Responses of Picea meyeri to Climatic Factors Revealed by Tree Ring Isotopes and Water Use Efficiency on Luya Mountain of North-Central China

Abstract

Along with tree ring width, carbon isotopes are also good proxies for climate change. Water use efficiency (WUE) can be calculated more quickly and accurately based on carbon isotopes. In this study, according to the principle of dendroclimatology, the sequence of δ¹³C and WUE of tree rings of Picea meyeri are built. Pearson correlation analysis and multiple regression analysis are used to explore the response of carbon stable isotopes of Picea meyeri to climate change, which revealed the relationship between δ¹³C of Picea meyeri and climatic factors. Based on δ¹³C, we calculated the WUE of Picea meyeri and analyzed its response to climate change. The main conclusions are as follows: (1) The δ¹³C of Picea meyeri decreases year-by-year from 1957 to 2020, in the range from −23.89‰~−21.67‰, and the average value is −22.67‰. The water use efficiency of Picea meyeri increases in the range from 17.26~61.31, with an average of 39.45. (2) The δ¹³C of Picea meyeri is negatively correlated with temperature, which has the highest correlation with the temperature of the growing season (c5–c9), and its coefficient is higher than that of the mean temperature of each month. (3) There is a significant positive correlation between WUE sequence and temperature. Meanwhile, due to the effect of precipitation and temperature, the Picea meyeri is subject to drought stress to some extent. Above all, temperature is the main climatic factor affecting the δ¹³C and WUE of Picea meyeri on Luya mountain.

1. Introduction

Picea meyeri is a unique and special tree species in China, which is distributed mostly throughout north-central China. The Taiga Forest of Picea meyeri, Larix principis-rupprechtii and Pinus tabulaeformis are distributed throughout the area of Luya Mountain above 2000 m [1]. At present, studies on tree-ring δ¹³C are concentrated in northwest China and southwest China, especially the upper limit of the forest in the arid region of northwest China [2,3,4,5,6]. There are few studies on tree-ring δ¹³C in north-central China. The previous research on the response of Picea meyeri to climate change on Luya mountain are based on the width of the tree ring and earlywood and latewood [1,7,8,9,10].

Compared with tree ring width, the stable isotopes of tree rings are less affected by age change and sensitivity to climate change [11]. Only some tree species have “juvenile effect”, which δ¹³C increases with age in the first 20 years [2], and no “juvenile effect” was observed in δ¹⁸O, δD [12]. Earlier studies mostly used whole wood directly for stable carbon isotope ratio (δ¹³C) determination. Some results show that the expression of climate information by δ¹³C is affected by different wood components and the change in proportion [13,14,15]. However, the C in α-cellulose remained relatively stable after fractionation in the carbon fixation process of photosynthesis, and it also has a single composition, stable structure and more climate information, which gradually replaced whole wood for isotope determination [16,17].

Earlier studies focused on atmospheric CO₂ concentration variation calculated by δ¹³C [18], and subsequent studies mostly focused on the response of δ¹³C to climate change [19]. Carbon isotope studies in China developed later than other countries. Li et al. obtained the evidence of the change in atmospheric CO₂ concentration from δ¹³C of Pinus tabuliformis in the 1990s, which was the first study on tree ring carbon isotopes in China [20]. Since then, studies on stable carbon isotopes of tree rings in China have developed rapidly. Chen et al. analyzed the effect of atmospheric CO₂ on tree rings δ¹³C of Picea schrenkiana [3], Picea crassifolia [4] and Sabina przewalskii [5] and their response to climate change. Xu et al. [21] found that Pinus koraiensis δ¹³C in northeast China showed a significant negative correlation with low cloud cover during the period of strong light from May to July. Meanwhile, power spectrum analysis showed that the El Niño–Southern Oscillation and East Asian monsoon had a positive periodic effect on Pinus koraiensis δ¹³C [21]. Liu et al. [22] concluded that in the Greater Khingan Mountains region, temperature and humidity from July–September have a stronger influence on Larix gmelinii δ¹³C than that of Pinus sylvestris.

