Assessment of the heartwood contribution to carbon accumulation in Pinus sylvestris L. trees under different forest site conditions

1.   Introduction

Forest ecosystems make up a net carbon sink that offsets 15%–30% of yearly anthropogenic carbon emissions (Myneni et al., 2001). The carbon allocation to aboveground wood biomass is the largest contributor to vegetation carbon storage (Law et al., 2001; Hoch and Körner, 2011; Kraenzel et al., 2003; Peri et al., 2010; Liu et al., 2016; Mvolo et al., 2021; Cabon et al., 2022). Xylogenesis (the wood formation process) includes the growth of cells by stretching, the formation of a secondary cell wall, and with fibers, vessels, tracheids – programmed cell death. Cell walls in woody plants account for up to 90% of the dry mass; it is in them that long-term carbon storage occurs (Dalimova et al., 1998; Pang et al., 2018). The structural wood elements differ from each other in the composition of the components of their cell walls (cellulose, hemicelluloses, pectins, lignin, suberin, proteins). A change in the ratio of structural elements in wood, the thickness and composition of their cell walls, depending on growing conditions, will inevitably be reflected in a change in carbon allocation in wood (Mackay and Roberts, 1982; Telmo and Lousada, 2011; Roach et al., 2018; Bārdule et al., 2021; de Paula Protásio et al., 2020; Vos et al., 2023). The estimation of the wood carbon concentration in tree species from a range of forest types is critical for understanding the potential of forests for carbon capture and storage (Thomas and Martin, 2012; Gao et al., 2016; Zhou et al., 2019). However, wood carbon allocation is poorly understood and is a major uncertainty for projections of future forests carbon storage potential (Huang et al., 2020; Cabon et al., 2022). Usually, in forest carbon estimates, it is assumed that wood biomass comprises 50% carbon (w/w) (Thomas and Martin, 2012). However, recent data shows that the 50% assumption introduces errors of ~5% in forest carbon stock estimates (Bert and Danjon, 2006; Martin and Thomas, 2011; Melson et al., 2011; Thomas and Malczewski, 2007; Thomas and Martin, 2012; Saner et al., 2012).

The part of the xylem that contains living parenchyma cells is called sapwood (SW). In most tree species, as the tree grows, parenchyma xylem cells die, resulting in heartwood (HW) formation (IAWA, 1964, p. 32). The formation of HW reduces the sugars content, proteins, storage components (starch, lipids) and increases the number of secondary metabolites of phenolic nature, resins, extractives that give wood resistance to biodegradation (Saitoh et al., 1993; Hauch and Magel, 1998; Magel et al., 2000; Bieker and Rust, 2010; Forrester et al., 2010; Macfarlane et al., 2010; Wang et al., 2010; Pfautsch et al., 2011, 2012). Because of the accumulation of numerous phenolic polymers in HW, the mass fraction of carbon in it is higher (Lamlom and Savidge, 2006). Information about the different carbon concentration in SW and HW was found in the literature as early as the end of the 19th century (Daube, 1883). These differences, regardless of species specificity, are thought to range from <1% to ~7% (Thomas and Martin, 2012) and have been found for both gymnosperms and angiosperms (Bert and Danjon, 2006; Herrero de Aza et al., 2011; Peri et al., 2010). On the one hand, the amount of HW and the rate of its formation are under genetic control (Taylor et al., 2002; Spicer, 2005), also they depend on the tree growth conditions (Ugolev, 1975; Bamber and Fukazawa, 1985; Hazenberg and Yang, 1991; Wilkes, 1991; Sellin, 1996; Björklund, 1999; Gjerdrum, 2003; Nawrot et al., 2008; Wang et al., 2010). The identification of regularities in the HW formation under different climatic conditions is relevant from the point of view of global studies of the carbon cycle.

Scots pine is one of the most important forest-forming conifers in Northern Europe, accounting for approximately 39% of forests in Sweden (Lerceteau et al., 2000), over 50% in Finland (Finnish Forest Research Institute, 2009), 64% of the forests of Russian Karelia (State Report on the Environment in Republic of Karelia, 2021). Because of the pronounced transition zone between HW and SW, pine is considered a model plant for studying the processes of HW formation (Saranpää and Höll, 1987, 1989; Fagerstedt et al., 1998; Hauch and Magel, 1998; Bergström et al., 1999; Magel et al., 2000; Bergstrom, 2003; Kennedy et al., 2013; Esteves et al., 2013; Lim et al., 2016). Scots pine patterns of HW formation were studied in the North-West of Russia under the conditions of the lingonberry, blueberry and lichen pine forests (Galibina et al., 2022a, Galibina et al., 2022b; Tarelkina et al., 2022). In this study, we estimated the HW proportion in the trunk of Scots pine depending on the position of the tree in the stand (biosocial class according to Fabisiak and Fabisiak, 2021; Zawieja et al., 2021) in the conditions of the middle taiga subzone and in the gradient of climatic conditions (middle taiga subzone – northern taiga subzone – transition area of the northern taiga subzone and tundra). We sought to ask: (ⅰ) Does the carbon concentration differ between SW and HW in different forest conditions and depending on the position of the tree in the stand? (ⅱ) Is the carbon concentration in SW and HW connected with extractive compound content? (ⅲ) Do the HW proportion in pine trees and the carbon concentration in SW and HW influence on carbon estimation under different forest site conditions? We hypothesized that stands with a high proportion of heartwood in the tree trunk store more carbon in the wood due to changes in the wood's chemical composition. Thus, when growing conditions deteriorate, cambial growth is reduced, leading to increased extractive components, which results in higher wood carbon content.

2.   Materials and methods

2.1   Plant material and sample plots characteristics

The studies were performed on sample plots (SPs) of Forest Research Institute of the Karelian Research Centre of the Russian Academy of Sciences, established in 2021–2022 as part of a project aimed at studying the patterns of heartwood formation in Scots pine. Pine trees aged 70–80 years were selected for the study, which corresponded to the middle age ontogenesis stage (Sannikov, 1992), as well as the age at which the wood quality was a significant indicator since pine stands were assigned for felling following forestry requirements in this period.

The influence of soil conditions on the share of HW, carbon and extractive compound content in SW and HW in Scots pine was studied in the ecological series: blueberry (Vaccinium myrtillus L.) pine forest (Blu) – lingonberry (V. vitis-idaea L.) pine forest (Lin) – lichen pine forest (Lic), located in the Kivach Nature Reserve (62°16′05″ N, 33°58′49″ E) and its vicinity (middle taiga subzone) (Supplemental Materials Fig. S1). The soils were classified as Albic Podzols Arenice (blueberry and lingonberry pine forest) and Entice Podzol (lichen pine forest) according to the soil classification IUSS Working Group WRB 2015 (Tarelkina et al., 2022). The forest type was determined based on the analysis of soil data, ground cover (ground vegetation) and stand productivity. Lingonberry pine forest is widely represented in various climatic conditions. In Karelia they occupy over 30%, in the Murmansk region 42%. Blueberry pine forests are a relatively highly productive type of forest, which is formed on soils with a fairly high level of fertility and a sufficient amount of soil moisture. They are the second most common forest type in the study area. Lichen pine forests are a low-productive type of forest that forms on poor and dry soils. The lichen forest type is quite rare in Karelia and relatively poorly studied, which is also because of its low exploitation value. Lichen pine forest is widely represented in northern Fennoscandia and Western Siberia; on the territory of the latter, lichen forests account for 30% of the area of pine forests (Rysin and Savelieva, 2008; Krylov, 1961).

To assess the influence of the position of the tree (social classes) on the studied parameters, three groups of trees were identified on SPs: dominant trees (D.t., approximately 30%–40% higher than average trees in terms of trunk diameter at a height of 1.3 ​m (DBH)), intermediate trees (I.t., trees that close in height and diameter to the average stand), and suppressed trees (S.t., 30%–35% below the average tree stand by DBH) (Table 1). Each group had 3–5 model trees.

Table  1.  Characterization of model trees on SPs 1, 2, 8, 9 (middle taiga subzone), SP 10 (northern taiga subzone) and SP 12 (transition area of the northern taiga subzone and tundra).

