Seasonal biochemical changes in Betula tortuosa Ledeb. annual rings in Alpine forest-tundra of Kuznetsk Ala Tau Mountains

Woody vegetation growth conditions have marked effects on hemicellulose, cellulose and lignin formation. In this study, the climatic responses of these major cell wall polymers in Betula tortuosa Ledeb. were analyzed. 35 annual rings (1980–2015) of Betula tortuosa Ledeb. trees growing in the alpine forest-tundra of Kuznetsk Ala Tau were studied using Fourier-transform infrared (FTIR) spectroscopy and pyrolysis–gas chromatography–mass spectrometry (Py–GC/MS). Analysis of the correlation of the resulting spectra and Py–GC/MS values with mean air temperature and precipitation showed that the polymeric composition of Betula tortuosa Ledeb. was mainly determined by June–August climate. The major factor limiting the development of the “unique” cell wall polymer composition of Betula tortuosa Ledeb. found in alpine forest-tundra of Kuznetsk Ala Tau was a deficit of heat. At the end of the growing season, precipitation had a largely negative impact on polymer formation in Betula tortuosa Ledeb. cell walls. The authors believe that combining FTIR spectroscopy and Py–GC/MS is an effective approach to quantifying the consequences of current climate trends for Siberian forest ecosystems.


Introduction
Future global climate change combined with and driven by the further increase in atmospheric CO 2 concentrations is expected to markedly affect plant growth and performance and, as a consequence, the composition and spatial distribution 1 3 of species in terrestrial ecosystems. One key for understanding the effect of climatic variables on the long-term plant performance is to characterize and interpret physiological information laid down in natural archives in the past. For woody plants with secondary growth, annual rings in the xylem (generally referred to as 'tree rings') provide a datable archive that records information about the environmental and ecophysiological conditions (Gessler et al. 2014). Annual rings are, in some respect, "fingerprints" of past climate variability that show its effects on wood biomass production. Since tree rings originate from seasonal cambial activity and as a result from such seasonality they are a phenological phenomenon related to photoperiod and the biological clock (Lüttge and Hertel 2009). Their width reflects past climate, as well as current atmospheric CO 2 and O 2 , air temperature, and precipitation patterns causing droughts and floods (Bouriaud et al. 2005;Kahle et al. 2008). Considering these controls as a background, Locosselli and Buckeridge (2017) proposed a new term-dendrobiochemistry that allows a broader interpretation, providing an integral picture of the relationship between environmental conditions and treering features. Among these features are physiological and biochemical parameters that can be measured analytically.
The mechanisms of climate influence on forest ecosystems, particularly of the influences of past climate changes similar to today's on woody plant growth, are insufficiently studied. Studying these mechanisms is highly important for developing a concept of wood synthesis and structure. An appropriate approach is to use rapid methods, such as pyrolysis-gas chromatography-mass spectrometry (Py-GC/MS) and Fourier-transform infrared (FTIR) spectroscopy. The latter is widely used to analyze molecular structures, to identify compounds of biological samples, and to study complicated polymers (Falcão and Araújo 2013;Acquah et al. 2015).
FTIR spectroscopy is a modern technique for determination of molecular structures, identification of compounds in biological samples and investigation of complex polymers (Arndt et al. 1999;Kacuráková and Wilson 2001). It is wellestablished for analysis of the chemical composition of wood (Pandey and Pitman 2003;Xin et al. 2017;Tintner et al. 2020). The detected vibrations of the molecules in the wood sample, caused by IR (infrared)-light, enable drawing conclusions on its chemical composition (Pandey 1999). Absorption bands of substances are generated during each analysis and can directly be assigned to a certain molecular bond (e.g., C = O) or to frequencies of functional groups (e.g., G/S lignin) (Pandey and Pitman 2003). FTIR spectroscopy offers a large potential for wood analysis, because it is a fast and quite simple method for determination of the chemical composition of complex samples. The changes in crystalline structures of cellulose in cell walls, the composition of the chemical structures of various tree species as well as the chemical composition of wood, its primary, composed and degradation products can be analyzed (Naumann and Polle 2006). The method allows the detection of the influence of binders on wood quality and the accumulation of additives. FTIR spectroscopy enables quick analysis of wood components. Due to qualitative differences in the spectra it is possible to assign the chemical structures of timber to a single wave number, such as that for lignin, extractives or polysaccharides.
It is possible to distinguish, for example soft or hardwood due to their different composition of lignin and holocellulose (Hori and Sugiyama 2003).
Pyrolysis-gas chromatography/mass spectrometry (Py-GC/MS) is an analytical technique which is able to provide useful information concerning the structure of oak wood components, assuming that pyrolysis products represent, to a greater or lesser degree, the structural units forming the macromolecule. Py-GC/MS is based on depolymerization of the macromolecules by heat followed by identification of the fragments by mass spectrometry. For the polymer subunits, which are not or only slightly volatile, the increase in temperature leads to their thermal degradation and then to the identification of the products formed (Nonier et al. 2006). Influences of climatic factors on tree-ring aromatic and carbohydrate components were investigated by Py-GC/MS method. The advantages of this method are a fairly rapid analysis, small amount of wood needed for the analysis, simple sample preparation that does not require lignin, cellulose and hemicelluloses to be extracted before estimating their quality and mass fractions in the wood under study (Scholze and Meier 2001).
The combination of the physical and chemical methods used in this study is successfully applied to a range of other tree investigation fields including dendroecology, botany, tree physiology, tree biochemistry, impacts of climatic conditions on wood, and silviculture. The authors believe that, on the whole, the information presented in this paper will enable to extend this range and will provide the basis for the next step in evaluating climate influences on chemical composition of tree stem wood and in predicting quality of wood developed under current climatic trends. In this case study, the sensitivity of the biochemical structure in Betula tortuosa Ledeb. tree rings to air temperature and precipitation in the alpine forest-tundra ecotone of Kuznetsk Ala Tau was assessed.