The δ¹³C can also be used to study the WUE of trees, which provides evidence for the water–carbon coupling of forest under climate change [23]. In ecology, WUE is expressed as the ratio of net primary productivity (NPP) to total evapotranspiration (ET) of an ecosystem [24]. WUE can be used as an important parameter to characterize the photosynthesis and transpiration of plants, which has attracted much attention in the academic world. However, traditional methods to calculate WUE, such as harvest, photosynthesizer and microweather method, are limited due to unreliable data and more destruction to vegetation [25]. Using stable carbon isotopes to calculate WUE directly shows the quantitative relationship between photosynthetic intensity and stomatal conductance [26]. Most of the current studies on WUE use leaf δ¹³C and a few use tree ring δ¹³C, which have proved that the increase in temperature in the growing season would lead to the increase in WUE, which would promote the growth of trees to some extent [27,28]. However, the WUE of Populus euphratica is negatively correlated with temperature in arid areas [29]. Some shrubs, such as Encelia frutescens and Ambrosia salsola, had significant effects on annual precipitation and CO₂ concentration [30].

In the context of climate warming, it is very important to study the response of trees to climate change using multiple indicators. The study of the Picea meyeri response to climate change on Luya Mountain by using the stable isotopes of tree rings is yet to be reported. δ¹³C of tree rings is different from that of tree ring width, which contains more accurate and precise climate information. δ¹³C can be used to calculate WUE accurately and quickly. The δ¹³C and WUE of Picea meyeri are measured and analyzed in this paper, which can be used to explore the response to climate change. This study has certain reference value for studying forest dynamics and forest management in north-central China under the global warming trend.

2. Materials and Methods

2.1. Overview of the Study Area

Luya Mountain belongs to Guancen Mountain, located in the northern section of Lüliang Mountain. Its main peak, Heyeping, is 2783 m above sea level, and its geographical coordinates are 38°35′ N~38°45′ N, 111°50′ E~112°05′ E. It is the source of Fen River (Figure 1), which is the second largest tributary of the Yellow River. Luya Mountain is located on the precipitation line, at which 400 mm of annual precipitation falls, belonging to the warm temperate zone and the boundary of the semi-humid and semi-arid zones. The mountain is northeast–southwest-oriented, which becomes the natural barrier to the East Asian summer monsoon. Luya Mountain is high and steep, and the vertical zonality of vegetation and soil is obvious. The main tree species are Pinus tabuliformis, Larix principis-rupprechtii and Picea meyeri. The distribution of forest and shrub is over 1850 m and Picea meyeri is the construction species. There is subalpine shrub meadow (2450–2772 m), cold-temperature coniferous forest (1750–2600 m), mixed coniferous broad-leaved forest (1700–1850 m), deciduous broad-leaved forest (1350–1700 m) and a forest steppe (1300–1500 m) from top to bottom. The soil types are subalpine meadow zonal soils, brown zonal soils, leached cinnamon soils, cinnamon soils and chestnut cinnamon soils [9,31,32].

According to the observation data of the Wuzhai (38°43′ N, 111°35′ E, 1396.7 m) and Kelan meteorological stations (38°55′ N, 111°49′ E, 1401.0 m), which are the closest to Luya Mountain, we use the average data of the two stations to represent the meteorological data of the study area (38°44′ N, 111°50′ E, 2501 m) (Figure 2). This region has formed a typical temperate continental monsoon climate, with obvious temperature differences between winter and summer, and the precipitation is mostly concentrated in summer. Here, summer is hot and rainy, and winter is cold and dry. Meteorological disasters mainly include frost, and spring drought is more serious in some areas [33,34].

