be considered as the main cause of this phenomenon, and more provenances may suffer from mild drought stress in the Yangqu test site.

PItotal and the peak location of the precipitation—PItotal fitted curves were varied between the two test sites. As shown in Fig. 7, the internal factors causing this phenomenon were the difference in leaf chlorophyll content and the probability of electron transfer to PQ and final PSI acceptors. Because photosynthetic pigment play an essential role in absorption and transfer of light energy50, and φPo, ψEo, δRo, and RC/ABS were key indicators of the total PSII photochemical activity. However, other internal factors such as morphological and physiological traits may also affect the PSII photochemical activity between provenances37. Precipitation and soil conductivity have significant effects on PSII photochemical activity of plant leaves51. Figure 8 also showed that the external factors resulting in this phenomenon are the different precipitation and upper soil conductivity in the two test sites.

Conclusion

We have shown that the differences in PItotal of Q. liaotungensis seedlings between provenances do exist, and the different trends of PItotal can be fitted by the Sine function in the two test sites. These results helped to screen the provenance of Q. liaotungensis with high and stable photosynthetic efficiency by using PItotal. In the future, it is necessary to eliminate genetic differences within provenance, increase the number of provenances, and incorporate more environmental factors to improve the accuracy of the OJIP-test.