Description of the study site
Kuznetsk Ala Tau Mountains encompass several mountain ridges of a maximum of 2200 m above sea level (a.s.l.) that stretch as long as up to 300 km north-southward. Alpine forest-tundra is a zone where climate strongly controls woody vegetation development through limiting its growth by the lack of heat (Holtmeier 2009 (Fig. 1).
The study region is in continental climate with a long and cold winter season and a warm summer season. On west-facing slopes, annual precipitation is 600-800 mm, whereas the central, windward, part of the mountains receives up to 1500 mm/ yr. From the 1980-2015 climatogram built using data from Nenastnaya weather 1 3    station it is clear that average air temperatures of January and July are − 15.3 °C and + 13.4 °C, respectively (Fig. 2). The long-term data recorded at this weather station showed that average annual air temperature has tended to increase since early 1970s (from − 2.6 ± 0. 5 °C in 1950-1970 to 1.5 ± 0.3 °C 1990-2010). Nenastnaya weather station is about 12 km from the site.

Wood sampling
To investigate physical and chemical heterogeneity of tree rings, five healthy trees were selected with the following characteristics: the trees are healthy; the trees show no signs of weakening; the leaves are green; the crown of the trees is dense; the average stand age is sixty years; the average stand height and tree diameter at a height of 1.3 m, i.e., at breast height (d.b.h.), are 4.6 m and 15 cm, respectively.
In a previous dendrochronological study (Petrov et al. 2019) it was shown that 110 analyzed trees from five samples for this region revealed a high synchronicity of ring width series [inter-series correlation coefficient r > 0.6 and expressed population signal (EPS) ~ 0.99]. The tree rings were separated using a microscope at tenfold magnification. After this preparation, 1980-2015 wood samples were conditioned to ambient humidity, which was about 65% at an air temperature of 20 ± 2 °C. The correlation of the tree-ring physical and chemical parameters of interest with climatic variables was estimated based on Spearman's correlation coefficient. To identify the dynamics of mean monthly air temperature, mean monthly precipitation, and of the physical and chemical parameters, a 5-year moving average was used that enabled to remove noise (Shiyatov 1986;Loader et al. 2019).