2.2. Sample Collection and Treatment

For sample collection, 52 Picea meyeri core samples were collected from 26 trees in the area of Luya Mountain (38°44′ N, 111°50′ E, 2501 m); two cores from the diameter at breast height (1.3 m) and the base (0.3 m) of each tree. After natural drying, fixing and fitting, LinTab™6 (accuracy: 0.001 mm) was used to measure the ring width, and COFECHA [35] was used for cross-dating. The correlation coefficients of seven trees reached the requirement of the carbon isotope study of tree rings that are used for stable carbon isotope analysis [16]. The sample sizes and the standard chronologies are shown in Figure 3.

The average diameter at breast height of Picea meyeri is 47.77 cm, and the average tree age is 77 years. In order to avoid the “juvenile effect” [2], we eliminated the first 15 years. Therefore, the valid years of the isotope sample are from 1957–2020. With reference to the methods of previous studies stated below [36,37,38], tree ring δ¹³C was analyzed and determined. Traditional isotope extraction methods need to peel every ring into a centrifuge tube before the experiment, which takes a significant amount of time and effort. With an optimized method from Kagawa et al. [36], experimental efficiency was improved. The specific steps are showed in Appendix A.

After the treatment, the appropriate number of samples were weighed and sent into the combustion tube (1000 °C) (Thermo 253 Plus mass spectrometer, Thermo Fisher Scientific, Bremen, Germany) by automatic sampler under the special ultra-clean tin cup tightly wrapped. The samples were fully burned in the combustion tube filled with chromium trioxide catalyst under the oxygen-rich condition, and the C element was converted to CO₂. The samples were detected the ¹³C and ¹²C ratio by isotope mass spectrometer, and compared with the international standard material (Pee Dee Belnite or PDB) [16] to calculate the δ¹³C value of the sample; the experimental systematic error was less than 0.2‰.

2.3. Calculation of WUE

According to the carbon isotope fractionation model equation established by Farquhar et al. [39,40]:

$${\Delta }^{13}\mathrm{C}=\mathrm{a}+\left(\mathrm{b}-\mathrm{a}\right)\left({\mathrm{C}}_{\mathrm{i}}{/\mathrm{C}}_{\mathrm{a}}\right)$$

(1)

$${\Delta }^{13}\mathrm{C}=\frac{{\mathsf{\delta }}^{13}{\mathrm{C}}_{\mathrm{a}}{-\mathsf{\delta }}^{13}{\mathrm{C}}_{\mathrm{p}}}{1+\raisebox{1ex}{${\mathsf{\delta }}^{13}{\mathrm{C}}_{\mathrm{p}}$}\!\left/ \!\raisebox{-1ex}{$1000$}\right.}$$

(2)

In Equations (1) and (2), ∆¹³C represents the discrimination value of stable carbon isotopes in tree rings; C_i and C_a is the tree intercellular and atmospheric CO₂, respectively; a is the diffusion fractionation coefficient for CO₂, which is approximately 4.4‰ [12]; b is the carboxylation fractionation coefficient, which is approximately 27‰ [12]; thus, C_i can be expressed as:

$${\mathrm{C}}_{\mathrm{i}}={\mathrm{C}}_{\mathrm{a}}\left[\left({\mathsf{\delta }}^{13}{\mathrm{C}}_{\mathrm{p}}{-\mathsf{\delta }}^{13}{\mathrm{C}}_{\mathrm{a}}+\mathrm{a}\right)/\left(\mathrm{a}-\mathrm{b}\right)\right]$$

(3)

WUE can be expressed as the ratio of the assimilation rate (A) to stomatal conductance (g), and WUE can be calculated using Equation (4):

$$\mathrm{WUE}=\frac{\mathrm{A}}{\mathrm{g}}=\frac{{\mathrm{C}}_{\mathrm{a}}-{\mathrm{C}}_{\mathrm{i}}}{1.6}=\frac{{\mathrm{C}}_{\mathrm{a}}(\mathrm{b}-{\Delta }^{13}\mathrm{C})}{1.6(\mathrm{b}-\mathrm{a})}$$

(4)

In Equation (3), δ¹³C_p and δ¹³C_a represent stable carbon isotope values for trees and atmosphere, respectively. In Equation (4), A is the assimilation rate of trees, g is the stomatal conductance of leaves, and 1.6 is the diffusion ratio of water vapor and CO₂ in the air.