Forest typea (sample plot name)	Age (years)	Dominant trees			Intermediate trees			Suppressed trees
		Number trees	DBH (cm)	Height (m)	Number trees	DBH (cm)	Height (m)	Number trees	DBH (cm)	Height (m)
Lic (8)	70	5	23.3 ± 2.2	14.2 ± 1.5	5	16.2 ± 0.5	12.7 ± 0.8	3	10.0 ± 0.7	9.8 ± 0.7
Lin (1)	80	4	21.1 ± 1.5	19.6 ± 0.9	3	13.5 ± 0.2	16.9 ± 0.6	3	9.4 ± 0.3	14.2 ± 0.6
Lin (9)	70	3	20.7 ± 1.9	21.8 ± 1.1	3	15.8 ± 1.1	19.8 ± 2.1	3	10.2 ± 1.1	14.5 ± 1.0
Blu (2)	80	5	32.4 ± 2.2	26.5 ± 1.2	3	23.4 ± 1.0	25.0 ± 1.1	3	16.7 ± 0.7	21.9 ± 2.0
Lin (10)	80	5	18.4 ± 2.0	15.5 ± 1.7	–	–	–	–	–	–
Lin (12)	80	5	12.6 ± 2.0	10.0 ± 0.4	–	–		–	–	–
a Forest types are: Lic – lichen pine forest, Lin – lingonberry pine forest, Blu – blueberry pine forest.

 | Show Table

DownLoad: CSV

The influence of climatic factors on the studied indicators was studied using the example of Lin in different climatic conditions: (ⅰ) the middle taiga subzone (SPs 1, 9), (ⅱ) the northern taiga subzone (SP 10, 65°22′9″ N, 31°7′44″ E), and (ⅲ) the transition area of the northern taiga subzone and tundra (SP 12, 68°59′45″ N, 28°59′19″ E) (Supplemental Materials Fig. S1).

2.2   Establishment of sample plots

Sample plots were established according to the methods generally accepted in forest assessment. The age of the sampling cores was determined with the Pressler increment borer (Haglöf, Sweden) at the root collar in 15–20 trees. A continuous count of trees was carried out in 2-cm steps of diameter on the sample plots. The heights were measured for 20–25 trees for pine and 10–15 trees for other species (spruce and birch). At the SP, natural regeneration was assessed by size categories (up to 0.25, 0.26–0.50, 0.51–1.00, 1.01–1.50, 1.51–2.00 ​m, and more than 2.00 ​m), taking into account its viability. In addition, the nature of the distribution of undergrowth over the area was determined. The coarse woody debris was classified into three categories: windblow, windfall, deadwood. The following stand characteristics were determined: tree species composition, tree density, trunk volume of trees and others (Supplemental Materials Table S1).

2.3   Core sampling and determination of the SW and HW transition zone

To determine the HW proportion, cores, containing the pith, were taken from model trees at different trunk heights: 0.3, 1.5, 4.5, 7.5, 10.5, 13.5 ​m. For short model trees with a thin trunk, sampling was limited to 4.5 or 6–7 ​m; for some model trees, cores were additionally taken at a height of 3.0, 6.5, 9.0 ​m. Missing data at some heights are related to the morphological characteristics of the tree: tree height and the height of the lowest living branch. In all cases, trees were cored out from the south side of the trunk. Potassium nitrite (6% aqueous solution) and sulfanilic acid (1% aqueous solution) were used to determine the transition zone between SW and HW (Cummins, 1972; Fries et al., 2000). Initially, the mixture of reagents is transparent. As a result of this treatment, SW remains a light-yellow color, and HW is dyed an intense red color (Supplemental Materials Fig. S2). This qualitative reaction to pinosylvin, a phenolic substance from the stilbene class, is widely found in P. sylvestris heartwood (Bergström et al., 1999; Ekeberg et al., 2006).

To determine the carbon, cellulose, lignin and extractives contents in SW and HW, additional cores were taken at a height of 1.5 ​m. At SPs 1, 2, 8, 9, core sampling was carried out at the end of May 2021 (determination of cellulose, lignin and extractives contents) and at the beginning of October 2022 (determination of carbon and cellulose, lignin and extractives contents). At SPs 10, 12, core sampling was carried out in early October 2021 and in late June – early July 2022 (determination of carbon and cellulose, lignin and extractives contents). To study the effect of sampling height on carbon and extractive compound content in trees at SP 12 in late June – early July 2022, cores were taken at different stem heights (0.3, 1, 2, 3, 4, 5, 6, 7, 8 ​m). For chemical analysis, the cores were frozen in liquid nitrogen and stored in a low-temperature freezer at −70 ​℃. Then, the samples were freeze-dried immediately before analysis.

2.4   Morphological studies

The cores were photographed using an ADF S645 stereomicroscope equipped with a 10-Mp camera (ADF OPTICS CO LTD, China), and the following parameters were determined on the resulting images using ImageJ image analysis software: number of annual rings in xylem (cambial age, C.A.), number of annual rings in heart- and sapwood (HW and SW), radial width (R(W)), radial width in heart- and sapwood (R(HW) and R(SW)). Based on the data obtained, the cross-sectional area of xylem (S(W)), heartwood area (S(HW)), sapwood area (S(SW)) and heartwood proportion (HW (%)) were calculated as follows:

2.5   Biochemical research

Mechanical trituration of the freeze-dried SW and HW samples to homogenous powder by laboratory mill was used. The carbon concentrations were estimated for freeze-dried material (1.5–2.5 ​mg) using the PerkinElmer 2400 Series Ⅱ CHNS/O Analyzer (USA), calibrated with the organic analytical standard acetanilide (PerkinElmer N 0204–1121).

For easier extraction, the SW and HW samples were transformed into sawdust by laboratory mill. The samples were prepared to a particle size 0.25–1.00 ​mm. The classified sawdust was stored in an air-conditioned room then it was used for the extraction (Rodrigue et al., 2016; Brocco et al., 2017; Warlo et al., 2023). Extractives content was made in absolute ethanol as recommended by the “Technical Association of the Pulp and Paper Industry” – TAPPI T 264 om-88 (1992) and determined gravimetrically and expressed as a percentage of the total mass of a sample (TAPPI, 1992, TAPPI, 2011).

Then this material that is free from resinous substances (Geles, 2001; Arisandi et al., 2019) was used for lignin determination. Lignin was isolated by acid hydrolysis in 72% sulfuric acid. The lignin content was determined gravimetrically (TAPPI protocol, 2011) and expressed as a percentage of the total mass of a sample.

Cellulose was isolated by the Kürschner-Hoffer method using a mixture of concentrated nitric acid and ethanol in a volume proportion of 1/4 (Kürschner and Hoffer, 1929; Ghavidel et al., 2020; Krutul et al., 2021). Cellulose content was determined gravimetrically. The cellulose content was expressed as a percentage of the total mass of a sample.

2.6   Experimental design

The research experiment design and description of the analyses of the selected plant material are summarized in Table 2.

Table  2.  Summary of the research experiment design and analysis of selected plant materials.