FTIR spectroscopy
Tree-ring IR spectra (8000-350 cm −1 ) were obtained with the help of VERTEX 80 V Fourier spectrometer (0.2 cm −1 spectral resolution and ± 0.05 cm −1 wave number repeatability), Bruker Optics, Germany. For acquiring IR spectra, wood samples were pressed in thin potassium bromide (KBr) tablets. The samples were weighed with XFR-125E laboratory scales. The sample weight for FTIR spectroscopy is about 1.5 mg. Then, 1.5 mg of wood mixed with 100 mg of potassium bromide was ground in a mortar, the ground material was put in a press-tool, the air was evacuated from the tool, and the material was pressed. The IR spectra were analyzed using OPUS software (https:// www. bruker. com/ en. html). IR absorption bands identification was carried out according to the literature (Faix and Böttcher 1992;Pandey and Pitman 2003;Rodrigues et al. 1998;Emandi et al. 2011).

Py-GC/MS
Pyrolysis-gas chromatography/mass spectrometry (Py-GC/MS) system was employed to separate and identify the pyrolysis volatiles. For this purpose, a Frontier EGA/PY-3030D-type pyrolyzer was directly attached to a gas chromatography/mass spectrometry (QP 2020, Shimadzu). The sample weight for Py-GCMS is about 1 3 80 μg. In the characterization process, the pyrolysis temperature of 600 °C was used. The chromatographic separation of the volatile products was performed using an Ultra Alloy capillary column (30 m × 0.25 mm × 0.25 m film thickness). Prior to the chromatographic separation, the temperature of the chromatographic column was progressively increased as follows: (i) 50 °C for 5 min; (ii) from 50 to 240 °C at a rate of 4 °C/min; (iii) the capillary column was maintained at 300 °C at a rate of 10 °C/min for about 10 min. Helium was used as carrier gas at a constant flow of 1.0 mL/min. The GC inlet was 250 °C and a split ratio of 50:1 was used. Mass spectra were recorded under electron ionization (70 eV) with them/z range: 35-450 au. Peak identification was carried out with the mass spectral library and literature (Hidayata et al. 2018;Mattonai et al. 2019;Ghalibaf et al. 2019).

Results and discussion
The use of the Spearman's correlation coefficient is explained by the fact that not all data series that were obtained by FTIR spectroscopy and Py-GC/MS had a normal distribution. In addition, the sensitivity of the main polymer components to air temperature and precipitation was observed at the low-frequency level of variation, while the sensitivity was not manifested at the high-frequency level variation. Figure 3 shows information about continuousness of the biochemical tree-ring series obtained by FTIR spectroscopy of Betula tortuosa Ledeb. tree rings. Table 1 presents descriptive statistics of the studied variables of Betula tortuosa Ledeb. tree rings .