2.4. Data Collection and Processing

SPSS19.0 software was used to analyze the correlation between δ¹³C and WUE of Picea meyeri and mean temperature (Tmean), mean precipitation (Pmean), maximum temperature (Tmax), minimum temperature (Tmin), relative humidity (RH) and vapor pressure deficit (VPD). The meteorological data of the Wuzhai and Kelan meteorological stations, which are the closest to the sampling point, are selected as reference stations. The data were provided by the China Meteorological Data Service Center (http://data.cma.cn/ accessed on 3 June 2022). CO₂ concentration in the calculation of WUE of plant data was from McCarroll et al. [12] and the NOAA website (https://gml.noaa.gov/ accessed on 22 March 2022).

3. Results

3.1. Picea Meyeri δ¹³C Value and WUE Sequence Characteristics

The variation range of the stable carbon isotope value sequence of tree rings (1957–2020) is −23.89‰~−21.67‰, with an average value of −22.67‰ (Figure 4). The δ¹³C value of tree rings decreases year-by-year with an average reduction of 0.03‰ per year. The highest δ¹³C value is −21.67‰ in 1957 and the lowest value is −23.89‰ in 2011. According to the test of the outlier, δ¹³ C from 1957, 2011 and from 2017–2020 all show anomalies, and after correction by quadratic polynomial, no longer show anomalies. The linear regression equation of this sequence shows R² = 0.481, p < 0.01, indicating a significant decreasing trend.

Trees use CO₂ through photosynthesis to feed their growth. During the process, the CO₂ produced by the respiration of surrounding trees can cause tree ring δ¹³C to change in value. The burning of chemical fuels causes the δ¹³C value of the atmosphere to decrease, making the δ¹³C of plants much lower than the former. Liu et al. indicate that atmospheric CO₂ has nothing to do with climate change, thus, atmospheric CO₂ should be excluded in the research [41]. McCarroll’s equation was used to correct the carbon isotope sequence [12], as shown in Equation (5), and the result of the calibration is shown in Figure 5.

$${\mathsf{\delta }}^{13}{\mathrm{C}=\mathsf{\delta }}^{13}{\mathrm{C}}_{\mathrm{p}}{-(\mathsf{\delta }}^{13}{\mathrm{C}}_{\mathrm{a}}+6.4)$$

(5)

In Equation (5), δ¹³C_p and δ¹³C_a are the tree ring measured δ¹³C and atmospheric background δ¹³C value, respectively.

The variation range of WUE is from 17.26–61.31 from 1957 to 2020, with an average value of 39.45, which is consistent with the WUE of the evergreen coniferous forest [42]. The value of WUE of Picea meyeri increases year-by-year, with an average increase of 0.69 per year. The highest value is 61.31 in 2019 and the lowest value is 17.26 in 1957. Figure 6 shows that the WUE is stable from 1957 to 1985 with small fluctuations, and the average value is 30.89. The fitting equation shows that WUE increases slightly from 1957–1985 (R² = 0.331, p < 0.01) and significantly after 1985 (R² = 0.867, p < 0.01), which may be similar to the slight increment in 1990 and 2005. The WUE sequence generally has a significant increasing trend.

3.2. Response of Stable Carbon Isotope Sequence to Changes in Temperature and Precipitation

The above analyses show that the variation trends of δ¹³C and WUE of Picea meyeri are the opposite. In order to investigate the possible “lag phenomenon”, correlation analysis should be made between the two sequences and the meteorological factors of the previous year. The Pearson correlation analysis results of the two sequences and climate factors, such as average annual temperature and annual precipitation, are shown in

Table 1. The relationship between δ¹³C and WUE of Picea meyeri and annual meteorological factor on Luya mountain.