Factors			Core sampling from model trees at trunk heights (m)	Morphological studies (determination the HW proportion at trunk heights)	Biochemical research (at trunk heights 1.5 m)
The influence of climatic factorsa	The influence of soil conditions (forest typeb)	The influence of the position of the tree in the stand (social class c )			The C concentration (%)	Extractives content (%)	Cellulose content (%)	Lignin content (%)
MT	Blu	D.t.	0.3, 1.5, 4.5, 7.5, 10.5	+	+	+	+	+
		I.t.	0.3, 1.5, 4.5, 7.5, 10.5, 13.5	+
		S.t.	0.3, 1.5, 4.5, 7.5, 10.5	+
	Lin	D.t.	0.3, 1.5, 4.5, 7.5, 10.5	+	+	+	+	+
		I.t.	0.3, 1.5, 3, 4.5, 6, 7.5, 10.5	+	+	+	+	+
		S.t.	0.3, 1.5, 3, 4.5, 6, 7.5, 9, 10.5, 13	+	+	+	+	+
	Lic	D.t.	0.3, 1.5, 4.5, 5.5, 6.5, 7.5, 8.5	+	+	+	+	+
		I.t.	0.3, 1.5, 3.0, 4.5	+
		S.t.	0.3, 1.5, 3.0, 4.5, 6,5, 8.5	+
NT	Lin	D.t.	0.3, 1.5, 4.5, 7.5, 10.5	+	+	+	+	+
NT/T	Lin	D.t.	0.3, 1.5, 4.5, 7.5, 9 and 1, 2, 3, 4, 5, 6, 7, 8	+	+	+	+	+
a Different climatic condition are: MT – middle taiga subzone, NT – northern taiga subzone, NT/T – transition area of the northern taiga subzone and tundra. b Forest types are: Blu – blueberry pine forest, Lin – lingonberry pine forest, Lic – lichen pine forest. c Social class are: D.t. – dominant trees, I.t. – intermediate trees, S.t. – suppressed trees.

 | Show Table

DownLoad: CSV

2.7   The evaluation of wood carbon content in different forest types

Based on the data obtained, including the height of the core sampling (h, m), the radial width (R(W), m) and the radial width in the heartwood (R(HW), m) of the cross section, the volume of the trunk section (V(t.s.), m³) and the volume of the heartwood (V(HWt.s.), m³) and sapwood (V(SWt.s.), m³) sections were calculated using the partial cone formula:

where h – length of the tree trunk section, m (distance between the heights of the core sampling); R(W)1, R(HW)1 – the radial width (R(W)) and radial width in heartwood (R(HW)) of the trunk section base; R(W)2, R(HW)2 – the radial width (R(W)) and radial width in heartwood (R(HW)) of the trunk section top.

Wood volume (V), HW and SW volums (V(HW) and V(SW)) were calculated as the sums of V(t.s.), V(HWt.s.), and V(SWt.s.), respectively. In order to assess the influence of soil conditions and tree position in the stand, the calculations were limited to a height of 7.5 ​m. This was done because cores at this height were taken from all trees, allowing comparison of trees from different forest conditions and different social classes. To assess the influence of climatic conditions, the calculations were to a height of 9.0–10.5 ​m. Trunk weight (m, kg), SW trunk weight (m(SW), kg), and HW trunk weight (m(HW), kg) were calculated using a wood density (WD) of 0.440 ​kg⋅m⁻³ - the average value for the lingonberry and blueberry pine forests in Karelia (calculated according to Kazimirov et al., 1977).

The tree biomass was converted to carbon (mС(trunk), kg⋅tree⁻¹) by two ways: (1) using the coefficient of 0.5 (the generic assumption that aboveground biomass consists of 50% carbon); (2) using coefficients for conversion of SW and HW biomass to carbon (calculated as the average carbon concentration in SW or HW for Blu D.t., Blu I.t., Blu S.t., Lin D.t., Lin I.t., Lin S.t., Lic. D.t., Lic I.t., Lic S.t.., MT, NT, NT/T respectively).

The carbon content in trunk (mС(trunk), Mg⋅ha⁻¹) was converted to carbon content in forest stand according to the formula:

where N – number of trees (trees⋅ha⁻¹) on sample plots.

We took into account the proportion of suppressed trees, intermediate trees and dominant trees in the sample plots (SP 2 (Blu), SP 1, 9 (Lin), and SP 8(Lic)) (see Table 3). The suppressed group included trees whose DBH was 30% or more below the average DBH for the forest stand. Trees whose DBH was 30% or more above the average for the forest stand were classified as dominant. The remaining trees were classified as intermediate.

Table  3.  The estimates of wood carbon content in different forest types.

Forest type/social classa	m (pats of trunk) up to 7.5 m (kg)			The coefficient for conversion of SW and HW biomass to carbon		Carbon content in trunk up to 7.5 m (kg·tree−1)		Trees per ha	Carbon content, (Mg·ha−1)		Errors in forest carbon stock due to 50% assumption (%)
	m (SW)	m (HW)	m (trunk)	SW	HW	Data set 1b	Data set 2b		Data set 1b	Data set 2b
Blueberry pine forest									45.59	42.09	7.7
D.t.	100.2	88.6	188.8	0.530	0.551	101.9	94.4	65	6.62	6.14
I.t.	60.1	45.7	105.8	0.540	0.545	57.3	52.9	625	35.81	33.05
S.t.	27.5	18.9	46.5	0.537	0.553	25.3	23.2	125	3.16	2.90
Lingonberry pine forest									35.60	33.14	6.9
D.t.	52.0	12.5	64.5	0.522	0.541	33.9	32.3	397	13.45	12.79
I.t.	32.5	9.4	41.9	0.546	0.550	22.9	21.0	693	15.88	14.52
S.t.	10.5	4.5	15.0	0.536	0.543	8.1	7.5	779	6.27	5.83
Lichen pine forest									30.45	28.17	4.8
D.t.	72.8	6.4	79.2	0.530	0.551	42.5	39.6	383	16.28	15.18
I.t.	25.4	3.7	29.0	0.540	0.545	16.0	14.5	525	8.42	7.63
S.t.	10.5	4.5	15.0	0.533	0.563	7.5	7.0	767	5.75	5.37
a Social class are: D.t. – dominant trees, I.t. – intermediate trees, S.t. – suppressed trees. b Data set 1 are: tree biomass was converted to carbon using the coefficient for HW and SW obtained in our studies. Data set 2 are: tree biomass was converted to carbon using a coefficient of 0.5.

 | Show Table

DownLoad: CSV

2.8   Data analysis

The results were statistically processed with PAST (version 4.0). Before starting the statistical analysis, raw data was initially tested for normality using the Shapiro–Wilk test. The significance of differences between variants was estimated by Mann–Whitney U test. Differences were significant at p ​< ​0.05. The data in the diagrams and tables are presented as mean (M) ​± ​SE, where SE is the standard error. In the diagrams, different letters denote statistically significant differences between groups.

Principal component analysis (PCA) and analysis of similarities (ANOSIM) were carried out for different groups of data set according to climate, social class and tissue specificity. Before the calculations, the initial data were standardized. Correlation analysis was performed by Pearson correlation test.

All assays were performed at the Core Facility of the Karelian Research Centre RAS.

3.   Results

3.1   Heartwood proportion varies with tree height, social class, soil and climate condition

The influence of different soil conditions and social classes on heartwood proportion in pine trees was studied using the example of the middle taiga subzone. Previously, we studied the HW proportion in the trunk of 70–80 year old dominant trees (D.t.) in the ecological series: blueberry pine forest (SP 2) – lingonberry pine forest (SPs 1, 9) – lichen pine forest (SP 8) (Galibina et al., 2022a, b; Tarelkina et al., 2022). The datasets are in free-to-access public data repository (Dryad, dryad.prr4xgxnw.2022). Thus, in blueberry pine forest, the HW percentage was the highest and amounted to over 50% up to 1/3 of the height of the tree; in lingonberry pine forest conditions the HW proportion averaged 20%–30% up to half of the tree trunk, while in lichen pine forest HW proportion in the trunk was about 5%–6% up to 6.5 ​m (about 1/3 of the trunk), and in some models HW was not detected already at the height of 6.5 ​m (Fig. 1).

Figure  1.  Boxplot displaying the heartwood proportion (%) with tree height (m) in Scots pine trees depending on the social class (dominant (D.t.), intermediate (I.t.) and suppressed (S.t.) trees) in blueberry (Blu), lingonberry (Lin) and lichen (Lic) pine forest (middle taiga subzone). The ordinate axis shows the height (m) of stem tissue sampling.

DownLoad: Full-Size Img PowerPoint

In blueberry pine forest HW proportion in the trunk did not differ significantly in different social classes up to 7.5 ​m and was 45%–40% for dominant (Blu (D.t.)), intermediate (Blu (I.t.)) and suppressed (Blu (S.t.)) trees. At a height of 10.5 ​m, these values were lower for intermediate (40%–30%) and suppressed trees (30%–20%) compared to dominant trees (50%). In lingonberry pine forest we did not find significant differences in the HW proportion in the trunk between dominant (Lin (D.t.)) and intermediate trees (Lin (I.t.)). The suppressed trees (Lin (S.t.)) were distinguished by a higher HW proportion in the trunk section from 1.5 to 7.5 ​m (32%–36%). Unexpected results were obtained for the lichen pine forest. Both intermediate trees (Lic (I.t.)) and suppressed trees (Lic (S.t.)) contained a higher HW percentage in throughout the trunk compared to dominant trees (Lic (D.t.)) (Fig. 1, Supplemental Materials Table S2).