FTIR spectroscopy
The results of the processing of 35 tree rings  with FTIR spectroscopy were represented as IR spectra. The absorption bands that revealed the correlation between mean monthly air temperature, as well as the correlation between mean monthly precipitation and IR spectrum characteristics of Betula tortuosa Ledeb. tree rings are shown in Fig. 4. Figure 4 shows, by way of example, IR spectra of the "fingerprints" (1800-700 cm −1 ) of tree rings formed in 1985.
The correlations between the spectra and mean monthly air temperature are presented in Fig. 5. The high intensities of the bands at 1375 cm −1 (of C-H bonds in cellulose) and at 1325 cm −1 (lignin of type S) exhibited significant negative correlation (p < 0.001) with May air temperature (− 0.59 and − 0.66, respectively). Bands at 1325 cm −1 (lignin of type S) and at 1268 cm −1 (lignin methoxyl group) had significant negative correlations (− 0.57 and − 0.64, respectively) with August air temperature. September air temperature was negatively correlated with the intensities of the bands at 1738 cm −1 (C = O bonds in hemicellulose) and at 1506 cm −1 (C = C bonds in lignin) (− 0.61 and − 0.73, respectively).
The mountain forest-tundra ecotone is a classic testing area for estimating the impact of air temperature on radial increment of woody species (Shiyatov 1986;Schweingruber 1996;Holtmeier 2009  Ledeb. tree rings  spectral characteristics were influenced mostly by the air temperature of May and September. In May, air temperature in Kuznetsk Ala Tau rises above zero to trigger changes in cell polysaccharide composition and in the activity of enzymes modifying cell wall. These changes result in intermolecular bonds in the LCC growing weaker. Air temperature signals are important regulators of daily and seasonal plant growth (Penfield 2008). Interestingly, cellulose crystalline part is reduced with increasing air temperature, which suggests an inverse relationship between the synthesis rate of cellulose and its crystallinity level (Guerriero et al. 2014). In August-September, lignin intermolecular bonds become stronger (Fig. 5). This may be related to the accumulation of phenolic compounds at the end of the growing season. High contents of phenolic compounds were found in autumn, when trees were entering hibernation (Prusakova et al. 2008). It was found that, at the end of the growing season, after a decrease in air temperature, the bands corresponding to hemicelluloses increased in intensity. It is not unlikely that, in the process of plant adaptation to lower air temperature, cell wall polysaccharide composition changes and the activity of the enzymes modifying cell wall, hence, increases to strengthen, as a result, the cell wall (Rajashekar and Lafta 1996). Moreover, plant adaptation to cold weather involves increasing content of hemicellulose that may contribute to cell wall rigidity and prevent cell collapse from dehydration (Weiser et al. 1990; Kubacka-Zebalska and Kacperska 1999). The correlation coefficients significant at p < 0.001 between mean monthly precipitation and the spectral values of the main Betula tortuosa Ledeb. wood polymers are shown in Fig. 6. The intensities of the bands at 1375 cm −1 (C-H bonds in cellulose) and at 1325 cm −1 (lignin of type S) had a negative correlation (− 0.61 and − 0.64, respectively) with June precipitation. However, the band at 1325 cm −1 (lignin of type S) showed positive correlation (0.60) with July precipitation.
In Kuznetsk Ala Tau, precipitation suppresses woody plant growth (Bocharov and Timoshok 2011). It was found that lignin and cellulose correlate negatively with June precipitation. For July, however, the correlation was positive. Precipitation in June and July is minimal. Although cellulose response to water deficit is not yet fully understood, cellulose synthesis may decrease with decreasing available water, which is proved by precipitation lack-induced reduction in cellulose content found Spearman's correlation coefficient (at p < 0.001) between mean monthly air temperature and IR spectra of Betula tortuosa Ledeb. tree rings for certain plant species (Piro et al. 2003;Bray 2004). According to Clifford et al. (1998), Hessini et al. (2009), and Tardieu (2011, cell wall elasticity modulus values increase in response to water stress in certain plants, which suggests that rigid cell walls may be necessary to support cell integrity during post-water stress rehydration. Lignification is a complicated process, in which many phenolic substrates and enzymes are involved (Wang et al. 2013). It may occur early, when cells begin to actively prepare themselves for the cold season. From the authors` viewpoint, lignin molecular bond strengthening found for June in this study was due to early lignification. In July, under a certain increase in precipitation, molecule interactions in lignin grew more intensively. Very little is known nowadays on the enzymes and signaling processes regulating the reaction of aromatic component to increasing available water during lignification.
Judging by the obtained correlation coefficients, the intensity of the band of Betula tortuosa Ledeb. tree rings lignin-carbohydrate complexes was more "sensitive" to the growing season air temperature than to precipitation.