	Tmean	Pmean	Tmax	Tmin	RH	VPD
δ 13 C	−0.711 **	−0.005	−0.676 **	−0.162	0.347 **	0.263 *
WUE	0.718 **	−0.061	0.457 **	0.424 **	−0.342 **	−0.254 *

* and ** symbol represent significant correlation at the 0.05 and 0.01 level, respectively.

. The results of the mean temperature (Tmean), maximum temperature (Tmax), minimum temperature (Tmin), mean precipitation (Pmean), relative humidity (RH) and vapor pressure deficit (VPD) in p9 (previous September) to c12 (current December) and the growing season (c5–9, May–September of the current year) are shown in Figure 7.

The temperature subtly changed from 1957–1985, but showed an obvious upward trend from 1985–2020, with an average of 5.40 °C (Figure 8). There is no obvious change in precipitation in the region, with an average annual precipitation of 433.83 mm (Figure 8). The correlation of δ¹³C and annual mean temperature is found with Pearson correlation, reaching −0.711 (p < 0.01) (Table 1). The correlations with annual Tmax, RH and VPD are also significantly correlated (Table 1). Similar to the monthly scale correlation, the negative correlation between δ¹³C and annual mean precipitation is also not significant (Table 1).

The coefficient of autocorrelation of the δ¹³C sequence is 0.866, which shows that there is a significant “lag phenomenon” of climate on δ¹³C, that is to say, the growth of Picea meyeri is not only related to the climate factors of the current year, but also influenced by the climate change of the previous year. Figure 7 shows that the δ¹³C of Picea meyeri is negatively correlated with the mean temperature of each month from 1957 to 2020. At the same time, the correlation between δ¹³C and the temperature of the growing season (c5–9) month is obviously higher (r = −0.591, p < 0.01) than that of any single month of the growing season. The correlation between δ¹³C and precipitation is rather low and not significant. It is only negatively correlated with precipitation in October of last year, even though the correlation coefficient reached more than 0.200, it is still not significant. It is significantly negatively correlated with precipitation in February of the current year (r = −0.267, p < 0.05), but not significantly correlated with other months or the growing season months. The correlation of δ¹³C and Tmax and Tmin of each month also reaches a significant level, showing a significantly negative correlation with Tmax of each month (p < 0.05); a slightly negative correlation with Tmin is found, but it only reaches significance in June, and no significant correlation with other months. The correlation between δ¹³C and relative humidity is poor, and it is only significant with December of last year and March of the current year (r = 0.343, p < 0.01; r = 0.330, p < 0.01, respectively). The VPD of only September of last year and the current year are significantly positively correlated with δ¹³C (r = 0.340, p < 0.01; r = 0.325, p < 0.01). No significant correlation is found in other months.

3.3. Relationship between δ¹³C and Climate Factors

Figure 7 indicates that the correlation between δ¹³C of Picea meyeri and climate factors is significantly correlated with Tmean, growing season temperature and Tmax, but is not significantly correlated with Pmean, Tmin, VPD and RH. To explore the influence of δ¹³C of Picea meyeri, climate factors that have the greatest influence are selected to establish the multiple regression equation. The result in SPSS shows that R = 0.760, R² = 0.578 and adjusted R² = 0.541.

$$\mathrm{Y}=-23.578-0{.654\mathrm{Tmean}}_{5-9}-0{.115\mathrm{Tmax}}_2-0{.040\mathrm{Pc}}_2+0{.019\mathrm{Rc}}_3+0{.030\mathrm{Vp}}_9$$

(6)

In this equation, Y is the Picea meyeri tree ring δ¹³C value, Tmean_5–9 represents the mean temperature of the growing season, Tmax₂ represents the maximum temperature in February, Pc₂ represents the mean precipitation in February, Rc₃ represents relative humidity in March and Vp₉ represents the vapor pressure deficit in September last year.

3.4. Response of WUE to Climate Change

The inter-annual variation in precipitation of the study area from 1957–2020 and the correlation analysis of WUE show that the annual precipitation is positively correlated with WUE, with the Pearson correlation coefficient of −0.718 (p < 0.01). The water use efficiency of Picea meyeri increases with the increase in precipitation. In addition, WUE is significantly correlated with Tmax, Tmin, RH and VPD (Figure 9).