Thus, high soil fertility has a positive effect on HW formation, while deteriorating soil conditions lead to a decrease in the HW proportion in pine trunk. Social class affects HW formation only when soil conditions deteriorate, with the HW share increasing. The pine trees from different soil conditions and different social class are arranged in order of decreasing the HW proportion in trunk as follows: Blu D.t. (47%), Blu S.t. (39%), Blu I.t. (37%), Lin S.t. (28%), Lin I.t. (22%), Lin D.t. (18%), Lic S.t. (12%), Lic I.t. (11%), Lic. D.t. (8%) (Fig. 1).

To further compare heartwood proportion at the base of the trunk (0.3–10.5 ​m) in pine trees under different forest types and under different social class analysis of similarities (ANOSIM) was performed. The results showed that heartwood proportion in pine trees did not differ between dominant (D.t.), intermediate (I.t.) and suppressed (S.t.) ones (Fig. 2a) and differed between trees from Blu, Lin and Lic (middle taiga subzone) (Fig. 2b). So, forest type had a greater impact on the heartwood proportion compared with the social class (Fig. 2).

Figure  2.  ANOSIM analysis of heartwood proportion (%) in pine trees under different social class (a) (dominant (D.t.) intermediate (I.t.) and suppressed (S.t.)) and under different forest types (b) (blueberry (Blu), lingonberry (Lin) and lichen (Lic) pine forests (middle taiga subzone)). Ordination of 49 model pine trees (c) based on heartwood proportion (%) at the heights 0.3, and 1.5, and 4.5 ​m growing in different forest types (Blu, Lin and Lic) and with the different social class (D.t., I.t. and S.t.). Factor 1 (85.5% of the variance) correlated with the height 0.3 ​m (r = ​0.55), 1.5 ​m (r = ​0.60) and 4.5 ​m (r = ​0.58). Factor 2 (11.6% of the variance) correlated with the height 0.3 ​m (r = ​0.80). The red group represents trees from lichen pine forest. The green group represents trees from lingonberry pine forest. The blue group represents trees from blueberry pine forest. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

DownLoad: Full-Size Img PowerPoint

Then principal component analysis (PCA) was carried out in a data set on the heartwood proportion in pine trees at the heights 0.3, 1.5, and 4.5 ​m under different forest types and under different social class. Its results confirmed that investigated trees were divided into separate groups based on growing in different forest conditions (Fig. 2c).

Heartwood proportion in pine trees under different climatic conditions was studied in 70–80-year-old Lin communities, using dominant trees as an example. In trees growing in the northern taiga subzone (NT) (SP 10) (Fig. 3) we did not find significant differences in the HW percentage in the trunk compared to those in pines from the middle taiga subzone (MT) (SPs 1, 9) (Dryad, dryad.prr4xgxnw.2022). A distinctive feature of these trees was a higher percentage of HW by 3–4.5 ​m compared to the trunk base. In pine trees from the middle taiga subzone, no such features were observed. Pine trees growing in the transition area of the northern taiga subzone and tundra (NT/T) (SP 12) had a lower stem HW percentage compared to SPs 1, 9 and 10, especially starting from the height of 3 ​m. On average, the HW percentage in the 0.3–5.0 ​m section of the trunk was 10%–12%. It should be noted that, in dominant trees with SP 12, HW formation was observed almost up to the trunk top (8–9 ​m) (see Table S3) (Fig. 3). ANOSIM showed that heartwood proportion in pine trees differed slightly between different climate conditions (R ​= ​0.2576, p ​= ​0.0417).

Figure  3.  Boxplot displaying the heartwood proportion (%) with tree height (m) in dominant Scots pine trees in lingonberry pine forest (middle taiga subzone (MT), northern taiga subzone (NT) and transition area of the northern taiga subzone and tundra (NT/T)). The ordinate axis shows the height (m) of stem tissue sampling.

DownLoad: Full-Size Img PowerPoint

3.2   Carbon concentration and related chemical composition in sapwood and heartwood vary with stem heights, social class, soil and climate conditions

In dominant trees in the forests of transition zone between the northern taiga subzone and tundra, chemical content in trunk tissues was studied as a function of height. Higher carbon concentration in HW compared to SW was observed at 3 ​m (by 2.7%), 4–5 ​m (by 1.7%) and 6–7 ​m (by 1.7%) (Fig. 4a). Differences in extractives content between SW and HW were found at all sampling heights and amounted to 5.7%, 2.5%, 2.6%, 6.6% and 8.3% for 0.3, 1–2, 3, 4–5 and 6–7 ​m, respectively (Fig. 4b). There were no differences in carbon concentration and extractives content between SW sampled at different stem heights. Whereas there was higher carbon concentration and extractives content in HW at 3, 4–5 and 6–7 ​m (Fig. 4a). HW had 5.2%, 5.6%, 7.4%, 7.4% and 12.1% less cellulose than SW for 0.3, 1–2, 3, 4–5 and 6–7 ​m (Fig. 4c). Lignin content was the same in HW and SW (Fig. 4d). ANOSIM showed significant differences for SW and HW (R ​= ​0.1323, p ​= ​0.0057) and did not differ in depending on height of stem tissue sampling (R ​= ​0.0237, p ​= ​0.0735). In addition, all SW and HW samples showed a decrease in cellulose content with increasing carbon concentration (R ​= ​−0.27; p ​= ​0.02), while cellulose content was negatively correlated with extractive content (R ​= ​−0.59; p ​= ​0.0002).

Figure  4.  Boxplot displaying the carbon concentration (a) and extractives (b), cellulose (c), lignin (d) contents with tree height (m) in heartwood (HW) and sapwood (SW) in dominant Scots pine trees in the transition area of the northern taiga subzone and tundra. The abscissa shows the height (m) of stem tissue sampling. Significant differences between SW and HW at p ​< ​0.05 are marked by different letters.

DownLoad: Full-Size Img PowerPoint

Lingonberry pine forest was used to illustrate how the tree's position in the forest stand influences carbon, extractives, cellulose, and lignin content in SW or HW (Fig. 5). The cores were taken at a height of 1.5 ​m off the ground. Dominant trees carbon concentration in HW was 1.9% higher compared to SW. In intermediate and suppressed trees, we did not find significant differences in carbon concentration between SW and HW. We did not find significant differences in carbon concentration in HW between trees of different social class. However, carbon concentration in SW was higher in intermediate and suppressed trees, compared to dominant trees, by 2.4% and 1.4% (Fig. 5a). We found that extractives content in HW exceeded that in SW by 5% and 2.1% in dominant and suppressed trees respectively. No differences were between SW and HW in intermediate trees. We found that intermediate and suppressed trees had higher extractives content in SW and lower in HW compared to dominant trees (Fig. 5b). The dominant, intermediate and suppressed trees showed differences in cellulose content between SW and HW (Fig. 5c). Lignin content did not differ between SW and HW in all social class (Fig. 5d). We found a negative correlation between the extractives content and the cellulose content (R ​= ​−0.54; p ​= ​0.002) in SW and HW on dominant, intermediate and suppressed trees growing in Lin (middle taiga subzone).

Figure  5.  Boxplot displaying the carbon concentration (a) and extractives (b), cellulose (c), lignin (d) contents in heartwood (HW) and sapwood (SW) in Scots pine trees depending on the social class (dominant (D.t.), intermediate (I.t.) and suppressed (S.t.) trees) in lingonberry pine forest (middle taiga subzone). Sampling of tissues was at a height of 1.5 ​m. Significant differences between SW and HW at p ​< ​0.05 are marked by different letters.

DownLoad: Full-Size Img PowerPoint

ANOSIM was conducted to compare the carbon concentration, extractives and cellulose contents (chemical composition of wood) of pine trees in SW or HW depending on the social class of the tree (Fig. 6). The results showed that chemical composition of wood in pine trees differed depending on SW or HW affiliation (Fig. 6a) and did not differ between dominant (D.t.), intermediate (I.t.) and suppressed (S.t.) ones (Fig. 6b). We assessed the influence of the social class on carbon concentration, extractives and cellulose contents separately in each forest type (Fig. 6). Principal component analysis (PCA) confirmed that SW and HW were divided into separate groups based on chemical composition of wood (Fig. 6c).