Py-GC/MS
The pyrogram of each tree-ring had over forty peaks. Figure 7 and Table 2 show the main compounds identified that revealed the correlation between mean monthly air temperature, as well as the correlation between mean monthly precipitation and the relative content of the products of component pyrolysis of Betula tortuosa Ledeb. tree rings. Figure 7 shows, by way of example, the pyrogram of the tree ring formed in 1995. The correlation between the relative content of the products of carbohydrate component pyrolysis in the total pyrolyzate and mean monthly air temperature is presented in Fig. 8. 1.4:3.6-Dianhydro-.alpha.-d-glucopyranose exhibited a negative correlation (− 0.77) with April air temperature. Furan, 2.5-dihydrofuran, 1.4:3.6-Dianhydro-.alpha.-d-glucopyranose were negatively correlated with air temperature (− 0.63, May; − 0.60, June). It is noteworthy that 4-Cyclopentene-1.3dione had a negative correlation (− 0.59) with air temperature of August and Furan, 2.5-dihydrofuran showed a fairly close correlation with September temperature, but it did not go over the significance level.  Table 2 Betula tortuosa wood Py-GC/MS products divided into categories C-carbohydrates, Gguaiacyl lignin, S-syringyl lignin (Pyrogram, Fig. 7 Low positive air temperatures of April had a positive impact on the relative contents of the products of pyrolysis of carbohydrate components, namely of 1.4:3.6-Dianhydro-.alpha.-d-glucopyranose, in the total pyrolyzate. This might be due to the effect described by Krasnobayev and Voronin (2011). They used the method of conductometry in their study to find negative effects of thaw spells on forest stands of Ural Mountains. During thaws, cambium layer electric conductance is close to that in non-sleeping trees. Spring thaw influences occur at the metabolic level and are induced by the alternation of above zero and below − 12 °C air temperatures (Krasnobayev and Voronin 2011). The current study showed that low above-zero air temperatures of May and September differed in influence on the Furan, 2.5-dihydrofuran relative content: In May, air temperature enhanced decreasing Furan, 2.5-dihydrofuran, whereas September air temperature enhanced its accumulation in cell walls. Furan participates in plant cell protection from oxidation stress (Stitt and Hurry 2002). Slightly increasing (low positive) air temperatures early in the growing season lead to decreasing Furan, and this decrease may then induce changes in gene expression and plant metabolism (Chinnusamy et al. 2007;Chang et al. 2010). For August, increasing relative content of 4-Cyclopentene-1.3dione was found, which belongs to the group of Cyclopentenoids. The compounds of this group are secondary metabolites important for plant growth and resistance to external stressors (Sevcikova et al. 2014). It was also identified that relative contents of 1.4:3.6-Dianhydro-.alpha.-d-glucopyranose decreased with increasing air temperatures in June. Low 1.4:3.6-Dianhydro-alpha.-d-glucopyranose content observed at low positive air temperatures apparently effected the strength of intermolecular bonds in cellulose and these effects were confirmed by FTIR spectroscopy (see above).

Fig. 8
Spearman's correlation coefficient (at p < 0.001) between the relative content of the products of carbohydrate component pyrolysis in the total pyrolyzate and mean monthly air temperature No significant correlation, at p < 0.001, between aromatic components and mean monthly air temperature was found, whereas eugenol and syringylacetone correlated significantly negatively with spring air temperatures (− 0.60, April and − 0.59, May, respectively). Sinapyl alcohol showed significant negative correlations, at p < 0.001, with June (− 0.71) and September (− 0.64) air temperatures (Fig. 9).
Phenolic compounds, namely eugenol, syringylacetone, and sinapyl alcohol, differed in sensitivity to air temperature variations during the growing season. In April, with increasing air temperature, relative eugenol content increased, which could be due to the effect described above (Krasnobayev and Voronin 2011). In May and June, relative syringylacetone and sinapyl alcohol decreased, respectively. However, relative sinapyl alcohol content increased in September. During the period of shoot growth in spring, phenolic compounds are actively accumulated. According to Prusakova et al. (2008), phenol accumulation is common in fall, before plant hibernation. The maximum of phenol content in fall may partly account for defoliation, bud breaking suppression, and a general decrease in plant metabolism. Prusakova et al. (2008) observed high contents of growth inhibiting phenolic compounds in buds of a number of plants.