As shown in Figure 9, WUE has a significant positive correlation with temperature. Similar to δ¹³C, the growing season temperature has a significant cumulative effect on WUE (r = −0.526, p < 0.01), which is higher than that of any single month of the growing season. However, the relationship between WUE sequence and monthly precipitation is not statistically significant, only significantly correlated with precipitation in February of the current year (r = 0.345, p < 0.05). It is negatively correlated with precipitation in summer, but the correlation is still not significant. The correlation between WUE and RH is poor, and only has a significant negative correlation with the RH in December and March of last year. There is a significant negative correlation with VPD in September of last year and the current year.

4. Discussion

4.1. Analysis of the Main Climate Impact Factors of δ¹³C of Picea meyeri

Photosynthesis in plants absorbs and uses ¹²C, allowing carbon isotope fractionation. Temperature and precipitation affect tree ring δ¹³C by affecting photosynthesis. Thus, it is considered that the tree ring δ¹³C can record long-term changes in ambient temperature and precipitation. Higher temperatures early in the growing season promote tree photosynthesis, resulting in lower fractionation of carbon isotopes and tree ring δ¹³C increase, that is, air temperature and δ¹³C is positively correlated. This result can be proved by the studies of Pinus tabulaeformis on Helan Mountain [41], Pinus massica on Zijin Mountain in south China [43], Larix sibirica on the Altai Mountains [44], Picea likiangensis on the Tibetan plateau [45] and Abies pindrow in central Himalaya [46].

However, if high temperature leads to relative drought or drought stress [8], the correlation of δ¹³C and temperature is negative. The δ¹³C of Picea meyeri selected in this study is negatively correlated with the temperature of each month and the combined temperature of each month (Figure 6), and the correlation of the temperature of the growing season is significantly negative (r = −0.591, p < 0.01), which is better than that of the temperature of each month, indicating that the temperature of the growing season has a significantly cumulative effect. On the annual scale, there is a significant negative correlation between δ¹³C and temperature (r = −0.775, p < 0.01) (Table 1). This result is contrary to that of Pinus tabulaeformis on Helan Mountain and other studies mentioned above [41,43,44,45,46]. The reason may be that the quantity and activity of the enzymes of photosynthesis decrease when the temperature is too high [47]. In addition, high temperature increases evapotranspiration, which leads to the decrease in soil water. To solve the drought stress, the stomata conductance decreases and CO₂ entering cells decreases, which weakens photosynthesis and indirectly restricts the growth of trees [48]. The significant negative correlation between δ¹³C and the Tmax can also explain the increase in temperature having an inhibitory effect on δ¹³C. Additionally, the effect of Tmax is obviously higher than the Tmin and the Tmean. Yang et al. used the Dendrometer to monitor the growth of Picea meyeri, and the result showed that the growth of Picea meyeri is in a significant negative correlation with the temperature in sunny conditions [1]. In addition, temperatures that are too high decrease soil water, which indirectly inhibits the growth of trees, showing that the increase in temperature weakens the assimilation effect of Picea meyeri, so the δ¹³C is negatively correlated with temperature.

Precipitation is an important factor affecting the growth; however, Picea meyeri in the treeline is restricted mostly by temperature [1,7,8]. In this study, the correlation between δ¹³C and monthly precipitation is quite low and does not reach the significance level of 0.05, which is only significant with precipitation in February (r = −0.267, p < 0.05). When Picea meyeri start to grow, the increase in soil moisture in spring leads to an increase in stomatal conductance, allowing the entry of intracellular CO₂, which contains mostly ¹²C, leading to a low δ¹³C value. In the annual scale, the nagetive correlation between δ¹³C and precipitation is very low and not significant (Figure 8). This is because at the treeline, precipitation is far from sufficient to satisfy the growth of trees and the fractionation of carbon isotopes. Therefore, when there is no significant increase in precipitation, tree ring δ¹³C is still negatively correlated with precipitation [6].