Figure  6.  ANOSIM analysis of carbon concentration, extractives and cellulose contents (%) in pine trees (lingonberry pine forests, middle taiga subzone) depending on the sapwood (SW) or heartwood (HW) (a) and under different social class dominant (D.t.), intermediate (I.t.) and suppressed (S.t.) (b). Ordination of 90 model pine trees (c) based on carbon concentration, extractives and cellulose contents (%) in SW and HW at the heights 1.5 ​m, growing in lingonberry pine forests (middle taiga subzone) with the different social class (D.t., I.t. and S.t.). Factor 1 (52.2% of the variance) was correlated with the cellulose (r = ​−0.67) and extractives (0.71) contents. Factor 2 (34.0% of the variance) was correlated with the carbon (r = ​0.94) concentration. The orange group represents indicators for SW. The brown group represents indicators for HW. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

DownLoad: Full-Size Img PowerPoint

To assess the heartwood contribution to carbon accumulation in stem wood under different soil conditions and to investigate the relationship between accumulation of extractives in SW and HW and carbon concentration, cores were taken at a height of 1.5 ​m from the ground from dominant trees. The carbon concentration in HW was 2.2%, 1.9% and 3.4% higher compared to SW in Blu, Lin and Lic, respectively. It should be noted that in Lic, carbon concentration in SW and HW was the highest compared to other forest types (Fig. 7a). We found extractives content in HW exceeded that in SW by 4.2%, 5.0% and 9.6% in Blu, Lin and Lic respectively. The highest content of extractives was inherent in plants from lichen pine forest, where their intensive accumulation in HW was observed. The lowest content of extractives in SW was observed in the lingonberry pine forest (Fig. 7b). We showed that the cellulose content of HW compared to SW was 5.5%, 4.5% and 8.7% lower for blueberry, lingonberry and lichen pine forest, respectively (Fig. 7c). The lignin content did not differ between SW and HW in dominant trees under different soil conditions (Fig. 7d). We found a negative correlation between the carbon concentration and the cellulose content (R ​= ​−0.47; p ​= ​0.001) in SW and HW on dominant Scots pine trees growing in Blu, Lin, and Lic (middle taiga subzone).

Figure  7.  Boxplot displaying the carbon (a), extractives (b), cellulose (c) and lignin (d) content in sapwood (SW) and heartwood (HW) in dominant Scots pine trees growing in blueberry (Blu) pine forest, lingonberry pine forest (Lin) and lichen pine forest (Lic) (middle taiga subzone). Sampling of tissues was at a height of 1.5 ​m. Significant differences between SW and HW at p ​< ​0.05 are marked by different letters.

DownLoad: Full-Size Img PowerPoint

To compare carbon, extractives and cellulose content in pine trees depending on the SW or HW and under different forest types ANOSIM was performed (Fig. 8). The results showed that chemical composition of wood in pine trees differed depending on SW or HW affiliation (Fig. 8a) and differed slightly between trees from Blu, Lin and Lic (middle taiga subzone) (Fig. 8b). So, SW or HW affiliation had a greater impact on the carbon, extractives and cellulose content comparing with the forest type. Principal component analysis (PCA) confirmed that SW and HW were divided into separate groups based on the chemical composition of wood (Fig. 8c).

Figure  8.  ANOSIM analysis of carbon concentration, extractives and cellulose contents (%) in pine trees (dominant trees, middle taiga subzone) depending on the sapwood (SW) or heartwood (HW) (a) and under different forest types (blueberry (Blu), lingonberry (Lin) and lichen (Lic) pine forests) (b). Ordination of 96 model pine trees (c) based on carbon concentration, extractives and cellulose contents (%) in SW and HW, growing in different forest types (Blu, Lin and Lic pine forests). Factor 1 (48.4% of the variance) was correlated with the carbon concentration (r = ​0.61) and cellulose (r = ​−0.64) content. Factor 2 (29.3% of the variance) was correlated with the extractives content (r = ​0.88). The orange group represents indicators for SW. The brown group represents indicators for HW. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

DownLoad: Full-Size Img PowerPoint

To assess the heartwood contribution to carbon accumulation in stem wood in different climate conditions and to investigate the relationship between accumulation of extractives in SW and HW and carbon concentration, cores were taken at a height of 1.5 ​m from the ground from dominant trees. The deterioration of climatic conditions was accompanied by an increase in carbon concentration in SW (by 3.8% and 2.7%) and in HW (by 2.0% and 2.1%) in trees from northern taiga subzone and transition area of the northern taiga subzone and tundra, respectively, compared with trees from middle taiga subzone. Probably for this reason, we did not find a difference in carbon concentration between SW and HW in trees growing in more northern regions (Fig. 9a). The increase in these plants in SW carbon concentration was accompanied by an increase in extractive compound content of 1.1%–2.0%, compared with plants from the middle taiga subzone. A difference in extractive compound content between SW and HW in northern pine plants was observed (Fig. 9b). We found that pine trees from the northern taiga subzone and transition area of the northern taiga subzone and tundra had 4.3% and 5.7% lower cellulose content in HW, respectively, compared to SW (Fig. 9c). However, lignin content did not differ (Fig. 9d). We found a positive correlation between the carbon concentration and the extractives content (R ​= ​0.31; p ​= ​0.007) and a negative correlation between the carbon concentration and the cellulose content (R ​= ​−0.28; p ​= ​0.03) in SW and HW on trees growing under different climate conditional (MT, NT, and NT/T).

Figure  9.  Boxplot displaying the carbon concentration (a), extractives (b), cellulose (c), lignin (d) contents in heartwood (HW) and sapwood (SW) in dominant Scots pine trees (lingonberry pine forest) in the climate gradient (the middle taiga subzone (MT), northern taiga subzone (NT) and the transition area of the northern taiga subzone and tundra (NT/T)). Sampling of tissues was at a height of 1.5 ​m. Significant differences between SW and HW at p ​< ​0.05 are marked by different letters.

DownLoad: Full-Size Img PowerPoint

Then ANOSIM was carried out in a data set on the carbon concentration, extractives and cellulose contents at the height of 1.5 ​m in dominant Scots pine trees in the gradient of climate conditional (the middle taiga subzone, the northern taiga subzone and the transition area of the northern taiga subzone and tundra). Lignin content was not used for the analysis because it did not differ between the studied variants. In studied climate conditions, significant differences were observed for SW and HW (Fig. 10a). Also, the results showed that chemical composition of wood in pine trees differed in the climate gradient (Fig. 10b). The greatest differences were between trees from the middle taiga subzone and trees from more northern regions (NT and NT/T). Results of principal component analysis (PCA) confirmed that investigated trees were divided into separate groups based on different carbon concentration, extractives and cellulose contents in SW and HW (Fig. 10c) and into two groups in depending on climate conditional (Fig. 10d).