Fig. 9
Spearman's correlation coefficient (at p < 0.001) between the relative content of the products of aromatic component pyrolysis in the total pyrolyzate and mean monthly air temperature In July, the area of interest received little precipitation and it increased gradually to the end of the growing season (Fig. 2) resulting in increasing relative contents of 1.4:3.6-Dianhydro-.alpha.-d-glucopyranose in Betula tortuosa Ledeb. cell walls (Fig. 10). The authors agree with Ricardi et al. (2014) in that 1.4:3.6-Dianhydro-.alpha.-d-glucopyranose enhances cellulose synthesis in a low-precipitation period and may maintain, thereby, the functional integrity of cell walls that proved, in turn, continuous cell growth. Accumulating 1.4:3.6-Dianhydro-.alpha.-d-glucopyranose in September, with precipitation reaching its maximum and decreasing air temperature, might be related to the need for a woody plant to develop mechanisms of resistance to frost stress.
The correlation found between Betula tortuosa Ledeb. wood aromatic component and mean monthly precipitation is presented in Fig. 11. The response of the aromatic component to precipitation of July and September appeared to be similar to that of carbohydrate component. Eugenol and 1,2-Benzenediol, 3-methoxy had positive correlations, at p < 0.001, with July precipitation (0.57 and 0.61, respectively), Syringylacetone correlated positively (0.63) with August precipitation, and sinapyl alcohol-with September precipitation (0.70).
Phenolic compounds also appeared to be sensitive to mean monthly precipitation. Py-GC/MC allowed us to find that 1.2-Benzenediol, 3-methoxy and eugenol relative contents increased in July, syringylacetone showed an increase in August, and sinapyl alcohol increased in September. Very little is known today about water influence on lignification. It can only be presumed that fairly low precipitation characteristic of July activates 1.2-Benzenediol, 3-methoxy and eugenol, which may contribute to cell wall rigidity and that, in a while, cells

Fig. 10
Spearman's correlation coefficient (at p < 0.001) between the relative content of the products of carbohydrate component pyrolysis in the total pyrolyzate and mean monthly precipitation become a bit longer and, hence, more compact and less water permeable. This may help the leaves and roots of woody plant retain turgor at a low water potential and prevent apoplastomes from losing water (Hura et al. 2012(Hura et al. , 2013Fan 2006). Increasing precipitation and decreasing air temperature (August-September) coincide with increases in syringylacetone and sinapyl alcohol, which may correspond to the period of a woody plant entering hibernation (Prusakova et al. 2008).

Conclusion
FTIR spectroscopy and Py-GC/MS method used in this study allowed us to develop new information on spectral and pyrolytic features of Betula tortuosa Ledeb. tree rings. It was found that climate conditions had significant influences on the strength of molecule interactions in the LCC and on the proportions of pyrolysis products in each tree ring.
-In April-May, a period of the pre-growing season cambium reactivation and of the beginning of woody plant growth with slightly increasing air temperature, the relative contents of the products of pyrolysis of carbohydrate and aromatic components increased and molecule interactions grew stronger in the LCC. These products were found to decrease during active tree growth. However, low precipitation in the growing season resulted in increasing product contents and in strengthening of molecule interactions in the LCC; Fig. 11 Spearman's correlation coefficient (at p < 0.001) between the relative content of the products of aromatic component pyrolysis in the total pyrolyzate and mean monthly precipitation 1 3 -At the end of the growing season, molecule interactions in the LCC grew weaker with decreasing air temperature, whereas the contents of the products of pyrolysis of carbohydrate and aromatic components increased due to heavy precipitation; -The tree-ring aromatic component exhibited a greater sensitivity to changing climatic factors compared to the carbohydrate component; -The correlation between mean monthly air temperature and IR spectra, as well as the correlation between the relative content of pyrolysis products of annual birch rings was negative.
This study found FTIR spectroscopy and Py-GC/MS to be effective for detailing features of Betula tortuosa Ledeb. Tree rings based on their physicochemical characteristics, due, among other things, to the climate conditions of the growing season. Applying these methods will enable to quantitatively forecast the consequences of the current climate trends.