4.2. Analysis of the Main Climate Impact Factors of WUE of Picea meyeri

The trend of WUE of Picea meyeri on Luya mountain is consistent with that of the Larix sibirica of the Altai Mountains in North China [44], Robinia pseudoacacia in central China [49], Pinus sylvestris and Picea abies in central Europe [50], Pinus nigra in northeast Spain [51] and other tree species which show a significant increasing trend. The water use efficiency of Picea meyeri is negatively correlated with temperature, and the coefficient is extremely low. The reason is that Picea meyeri at the treeline is so sensitive to temperature limitation that it has a strong response to climate warming [52]. The temperature in the Luya mountain area shows a significant rising trend, and the precipitation decreases, which is sufficient to prove that there is a warming and drying trend in this area (Figure 7). The correlation between the average annual temperature of Luya Mountain and WUE is greater than that of the growing season, which is possible because the increase in temperature in autumn and winter prolonged the end date of cambium activity [53], which improved the WUE of Picea meyeri. In the early spring, the temperature gradually increases and the snow melts, so the period of dormancy is gradually stopped. The WUE is in a significant positive correlation with the precipitation and temperature in February, which is the early growth season of Picea meyeri. The increasing temperature can effectively reduce the inhibitory effect of low temperature on the roots and reduce the dormancy, photosynthesis is enhanced, and WUE is in a significant positive correlation with temperature. In the period, temperature is the limiting effect of WUE. From 1957 to 2020, the temperature in the study area increased significantly, while WUE showed a consistent increasing trend. At the annual scale, WUE showed a significantly positive correlation with temperature and a very low correlation with precipitation (Table 1). Studies show that NPP and ET are much more affected by temperature than precipitation [24].

Due to drought stress [8,54,55], the unnecessary water loss is reduced by the reduction in the stomatal conductance of Picea meyeri to maintain its physiological activity. Meanwhile, with the reduction in the stomatal conductance, intercellular CO₂ reduces, leading to a high WUE, showing that the growth of the Picea meyeri in the treeline might be in a good condition in climate warming conditions. In this study, the correlation is so poor with precipitation that it is only positively correlated with the precipitation in February. The increase in precipitation in spring would lead to the increase in WUE, which might be because the precipitation before the growing season strengthens the carbon fixation ability of Picea meyeri. In addition, according to Fan et al. [56], global warming has affected the distribution of willows, which can predict the expansion of distribution in high latitudes (Arctic regions). Moreover, Liu et al. [57] used the TWINSPAN (Two Way Indicator Species Analysis) method to classify species and niches, which indicated that the distribution of spruce and fir forests showed obvious stand replacement in high altitudes (Changbaishan National Nature Reserve, Antu, China). In that case, we assume that Picea meyeri on Luya mountain might expand to higher altitudes, leaving lower niches for the deciduous tree species. Meanwhile, the RH and VPD are negatively correlated with WUE, and the correlation is only in a few months. Further study shows that neither precipitation, RH or VPD are the limiting factor of WUE, but temperature factors are the limiting factor of WUE.

5. Conclusions

The δ¹³C of Picea meyeri from 1957 to 2020 showed a significant decreasing trend. WUE showed an increasing trend, and it increased more significantly from 1985 to 2020 than that from 1957 to 1984. Therefore, the growth of Picea meyeri is good in climate warming conditions. Tree ring delta δ¹³C has a significant negative correlation with Tmean and Tmax in the growing season, and an insignificant relationship with Pmean, RH and Tmin. WUE has a significant positive correlation with Tmean, and a low correlation with Pmean. It is not difficult to see that the impact of temperature in Luya Mountain is stronger than precipitation in both δ¹³C and WUE. Therefore, the main impact factor of δ¹³C in tree rings of Picea meyeri is temperature rather than precipitation, and Picea meyeri is also highly affected by drought stress.