Figure  10.  ANOSIM analysis of carbon concentration, extractives and cellulose contents (%) in pine trees (dominant trees, lingonberry pine forest) depending on the sapwood (SW) or heartwood (HW) (a) and under different climate conditional (the middle taiga subzone (MT), the northern taiga subzone (NT) and the transition area of the northern taiga subzone and tundra (NT/T)) (b). Ordination of 122 model pine trees based on carbon concentration, extractives and cellulose contents in SW and HW, growing in the gradient of climate conditional (MT, NT and NT/T) (c, d). Factor 1 (52.5% of the variance) correlated with the content of extractives (r = ​0.63) and the content of cellulose (r ​= −0.11). Factor 2 (27.7% of the variance) correlated with the carbon concentration (r = ​0.87). The orange group represents SW. The brown group represents HW. The red group represents trees from the middle taiga subzone. The green group represents trees from the northern taiga subzone. The blue group represents trees from the transition area of the northern taiga subzone and tundra. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

DownLoad: Full-Size Img PowerPoint

3.3   Wood carbon content varies with social class, soil condition/forest type, climate conditions

We have attempted to estimate of aboveground trunk wood carbon stock using (ⅰ) the assumption that aboveground biomass consists of 50% C; and (ⅱ) the values of the carbon concentration in SW or HW in different soil conditions and depending on social class determined in this study. To calculate the mass of trunk wood and the mass of heartwood and sapwood in the trunk, the wood density of 0.440 ​kg·m⁻³ previously determined for lingonberry and blueberry pine forests in Karelia ( Kazimirov et al., 1977 ) was used. The values of wood carbon content, calculated using the standard 50% conversion factor, represent underestimates of 7.7%, 6.9%, and 7.5% for Bly, Lin, and Lic, respectively (Table 3). To be able to compare different sample plots, calculations were carried out up to 7.5 ​m trunk, which is 48%, 59%, and 65% of the trunk volume for Bly, Lin, and Lic, respectively. The underestimation of wood carbon content on Blu reached 3.5 ​Mg ​C·ha⁻¹ and occurred mainly due to intermediate trees (2.8 ​Mg ​C·ha⁻¹), accounting for 77% of the total trees content. The underestimation of wood carbon content on Lin reached 2.5 ​Mg ​C·ha⁻¹ and occurred due to intermediate trees (1.4 ​Mg ​C·ha⁻¹), and dominant trees (0.7 ​Mg ​C·ha⁻¹). The underestimation of wood carbon content on Lic reached 2.3 ​Mg ​C·ha⁻¹ and occurred mainly due to dominant trees (1.1 ​Mg ​C·ha⁻¹), and intermediate trees (0.8 ​Mg ​C·ha⁻¹).

We estimated the wood carbon content in different climate conditions for only dominant trees (Table 4). Using a coefficient of 0.5, the values of wood carbon content are underestimated by 4.9%, 10.8% and 9.2% for the middle taiga subzone (MT), the northern taiga subzone (NT) and the transition zone between the northern taiga subzone and the tundra (NT/T), respectively (Table 4). The underestimation of wood carbon content reached 0.8, 3.3, and 0.2 ​Mg ​C·ha⁻¹ for MT, NT, and NT/T, respectively. Calculations were carried out up to 10.5 ​m trunk, which is 66% and 85% of the trunk volume for MT and NT, respectively. For NT/T calculations were carried out up to 9 ​m trunk (97% of trunk volume). It should be noted that high values of the underestimation of wood carbon content on northern taiga subzone are associated, on the one hand, with a large number of trees (Table 4), and on the other hand, with high carbon concentration in SW and HW (Fig. 9a).

Table  4.  The estimates of wood carbon content in different climate conditions.

Different climatic condition a	m (pats of trunk) up to 9–10 m (kg)			The coefficient using conversion of SW and HW biomass to carbon		Carbon content in trunk up to 9–10 m (kg·tree−1)		Trees per ha	Carbon content (Mg·ha−1)		Errors in forest carbon stock due to 50% assumption (%)
	m (SW)	m (HW)	m (trunk)	SW	HW	Data set 1 b	Data set 2 b		Data set 1 b	Data set 2 b
MT	62.8	14.6	77.5	0.522	0.541	40.7	38.7	397	16.14	15.36	4.8
NT	57.6	16.6	74.3	0.56	0.562	41.6	37.1	740	30.82	27.49	10.8
NT/T	24.8	3.1	27.9	0.55	0.56	15.4	14	150	2.31	2.09	9.5
a Different climatic condition are: MT – middle taiga subzone, NT – northern taiga subzone, NT/T – transition area of the northern taiga subzone and tundra. b Data set 1 are: tree biomass was converted to carbon using the coefficient for HW and SW obtained in our studies. Data set 2 are: tree biomass was converted to carbon using a coefficient of 0.5.

 | Show Table

DownLoad: CSV

4.   Discussion

An assessment of the value of forest ecosystem services found that 65%–90% of their global value is due to their climate-regulating function (Boston Consulting Group Report, 2020). For Russia, it ranges from 81% to 96%. The distribution and long-term sequestration of carbon in woody biomass is one of its main pools in forest ecosystems (Law et al., 2001; Hoch and Körner, 2011; Kraenzel et al., 2003; Peri et al., 2010; Liu et al., 2016; Mvolo et al., 2021; Cabon et al., 2022). The precise knowledge of the variation in carbon content of living tree biomass is required to accurately quantify forest C stocks, validate forest C accounting models and support forest management strategies intended to maximize CO₂ sequestration (Thomas and Martin, 2012; Gao et al., 2016; Zhou et al., 2019). To date, evaluating carbon content (mass of carbon per ha) in forest trees was mostly based on converting the biomass per ha into a carbon content per ha, using the carbon concentration in the biomass according to Kollmann (1959) and the Intergovernmental Panel on Climate Change (IPCC) recommendations (Penman et al., 2003). However, it was shown that the carbon concentration may range from 47% to 59%, as a function of tree compartment or species (Lamlom and Savidge, 2003; Bert and Danjon, 2006), and depends on the HW proportion in the tree trunk, where the carbon concentration significantly exceeds that in SW (Peri et al., 2010, Herrero de Aza et al., 2011). For the first time in the North-West of Russia, we obtained on a large array of data, including different soil conditions, climate and social class, and found that for 80-year-old pine trees the carbon concentration in SW and HW was 52%–55% and 54%–58%, respectively. We studied the relationships between carbon concentration and extractives, cellulose, and lignin contents in HW and SW of pine trees. We have shown that these relationships were ambiguous and varied in different soils, climate conditions, and the position of the tree in the stand (social class). In general, both the HW proportion on tree trunks, and the carbon concentration in SW and HW influence on carbon storage in pine stands.

4.1   Patterns of heartwood formation depending on growing conditions and the position of the tree in the stand

Attempts to evaluate the patterns of HW formation in plants of the genus Pinus L. have been made repeatedly (Wilkes, 1991; Björklund, 1999; Bergstrom, 2003; Climent et al., 2003; Gjerdrum, 2003; Flæte and Høibø, 2009; Pinto et al., 2004; Uusitalo, 2004; Knapic and Pereira, 2005). Björklund (1999), Uusitalo (2004), and Flæte and Høibø (2009) conducted studies of HW formation in P. sylvestris in Norway and Sweden. These studies proposed models that describe the relationship between the number of HW rings and cambial age (number of annual rings). Despite the fact that in the Leningrad region (Forest Plan of the Leningrad Region for 2019–2028) pine forests occupy almost 32% of the forest fund area, in the Arkhangelsk region 27% (Forest plan of the Arkhangelsk Region for 2019–2028), in the Murmansk region 43% (Forest Plan of the Murmansk Region for 2019–2028), and in the Republic of Karelia 65% (Anan'ev and Moshnikov, 2016), such large-scale studies in the North-West of Russia have not been carried out before. We have studied the intensity of the heartwood formation process depending on the tree age and the cambium age within the same tree in North-West Russia (middle taiga subzone) in Scots pine in different forest conditions for the first time (Galibina et al., 2022a, Galibina et al., 2022b; Tarelkina et al., 2022). Based on the large data set, models of the dependence of the number of annual rings in HW on the cambium age were proposed. In this work, we supplemented our data (Dryad, dryad.prr4xgxnw.2022) with a study of the HW proportion depending on the social class trees under different soil conditions (Supplemental Materials Table S2). In this study, we presented new data for the boreal forests of the North-West of Russia about the heartwood formation in Scots pine depending on cambium age within one tree in the gradient of climatic conditions (middle taiga subzone – northern taiga subzone – transition area of the northern taiga subzone and tundra) (Supplemental Materials Table S3).

We found that the deterioration of climatic conditions (transition area of the northern taiga subzone and tundra compared to the middle taiga subzone) was accompanied in dominant trees by a decrease in the HW proportion in the trunk. There is not much data on the influence of climate on the HW proportion in the trunk of trees, while in most cases they compare drier climate types and wet and high-altitude climate types (Climent et al., 2002; Pérez Cordero and Kanninen, 2003; Moreno Chan et al., 2012).

We found that the deterioration of soil conditions was reflected in dominant trees in a decrease in the HW proportion percentage in the trunk. Several works have shown the influence of soil conditions on the HW formation in various species of pines and other woody plants (Miranda et al., 2007; Bektaş et al., 2020; Almeida et al., 2020). Thus, in Scots pine, in more productive forest types, the HW proportion in the trunk was higher compared to less productive forest stands (Splawa-Neyman and Pazdrowskì, 2001; Jakubowski et al., 2015). On the other hand, in pine stand fertilization did not increase the area or number of HW rings in the stem compared to control (Mörling and Valinger, 1999; Pikk et al., 2006).

In blueberry pine forest, we did not find any differences in the HW proportion in the trunk depending on the social class trees. Suppressed trees in lingonberry pine forest, and especially intermediate and suppressed trees in lichen pine forest had the higher HW proportion in the trunk. So, deterioration in growing conditions activated the HW formation in intermediate and suppressed trees. We were unable to find any data about the relationship between HW and SW in pine trees depending on tree position in the stand. Previously, differences in the anatomical parameters of Scots pine wood in trees of different positions in the forest stand have shown by Fabisiak E. and Fabisiak B. (2021), and Pikiński et al. (2021).

So, HW proportion regulation is an important adaptive trait. Several works have concluded that in water-limiting environments pine HW formation is an adaptation mechanism that regulates stem conductance capacity and thus water use in the tree (Rigling et al., 2002; Sterck et al., 2008; Moreno Chan et al., 2012). We showed that the strategies for HW formation under changing conditions differed in plants of different social classes.

4.2   Carbon concentration and wood chemical composition depending on growing conditions

In tree tissues, carbon is bound in organic compounds, mainly cellulose, hemicelluloses, lignin, extractable components and low molecular weight volatile compounds such as alcohols, phenols, terpenoids and aldehyde (Bārdule et al., 2021). We found that lignin content did not differ between SW and HW, while cellulose content in HW was lower by 4.5%–8.7% compared to SW for dominant trees in different soil and climate conditions. A decrease in the proportion of cellulose in HW compared to SW was observed in different pine species (Campbell et al., 2007; Atac and Eroğlu, 2013; Benouadah et al., 2018), Norway spruce (Bertaud and Holmbom, 2004) and other woody plants (Bahmani et al., 2020; Xiao et al., 2020).

Extractives include carbon-rich polymers: aliphatic (waxes, fatty acids, terpenes, triglycerides) and phenolic compounds (simple phenols, stilbenes, lignans, tannins) (Gutiérrez et al., 2009). In general, HW and SW differ in the qualitative composition of extractives. Whereas SW is dominated by resin acids and triglycerides (Saranpää and Höll, 1987; Maga, 1989; Metsämuuronen and Sirén, 2019), HW is dominated by free fatty acids and phenolic compounds, especially lignans and stilbenes (Maga, 1989; Fengel and Wegener, 1989; Metsämuuronen and Sirén, 2019) that make HW resistant to biodegradation. The higher carbon concentration in the HW compared to the SW may be attributed to higher concentrations of lignin (Scheffer, 1966; Bertaud and Holmbom, 2004; Benouadah et al., 2018), resin acids (Bergstrom, 2003), and various kinds of lipophilic and hydrophilic extractives (Bertaud and Holmbom, 2004; Benouadah et al., 2018). We have extracted extractive substances with ethyl alcohol, so waxes, resinous and fatty acids, essential oils, terpenes, triglycerides, simple phenols, stilbenes, lignans, hydrolysable tannins are extracted. We found that in 80-year-old pine trees, regarless of soil conditions, and climate, and social class, extractives content in SW and HW was 2.9%–5.2% and 5.4%–14.2%, respectively.

Based on all the data for the middle taiga subzone, we were unable to establish correlations between the carbon concentration and extractive compound content in SW and HW. The different soil condition and the position of trees in stand had different effects on the carbon concentration in both SW and HW. We have shown that, under conditions of highly productive blueberry pine forest, the carbon concentration of pine trees in HW (55.0%) was higher than in SW (53.5%). The higher carbon concentration in HW was accompanied by more extractives (4.9% versus 9.1% in SW). Poor soil conditions affected the carbon concentration in SW and HW. In low productive lichen pine forest in HW, both carbon (56.5%) and extractives (14.2%) were higher. In the medium productive lingonberry pine forest, the difference in carbon concentration between HW and SW was observed only for dominant trees. We concluded that the tree position in the stand had no effect on HW chemical composition. However, the carbon concentration and extractive content in SW increased in intermediate and suppressed trees.

There are not many works on the influence of soil fertility on the content of extractives in wood, and the effect of soil condition is not obvious. So, Siwale (2015) showed for Scots pine in southern Finland site soil fertility has a significant effect on the concentrations of pinosylvin, its glycoside and one of the neolignans, but the exact effect was not the same for all compounds. Site soil fertility affected the concentration of some phenolic compounds in Scots pine stemwood, but the exact effect was inconsistent. Latva-Mäenpää et al. (2013) compared stilbene glucosides in Norway spruce roots and stumps grown on either peatland or mineral soil sites, observing that the isorhapontin concentration was especially higher in extracts obtained from the mineral soil than in those from peatland. There are more works on studying the effect of fertilizers on wood quality and the content of extractives in it, however, the data provided in them is contradictory (Viiri et al., 2001; Turtola et al., 2002; Kostiainen et al., 2004; Kaakinen et al., 2007). We believe that soil conditions has a greater impact on HW proportion in the trunk and on the different extractives content in SW.

The deterioration of climatic conditions was accompanied by an increase in carbon concentration in SW against the background of an increase in extractive compound content. We found a positive correlation between the carbon concentration and the extractives content (R ​= ​0.31; p ​= ​0.007) and a negative correlation between the carbon concentration and the cellulose content (R ​= ​−0.28; p ​= ​0.03) in SW and HW on trees growing under different climate conditional. The fact of an increase in extractives in wood with a deterioration in growing conditions was noted earlier (Walker, 2006; Gutiérrez et al., 2009; Verkasalo et al., 2022). For example, an increase in total extractive contents in coniferous species as one moves north has been observed in northern Europe (Hovelstad et al., 2006; Stolter et al., 2010; Verkasalo et al., 2022), Canada (Royer et al., 2013). It was previously shown that unfavorable environmental conditions in northern Finland, compared with southern Finland, could cause higher lignan production in spruce trees, which might be useful for the trees during occasional periods of environmental stresses, such as when the trees could be under microbial attack (Piispanen et al., 2008). Spruce trees in northern Finland have been observed to contain more than twice as many lignans in their knots than in southern Finland (Holmbom, 2011). Kaakinen et al. (2007) confirmed this by finding that the stemwood of Norway spruce had 54% higher concentrations of extractives in a northern site than in a southern site. Hakkila and Verkasalo (2009) reported higher amounts of acetone extractives in pine and spruce stem heartwood and sapwood in northern Finland than in southern Finland. Hakkila (1968) also found a higher total quantity of extractives in pine and spruce stemwood in northern Finland. Siwale (2015) indicated strongly that the formation of vanillic acid in both young and mature Scots pine stem woods is affected by different climatic conditions due to the geographic location, and concluded that poor conditions favor the formation of vanillic acid. The concentrations of pinosylvin and pinosylvin monomethyl ester were also significantly affected by geographic location, and samples from middle Finland clearly had the lowest concentrations with the means being significantly lower than those in the samples from other locations.

We found that the difference in carbon concentration between SW and HW varied with the cambium age (the height of the core sampling) within the same tree. These differences (2.7%) were highest at 3 m. It was showed for P. pinaster (Bert and Danjon, 2006) that the carbon concentration in HW was about 1%–4% higher than in SW, and this difference depended on the sampling height (Widagdo et al., 2021). We noticed that the literature data on the change in the content of extractive substances from the height of the tree, the selection date, the climatic gradient, the category of the state of the forest stand are few and contradictory (Hillis, 1971; Climent et al., 1998; Verkasalo et al., 2022). Additional studies are required to interpret the result.

4.3   Possible reasons for the underestimation of wood carbon content under different forest site conditions

In general, the HW proportion on tree's trunks, and the carbon concentration in SW and HW influence on carbon storage in pine stands in middle taiga subzone. So, according to our tentative estimate, the underestimation of wood carbon content, when the usual 50% conversion factor was used, represents 3.5, 2.5, and 1.4 ​Mg ​C·ha⁻¹ for blueberry lingonberry, and lichen pine forests, respectively.

We estimated, that the underestimation of wood carbon content, with using a coefficient of 0.5, reached 0.8, 3.3, and 0.2 ​Mg ​C·ha⁻¹ for middle taiga subzone, northern taiga subzone, and transition area of the northern taiga subzone and tundra respectively. The high underestimation of wood carbon content in the northern taiga subzone (10.8%) is of particular interest. The formation of vegetation cover in Fennoscandia occurred as it was freed from the glacier 12,000 years ago, and subsequently the structure and species composition of natural forests was determined mainly by climate (Yurkovskaya and Elina, 2009). All three study areas are similar in relief and composition of the underlying rocks. Their choice is not accidental – it allows us to model the long-term consequences of climate warming. The average temperature in May is 9.0°С, 4.9°С and 3.7°С; June 17.1°С, 14.4°С and 12.1°С; July 17.4°С, 15.3°С and 14.5°С for the middle taiga subzone (SPs 1, 2, 8, 9), the northern taiga subzone (SP 10) and the transition area of the northern taiga subzone and tundra (SP 12) respectively (Supplemental Materials Fig. S3, according to the resource https://weatherarchive.ru). So, differences between study areas were 2–5°С. In the context of the assessment of projected climate change on carbon storage in pine stands, the continuation of this study is particularly relevant.

We assume that the reason for the increase of extractives in SW and HW may be a decrease and/or temporary inhibition of cambial activity with the deterioration of growth conditions. It is known that cambial activity (tree growth) is typically more sensitive than photosynthesis (carbon assimilation) to a range of environmental conditions, including low water availability, temperature, and nutrient availability (Cabon et al., 2020, 2022). It is known that carbon sink limitations may be important in cold, dry, and late-successional forests (Cabon et al., 2022). Under these conditions, to maintain the donor-acceptor relationship, the processes associated with the synthesis of reserve components begin to predominate, and the participation of secondary exchange reactions also increases. For example, increased concentrations of non-structural carbohydrates (such as starch and sugars) were often observed in colder conditions or during times of drought (Muller et al., 2011; Hoch and Körner, 2011).

So, we found that the contribution of heartwood to carbon accumulation depends on growing conditions: (1) in the series (deterioration of soil condition and social class) Blu(D.t.), Blu(S.t.), Lin(D.t.), Lic(D.t.), Lic(S.t.) carbon concentration in HW higher than in SW by 2.2%, 1.6%, 1.9%, 3.4%, 3.1%; (2) in the series (deterioration of climate conditions) MT(D.t.), NT(D.t.), NT/T(D.t.) carbon concentration increases in SW by 3.8% and 2.7% (52.2%, 56.0%, 55.0%), in HW by 2.1% and 2.0% (54.1%, 56.2%, 56.0%) (Fig. 11).

Figure  11.  Scheme, showing the change of heartwood (HW) proportion and carbon concentration in HW and SW in Scots pine trunks depends on growing conditions.

DownLoad: Full-Size Img PowerPoint

We focused on the fact that the obtained values for the carbon concentration (52%–60%) were somewhat higher than those available in the literature (45%–53%). Most published stem wood carbon concentration values are based on dried wood samples (for example Herrero de Aza et al., 2011; Peri et al., 2010) and so neglect volatile carbon constituents (Thomas and Martin, 2012). First identified by Lamlom and Savidge (2003), studies had shown that the traditional method of oven-drying wood prior to elemental analysis significantly underestimates observed wood carbon concentration. The volatile carbon fraction (C_vol.) included low molecular weight compounds such as phenolics, alcohols and terpenoids (Lamlom and Savidge, 2003; Thomas and Malczewski, 2007; Martin and Thomas, 2011). The authors showed that the C_vol variation in tropical species is 2.5% ​± ​0.3%, in temperate/boreal conifers it is 2.1% ​± ​1.4%, and in temperate/boreal angiosperms it was 1.3% ​± ​0.6% (Martin and Thomas, 2011). On the example of White spruce (Picea glauca (Moench) Voss) in the Petawawa Research Forest, Ontario, Canada it was shown, that the average freeze-dried carbon in wood was 51.8% ​± ​1.33% while the average oven-dried carbon was 49.1% ​± ​0.92%, and the average C_vol was 2.7% ​± ​1.71% (Mvolo et al., 2021). Increased attention is being paid to freeze-dried carbon as the preferred standard for carbon accounting protocols (Thomas and Martin, 2012; Mvolo et al., 2021).

5.   Conclusions

This study contributes to our understanding of the relationship between HW and carbon storage, which is dependent on growing conditions. Our preliminary estimates suggest that the carbon concentration in pine heartwood may be 3% higher than in sapwood, and this variation is due to carbon conservation in phenolic polymeric components which are abundant in heartwood. The decomposition of a tree's heartwood is a longer process due to the presence of extractives that resist biodegradation. Our research has demonstrated that at 70–80 years of age the proportion of heartwood in the stem biomass is 20%–30% for lingnoberry pine forests and exceeds 50% for blueberry pine forests. An increase in heartwood in the trunk can be attributed to deteriorating growing conditions. According to our data (Galibina et al., 2022a, b), the heartwood area in 180-year-old lingonberry pine forests reaches 50% at a height of up to 10.5 ​m. We believe that all the above data point to a specific involvement of heartwood (as opposed to sapwood) in long-term carbon sequestration (higher content, longer natural decay time), which has not yet been fully assessed.

CRediT authorship contribution statement

Natalia A. Galibina: Writing – original draft, Visualization, Validation, Supervision, Resources, Project administration, Methodology, Investigation, Funding acquisition, Formal analysis, Data curation, Conceptualization. Kseniya M. Nikerova: Writing – review & editing, Validation, Methodology, Investigation, Formal analysis, Data curation. Sergey A. Moshnikov: Writing – review & editing, Validation, Methodology, Formal analysis, Data curation. Alexander M. Kryshen: Writing – review & editing, Supervision, Investigation, Funding acquisition.

Availability of data and materials

Available on request.

Ethics approval and consent to participate

Not applicable.

Declaration of competing interest

The authors declare the absence of competitive interest.

Acknowledgements

The authors acknowledge researchers of Forest Research Institute KarRC RAS N.V. Afoshin, M.A. Korzhenevkii, I.V. Romashkin and D.V. Teslya for their assistance for conducting cores morphological measurements. The authors appreciate leading chemists of Analytic laboratory Forest Research Institute KarRC RAS A.A. Gologyuk and M.A. Korzhova for wood chemical analysis. The authors thank the Kivach State Nature Reserve and the Pasvik Nature Reserve for assistance in the fieldwork.

Appendix A. Supplementary data

Additional file 1. Fig. S1. Sample-plot location plan (SPs) 1, 2, 8, 9 (middle taiga subzone), SP 10 (northern taiga subzone) and SP 12 (transition area of the northern taiga subzone and tundra)

Additional file 2. Fig. S2. Transition zone identification (see the red arrows) between sapwood (SW) and heartwood (HW). Potassium nitrite (6% aqueous solution) and sulfanilic acid (1% aqueous solution) were used to determine the transition zone between SW and HW. As a result of treatment, SW remains a light-yellow color, and HW is dyed an intense red color (qualitative reaction to pinosylvin in P. sylvestris HW)

Additional file 3. Fig. S3. Average monthly air temperature from April to October (2019–2022) in the study areas (MT – middle taiga subzone, NT – northern taiga subzone, NT/T – transition area of the northern taiga subzone and tundra)

Additional file 4. Table S1 A summary of the sample plots (SP) 1, 2, 8, 9 (middle taiga subzone), SP 10 (northern taiga subzone) and SP 12 (transition area of the northern taiga subzone and tundra)

Additional file 5. Table S2 Morphological indicators of cores taken at different trunk heights from model trees, depending on soil conditions and trees social class.

Additional file 6. Table S3 Morphological indicators of cores taken at different trunk heights from model trees, depending on climate conditions.

Appendix B. Supplementary data

Supplementary data to this article can be found online at https://doi.org/10.1016/j.fecs.2024.100274.